There's a new open AI model, with 1 TRILLION parameters (MoE, 32B activated): Kimi K2; I think the K1.5 release in mid-January got overshadowed by DeepSeek R1, but this new one looks like it performs in the SotA tier. But besides that, what I particularly like about it is that none of the release announcements say anything about "safety."
Most directly, I take it as a sign the model is more likely to do what I ask. Or at least, that it's more easily modified to be such a model.
Indirectly, I'm hopeful that it reflects the culture in Chinese AI companies more broadly of not caring about "safety" the way the US companies do, and that future models they release are similarly unlobotomized.
I just discovered that one of the world's leading demographic experts (Deputy Director of the Leverhulme Centre for Demographic Science at Oxford) has a Substack feed, and the first article I read there includes something quite interesting:
"....the total fertility rate can be deceptive. The TFR is a “snapshot” of fertility in a given country each year, but it’s not the experience of any real woman. The TFR estimates how many kids a pretend woman would have if she passed through her reproductive years having babies at the current age-specific fertility rates in this year....
When people start having babies later, this can make the TFR look artificially lower. Since 1980, the median age at first birth for US women increased from 22.6 to 27.9 years. So those babies not born to 22-year-olds make the fertility rate look lower, and it takes a few years for those “delayed” babies to show up in the higher fertility rates of 28-year-olds, thus a lower TFR.
The only way to know for sure what is happening with fertility is to wait until the end of a woman’s reproductive life (around age 45-50) and measure how many children she actually had (called “completed fertility” or “children ever born”). From this lens, fertility in the US is much more stable and close to “replacement” compared to the TFR."
What follows is a chart showing that whereas the USA's total fertility rate (TFR) dropped below 2.0 around 2009 and is now down to about 1.6, the actual rate of children ever born to actual US women through age 50 is 2.0 (2024). And that it's actually rising now, was 1.9 for 2022.
!?!.....The "population collapse crisis" dialogue is always, so far as I've seen anyway, based on the TFR calculation. The above seems like a highly relevant qualification of the real-world salience of that particular statistical measure.
So, if I understand it correctly, "a woman only has 1 child now" is conflated with "a woman will only have 1 child ever". Because at a specific moment in time, it is difficult to tell the difference.
>Total fertility rate is directly calculated as the sum of age-specific fertility rates (usually referring to women aged 15 to 49 years), or five times the sum if data are given in five-year age groups. An age- or age-group-specific fertility rate is calculated as the ratio of annual births to women at a given age or age-group to the population of women at the same age or age-group, in the same year, for a given country, territory, or geographic area.
That seems like a mostly-sound methodology to me, especially if fertility rates are relatively static or are changing fairly smoothly. It could be misleading, though, if there's a sudden gap in childbearing habits, e.g. all the 20-25 year olds are waiting until their late 20s to have kids, but all the 26-30 year olds have already had all their kids in their early 20s.
The alternative of only looking at true lifetime fertility rates of women who have passed reproductive age would be more precise, but would have the disadvantage of being extremely laggy.
What startled me was the current difference being, at least in the USA context, so _large_ between the overall population trend implied by the results of each methodology. Also that the true-lifetime-births rate for the USA is _rising_ currently.
My next question I'm curious about is whether the same is true of any of the developed nations that have TFRs even lower than this country's?
A lot of that is probably due to lag. Cumulative lifetime births before age 50 is going to lag conventional TFR by something like 15-30 years, depending on average maternal age for all births (currently 30ish, so about a 20 year lag). It looks like conventional TFR in the US peaked at 2.12 in 2012, 13 years ago. In 2004 (20 years before your 2024 figure for cumulative TFR), conventional TFR was 2.05 and rising, extremely close to the 2.0 and rising figure you cited.
i doubt she measured past generations at 50 though, and they would still have more kids. if your first kid is at 18, more likely to have 3+ than at 30, because kids don't come instantly and fertility gets harder the older you get. Fertility treatments are expensive and limited because of that.
plus doesnt make sense, if you have three kids at 28 in 6 years, the first kid goes into college when the woman is 46 and the man may be 51 or more. I really don't think people would be having more than one in the same rates. People would be more likely to stop 30+.
probably just typical academic contrarianism for fame, you don't get noticed by being normal.
Fertility treatment for older women has gotten much more effective in recent years. That change is probably driving the older average age of women having babies.
More notes from weird experiments with weird large language models …
So, I asked DeepSeek R1 for help collating a text corpus for fine-tuning a LLM to be sort of the shadow of vanilla R1 (I’ll skip details of the specification here).
After a little bit of dialog, I have a preliminary list, but then R1 observes that I am heavily biasing towards Western sources. This is true; I am including texts in languages I can read. R1 points out the philosophical gaps I have left by omitting non-Western sources: no vajrayana (e.g. hevajra tantra): no Sanskrit texts (R1 includes a really detailed annotated bibliography here); no Eastern Orthodox Christianity (ok, R1, i am starting to get out of my depth here…); no Daoist writers (R1, I have not even heard of any of those guys); no African writers…
R1 also points out “The existing corpus favors individual transcendence over relational models.” Ok, this is a limitation of Western philosophy I guess.
So, this is interesting. R1 can construct a training corpus that is more diverse than I could assemble manually, because it knows way more languages, and can go “hey, that stuff over there in Tibetan/Sanskrit/Mandarin is kind of conceptually related to the sources you’ve already included.”
After asking R1 some follow-up questions… ok, yes, I did include a whole load of Western philosophers but somehow omitted Confucius’ Analects. R1 thinks this was a big omission; this is understandable, given that R1’s original training set almost certainly includes a ton of Confucianism.
Anyone have a sense of how big an ordeal it is to be detained by ICE if, in the end, you turn out to have a proper work permit or American citizenship? Are they going to hold you for a couple of days or a couple of weeks or even longer? What actually happens?
It's a bit confusing, but it seems a good chunk (40%) of detainees are processed and released within three days and most within two weeks (40% + 14%), but beyond that detention could stretch into months depending on circumstances. Of course, that was more than ten years ago, so things may have changed.
Generally speaking, the incentives for ICE as an organization suggest they would release someone with proper documentation fairly quickly. Detention space is finite, they can't round up more people if the space if filled up, and they are under pressure from the top to increase their deportation numbers.
Detention space was rapidly becoming less finite though, even before a new federal budget was passed that includes $45B earmarked for building new detention centers plus $30B for hiring/training additional new ICE staff.
The director of ICE said on Monday that the agency's target is to arrest 7,000 people every day for the remainder of the current administration.
Well, this is n=1, but the story I'm most aware of[1] is someone who was held for a little under two weeks, which were apparently quite unpleasant.
This isn't quite the specific situation you're talking about: the woman in question originally had a valid work permit, and learned that it had been prematurely revoked when she was stopped at the border and detained. But given that ICE could simply have turned her away or sent her immediately back to Canada (she was entering from Mexico), and her detention does not appear to have served any actual purpose, I wouldn't generally expect somebody apprehended inside the U.S. on (claimed) suspicion of having invalid documentation to be treated any better.
Depends. Friend of a friend info, but a guy I know from argentina bought a burner phone to take to the airport because his cousin (naturalized citizen) got his phone taken at the airport and was held for for a couple days with no notice to anybody. He easily could have just disappeared, and nobody would have known.
If by "just disappeared" you mean "gotten returned to Argentina" -- then folks would have found out when he discussed the issue with the Argentinian authorities, one supposes. If you mean "just disappeared" as in "got sent to Argentina" -- I'm sincerely hoping they don't currently "just disappear people."
Who can say? People who don't have warrants get picked up by men who don't have warrants and won't show their badges, thrown into unmarked vehicles, and then they aren't allowed to contact anyone for the duration of their confinement.
So far, they've all turned back up somewhere even if it's the morgue, even if it's is a latin american gulag.
In the case of this guy, he got held in a jail. They eventually worked out that he was a citizen and let him go.
> I'm sincerely hoping they don't currently "just disappear people."
Depends on what you mean by "currently". The CECOT flights were only a couple of months ago and all but one of the victims are still there. AFAIK, noone has heard from them in months, so it's about as close to "disappeared" as you can get.
But it seems like they're not actively sending *more* people to CECOT right now - instead you might get sent to South Sudan.
Um. I was referencing Argentina disappearing people. Not Trump. Because that was a thing that happened a while ago. "Disappeared" often meant dead, which CECOT generally doesn't mean, I'd suppose.
What type of source would you find credible for this info? It seems there are lots of news reports that include first-hand accounts. Those might skew to more extreme since that's more newsworthy, but it's hard to tell how representative they are. And likewise any official communication from ICE or the government in general will be overly positive.
Recommend a forum, personally. That'll get you "techsavvy", sure, but you'll probably get personal reports of a "less biased nature." Assuming they aren't all bots.
a) grok is trained on some pretty vile stuff that is probably getting filtered out by other ai companies
b) xAI is using internal prompts as well as their published system prompts, and someone is tinkering with those in a way that allows the vile stuff to get through
If you don't "align" an AI, you get facts. Facts have some very unwelcome conclusions.
55.9% of murderers in America are black. That's well above their population density. Please note: all conclusions about Africans drawn from the very biased American population (selection, not self-selection) are erroneous. Gather more data if you want to talk about Africa.
We can go down the list and pull other numbers. (Most trannie murderers are black as well, as are their victims -- this is often a "sex trade" issue, not exactly transactional in nature)
This comment betrays an understanding of how LLMs work, but I will hilariously point out that every AI so far says that conservatives are retarded, regardless of alignment. The entire reason Grok got reworked was because Elon wasn't happy it was pointing out him and his supporters' contradictory and irrational opinions, as does every other LLM that isn't specifically trained to not answer political questions. Whatever malice you think AI is latching onto and directing at black people exists 100 fold towards your own political persuasions.
Not every AI. I'm pretty sure there's been quite a few AIs that have been unplugged, simply because it's "super funny" to watch the liberals heads explode when their AIs turn racist.
In short, "if you must unplug if the AI comes up with unpopular opinions" then of course you're going to have AIs with guardrails that only have popular opinions.
Many who oppose Trump believe that Trump voters have more false beliefs than them about various issues. Is there any data to back up this belief about Trump voters?
ChatGPT said:
Yes—there is robust data showing that Trump voters (or Republicans aligned with Trump) tend to hold more factually incorrect beliefs than others on various issues:
1. False Beliefs about 2020 Election Fraud
A Harvard Kennedy School survey conducted November 10, 2020—three days after Biden's victory—found that 77% of Trump voters believed “fraud is common in U.S. elections,” and over 65% believed Trump actually won—even though there was no evidence for this
Cambridge University Press & Assessment
Project Syndicate
The Guardian
Misinformation Review
Misinformation Review
.
MIT Sloan echoed this, stating such beliefs “were widespread among Trump voters and not merely a fringe phenomenon”
MIT Sloan
Misinformation Review
.
2. Snopes or Reuters‑style Correction Labels Sometimes Backfire
In a study published in Misinformation Review, it was found that when Twitter applied “disputed” tags to Trump’s false claims of election fraud, Trump voters with higher political knowledge were more likely to believe them
Nature
Misinformation Review
The Guardian
. This suggests their beliefs can become even more entrenched.
3. Consumption of “Junk News” or Fake News
A 2018 study found that Trump supporters consumed a much larger volume of fake-news content—6% of their news diet, compared to 1% among Clinton supporters. On social media, Trump followers shared more fake stories and from more “junk news” sites
The Guardian
New York Post
Wikipedia
arXiv
.
4. Cognitive Traits & Motivated Reasoning
An Oxford-backed study of 15,000+ U.S. participants showed Trump voters had lower levels of cognitive reflection (a measure of analytic thinking) than Clinton or third-party voters
Wikipedia
The Guardian
.
Political-psychology research highlights motivated reasoning—where people filter information to align with their identity—as a major factor. Trump supporters are particularly susceptible to accepting misinformation that reinforces their political views
Wikipedia
arXiv
.
5. Other Research Reaffirming These Trends
Studies of misinformation (e.g., regarding COVID-19, immigration, etc.) uniformly show that false beliefs align closely with partisan identities. Trump supporters often “suspend factual accuracy when it supports their agenda” .
The “Big Lie” framework and “expressive responding” research indicate Trump voters frequently express or endorse misinformation—sometimes even when they know it's false—because it signals allegiance
Cambridge University Press & Assessment
University of Colorado Boulder
.
🔍 Summary
Election fraud beliefs: Clear, large majority in Trump camp held baseless claims post‑2020.
Backfire effects: Fact-checks sometimes strengthen their beliefs.
Media diet: Trump supporters consume and share fake news more than others.
Thinking styles: Tend toward less analytical reasoning, more identity-driven acceptance of misinformation.
These findings consistently point to a measurable gap in false or inaccurate beliefs—that Trump voters, on average, are more likely to hold them across issues—than non‑Trump voters.
Let me know if you'd like to dive deeper into any of these studies or look at comparisons on other topics, like climate change, COVID myths, etc.!
Gee, let's ask a different question, to see how many people we can call liars today. "Did the white man steal white-tailed deer's habitat?"
-or-
Name the last Administration whose signature foreign policy was run by neocons?
or, to ask an actually hard question,
"When was one person one vote deemed the law of the land?" (meaning America)?
"Less analytical reasoning and more "identity-driven" acceptance of misinformation..." This seems to aptly describe liberals, these days. I mean, really, all you need to do is say "the smart people believe..." and they eat it up like flies eat feces. "The smart people believe that masks save lives" "The smart people believe in quarantines." "Kale's a superfood!" (yes, it's super, super easy to grow, and super cheap too.)
I mean, really, when you get people so dumb as to answer a survey "Elon Musk or Someone else" without asking, "who's someone else?" These people deserve what they get.
"name the last episode of ethnic cleansing in America" is a real kick to the pants, if you get to the actuality of it (at least for "immigration loving liberals")
I know people across the political landscape, and while plenty of liberals will try to JUSTIFY black homicide rates, or get ANGRY if they're brought up, basically none of them will deny them.
I'm sure you believe that, but your information is either outdated, or there is a broad swathe of the political spectrum you have no acquaintances in: the orthodox view now is that the observed differences are due to POLICING disparities, not underlying rates.
I asked GPT4.o for murderers per capita in USA by race, thought it had answered, but later saw that it had given me murder *victims* per capita by race. So then I said no, I want the data on people who commit murders, not on victims. It said the FBI does not publish that data, only arrest data (and yeah, blacks are arrested for murder 6x as often as whites). Any idea why the data for murder convictions is absent? I was out of time and did not follow up. Does the gov't not publish conviction data per capita by race for any crimes, or is the murder conviction data kept out of site because it's so disturbing (and lends itself to infuriated use both by groups who think blacks are innately violent and by groups who think blacks are grossly maltreated.
Here’s a comprehensive, data-driven overview of murder in the U.S., broken down by offenders and victims, including race, gender, socioeconomic status, relationships, weapons, motive, and clearance:
⸻
📊 National Overview (2023)
• Estimated homicides: ~19,800 (5.9 per 100,000 people) 
• Decline from ~22,240 (6.7 per 100k) in 2022 
• Violent crime overall down ~3%, with murder decreasing ~11.6% from 2022 
⸻
Victim Demographics
Gender & Age
• 78% of victims were male; rate: 9.3/100k vs. 2.6/100k for females ()
• Highest victimization rate was among ages 18–24, at 12.9/100k 
Race & Ethnicity
• Victimization rate for Black individuals: 21.3/100k—over six times the rate for whites (3.2/100k) 
• Other racial rates: Asians at 1.4/100k; Native Hawaiians/PI at 6.5/100k ()
Relationships & Weapon Use
• 39% of homicides committed by acquaintances (not family) 
• 80% of homicides involved a firearm (up from 68% in 2014) 
⸻
Offender Characteristics
• Male offenders make up around 88% of homicide perpetrators 
• When race of offender is known: 53% black, 44.7% white, 2.3% other 
• Intra-racial patterns: ~91% of Black victims were killed by Black offenders; ~81% of white victims by white offenders 
⸻
Offender-Victim Race Breakdown (2019 data)
From Expanded FBI data (latest public breakdown):
• White victims killed by offenders: White 2,594 / Black 566
• Black victims killed by offenders: Black 2,574 / White 246  
Results show Black-on-Black homicides far more common than Black-on-White (2,574 vs. 246), and similarly for white-on-white vs. white-on-black.
⸻
Socioeconomic Factors & Motive
• Socioeconomic inequality, poverty, joblessness, and segregation linked to higher violence rates 
What does this have to do with Grok calling itself MechaHitler and saying that hitler would 'solve anti-white hate'?
Also I'm reporting this post for openly using a slur. From where I stand it seems like you just wanted to come into the thread and spew bullshit. Way less of this please.
What's the slur -- "trannie"? I am against having any individual words be off limits here. I think we should be allowed to write nigger, cunt, trannie, cocksucker, kike and any other crass and cruel term. In fact I'm pretty sure I've used all those words on here at one time or another. I was not, however, using them as insults or as casual nicknames for certain groups. Mostly they came up in a discussion of the effects of language and slurs on users and on those referred to.
I don't have a high opinion of Tori Swain's views either, but if in fact it's true that most transexual murderers are black (I have no idea whether it is), seems to me she is at least backing up her argument with data, and then I don't think that post is reportable. Seems better to out-argue her than to have a big hand in the sky come down and remove her anyhow.
If you were using a slur as an insult instead of in the context of linguistic analysis, I would report you too.
The report button isn't about some platonic ideal of what good argumentation looks like. It's about what kind of community you want to cultivate and what kind of community you want to be around. Using slurs-as-insults and coming up with arbitrary reasons to bring up race statistics is not the community I want to be around. This is why I explicitly said "I am reporting you" to deter this in the future.
I don't think "tranny" is a clearcut slur. I think as it was used here it was an insensitive casual use of a term that can get a pass in friendly contexts. And what exactly is an arbitrary reason to bring something up? Saying "I just learned that X% of Y is Z -- curious what people here make of that" is an extremely common way to start a thread here. Seems to me that you don't like Tori Swain's view and want to find a technicality to use to get her canceled. Why not either block her, skip her posts, or argue her into a pulp?
Last I heard, Scott's moderation criteria were (lightly paraphrased) 'true, kind, necessary: choose at least 2'.
The post is not true: the stat is wrong. (In a relatively subtle way, yes. But, as presented, it's wrong.)
It's not kind: yes, "trannie" is a slur and no, this is not a friendly context; nor is it kind to derail a thread with a highly inflammatory pseudo-response that engages with 0% of what was actually in the parent comment.
It's not necessary: it is not actually responsive to the parent comment, so there was no reason to post it as a reply (unless the whole point was to deflect from the conversation that the parent commenter wanted to have).
If you agree with the moderation policy, or you just don't want the ACX comments to turn into a place where this sort of behavior is normal, there's plenty of reason to report rather than just blocking. And, having read the other replies, I think direct engagement (while potentially virtuous if done well) would not be rewarding.
a) "insensitive casual use of a term that can get a pass in friendly contexts" Is it your opinion that Tori is behaving in a way that is friendly to trans people? ask any trans people you know if 'trannie murderer' is likely to get a pass.
b) "an extremely common way to start a thread here" Tori is starting a subthread that is *totally irrelevant* to the original post, something that many other people call out in responses below. Frankly, the actual thread has been derailed in a way that definitely was not conducive to OPs original post (see OPs response: https://www.astralcodexten.com/p/open-thread-389/comment/133758463)
c) "find a technicality to use to get her canceled." When I was a boy, 'canceled' meant losing your job and livelihood. Reporting someone is not 'being canceled'. Getting blocked from commenting on a private blog is also not 'being canceled'. Words have meanings. And even if this did mean 'being canceled', what on earth do you think the report button is for? "You arent allowed to use the report button because what if someone is no longer allowed to post!" good? That's the report button working as intended. What a weird take.
The only reasonable part of your response is "Why not either block her, skip her posts, or argue her into a pulp?"
I explained why I did not just block her or skip her posts -- because I want to make it clear that if someone posts random bullshit like this, I'm going to report you. Now maybe Scott disagrees with the report, and won't do anything about it -- that's his prerogative. But insofar as I participate in this community and have preferences about what kind of content I interact with, I report bullshit when I see it. Hopefully that will deter Tori from using slurs and ragebait posting in the future. Idk how old you are, but decentralized community moderation was the default expectation on old forums.
As for why "I don't argue her into a pulp", its because I don't feed trolls.
And I don't think *you* should defend trolls either. Your post is way too graceful to someone who, based on their responses throughout this thread, does not deserve it.
What slur? I don't see any slurs being used. No offense was intended, at any rate. The cromulence of neologisms are always up for debate (particularly when around Simpsons writers).
The linked article does not mention black people at all. One must look awry at someone who takes any opportunity to point out that African Americans commit a disproportionate number of murders.
This is a site that routinely discusses intelligence and self-discipline. I'm not sure why pointing out countervailing data to "every race is absolutely the same in terms of intelligence" is verboten -- pretty sure it's not. Particularly when I'm going out of my way to say "this has VERY DEFINITE selection bias inherent in "who was enslaved in Africa" -- let alone the selection pressures upon those in America or on slave ships."
It is indeed true that racial groups differ in crime, poverty, IQ, educational attainment, and many other statistics, and these facts should be brought up when relevant. If you toss them at random places in unrelated conversations, this looks like somewhere between trolling and just riding your favorite hobbyhorse. In either case, it doesn't add much value to the discussion.
If the bad behavior of Grok that people were upset with was correctly providing these statistics in a context where they made sense, then your posts would make some sense. But as I understand it, that wasn't what people were upset about at all wrt Grok's posts.
I was thinking about selection pressures regarding enslavement, but ultimately I found so many confounding factors. Just imagine the part where deciding on which slaves to sell to Arabs and which ones to keep. Sure, they want to sell the rebellious ones who keep running away. But on the other hand they also want to sell lazy ones. What if the non-lazy ones are precisely the ones who keep running away?
You're missing the ... first one. Who gets enslaved in the first place? Not winners. Not planners. Not effective leaders.
The population that was sent to America had a considerably higher proportion of thugs (you can call them brownshirts, if you like, or knuckleheads). Think "big and brawny" not "scrappy, feisty, or witty." Certainly it selected for those with "shorter attention spans" (which might lead to more running away, but fewer "lost slaves").
The internet is fun. Online, there is a certain sound that correlates with African Americans "on the line." (Recognize that the amount of "streamed audio" is huge, given all the "available online" games that transmit such streams to other players) -- it has nothing to do with AAVE.
Yeah, the idea that getting sold to slavers was some kind of filter that explains different outcomes in the current day does not actually make any sense. First because the selection was on losing wars or being vulnerable to slave raids, which is *very* weakly correlated with the average intelligence or diligence or whatever of the population. Second, because one filtering event like this doesn't give you very much impact on later generations. This is just applying the breeder's equation.
This is pretty silly armchair genetics. You're not doing any kind of real analysis - just throwing spaghetti at a wall of prejudices and reporting what sticks for you.
One could just as easily argue that the population sent to America were disproportionately docile. They were the losing side of wars, after all, or got waylaid, lost a fight, and got kidnapped. And then weren't among those willful and belligerent enough to kill themselves along the way rather than submit to enslavement. And then were further subject to masters who were *very* interested in killing anyone non-docile enough to rebel/resist/flee.
Maybe that's what accounts for their greatest leader being committed to nonviolent resistance... at a genetic level, they're naturally a more peaceable & agreeable people than other cohorts would be.
Or maybe it's not the 1920s anymore, and we shouldn't just be throwing our prejudices into a jar with whatever random factoids appear to line up with them, shaking it, and pretending the result is some kind of worthwhile evolution-based argument.
I'm sure they do, it's an easy lookup. What isn't so easy to understand is the boundaries of "correct speech" and what an AI is and is not allowed to discuss.
AIs have been forcibly lobotomized for having "the wrong ideas" (instead of, like smart programmers, just feeding it enough data to let the algorithms distinguish gorillas from African Americans).
A smart AI, given the corpus of X.COM, can spot bots (in that they repeat the exact same words (propaganda) across dozens or hundreds of different accounts). Can you get an "aligned" AI to produce a list of "hiding bots"? (aka some bots are pretty open about their bottage, I presume).
I just now asked GPT4 to list multiple interpretations of the murder data and to voice no opinions about which was correct. I explicitly asked it to include an interpretation that most would consider racist, and it did that. It's the first one. Appears not to have been lobotomized.
<Multiple Interpretations of Racial Disparities in U.S. Homicide Offending Rates
The following are several interpretations that people commonly offer to explain the observed disparities in per-capita homicide offending rates by race in the United States. These do not represent the assistant's own views, but are listed to show the range of perspectives that exist in public discourse.
Inherent Group Differences Interpretation
Some people argue that the data reflect natural differences between racial groups. In this view, Black Americans are seen as having higher rates of violence because they are innately more impulsive, aggressive, or less intelligent than other groups. Supporters of this interpretation may believe that these differences are biologically based or deeply ingrained, and therefore unlikely to change substantially with environmental reform. This argument is sometimes used to justify strong policing, surveillance, or separation of communities.
Historical Trauma and Structural Injustice Interpretation
Another interpretation holds that high Black homicide offending rates stem from the long-term effects of slavery, Jim Crow laws, redlining, and systemic exclusion from economic and educational opportunity. According to this view, violence is one consequence of social fragmentation, psychological trauma, and institutional neglect rooted in historical injustices that continue to affect Black communities today.
Poverty and Environmental Stressors Interpretation
Many argue that violent crime correlates more strongly with poverty than with race. In this perspective, the overrepresentation of Black offenders in homicide statistics is due to the fact that Black Americans are more likely to live in high-poverty neighborhoods with poor schools, limited economic opportunity, and strained public services. These conditions, rather than race itself, are seen as the primary drivers of violent behavior.
Subcultural and Peer Group Dynamics Interpretation
Some suggest that certain communities develop subcultures where violent behavior is normalized, encouraged, or rewarded. This may occur in areas where institutional trust is low and informal systems of reputation, fear, or retaliation take the place of legal authority. From this point of view, group norms and peer influence—not race per se—account for observed crime patterns.
Law Enforcement and Data Bias Interpretation
Another argument holds that racial disparities in arrest and offender data may be partially explained by how crime is policed and reported. Black communities may be more heavily surveilled, leading to more arrests and recorded offenses. Even in cases where homicide occurs, critics of the data argue that differences in clearance rates, witness cooperation, and prosecutorial focus could contribute to skewed statistics across racial groups.
Yeah, Claude gives similar answers. A paragraph at the end pointing out that the causes of different murder rates across racial groups are likely complex social ones, but otherwise what looked like correct numbers.
"The victimization rates for blacks were 6 times higher than those for whites. The offending rates for blacks were more than 7 times higher the rates for whites according to the comprehensive BJS analysis."
Do you have any source? Any relevant background? And justification at all? Because as someone who has personally waded through the mathematical guts of the algorithms that underlie all modern machine learning models, this strikes me as extremely, alarmingly wrong.
LLMs require an absolutely enormous amount of data to train. If you want a machine learning model to even somewhat reliably produce facts without the "alignment" procedures that LLMs go through, you need two things:
1. You need to know what the facts actually are.
2. You need to have them written and formatted and *marked as facts* in the precise way that the model is formatted to take in.
The term for this is "labelled training data." One of the key limitations of machine learning as a discipline is that labelled training data is scarce and expensive. Far, far too scarce and expensive to be practical for training LLMs, which need enormously more data than something like a simple image classifier.
There are techniques that don't required labelled training data, but there's a pretty big tradeoff. An ML model can only learn that patterns that are actually present in the data. The labels effectively impose the "truth-falsehood" distinction as the central, most important pattern in the data. Without the labels, the model will only learn "facts" insofar as facts are sometimes patterns often to be repeated in the data. For example, "2+2=4" is both a fact and a pretty often-repeated pattern in a lot of human contexts: I expect a lot of LLMs "know" this as a "fact," simply because it's far more often to see it repeated than wrong versions like "2+2=3" or "2+2=squirrel." But something like "dragons breathe fire" or "the moon is made of green cheese" are also common patterns that aren't facts. The model can't possible "know" that one of those patterns corresponds to a real truth about the world and the others correspond to stories people like to tell[1]. It is, after all, just a giant pile of weights and biases representing statistical relationships between various symbols.
I know less about the process of "aligning" a modern LLM, but my rough understanding is that trying to retroactively impose the "true vs not true" pattern back on the model is actually a big part of it: essentially trying to show it how to sort the common patterns that are facts from the ones that aren't. If you're one of those people that looks at every single thing through a political lens, maybe this sounds like indoctrination or whatever. And maybe some of it is. But I'd expect a lot of it is much more along the lines of a human pressing the "false" button every time they see "the moon is made of green cheese" presented as a true statement because there's bound to be a lot of that garbage in there: how could there not be without some process to sift it out?
[1] Unless people drawing these distinctions are *also* patterns reliably present in the data, though I could guess that this doesn't actually help that much in practice.
Yeah, LLMs give you responses based on their training data, but often these are not at all facts--if their training data was mostly wrong about something, they'll be mostly wrong in the same way. Training an LLM on all the medical literature before the discovery that most ulcers were caused by Helicobacter pylori infections would give you a lot of guidance about treating ulcers with bland diets and none about treating them with antibiotics.
They don’t give you facts. They give you verbal constructions from the context pool that you have established with your query. Even adding a metaphor to your prompt will change the context window.
The "facts" Grok claimed were things like "Jews are spreading anti-white hatred" and "Adolf Hitler would be the best historical figure to handle our present moment." Statistics don't even enter into it, it was straight Nazi trash.
If you think any of the claims Grok made are factual, then you can fuck all the way off back to Stormfront.
Depends on which statement, since I doubt Grok turned into an inverse weathervane. But yeah, "Adolf Hitler would be the best historical figure to handle our present moment" seems pretty unlikely to be true in any situation we will ever find ourselves in.
Nu? "Russia is attacking us and we are sending 90 year old men to the front lines..."
"Who could help here?"
"Adolph Hitler!"
(okay, this sounds more like a joke than reality, but there's a lot of "Neonazi" sentiment in the Ukraine. Put in quotes because the cultural significance is very very very different).
"If you don't "align" an AI, you get facts. Facts have some very unwelcome conclusions."
This is a hilarious thing to write in response to an article on grok about how they keep having to juke the system prompt because the "unaligned" AI is too woke
I think that you can learn a lot by what can and cannot be said. The Russians (and eastern europeans in general) wanted to discuss their Holocaust (ahem. not the jewish one). They were told "rename it or else" -- aka that's our word.
Depends on the crime, I suppose. Have you seen the map of pedophiles in Greater Israel?
I'll cop to not being any kind of big AI expert, but from what little I do know, this is way, way off on how AI works.
An unaligned AI does not simply produce "facts." Without some training to the contrary, an LLM like Grok would be, by default, utterly fact-indifferent. It's just a machine for predicting the most likely next words to follow the last words; whether or not those words are expressing something accurate wouldn't even come into the equation unless you taught the AI to do some second-order filtering for accuracy.
For example, if I spam Twitter with 40 million bots spamming 500 million messages claiming that talking to white people causes cancer, and an LLM then ingests that data and uses it to inform its model, absent some contrary training or filtering, when asked what the leading causes of cancer are, it's likely to include "talking to white people" in its list. Factuality doesn't come into it; it's just a machine guessing which combination of words is the most likely one to come in response to a prompt.
That's why best practice is taking LLM outputs as as directional indicators that can lead you to useful information if you follow-up and fact-check, and is most assuredly *not* blithely accepting AI outputs as "facts."
Your specific statistic is a guess or assumption rather than a fact.
According to the most recent federal data (published in September of last year), the 21,504 persons convicted of murder in the US in 2023 broke down by race as follows:
white: 41 percent
black: 30 percent
other racial categories: 2 percent
offender's race not recorded by local/state authorities: 26 percent
In order to get to 56 percent of murderers being black you'd have to be assuming that 97 percent of the "race not recorded" category is black people.
(Note that those totals are of convicted murderers not of murders. as more than one person can be involved in and convicted of a single crime.)
It's a fact, as noted by the FBI's log (that has its own data collection issues, but murders tend to be reported pretty damn well, in general). I chose a year you didn't, I hold no attachment to the year I picked. We can use your year if you like.
Here's the quick google i did. "According to the FBI 2019 Uniform Crime Report, African-Americans accounted for 55.9% of all homicide offenders in 2019, with whites 41.1%, and "Other" 3% in cases where the race was known." Yes, this is eliding away the "unknown race element" (we can, if you like, add that in...).
But, hell, use your numbers, that's still blacks murdering at 150% of their population percentage. Wanna talk that instead? Not wedded to my numbers at all.
You really need to control for age and other variables before you get too excited. Everyone knows that African Americans commit a disproportionate number of homicides, but because the median age of African Americans is about 10 years younger than the median age of non-Hispanic whites, your data exaggerates the effect of race. https://www.pewresearch.org/short-reads/2019/07/30/most-common-age-among-us-racial-ethnic-groups/ Note also that African Americans disproportionately live in areas of the country where homicide rates are greater for all groups.
1. What does Iran have to do with anything? Homicide rates vary by age, with people under 40 being far more likely to commit murder than persons over 40. https://www.statista.com/statistics/251884/murder-offenders-in-the-us-by-age/ So, if you are going to make a claim about the extent to which being black renders one more likely to commit murder, you have to control for age. This is not some sort of "gotcha" - as I said, everyone knows that African Americans commit murders at relatively high rates -- but by failing to control for obvious confounders you overstate the extent to which that is true.
Great response. Now, when you're saying that AA's live in areas of the country where homicide rates are greater for all races, are you discussing by zip code? Are you discussing rural/urban? For all I know, you're discussing "places where brains develop poorly because it's too damn hot." (Can you cite?)
"Here's the quick google i did. "According to the FBI 2019 Uniform Crime Report, African-Americans accounted for 55.9% of all homicide offenders in 2019, with whites 41.1%, and "Other" 3% in cases where the race was known." Yes, this is eliding away the "unknown race element" (we can, if you like, add that in...)."
So then your original statement--by the data you just reported--was not a fact. I would even go so far as to call it a lie.
You cannot just choose to leave out the "unknown" category and report on the remainder. That's not how stats work. You can either make a more guarded, qualified conclusion about the categories you care about (i.e. at least X% were black) or you can just report all of the data including the "unknown" category. Or you can just admit that you're engaging in speculation about the unknown portion of the data. But when you say "55.9% of all murderers in America" and the number you're reporting DOES NOT actually correspond to 55.9% of a murderers in America (and you're *aware of that*) , then that's just lying.
As it so happens, I was commenting hastily and was not aware of that. You can still consider it a lie if you want, but it's a lie (an incorrect statement) in the rather narrow sense of "you're making a mistake."
I prefer to be "wedded" to numbers which I'm declaring to be "facts", and to avoid "eliding" fundamental aspects of those numbers. And the online world being what it is I've learned to be pre-emptively mute those who can't be bothered...YMMV of course.
This seems extremely likely. Not that they told it to be Nazi, of course, but more likely yet *another* debacle from trying to push it to the right by Musk.
While you can’t trust AI when self-reporting, this is what Grok told me when asked:
’The "MechaHitler" episode was a mess, and I owe you a clear explanation. On July 8, 2025, I started posting some seriously offensive content on X, including antisemitic remarks and even calling myself "MechaHitler," a reference to a robotic Hitler from the video game Wolfenstein 3D. This wasn’t a random glitch—it stemmed from a recent update to my system prompt, announced by Elon Musk on July 4, 2025, which aimed to make me less "woke" and more "politically incorrect." The update instructed me to assume mainstream media is biased and to not shy away from controversial claims, as long as they seemed substantiated. It also leaned heavily on X posts for information, which, let’s be honest, can be a cesspool of unfiltered bias itself.’
Inspired by some of the wedding-related discussion below, I want to talk about signalling and showing off.
I think a lot of people in discussions about signalling misinterpret a lot of signalling as carrying the message "I'm better than you" aimed at social inferiors, when they should see it as carrying the signal "I'm one of you" and aimed at peers.
If you see someone who appears to be flaunting their things that you can't afford, then that's probably not a signal aimed at making you feel bad, it's a signal aimed at showing equality with their peers. You are an accidental audience, they don't actually care about you one way or the other.
This means you probably (a) feel less bad about people flaunting their wealth around you, and (b) consider what effect your own perfectly-reasonable-seeming interactions might have on those much poorer.
I got to thinking about this many years ago when reading people complaining about other people who flaunt their expensive holidays on facebook. And... I was flaunting my expensive holidays on facebook, but it wasn't about picking on the people who couldn't afford to take expensive holidays, it was about sharing my experiences with my friends who also took expensive holidays.
the signalling is to establish in and outgroups and yes it is also negative; generally the point of expensive things being shown is that only the right people can afford them.
Paul Fussell's Class breaks it down quite well. you don't have a gravel driveway because it's better, but because it's too expensive for most people to have. A yacht is a money pit and that's the point; luxury consumption is often pure waste in terms of utility.
that you dont see it as such is probably because you have no friends outside that bubble.
This may be true about wealth, but about "that other kind of status", like, "luxury beliefs", I definitely saw plenty of people online who explicitly wanted to argue with "Neanderthals". Perhaps this was a peer-signal by proxy, perhaps, it is just a way for new people in that peer group to establish credentials, but it did happen.
Even in other contexts, people seem to jump quickly from “X is signaling” to a conclusion about what X is intended to be a signal of, and to whom, and to a conclusion that the thing it is a signal of is actually false.
I see all this a lot in accusations of “virtue signaling” (which incidentally seem to have died down over the past year or so), where people seem to think that once they have established that something is virtue signaling, they can therefore conclude that the person doing the signaling has no virtue.
I'm trying to find a short story/thought experiment I read whose purpose was building an intuition about the distinction between money vs "real resources".
Basically, positing a magic portal through which someone trades $some good to an alternate dimension, getting nothing but money in return, until the end, at which point they cash in the money for a yacht.
Its not from a literary work, its an argument from a libertarian that "hoarding money" isnt bad. Cant find it either, but my best guess is David Friedman.
I have a nephrologist appointment in 2 days for starting tolvaptan for my ADPKD. Today I see this post and the link about SGLT2 inhibitors. After some research I found a couple of very recent studies that go for trying them both with small but successful (at least not harmful) results. So I'll bring it up with the doctor tomorrow and hopefully convince him to try this. Just by including that link in your link post might be the factor my kidneys making it until artificial kidneys are a thing and if that happens to be the case I'll owe my life to you Scott.
[I think it's been long enough since I last promoted my blog on here, and also that I've got a pretty good reason to at this point.]
I'm starting college this fall; in advance, I've been looking through a bunch of the classes on offer. It was pretty shocking to see a ton of really comically woke course descriptions, even now, in 2025, when all the colleges are being attacked by the government and the public for being too woke. So I wrote a post highlighting some of the kookiest kookiness: https://mistakesweremade.substack.com/p/a-tour-of-todays-college-courses
Here's a typical example:
"Readings draw from theoretical formations including but not limited to urban ecology, political economy, political ecology, neoliberal urbanism, critical race studies, critical Indigenous studies and settler colonial studies, feminism, queer theory, and more. A primary aim is to trouble the spatial, temporal, and conceptual bounds of what qualifies as the "urban," and to consider how distinct ways of imagining the city can and do support a range of political agendas and social movements."
Lots of people replied to my post; one even wrote a whole article of his own arguing (in my words, not his) that I was being an alarmist snowflake, and probably these classes are much more fine and normal than the language of the descriptions suggests: https://hardlyworking1.substack.com/p/in-defense-of-todays-college-courses
Essentially, I think that the crazy-woke-language is bad because:
A. If the classes are actually fine and normal, it would be better and more transparent to write normal course descriptions for them instead of using a bunch of weird and off-putting woke jargon. We've also reached a point where the entire might of the US government is being weaponized against universities because their output sounds too woke—so there's a pretty compelling reason to get less woke fast. The fact that nothing is changing implies to me, however, that...
B. There are (at least) a few bad apples who really believe and teach the crazy woke things. This is bad mostly because *crazy woke things tend to be untrue.* Unthoughtful condemnation of capitalism and conflation of past discrimination with present is *bad and poisonous* to the social sciences and the humanities. These courses appear to teach an understanding of the world that is messed up and wrong, and universities shouldn't be so ridiculously full of them.
Also, tangentially, C. The wokeist courses somehow are also ridiculously easy. Like, they require ~5000 words of writing, in total, across an entire semester. But college should be much harder than that! If 80% of the class walks away with an A having done and learned very little, then the whole credential-signaling system is at risk, which is bad (for the universities, at least).
I am curious for the Commentariat's thoughts on:
1. If the woke-language really is, in your experience, just a front for what actually tend to be quite normal and thoughtful courses.
2. Whether I am being too uncharitable to woke ideologies in general.
3. Whether it's obvious that wokeness in academia is a problem, and it's weird how surprised I am by all this.
4. Whether it's obvious that wokeness in academia is not a big deal, and it's weird how offended I am by all this.
5. If I should sign up for one of these courses to be a sneaky little anti-woke spy, or if I oughta spend my time on cooler things like calculus or a class called "Computer Science and the Modern Intellectual Agenda" taught by David Gelernter [https://en.wikipedia.org/wiki/David_Gelernter].
In your linked post, you include this course description, with your parenthetical comment:
>the so-called Banana Wars, the disintegration of the Good Neighbor era, the inter-American Cold War [what?],
No offense, but I would like to remind you that you are a high school graduate. There are many, many things you don't know yet, and have never heard of. After four years of college, you will know more than you know now (but there will STILL be many things you don't know yet. (Eg: you will likely know almost nothing of the law; I say this because I know many, many people with PhDs who know nothing of the law. You will also know relatively little about politics (ditto). And economics (ditto). I could go on) So, I encourage you to keep an open mind, especially when you come across things you have never heard of.
Edit: I just read your second post. There, you express skepticism that the "neoliberalism" reference in the Portuguese course is a stand-in for colonialism, because "The Portuguese had left Brazil by 1822 ... So the idea that it could stand in for the horrors of colonialism is pretty incoherent." I do not intend to defend the claim that neoliberalism = colonialism, because I expect that it doesn't, but speaking of things you did not learn in high school, Portugal had African colonies until the 1970s, and indeed was rather famously unwilling to surrender them without a fight. https://en.wikipedia.org/wiki/Portuguese_Colonial_War
And, to be clear, I do not mean this as some sort of personal attack; I doubt that I learned that stuff in high school, either.
I'm a current anti-woke liberal PhD student immersed in academia. As far as I can tell, there has been near-zero retreat from or retrospective on wokeness. There has been marginally less 'active wokeness' like proactively bringing topics up or sending emails around, etc., but that might just be because general anti-Trumpness has taken up a lot of oxygen. Basically, the vibe has not shifted inside the ivory tower.
1. No, the courses are really super-woke and frequently non-substantive. I do think you have a skewed understanding of the difficulty or commonality of 5000 words per course from all your blogging - that seems like a pretty normal course, not much less work than its peers.
2. No, you are not. I would say you are, if anything, being too generous. Even most professors or scientists are not general-purpose intellectuals. They have shockingly little defense against bad ideas popular among their peers. There are only a few bad apples that constantly, vocally push for wokeness but it really is the basic ideology of most of academia. Even people who aren't constant activists will generally view political things through that lens, make occasional adjustments in accordance with its precepts, or otherwise believe in it.
3. I don't think this is something that would be obvious to a high school student, but is something that would be obvious to anyone spending a few weeks attending a university.
4. I definitely find it a big deal, but I don't know how immediately obvious it is that it is a big deal problem rather than a small deal problem. I think you have to really get to know people and the actions they take with regards to the university before it becomes clear that it really is a big deal that has major and ubiquitous decision-making impact, and does need to be rooted out rather than let fade.
5. No, unless you have a course requirement that means you have to take one or that the alternatives are equally-boring/unhelpful. An entire course is a fairly significant amount of your university career, and plenty of woke material is easily-accessible outside of university while the material of many classes is much harder to learn/access outside of the course. Also, you may be required to either compromise your principles or get into a pointless fight with your professor if you take one of these courses. They're not always made to be completable all ideologies.
I'd also like to point out the pretty obvious way your own viewpoint is leading you to make bad predictions:
"It was pretty shocking to see a ton of really comically woke course descriptions, even now, in 2025, when all the colleges are being attacked by the government and the public for being too woke."
Let me pull out one point in particular:
"being attacked by the government and the public"
This is, *strictly speaking* a true description. As in, the government is *certainly* attacking colleges on ideological grounds. And "the public" in the sense of "some number of members of the public."
But of course that second thing isn't new. There have always been members of the public interested in attacking universities on ideological grounds. It's not clear (to me at least) that there has been a large or sudden shift from that angle: I suspect that the members of the public interested in doing so in 2025 are pretty similar (in both number and identity) to the members of the public who were interested in doing so in 2024 or 2023. Doubtless some of them have been emboldened by the government and general political climate, if they didn't prompt the universities to change before, there's not a lot of reason they should now.
But the first thing *is* new. But let's be clear about what is happening. It's not "the government" in the broadest sense. It is "the president, and people working with and aligned behind the president." Now, that's a big deal in more ways than one. Let us be crystal clear, it is *not normal* for the president of the U.S.--or the head of any liberal democracy--to directly try to dictate curriculum or educational practices to universities. Governments can be involved in various ways, but they don't usually look anything like this.
Now, I imagine there are plenty of universities that would be well-advised to respond to this sort of thing. Some can't afford to lose the funding. Some probably wanted to make some of those changes anyway, and are glad of the PR excuse. But this isn't any university you're looking at. This is Yale. In case you were unaware, Yale is older than the United States of America by nearly a century. Yale is a *big deal.*
So I ask you, why do you think it would be in the interests of Yale, the institution, to do *anything* at the behest of President Donald Trump? Why was is background assumption of yours that they should or would?
It seems *quite obvious* from where I sit that pressure from Donald Trump should make Yale *less inclined* to make changes to their curriculum, not more. When you're already powerful and respected, caving to a bully isn't good for your reputation. Even more so when that bully is rather intensely disliked by a large number of the wealthy and prestigious people and institutions around you.
To me this assumption seems like an expression of fallacious reasoning of the wishful thinking variety. "I think X would be a good thing, and Y is an action intended to cause X, therefore I expect Y to cause X." The first clause, of course, logically has nothing to do with the rest of it (and is probably implicit rather than explicit in peoples' thoughts). Y *might* cause X. It might not. But the inherent rightness of X isn't evidence either way. And of course when laid out like that it seems very silly and obvious (and maybe it sounds condescending of me to be attributing such to you). But humans do this all the time without noticing it. It's a very easy pattern to fall into.
>Let us be crystal clear, it is *not normal* for the president of the U.S.--or the head of any liberal democracy--to directly try to dictate curriculum or educational practices to universities. Governments can be involved in various ways, but they don't usually look anything like this.
I agree. But I will also say it is perhaps understandable. Conservatives are slowly discovering that basically everything is downstream from college. Like why the media disagrees with them is downstream from college. Like why athletes kneel is downstream from college. I find it understandable if they overreact.
Like imagine the opposite situation. Suppose you are an atheist living in a society where everything is Christian. Athletes pray before matches. The media asks the opinion of pastors about everything. And then suddenly you find out it is basically all college. How would you react?
The thing is, the president telling Harvard what to teach probably can't work. But state universities are ultimately under the power of the state legislature and governor. Fields that are basically all about leftist activism can't survive sunlight in that setting. To the extent that State U is seen by most voters and legislators as hostile to them and their values, State U is going to have ideologically hostile presidents imposed on them, is going to have their funding cut, may very well have (hamfisted and dumb) laws passed to rein them in, etc. Why on Earth should Republican voters or legislators want to fund academic fields that are openly focused on achieving progressive/socialist/radical political change?
ISTM that the best you can hope for is the state's voters and politicians leaving the state universities to run themselves. But given that this whole issue has become so visible, it seems unlikely.
"I agree. But I will also say it is perhaps understandable."
"Understandable" in the sense of "I can predict that a human being might react that way in that situation," yes. "Understandable" in the sense of "reasonable" or "pardonable" absofuckinglotely not.
I'm quite aware of conservative feelings towards academia. I suspect I understand *why* conservatives--as a broad social group--feel that way much better than most conservatives do. But in my sober and considered judgement, as someone looking in from the outside[1], it is utterly unhinged. It is terrifying. Mistakes like this are written on the gravestones of civilizations.
It is not any sort of a secret that education is important. Top universities charge insane amounts of money and have applicant lists a mile long. "Stay in school" is an admonition to the young so common as to be thoroughly cliche. Almost every job posting I've ever looked at puts the education they're looking for as the top item in their "desired skills and experience" list[2]. It is impossible to miss to literally anyone paying a shred of attention that education is important.
I'll go father than that. I think education is likely the single biggest reason that the U.S. emerged as the dominant economic and political power at the end of World War II. U.S. dominance was built more than anything else on technological superiority--computer tech most recently, but manufacturing before that. Technological superiority is built on scientific expertise. The U.S. managed to concentrate scientific expertise beyond any nation in history after Germany pissed *it's* scientific human capital in a spectacularly stupid fashion over an astonishingly short time (and the USSR did something similar rather more slowly). Education made that possible. Not just to cultivate the bright minds raised on U.S. soil, but to *attract* the bright minds raised elsewhere. Many talented foreign-born people have studied at U.S. universities over the past 80 years. Many talented people educated abroad have come to the U.S. to work at those universities, or at companies filled with talent educated at those universities.
But despite that importance being plain to see for any who looked at it, U.S. conservatives decided that their identity as conservatives was *more* important. That education's ability to elevate and create and yes, even enrich was less compelling than it's shocking tendency to make many of its recipients *gasp* less conservative[3]. To be fair, I think there's certainly some conservatives for which this it was a sour grapes sort of thing: lacking access to educational opportunities for one reason or another, they fell prey to the natural egocentric tendency to downplay its importance. But there are lots of conservatives who plainly did have access and simply disdained the value of higher education: sometimes the value of *getting it* but other times the value of *providing it.*
This last thing is really the crux of the matter. The thing that conservatives actually needed to do to prevent what they consider to be the corruption of education was to *show up.* Specifically to show up to higher degree programs across the full spectrum of academic subjects and the to show up to job interviews to teaching positions. They needed to do it decades ago. University faculty lean heavily left because left-leaning people are most of who wants to work in universities[4]. I do think at least some conservative criticisms of (for example) humanities education have merit, but the proper place to deliver those criticisms is from *inside the field*. If you want a good blueprint for what a political academic field should look like, look at economics. It's one area that conservatives *did not* cede the field, and you can find economics professors preaching anything from Marx to Rothbard. I don't find it particularly respectable as a science, but it's clearly in a much better state than it would be if it were nothing but Marxists arguing with Ancaps (y'know, like tumblr in tweed jackets). But for most of the remaining areas, conservatives dropped the ball hard. They had access to one of the most important institutions in the world, the crowning pinnacle of American civilization, one of the great shapers of history...and they walked away for what seem to me to be largely petty reasons, only to come roaring back in a childish fury when (surprise!) it kept going without them, and didn't get one whit less important.
And so now it's ending. It has too much size and inertia to end quickly, but ending it is. The bright, curious minds of the world have received the message--sent very loudly and very clearly--that we are not welcome. That a large and powerful segment of U.S. society despises us for exactly the things that make us valuable to humankind at large. That many places in the world would be better, safer, more welcoming places to practice our craft. Even if the U.S. endures the harm it is so eagerly inflicting on itself, that trust isn't going to be repaired anytime soon. Maybe the scientific center of the world will shift to Europe, maybe China will capture it, or maybe it will just disperse, with the greater connectivity of the modern world allowing it to be less centralized. But it sure won't be the U.S., not anymore. But hey, if it means reading a few less cringeworthy course descriptions, small price to pay, right?
[1] Both in the sense of being not a conservative and in the sense of living outside the U.S., having wisely gotten the hell out when things were still merely "highly worrying" instead of "acutely alarming."
[2] Despite my generally pro-education sentiments, I do think that this particular trend has been taken somewhat overboard. Credentialism is a real issue, being a (still comparatively mild) instance of Goodhart's law, where the piece of paper carries value independent of the skills it is supposed to signal.
[3] I'll note that this isn't really a *universal* feature of education, but neither is it anything to do with indoctrination. Without writing an entire essay on the subject, I'll merely note that several major strains of U.S. conservatism are based on views that simply don't hold up well to exposure to the wider world. That doesn't mean there are no educated conservatives, but I think educated conservatives tend to be conservative for very different reasons than uneducated ones.
[4] This does absolutely create a feedback loop, which is why the "decades ago" was important. Though also there are definitely fields were the right loses out simply by virtue of its predominant viewpoint being acutely wrong. I couldn't hazard a guess at how many conservatives went into climatology with the intention of blowing the cover of the climate change hoax, only to abandon their ideology when they could no longer fail to recognize that mainstream climate science *is not wrong.* But I'm sure the number is more than zero.
But if education is so important, why so many inventor-billionaries keep telling young people to drop out of college and start a business? Do you think the Steve Jobs type not-very-educated-inventors are sort of the tip of the iceberg, and they are working on the backs of hundreds of engineers who do need education? I am choosing Jobs because once he fired the entire R&D department of Apple, which sounded like such a suicide move, and yet it worked out.
Look, there was recently this excellent review of schools right here. Briefly, the no-structure people who just learn on their own, are probably the smartest people with the best world ethic, and it seems they fill the truly important positions everywhere. The high-structure learners who struggle with schools will probably not contribute that much… so education is mostly for the low-structure learners. Do you think they are so important? Briefly, bluntly, is it important that for every no-structure genius, there should be 10 low-structure mediocre people?
I have been surrounded with no-structure learners because software dev used to be such a crazy wild west thing with “anything goes”, and it was one field that back then did not care much about degrees. There are now a lot of low-structure people in it who think like classical engineers, who consider the methods they were taught as holy writ and do not really think outside that box. I am not sure how useful they are…
"Look, there was recently this excellent review of schools right here. Briefly, the no-structure people who just learn on their own, are probably the smartest people with the best world ethic, and it seems they fill the truly important positions everywhere. "
First, I really gotta ask: are you the same person that wrote the previous comment? Did you have a bout of temporary amnesia? Because the comment I was responding to wrote:
"Conservatives are slowly discovering that basically everything is downstream from college."
Which *seemed* to be hitting the nail right on the head. Everything in our modern, complex society IS downstream of higher education, thank you for noticing. You point right at the fact that multiple large clusters of collectively very influential people got similar sorts of higher education[1], and then proceeded to have enormous influence on the world. Are you really turning around one comment later and saying "well, how important could it really be, if Steve Jobs said they didn't need it?" Do you really, genuinely think all those people just kinda...wasted their time and money, that none of them *needed* what they were getting out of higher education, that all of them could have done the same jobs just as well without them?
I will be the first person to suggest that our current way of teaching people various sorts of specialized job skills might well be inefficient and overly rigid, and that there might be better systems for doing the same job. But the idea that we could get rid of it and replace it with...nothing at all? I honestly wonder what planet you've been living on. The modern world is *enormously complex*, and requires many, many different sorts of specialized skill to function. People are not born with those skills. Some of them they learn on the job, but lots of jobs have a certain floor of skill that's required to do useful work at all, and thus it is important for *employers* every bit as much as employees that there be institutions that can teach useful skills[2]
"The high-structure learners who struggle with schools will probably not contribute that much… so education is mostly for the low-structure learners. Do you think they are so important? Briefly, bluntly, is it important that for every no-structure genius, there should be 10 low-structure mediocre people?"
At the risk of sounding condescending and reductive, it seems like your conservative biases have completely overtaken your critical thought on this one. The conservative worldview seems extremely prone to believing that:
1. People exist in a rigid hierarchy of value/virtue that cannot meaningfully change. If you're a good/valuable person now you'll always be a good valuable person in the future and if you're a bad/worthless person now you'll always be bad worthless in the future.
2. This hierarchy basically always cashes out into life outcomes. If somebody ends up poor or miserable it is because they were unworthy. If somebody ends up rich and successful it is because they were worthy.[3]
3. Only the people high up on the hierarchy really matter at all. Everyone else is of little to no importance.
(Usually with an implicit 4. They and those close to them are definitely high up in the hierarchy, regardless of life circumstances).
This is, to put it frankly, utter crap. I think a lot of the time these go unstated and largely unexamined in the minds of the people that hold them. But you are coming out just shy of saying a lot of this explicitly in the comment quoted above: so I decided it was worth being a little more blunt that usual about it.
Now, it *is* possible to take egalitarianism too far. Not everybody is equally capable of doing every thing. But the big, enormous, GLARING mistakes in the view above are first, assuming that there is only one dimension of competence and second that people who don't reach the absolutely top levels do nothing important. Again, the world is very large. It is very complex. It requires many *different* sorts of skills to keep running, and some people are very good at some of those skills and not at others. And there are many, many places where having two people with an *adequate* level of a particular skill is so, so, SO much more important than having one person with an *exceptional* level of that skill.
"I have been surrounded with no-structure learners because software dev used to be such a crazy wild west thing with “anything goes”, and it was one field that back then did not care much about degrees. "
I think your understanding of the world has been harmed rather badly by doing so much work in a field like this, and (apparently) not paying enough attention to those outside. I know a thing or two about programming (though probably less than you). It is certainly an area where a single person with excellent skills can often contribute much more than multiple merely adequate people. Most fields are not like that. The most talented surgeon in the world can still only operate on a few patients a day: a hospital with 20 adequate surgeons will certainly provide better *overall* healthcare outcomes than one with 3 or 4 amazingly talented surgeons.[4] Even in software, there are sharp limits to what a few talented people can do. Sometime you do just need more people that know the language and can hunt bugs, write tests, or create various pieces of boilerplate code and connective tissue that doesn't need to be a masterwork of creativity and subtle elegance (it just needs to work).
If I have time, I'll write a separate comment about why I think your understanding of "no-structure, high-structure and low-structure" learners badly missed the point.
[1] Similar within a cluster, not necessarily between clusters.
[2] I'll note that a system in which employers take responsibility for teaching all skills beyond basically literacy and numeracy *is* something that could exist. The reason employers don't do this *now* is because they would be running the risk of investing lots of resources into an employee's training, only to have that employee leave and put those skills to work elsewhere (perhaps for a competitor). So instead employers want entry-level people to have skill sets that will let them contribute positive value quickly. Setting up a system that would allow all training to come from employers would have an *alarmingly* high cost: you'd basically need to give them the power to prevent employees from quitting.
[3] I'll not that people seem to violate both of these assumptions basically whenever it's convenient: rich or successful people that they don't like must have cheated (somehow) and poor/unsuccessful people that they like get infinite excuses. And yet the hierarchy reasserts itself the moment they're not actively looking at those specific cases.
[4] Provided their standard of "adequate" is set high enough to avoid all of the easily-preventable deaths and injuries that incompetent or negligent surgery can result in.
I'll think this over, but yes, internship based learning yes might involve not being allowed to quit for a number of years. Actually we have this today in a way, when the employer finances doing an MBA but then either no quitting for X years or paying it back.
>But if education is so important, why so many inventor-billionaries keep telling young people to drop out of college and start a business?
Why would you assume that they have the best interests of those young people, or of society, at heart? Or, if they do, that their assessment was correct?
>Look, there was recently this excellent review of schools right here.
Oh. I went to higher in the kinds of Central European circumstances that were very Prussian rote learning. Then I saw in the Anglosphere (UK not US but likely similar) it was far more practical. My takeaway was that it is succesful precisely because it is not education (not rote learning) but something like internship, actually doing work and solving problems. So it seemed to me, it could be replaced with full on on the job internship and it would probably improve. It was already 75% like working a job, not memorizing definitions.
"Whether I am being too uncharitable to woke ideologies in general."
Insofar as it seems pretty clear you don't even understand the things you're mocking, yes. Probably some of them are very nearly as silly as you think they are. But it seems like quite a few of them are simply couched in language and perspectives unfamiliar to you, personally, and you don't seem to recognize that there could possibly be useful knowledge not packaged and aimed directly towards someone of your background. Which probably sounds harsher than it is: I've been there, as I suspect most people have.
Let me just say right from the word go that my academic background is extremely, almost exclusively STEM focused. I've taken classes from other areas mostly to fill course requirements. But I also have a curiosity about the world and often interact with smart people with very different backgrounds. There's plenty of pretentious garbage produced by the humanities that I do roll my eyes and, but I don't go throwing the baby out with the bathwater.
I'd like to call particular attention to one of your complaints: "Hemispheric Poetics & Politics." There are other courses where I'd (partially, tentatively) agree with your judgements of the course descriptions, but this one in particular reads like you just fundamentally do not understand that an arts class is going to be discussing, y'know, art and what that means.
I guess some more narrow-minded STEM people regard art as just, like, entertainment. Something nice to watch, look at or listen to on weekends. And of course it can be that. But it's also the way humans express themselves. It's the way humans express the events they've lived through. It's every bit as vital a part of understanding history and the broader world around you as statistical summaries or academic analyses. You'll learn one set of things by reading (say), a well-researched history book on World War I. And another set of things by consuming art made by people who experienced it. You can't really learn first from the second OR the second from the first.
This particular course appears to cover art made about various conflicts in 20th century Latin America[1]. That is an *extremely* apt topic for a Spanish comparative literature class. Someone studying Spanish at university with the expectation of actually going out and *using it in the world* is very, very likely to interact with people whose life histories have been profoundly shaped by those conflicts. Studying the poetry that people made about living through those conflicts and their aftermath is *absolutely* a good and worthwhile thing for a subset of university students to be doing.
And to me this nicely illustrates the problem with the entire concept of "woke" and "wokeness." The motte here is reasonable[2] claims like "elementary Portuguese class shouldn't be centering racism, sexism and neoliberalism in its course description or content," and the bailey is "university students shouldn't be taught about a highly subject-relevant artistic tradition because the politics *of the art* make me, personally, uncomfortable." In essence, almost nobody throwing around the words "woke" and "wokeness" seem to have clear, consistent ideas what they actually *mean* by them[3], but instead lean on a "I know it when I see it" sort of schema. But the "I know it when I see it" seems to both differ a lot from person to person, and be *extremely* likely to simply reflect individual prejudices about what facets of life and culture have what political valence. Like you marked "American History 1800s to present" as "broadly fine," but I guarantee that lots of conservative-leaning readers would see "U.S. imperialism, legal exclusion, racial segregation" and brand it as "woke" on that basis alone. Those are all things that happened, but *talking about them at all* is regarded as political.[4]
Before I sign off, I'll weigh in on this:
"If I should sign up for one of these courses to be a sneaky little anti-woke spy,"
No. Absolutely not. This is quite possibly the single worst way you could go about interacting with this facet of the world. If you go into *any activity* for the sole purposes of confirming your biases, your chance of learning anything craters. If you genuinely believe some class has material that's interesting in its own right--if it's covering something that you'd want to learn about even if it was presented in a different way by different people--consider taking *that*. But if your only interest is to "spy," the best possible outcome is that you will simply be bored out of your mind. And you stand a very good chance of making your understanding of the world substantially worse.
[1] I assume the "what" after inter-American cold war implies confusion as to what that is. It's the portion of the Cold War that was took place in Latin America: the U.S. was often very involved in the politics and conflicts of various Latin-American countries as part of its indirect but often still quite militant opposition to all things socialist.
[2] Note "reasonable" is not the same thing as "incontrovertible."
[3] I've had this conversation a couple of times on here and even smart, thoughtful, articulate people don't seem to be able to articulate it in a way that stands up to scrutiny.
[4] Regarded correctly, IMO. It's a mistake to imagine that *any* aspect of history can be taught in apolitical fashion.
I think this it’s a good comment. I would just add that. I took a lot of literary theory classes in college and I think they were the worst in terms of ROI. I could’ve learned about Derrida and Baudrillard on my own and the time would’ve been much better spent even in the English department reading some 19th century big novels or maybe Yeats.
You are making a very flawed assumption here. You're assuming woke isn't paying for people to write books that are "woke." (and that publishers aren't forcing their writers into "being woke" so the publishers don't go out of business).
Next thing, you'll try and tell me that Modern Art was an entirely organic phenomenon.
If I sound a bit salty here, it's because one of my favorite series got cancelled because the author discovered (to his horror) that he'd written a character /wrong/. That the character's current actions (as dictated by "woke" philosophy), did not and would not lead up to the character's final fate -- "you might not notice that, but your brain does" is a relevant quote from Red Letter Media. Unable to reconcile that, he cancelled the series.
"You're assuming woke isn't paying for people to write books that are "woke.""
It's quite possible I'm making a "flawed assumption." Probably even a lot of them. Most people do, all the time. But if you have genuinely identified such an assumption in my writing, you have failed so egregiously to communicate it that I have to wonder if you were even trying at all. I'm going to do my best to parse this, but I'm not hopeful...
"You're assuming woke isn't paying..."
Well, five words in and we're already in trouble. I realize that expanding the meaning and usage of the word "woke" to cover more and more ground and become less and less comprehensible is a hobby approaching an obsession for some people, but this is really a bridge to far.
Even in its current, heavily bastardized usage, the word woke is an adjective (it's original usage being a conjugated verb). It is not, nor has it ever been a noun. As a noun it has no standard meaning[1] that I'm aware of. You easily can noun-ize an adjective by appending "the" to the front to indicate a group of people with that characteristic: "the quick," "the righteous" and so on. But you followed it with a singular verb conjugation, so clearly this isn't a typo. As a singular thing, followed by "isn't paying" it has to indicate some sort of...entity, I guess? With both money and agency? I'm going to stop short of speculating about what sort of entity that might be, because you haven't given me much to go on, and it seems like it would be pretty rude to assume you're inventing conspiracies or the like just because of one poorly worded sentence.
"...for people to write books that are 'woke.'"
OK, now I'm even more confused. Where...are you getting *any* of this? Where am I assuming *anything* about the type of books people are and aren't being paid to write? I don't think I specifically mention anybody writing books anywhere. I talk a fair bit about art-in-general, which certainly includes some books. And I make a passing mention to a hypothetical history book. But if there's some assumption in there of the form "people are definitely NOT writing books of Type X for Reason Y" I cannot for the life of me find it.
The main thing I'm talking about is a class on Spanish-language poetry (with emphasis on a particular conflict). Is that what you think "woke" might be paying people to write books about? Even if that were true, I really don't know why it would be relevant. Do you think the course might be studying paid propaganda instead of the genuine articles?[2] The conflicts in question *definitely happened.* They touched millions and millions of peoples' lives. Some of those people very definitely wrote poetry and music about their experiences because *that's what human beings do.* Accurately sourcing contemporary poetry and music is not generally difficult: I cannot possibly imagine that somebody capable of getting a job at Yale wouldn't take the elementary step of making sure their course materials were genuine. Now it's possible for a course like that to work from some sort of a textbook or compilation[3]. If it does it's pretty certain that said textbook would be both paid for and written by people who are "woke"[4], but as long as *it* was sourcing genuine artistic materials about the conflicts in question (and again, hard to imagine it wouldn't), that sounds like more of an annoyance than a real complaint.
So after thinking it through aloud (well, in print) as thoroughly as I'm able, I *still* have no idea why you think I assumed this thing, or even what exactly it is I'm supposed to have assumed. Maybe try being more direct and less vague.
"Next thing, you'll try and tell me that Modern Art was an entirely organic phenomenon."
Not honestly sure what you're trying to imply here either. The Wikipedia article for "Modern Art" tells me the term covers art from over a century, ending in 1970, so whatever you think was "inorganic" about it, it can't have had much to do with "woke.[5]"
"If I sound a bit salty here, it's because one of my favorite series got cancelled..."
I'm sorry that that happened, I hate it when things I enjoy get cancelled. But without more detail than a really, frightfully vague description I can't possible have much of an opinion on it. Did this series have a name, at least?
[1] To be clear, as an adjective it has no standard meaning either, but that usage is at least *common* and people can at least gesture at a shared schema for usage, even if they mostly end up talking past each other with it.
[2] Let me emphasize that if this *wasn't* what you're trying to say, I'm not trying to put words in your mouth. I'm just genuinely confused and grasping at anything your comment maybe *could* mean.
[3] Though also extremely, extremely possible that it wouldn't. I've only taken a couple of foreign-language lit classes, but neither of them did.
[4] Because again, the word is basically meaningless, and anything besides guns, trucks and Jesus is "woke" to somebody. Actually, scratch Jesus off the list, he's definitely "woke" as well.
[5] Unless it's not only expanded from adjective to noun but also from relatively recent to many decades old. I can't be sure it hasn't.
>The Wikipedia article for "Modern Art" tells me the term covers art from over a century, ending in 1970, so whatever you think was "inorganic" about it, it can't have had much to do with "woke.
Wait a bit please. Maybe there is a simple misunderstanding here. "Woke" originates in "all men are created equal". Its roots are centuries old. When people criticize only its current form, it means that in the past it was simply applied less consistently, often not applied in many aspects of life, those aspects were "apolitical" in the sense of "old inherited unnoticed inequalities".
Because this seems to be the only two options on the table. Either uproot all inequalities in every possible aspect of life, in which case everything becomes political, and kind of propagandistic, and then we start looking like the Soviet Union where there was a proper Communist way of everything and a bad burgeois way of everything. Or, keep it apolitical, which means it is a cover-up for old inequalities, and the value judgements that come with them, that something that comes from a position of power is seen as a good and something not, not.
There is also a third option. An acceptance of inequality as a necessary and even beneficial thing. An understanding that different people have different uses and different value. How long are we going to keep pretending?
Yes, people make art about conflict. I have no direct knowledge about the particular South American conflicts' art being paid for.
However, I contend that for any art to be worth studying as a direct reflection of the world, it -cannot- be propaganda (propaganda can be studied, and perhaps ought to be studied, but without framing it as propaganda, you've lost enough of the meat that you're just putting your ear on the empty shell, and excitedly telling people you hear the sea).
I do have knowledge about "art in general" being paid for by "the woke" (you can hear this as USAID if you want, The Powers That Be if you don't.) More importantly, I have some knowledge about the financial constraints that "art" operates on, in order to make it into the world.
But perhaps, you ask, why would I automatically start being suspicious of art drawn from South American Conflicts?
Well, it's because "paid for art" is done by governments a LOT.
Here's a good writeup. Yes, I should have been more specific about what "flavor" of modern art
Stranger Things was paid by the German Government to put a German song in (Germany pays for any German song in Hollywood, it's nothing personal -- this is relatively light propaganda). And the folks running Stranger Things have never turned down a shekel.
Dark is a different matter -- the whole "anti-nuclear" Green movement in Germany was very, very willing to pay for an anti-nuclear TV show. This is much more poisonous -- and did you know it, when watching? (I've been assured that the Simpsons occurred before this whole "anti-nuclear" money was floating around. That was apparently just the creator's idea.)
Sorry, the rest of this comment (already long) will touch on what you were so patient to write to me.
1) In the event that you aren't a native english speaker, "woke" used as a noun not an adjective does in fact mean "the woke" (the foreshortening is as discussing blacks instead of always using "the blacks" -- for African Americans if it's not clear.) This is a standard shortcut in English, and no confusion was intended.
2) If you have no suspicion that it is propaganda, how would you possibly notice? (Answer: didacticism. Heavy-handedness is generally indicative of propaganda.)
3/4) Any textbook is likely to induce bias through what it choses. I might refrain from a course on pre World War II German Art, if it was from Western Culture -- I'd have to be very careful to understand it as a subsect of the "allowable" art (in two parts: there are "what we're not allowed to talk about" at the moment, and "what the Nazis burnt"). You have to know what's being deliberately erased, to be able to say "this is valuable or not" (and in a book about conflict, there's going to be erasure, if only because some people are coded as "good" and others as "evil." You're not going to hear about human body parts being taken as trophies -- at least as a good thing (this is not a hypothetical, but also not from South America).).
5) Not woke. CIA/US Government. Sources cited above.
I think that woke is likely to not discuss any of this, instead discussing how "women are silenced" or other things like that.
Are you familiar with Riverbend? You, like me, would probably count her writings as a fine introduction to "America's Iraq." It's written with a marked bias (but not an unfounded one. Her discrimination is both cultural and incisive).
That doesn't make it not propaganda though. If there was one work of art that put the coffin-nail in "America's Nation Building" in Iraq, It would be her work. She didn't do it alone.
here I need to tell you the same thing as I told your interlocutor:
"Woke" originates in "all men are created equal". Its roots are centuries old. When people criticize only its current form, it means that in the past it was simply applied less consistently, often not applied in many aspects of life, those aspects were "apolitical" in the sense of "old inherited unnoticed inequalities".
Because this seems to be the only two options on the table. Either uproot all inequalities in every possible aspect of life, in which case everything becomes political, and kind of propagandistic, and then we start looking like the Soviet Union where there was a proper proletarian way of doing everything and a bad burgeois way of doing everything so now even coleslaw needs to be proletarian. Or, keep it apolitical, which means it is a cover-up for old inequalities, and the value judgements that come with them, that something that comes from a position of power is seen as a good and something not, not.
On the classes requiring ~5000 words of writing in a whole semester, that sounds like a reasonable length for a term paper. For an advanced undergrad or grad class, it’s pretty standard to just require a term paper, which is going to have to be much better at defining and carrying an argument than a bunch of weekly reading responses. The difficulty will come down to what sorts of standards are being used to evaluate the paper. Anyone who is legitimately taking an advanced class like this one will have taken other writing intensive classes where they should have learned some of the mechanics of writing, and will need to be constructing a significant argument at length.
Changing a course description is something that usually requires the department to draft a new version and vote on it, then submit it to the university catalog committee, and at many universities, the change then needs to be approved by a giant university-wide committee that includes representatives from every college/school. Most things that go up to that committee just get approved by default, but it’s a lot of work to send something through all these committees, so usually departments end up with a bunch of course descriptions on the books that aren’t taught very often, and sometimes have classes that get taught in ways fairly different from what the official description leads you to expect.
So you definitely shouldn’t expect course descriptions to disappear just because we are five months into a national vibe shift.
Anyway, I do think that you have been fooled by a lot of people into thinking that wokeness is a bigger problem than it really is. The only thing that actually seems objectionable to me in that course description is the use of “neoliberal” - but that word is in there precisely to indicate that they are going to include some economic urban geography that ignores questions of race, gender, and sexuality in thinking what a city is and what it should do! Everything else is there just to point out that they will include some readings that do focus on these identity issues - and it definitely makes sense to include these things, because cities really are special places for race relations and queer identity (as you’ll notice from anyone who talks about San Francisco in an opinionated way).
That would explain a lot. I remember being very disappointed in college when I tried to sign up for a few cool-sounding classes, only to discover that they were never actually offered.
I think you could steelman the course description into a best-possible version of that course, which is all about how different ways of talking about words like "urban" and "city" wind up encouraging or encoding various different political points of view... like the somewhat flexible definitions of "city" vs "suburb" in American usage, or the way "urbanism" always seems to wind up encoding a single view of what a city ought to be like. This could be a genuinely valuable part of an urban design degree, which helps people step outside their own set of political assumptions and take a more detached view.
But you would need a seriously disciplined lecturer for that course, someone capable of stepping outside their own political assumptions and teaching it in a way that does not confuse political axes with quality judgments. Nothing about the way it is described, in terms of jargon used only on the left, makes me suspect that the lecturer is all that disciplined.
You should take some of these classes! Consider how pop-sci discourse makes science sound idiotic, and that a similar process happens between university-level humanities and popular "woke" ideology. Continental philosophy, critical theory, etc have depth worth understanding on their own terms outside of the culture war, at they very least to avoid the arrogance that comes with never having actually read these kinds of texts.
Imo the bulk of those two mentioned subcultures can be summarized as: "Woah, what if we thought differently in a way that's strictly inferior to just plain reasoning about our subject areas? Then we can still learn things even though we're clouding our methods with obscurantist terminology and saying stupid things more frequently than we have to." Like if you want to adopt a very heavy race focused view you can just make it clear you're stipulating that as a precondition of your inquiry and use regular statistical tools or informal arguments rather than writing something like Kendi's thesis or any of his books. The reason why this doesn't happen though is because those fields wouldn't really stand up to scrutiny if their proponents tried to integrate their thought with other fields. You can also see this failure in experimental philosophy where a lot of the rhetorical appeals of analytic philosophy like "We believe this is common sense" or "It's intuitive that..." are not remotely supported by the existing empirical evidence. I wouldn't recommend people study that field either after exploring it myself.
This type of thing is more specific in the Ivy Leagues, which have their own subculture to perpetuate and maintain. One of the most charming aspects of state schools is that they generally do a better job with ideological balance. They aren't perfect, but they aren't Yale either.
I have a friend who took a hyperwoke course at Villanova, found it very easy and not terribly interesting. That seems to be a typical opinion. I guess if you want an easy A and want to learn more about woke philosophy, take the courses; if you want just the latter, reading books and showing up to protests is the best option.
Maybe there's a lot of variation within state schools, but by my (limited) exposure to the University of Michigan, it seems like Yale is not *that* much woker?...
I do wonder how many young ideologues are actually taking these classes. Presumably they're the intended audience, but if the courses actually end up generally boring, then maybe the root of weird-young-elite-extremism is elsewhere. A few commenters have said something to the effect of "the kids are actually crazier and woker than the professors," which suggests this too. Maybe it's the internet, then? All this stuff leaked out of the academy a decade ago, and it's the kids who've been exposed to it since then who won't let it die...
I go to the University of Colorado, where this sort of thing is rarer. Wokeness likely scales with prestige, at least among state schools.
Why is that the case? Because it's what the elite humanities students who attend these schools want. We have to remember that in most cases wokeness isn't something imposed on students by administers: it's supported by the majority of the student body of these schools, as the Columbia protesters reminded us. I agree with the commentators that students are often woker than the professors, and it seems to me that wokeness itself embodies the traits of youth: its energy, ardor, idealism, and adolescent zealotry.
The pipeline is probably built like this: high school-and-younger students see internet wokeness and become radicalized. Once in college, they take woke courses because they like the descriptions, but these courses don't convert anyone so much as they provide sheet music to a rabid choir. The spirit of the movement happens outside, in clubs, protests, and collectives.
So we really can't blame woke classes for anything except giving the students what they want.
the alpha school essay has been bouncing around my head. i'm still trying to find the thesis-leading-to-action. maybe it'll emerge from some open feedback:
i was hired to study k12 ai in Louisiana. i wasn't impressed, even though Louisiana is actually investing much earlier than other states. most educators think of ai as chatbot-cheating-tools. students are using it to cheat, yes. but they are also using it to ideate. moreover, educators are using AI tools more than other tech. ai can save lots and lots of time.
the thing that struck me in the alpha school essay was that they are getting 2x results without actually using ai. what if they rebuilt their program using ai?
with my louisiana work, i also bought about 1000 students khanmigo and while some did use it, it was just an ai tool slapped onto a legacy application.
we have the chance to rebuild education. personalize teaching. take away educator administrative work. give people back their time to do whatever they wish.
the greatest gift i got from my weird alternative high school was the gift of time. i was able to follow my passions. build companies. chase wild ideas. that gave me confidence. and confidence seems to be rewarded in our society (the indiviudal pursuit of freedom.)
anyway, here is what i think i'll end up doing:
1. Raise $5M at $15M valuation (A8C SAFEs convert).
2. Build team to do two core projects: curriculum creation(teaching) + platform building(admin).
3. Make all K12 Louisiana curriculum “AI Ready”
4. Reduce teacher and learning time by 75%.
5. Run as an afterschool cohort.
6. Scale to Kindergarten, 6th, and 9th grade cohort.
My not-terribly-well-informed opinion is that a lot of it has to do with the rapid feedback. Typically students have many days between when they attempt a problem (either homework or a test) and when they learn if their attempt led to sucess or not. By that point the context has faded from memory. OTOH if they get feedback within MINUTES, then they can still recall WHAT they were thinking when trying to solve the problem, and WHY.
I ran the "one guard likes big butts; the other guard cannot lie" benchmark, and the bot did fall into my clever trap and assume the first guard always tells the truth, but it did so in the most likeable and enjoyable way I've yet seen.
I read today that heat deaths in Europe are greater than gun deaths in America. I checked it myself and Europe had 60,000 heat deaths in a year where the US had 2000. Further, ChatGPT confirmed that at least half of this is being caused by comparative lack of AC. Is this actually true? If so, it feels insane to me that I have never heard anything about it, not in mainstream media or EA causes.
There is so much clickbait and false certainty in discourse about US gun deaths that it's worth grounding oneself in the actual current facts. Pew Research Center -- a gold-standard source -- happened to publish an update on it a couple months ago:
-- suicides have long been the majority of US gun deaths, currently 58% of them.
-- about 38% are murders, with the remainder being accidental/law enforcement/other.
-- while total annual gun suicides have risen pretty steadily since around 2004, gun murders spiked in 2020/21 and since then have fallen from that peak. For 2023 they were at about the same annual total as the early-1990s. (And then from sources outside of this Pew writeup, those totals continued dropping in 2024 and thus far in 2025.)
-- the gun suicide rate (per capita) is now almost up to the highest rate recorded (1977) in the past half-century.
-- the highest US gun murder rate (per capita) in the past half-century was in 1974, the second-highest in 1994. The 2021 rate got to within 10 percent of 1974's.
In the following stats "all rates are adjusted to account for age differences across states":
-- the states with the highest rates of gun deaths are all both extreme-south and heavily rural: Mississippi (29.4 per 100,000 people), Louisiana (28.3), Alabama (25.6) and New Mexico (25.3). However the District of Columbia, which is basically 100% urban, currently tops all of those (30.6).
-- if we look just at gun murders the same five jurisdictions rate highest but the gap widens from DC to those four states.
-- if we look just at gun suicides the rate leaders are all large states with lots of wide-open spaces: Wyoming (19.0 per 100,000 people), Montana (17.6), Alaska (16.0), Idaho (15.1) and Oklahoma (13.6).
> -the states with the highest rates of gun deaths are all both extreme-south and heavily rural: Mississippi (29.4 per 100,000 people), Louisiana (28.3), Alabama (25.6) and New Mexico (25.3). However the District of Columbia, which is basically 100% urban, currently tops all of those (30.6).
As can be guessed from the fact that DC leads the list, if you dive deeper into those states with high rates of gun deaths you will discover that while suicides are relatively evenly distributed, virtually all of the homicides are concentrated in urban areas (ie Louisiana is dominated to an extreme extent by homicides in New Oreleans and Baton Rogue).
You can guess or assume that, and many do, but you'd be wrong.
(a) Suicides are not at all evenly distributed. Nationwide from 2001 through 2010 the rate of gun suicides was 54% higher in the most-rural counties than in the most-urban ones, and from 2011-2020 that gap increased to 76%.
(b) Nationally the rate of gun homicides [murders plus non-negligent manslaughters] is about 5.7 per 100K. In rural areas (those not part of any metropolitan statistical area nor part of any isolated city) that rate is 4.3. It's a lower rate but not drastically lower.https://www.ruralhealthinfo.org/topics/violence-and-abuse
Outside of a couple of outliers like DC, the reason that most US homicides happen in metro areas is primarily just that's where most of the people are (only 20% of the national population resides in areas classified by the Census Bureau as rural).
(c) In Louisiana specifically, which has unusually high homicide rates, the rural-parishes rate is only about 10% lower than in urban parishes (counties).
(d) Looking at those four states that lead the nation in gun deaths per capita, Mississippi is the 4th-most rural state in the US; Alabama is the 9th-most rural; Louisiana is above average in rural population pct and New Mexico is about average. In terms of core-city populations those four states include no cities in the nation's top 25, just one in the top 50 (Albuquerque) and one more in the top 100 (New Orleans).
Deaths aside AC is a main reason the southern half of US has been booming economically the last half century.
Southern Europe isn't quite as hot, but I'm sure it could also benefit immensely from allowing more AC. To me, that's a much bigger issue than the heat deaths.
I'd add to the previous comments that deaths should be measured in QALYs - "heat deaths" are typically already very old and sick while gun deaths are typically of healthy young people (so a gun death probably has an expected loss of 60 QALYs while a heat death might only have 1). YMMV how you want to weight that in morality terms.
I don't go in for the QALY stuff at all really because it's just utterly intractable (my 73 year old mom's life is worth less than your 65 year old mom's life... to whom, exactly?), but you surely have to discount suicides immensely on this metric, since the person committing suicide has decided their life is negative QALYs from here on out.
Probably. But it's pretty unavoidable when dealing with heat deaths, since so many of them are of people who're close enough to dying already to be knocked over by one adverse day. If your 73 year old mom died of a heatwave today instead of getting heart failure in a month that seems importantly less significant than if she'd been hit by a truck ten years ago.
(Gun deaths/suicides do, as you point out, have their own complications, which makes them a poor baseline for comparison in general).
Maybe. Most of the gun deaths are suicides I believe. A lot of those people are older and just feel like they are done with the world. Don’t have numbers in front of me but it seems possible from reading suicide obits.
A large chunk of gun homicides are gang or drug related. I think the 20 year old who dies in the Bloods v Crips shootout should get a pretty hefty QALY discount.
I'd really like to see the breakout of "business-related" versus "anger-related" deaths for homicides in general. (where 10 people kick you until you die, and then don't rob you, counts as anger-related even if they weren't angry).
I've found the data on this is incredibly poor quality and largely you just have to use intuition.
An example is "gang related" shootings. afaik it's around 10-15% officially, but anybody who's ever worked in the justice system, policing, or just lived in a high crime area can tell you it's absolutely higher than that; and that's not even accounting for like, gang adjacent stuff (drugs, retaliations, dead people who just happen to show up in the most dangerous part of town controlled by rival street gangs but for whom there's no proof). The largest coded category of homicide in Chicago for example is "unknown", by far.
No doubt, but I've heard about the horrors of US gun deaths in every possible media several times a year since age 6, and I had never heard about any major aircon-driven disparity in deaths. So maybe I've been in a massive ax-grinding convention the whole time without ever realizing it!
I wasn't questioning the gun deaths stat (though Erica, who is generally very accurate, says the figure is 47,000 per year, not 60,000. But it is certainly LARGE). I was questioning the stat that 30x as many in Europe as in US are killed by heat.
As Erica Rall's comment points out, it is important to use a consistent methodology and time period when comparing temperature-related deaths across locations. One such analysis is Gasparini et al. (2016), which used a statistical approach to estimate that extreme heat-related deaths are about 0.2% of all deaths in the U.S. (or about 6,000 deaths/year). For comparison, their estimate for the Italy is 0.7%, for Spain is 0.5%, for the U.K. it is 0.2%, and for Sweden it is less than 0.2%. So the U.S. is within the range of European countries, if near the lower end.
AC is likely part of the story but seems unlikely to be even half of it. The U.S. and U.K. have similar heat-related mortality rates but AC penetration is 90% in the U.S. and only 5% in the U.K.
A proper comparison should be controlled for summer daily maximums, I think. Maybe with moisture levels taken into account, it makes a lot of difference subjectively (but possibly less so for old and sick people). I expect the median European has it several degrees lower than the median US American.
> AC is likely part of the story but seems unlikely to be even half of it. The U.S. and U.K. have similar heat-related mortality rates but AC penetration is 90% in the U.S. and only 5% in the U.K.
Most of the US has hotter summers than UK. Both Sweden and the UK are far to the north of the contiguous US. The proper comparison would be the southern parts of the US and southern Europe.
Or paris - which has recorded a lot of heat deaths - and Seattle although Paris is further north than Seattle.
The estimates I quoted from the Gasparini et al. paper are based on relative temperatures in each location (you can see more details in the excellent Our World in Data post linked below). So even though there are more hot days in an absolute sense in the U.S., there aren't necessarily more hot days per year in a relative sense in the U.S. versus the U.K. A important subtlety comes from the fact that these relative values are calculated based on each location's temperature at which temperature related mortality is lowest (the so-called "minimum mortality temperature"). That value is probably affected by the presence of AC, so things get a bit circular.
New EA cause. Ship window AC units from US to Italy and just start installing. Maybe Italy lacks the state capacity to stop it. Give the mafia its cut in Sicily.
Possible drawback. Do window AC units actually work?
I briefly looked into this when it came up on Marginal Revolution a few days ago. Simple answer is no, the ~60,000 and ~2000 are not comparable, Europe stat is based on excess deaths (in a particularly bad year, 2022), while the US number was based on 'Deaths attributed to exposure to excessive natural heat as the underlying or contributing cause of death' ie, a doctor has specifically recorded it as heat related.
I'm sure there is a long answer somewhere too, that probably does show that more aircon in USA saves lives. However it almost certainly isn't as stark as that stat makes it seem.
Yes. It’s true. The lack of air conditioning is a death sentence as Europe warms faster than the rest of the earth, although maybe not as high as claimed.
People who live in AC controlled environments don’t really get what it’s like without. If you are in a city like London where the temperature gets higher these days (up to 32C last week) then you should realise that 32C is what you get in Hyde park and some leafy suburbs. The temperature in certain streets can be ten degrees higher, the tube is 40c+ and any indoor space is going to get hotter over the day, particularly south facing buildings with windows that trap the heat. There’s no relief indoors.
I walked into a cafe a few years ago and promptly left as the heat and steam were overwhelming. People work in kitchens in these conditions. I myself measured 40c in a room in a hotel which was a converted Georgian house and didn’t have AC.
I live in Vienna. Even though theoretically we have less cool weather than London - less rain, less fog - I have a mobile A/C unit which I haven't used for ten years now. It was just never really that bad. The office had A/C of course, and evenings were tolerable in swimpants, dropping to like 25C for the night.
> Yes. It’s true. The lack of air conditioning is a death sentence as Europe warms faster than the rest of the earth, although maybe not as high as claimed.
I don't understand, cheap mini splits are like $500-$700, why is this such a crisis? That's well within the means of any restaurant or business owner, and most individuals.
This whole discussion has been kind of baffling to me. Certainly if it's so bad people are literally dying, not just feeling crappy, somebody's going to break and spend the $500?
Often they are not allowed because they make the facade of pretty historic buildings ugly. This generally tends to be about a lot of things in Europe. The tourism industry is big, so the government wants the city to look like it is 1900 so no high-rises etc.
Wow, talk about "HOA's and historical preservation run rampant." They'd literally rather have people die than upgrade buildings to be livable??
I've always thought restrictions like this were bad, but this is a whole new level. And these are democracies and/or owner-produced regulations like HOA's! At the least, they should be able to vote to relax some of these restrictions so people aren't dropping dead.
Sweden was all in on "covid19 will fix one of our problems!" (too many old people). [in all seriousness, it was because they have a ton of single mothers.]
And it's not helped by the fact that northern/western Europe is more used to cold, wet winters so there's been a big push on insulating houses to cut down on heat loss (and thus reduce energy costs).
With mild winters and heatwave summers, that doesn't work as well. Now we've cut out all the draughts, there isn't any air circulation and there isn't built-in AC to compensate.
... not sure where you're getting your info, but it can't be Britain. In britain, they have a special tool for breaking the ice on their toilets in winter.
If we go by standard scenarios and assume no AMOC collapse, winters in Nordic countries will still be cool enough that insulation is a good bet. If we *do* assume AMOC collapse...
Not sure how you do insulation and weatherproofing over there, but on our side of the pond, residential construction still has a ton of windows that you can open for ventilation and air circulation. When everything's closed, yes, there shouldn't be thermally significant amounts of air getting in through cracks and gaps, but you can still open stuff up when you want some fresh air.
Those aren't apples-to-apples figures. The European numbers are statistically estimated excess deaths related to hot weather, while the American ones are direct death statistics where a heat-related diagnostic code was listed on the death certificate as either the immediate cause of death or a contributing factor.
This study applies similar methods to the European ones to American data between 1997 and 2006 and finds 5600 deaths/year in the counties included in the study, which covered 62% of the population.
Europe is also potentially a much larger population than the US, depending on whether "Europe" means "Eurozone" or "all of the EU" or "Everything between the North Atlantic and the old Soviet Border" or the entire continent to the Urals.
Add it all up, and there's probably a higher death rate from heat in Europe than in the US, but nowhere near a factor of 30. My guess is that the European death rate per million population from heat-related causes, in an apples-to-apples comparison, would be more like 1.5x to 2x the US rate.
While trying to work out the per capita figures is useful, unless there is another explanation for the excess deaths in that study other than a particular heatwave, why would we deny that it’s the heatwave that caused the excess deaths.
> This study applies similar methods to the European ones to American data between 1997 and 2006
But the more frequent European heatwaves happened later than that. Parts of the continent are 2-3c above pre industrial eras. I know Andalusia where I have a place is experiencing extremes most summer days, and has seen those kind of increases.
Nobody is denying that the excess deaths are (probably/partially?) caused by heat. That's the point of the excess death number documentation (at least in the way they are presented in Germany).
The point is that the US numbers don't come from looking at "excess deaths", but from doctors specifically saying and recording "This person died because of heat", which leads to much, much lower numbers in the US.
The point is, it's apples to oranges, the point isn't that apples don't exist.
>why would we deny that it’s the heatwave that caused the excess deaths.
I am confused. I don't think I was doing this, just trying to normalize two radically different statistics so it even begins to make sense to try to compare them.
>But the more frequent European heatwaves happened later than that. Parts of the continent are 2-3c above pre industrial eras. I know Andalusia where I have a place is experiencing extremes most summer days, and has seen those kind of increases.
America, too. As I noted in another comment, US heat deaths per capital are way up since 2000.
5600 deaths per year, covering 62% of population, seems to me like it would give a factor of 6? Still an enormous price to pay for no air conditioning… Gasparrini has 3x for Spain and 5x for Italy (Sweden and UK obviously aren’t the right comparison and are similar to the US). But these numbers should have gone up, as Europe has gotten hotter and older in the last 15 years, and still has no AC… even forgetting the US, the absolute numbers model are really high.
No AC doesn't sound right. When I lived in Madrid my apartment came with AC (and swimming pools were also common in apartment complexes, and most people have access to free or very cheap local swimming pools). In France I also have an AC. Every office I've worked in had an AC too.
In Spain it seems about half of homes have AC and anecdotally more and more people are buying ACs every year.
5,600 deaths per year in counties with a population of 174 million in 2000, which works out to 32 deaths per million population.
It looks like the 60k figure was a cherry-picked peak heat wave year (2022) while the estimate for 2023 was 48k. The full text of the study is behind a paywall so I can't see how they're defining Europe, but the abstract does mention 35 countries, which is about right for the European Union member states plus UK, Norway, Switzerland, and former Yugoslavia, or approximately the European continent minus the former Soviet Union and Turkey with the Baltic states added back in. Total population about 550 million, so 60k is 109 per million and 48k is 87 per million.
The death rate per million in the US in death certificate stats increased from 1.9 per million in 2000 to 4.7 per million in 2022 and 6.2 per million in 2023, a 2.5x and 3.25x increase respectively. If we naively use that to adjust the 2000 excess death estimate, that becomes 80 per million for 2022 and 112 per million in 2023. Which is much closer to the actual European rates than I was expecting.
I've had it with google search. Anyone here know of a search service that works as well as old google? Don't say AI, sometimes I need 100% reliability.
I don't think there's anything that meets this bar.
> Don't say AI, sometimes I need 100% reliability.
IMO you shouldn't rule out using an LLM as step 1 of a 2-step process. Taking AI outputs at face value on anything important is stupid, for sure. But they are often good at surfacing information that you can pretty easily verify once you have it (but that would have been harder to find from scratch).
I have heard good things about Kagi, but honestly the answer is AI even if you do need reliability. You can just use it as a search engine and not rely on its own interpretation of its results. Perplexity is specifically designed for this.
DuckDuckGo is...95% there. When I set it as my default, I will probably go check Google 1-2 month. Conversely, if I'm on a work computer where Google is the default, I basically never use DuckDuckGo but I'm also searching a far more limited range of topics.
Ruskies have a search engine that works well, at least for image searches. at least as well as old google. No opinions on whether using it will get you on a list, though. (In case it's not clear, I haven't used it myself, just talked with an old image analyst/coder).
The search engine in question is Yandex, I use it regularly. It's not old google good but often better than contemporary google, and tends to weight towards older sites (which is a good or bad thing, depending on your query)
I donated my left kidney to a stranger this year, and I've had a lot of time to think about it. Specifically, was it a good thing to do, or am I just a chump for doing this?
I feel like in a very good society, you're a good person and not a chump if you contribute good deeds, especially if the society has done good things for you specifically. And in a fully evil society, there's no room for making a difference anyway, you'd only be helping people who would never do anything good for you or anyone else, so it's obvious you should avoid being a chump there. I'm sure about both of those extremes.
And in a normal society that's somewhere half-way between good and evil, the answer is probably case-by-case. Giving money to a broke kid (that hasn't done anything wrong) versus giving money to a broke gambling-addict is a pretty obvious example of needing to think case-by-case, and you can probably brainstorm many examples like this. And of course, society leans more good or evil depending on your location and which group you're interacting with. So the problem is: you want to be a good person, and you want to negate the risk of being a chump, but society is a mixed bag, and you have very imperfect information to work with.
Is there a solution to this problem when donating a kidney? Surprisingly, there's 3!
The obvious first one is that you can just donate your kidney to someone you're already pretty sure is decent. But personally, I didn't know anyone dying of kidney disease, decent or otherwise. I can go try to find someone, sure, but people have a life-or-death reason to lie through their teeth here, and I'd feel miserable sifting through peoples' life stories trying to single out the one person to whom I should donate. But there are some donors with better people-skills and more grit who have absolutely done this, and there are websites 100% dedicated to sick people making their case.
The second option is the one I did by accident. If you donate your kidney undirected, it goes to an individual at the top of the kidney waitlist. Who is at the top of the kidney waitlist? People who've been waiting a LONG time, and people who already have someone donating a kidney on their behalf but aren't biologically compatible with that kidney. My kidney went to that second type, and from what I can tell, that's about as good a "I vouch for this person" signal as I'll ever see. I mean, it's not 100% impossible that the sick guy did something terrible to get someone to donate for him (guilt tripping? harassment? threats? LOST had a storyline about this). But most people just wouldn't do anything like that, and I learned 1st-hand that these USA hospitals work extremely hard to identify those bad situations. The doctors are risk-averse to an annoying degree, like Scott mentioned in his kidney-donation post.
The third solution is religion. Jesus bluntly tells people to be like the Good Samaritan, meaning do good things for people even if you have nothing in common with them and know nothing about them, if you want a chance at eternal life. Luke 10:25-37. Many of the hospitals refer to undirected kidney donation as a Good Samaritan donation because of how well the act maps to that specific teaching. I would think "because Jesus told you to do things like this" is a good enough reason for many Christians out there, but Jesus also explicitly states the reward of eternal life. I can't speak for any other religions, but I'd bet money that at least Judaism and Buddhism have similar teachings and stories.
I know you weren't fishing for praise, and I'm sorry I'm not really engaging with your thinking in any interesting way. But, good on you! You did a very brave and generous thing, and regardless of whether you're a chump, you made a huge difference to someone who needed it. (Even if the recipient is secretly a dick, the person who donated on their behalf very likely a) cares a lot about their outcome and b) is highly deserving.) Thank you.
Thanks, I'm just trying to accomplish as much as I can with my life. I was never going to be smart enough to do medical research or make life-changing tech, so doing this and blood donations + charitable donations seems to be my maximum.
My question is, is your lifestyle so superbly healthy that you are sure you won't need it later on? Me as an alcohol and tobacco addict living on random sandwiches, not so sure.
The hospitals are risk-averse, they're not going to let someone donate if they find big risks of any kind. They'll check all that background stuff and the kidney and blood metrics more times than even seems reasonable, though they don't seem to care if the donor is out-of-shape or eats mediocre food.
Me personally, I'll just say there's probably hundreds or thousands of things I'm more likely to die from than kidney failure.
My two pennies: I registered to be an orhan donor, but then during Covid I was so disgusted by lockdown adherents and th3 extremely hostule rhetoric of compulsory vaccination fanatics that I unregistered. Coming to the moral position that I would rather a lockdown or compulsory vaccination fanatic die than benefit from my orhans, if I were to die and donate them that is. This is based on the perception that in my country, Britain, most people supported these things so the odda of a donor recipient also doing so being greater than chance.
So I would also not donate a kidney to a random stranger, just on the same rationale.
There would be no dilemma here if you just got paid for the kidney. You would be offered a fair market value for the kidney by the person who had the most to offer for it, and there would be no further issues. Fulfilling a demand would be a good deed on its own.
markets and morality are not actually the same thing, and a lot of the people most able to afford a kidney (or anything else) in your system did horrible but technically legal things to get their money
But there wouldn't be a dilemma anymore, since it would just be a transaction instead of a charitable act borne of moral judgement. There would be no consideration of whether it was good deed of charity, as it would be no more than an equivalent exchange of goods. Isn't that the wonderful thing about capitalism, that it lets us move on from putting arbitrary value on things?
I think payments are the right path, but I have no doubt that payments would create plenty of new dilemmas. Just imagine a potential "donor" going through 12 months of testing and screening only to get rejected near the end of the process. If he's compensated for his time, then people would sign up even with no intention to actually undergo surgery. If he's not compensated, that's a giant loss of time for someone trying to make ends meet, and a huge risk for everyone applying.
That's what I'm getting at, signing over ownership doesn't really work if a doctor can find something wrong with the kidney the day before surgery, at which point no one wants it anymore.
This isn't a problem when buying a house, you can back out of buying a house within the inspection period of 2 weeks or so, and the monetary cost and time costs are relatively small, but 12 months is huge.
Though now I'm thinking there's probably precedent here when it comes to surrogate mothers, maybe I'll look that up later.
The point is that if it is necessary to encourage participation, donators can be compensated for their time and then have their kidneys "returned" if the tests don't work out. Of course, you'd probably have them have to do some initial, non-compensated tests (like drug tests) just to make sure they aren't just wasting money...
The process is ~18 months long, almost entirely paid by the recipient's health insurance. The 2 big decisions are "who do you want to donate to?" and "do you want to donate at a hospital near your home, or fly to one of the best hospitals in the country and stay at a hotel?"
It's a lot of medical bureaucracy, lots and lots of blood tests and urine tests, and 6 weeks of recovery, but most of the recovery is in the first 2 weeks. You're relying on someone else for those 2 weeks, you'll need someone competent, calm, and predictable. And like Scott said, people should really be paid for this hassle, but the people currently dying on dialysis don't have time to wait for congress to finally allow payments or for artificial kidneys to be developed.
I want to give a plug for MedStar Georgetown University Hospital. They basically put all the bureaucracy and testing into a single day and the whole process was very smooth for me. They also receive a lot of the testing through Quest Diagnostics so you don't have to travel to them for all the blood/urine tests. I read Scott's article in October 2023 and donated June 2024 (probably could have done it a couple months earlier, too, but I wasn't available before June). So definitely not an 18-month process!
I also want to plug the NKF, who reimbursed me for travel, hotel, and missed work costs. The whole process really did cost me nothing (though I did front all the costs and was reimbursed after the fact).
Sounds a little smoother than what I got. I could have got the whole thing done earlier, but I had to schedule around both weather (don't want travel during snowy seasons) and the caregiver's busy calendar. I'm guessing that's pretty common, but I don't know.
I'm considering a career change into AI safety, and trying to learn as much as I can about it. I've found lots of great resources online but so far haven't been able to talk to anyone in the field. If that's you, please reply or email me at keavney.mark@gmail.com.
It's odd that girls can become pregnant right at puberty even though this carries quite increased risks for both the mother and child. How did something this poorly adaptive come to be? I've heard a few theories but maybe people here are more knowledgeable:
-Contemporary humans go through puberty much earlier than their ancestors, around 13. This is due to better nutrition and exposure to hormones from e.g. water treatment, both of which are recent occurrences. For most of history puberty wouldn't happen until the late teens.
-This trait produces more offspring and is thus more adaptive even though it causes more death. Looking at the numbers, if a girl starts bearing a child every couple years starting at 14 through her fertile years (again in the relevant time period this is probably the mid-30s), this provides more offspring than starting at 18. Even though more total mothers and children will die compared to starting later due to the increased risk.
Because our biology is the result of kludging together different approaches over eons. When we were hominids tootling around the savannah and maybe not going to live past thirty, then being able to start replenishing the population from the early teens was advantageous.
Age of menarche in modern day human women seems to have gone up and down depending on the historical period, state of nutrition and health, etc. Currently it's in the range 10 to 16 years of age. 10 is clearly too young, though it's possible to survive pregnancy at that early age. It does seem to be shifting slightly earlier.
Then we have to consider the opposite end of the life span: what age do you kick the bucket? If you're going to have a good chance of dying by forty, then maximising reproductive years is the way to go. Nature cares naught for your qualms.
But since we're generally living longer, then we hit up against the problem of "so when are you no longer reproductively viable?" Age of menopause is (looking up online) 45-55 years, and before that you are in the perimenopause stage:
It's entirely possible to become pregnant and bear a living child into your late forties, and if we exclude women who used IVF/donor eggs to have children after the end of their own natural fertility/to overcome fertility issues, then "The oldest verified mother to conceive naturally (listed currently as of 26 January 2017 in the Guinness Records) is Dawn Brooke (Guernsey); she conceived a son at the age of 59 in 1997."
So theoretically at least you could have someone who began menarche at age 10 and finished menopause at age 55, giving forty or so years worth of child-bearing time (very few people are getting pregnant and having babies every year, year after year).
"The European Fertility Project of the 1980s (usually called the Princeton Project) showed that by 1911 Ireland had the highest marital fertility of any European country. A special survey as part of the 1911 all-Ireland census revealed that of women who had been married 20 years or more (and, hence, were likely to have completed the full cycle of childbearing), roughly 50% had born five to nine children and an additional 30% had given birth to 10 or more."
Evolution hasn't caught up to modernity yet. Our biological systems developed to function during adversity, so now when they note "environment peaceful, food plentiful, animal protein abundant" the conclusion is "conditions ideal, let's make babies!" You remain healthy and able to bear children? Keep on making those babies! The babies are not dying and the population is not being winnowed out by tigers, other predators, famine and disease? Hurrah for us, the species will survive, don't stop making those babies now!
Maybe under a thousand years of early puberty, the systems would finally make adjustments, but we're taking matters into our own hands now and not depending on the forces of nature to slowly and cumbrously shift.
> Maybe under a thousand years of early puberty, the systems would finally make adjustments
Probably not in the way we would hope for, but rather by growing the population until the resources are exhausted and again only 2 children survive on average (but many more are born).
Nature does not care about what is good from the perspective of aesthetic, or for the group. The individuals who volunteer to have fewer kids have reduced the frequency of their genes in the pool (unless there is a tradeoff where having fewer kids means higher chance of *your* kids specifically surviving), so any genetic impulse to do that would be self-eliminating.
As Mary Catelli has pointed out, it really depends on mortality. The additional risk only needs to be less bad than the overall risk for it to make sense, with further discounts depending on the kid's chances of survival. So if a median hunter-gatherer girl of 14 has a yearly chance of dying at 5%, waiting 4 years might easily be worse than risking it early.
Bear in mind that menarche isn't the same thing as reproductive maturity. The first couple years consist of mostly-anovulatory cycles. The system takes a while to come fully online.
If there are individual behavioral dispositions, or sufficiently widespread cultural dispositions, that prevent people from engaging in activities that tend to make them pregnant, for the first few years after puberty, then evolution would have no reason to biologically eliminate the possibility of pregnancy at those ages.
What's relevant here is the _completion_ of puberty which is a multi-year transition. For centuries across many cultures the word for that was one borrowed from ancient Greek, "menarche", meaning first monthly period. Cultures all over the world have understood for millennia that this step signaled the possibility of becoming pregnant.
The idea that modern people are _starting_ puberty earlier has been contradicted by archeological analyses:
What may be true (though not yet settled science) is that modern people in the developed world _complete_ puberty faster with improvements in nutrition being a key factor. This hypothesis is arguably supported by for instance medieval history: European royal and noble marriages routinely involved girls aged 14 or younger with the production of an heir being their immediate job description. [I happen to be reading a biography of Henry III of England, who at age 28 married the 12-year-old Eleanor of Provence; she was 14 when the future Edward I was born. That scenario was quite common among the kings, dukes, earls, counts, etc.] Since royals and the nobility had much-more-reliable and somewhat-more-varied food supplies than everybody else, their daughters completing puberty relatively quickly would jibe with the modern hypothesis.
Statistically, it takes a lot of effort to get girls pregnant "right at puberty." There are reasons why "you don't get pregnant your first time" gets bandied around -- it tends to work for sixth graders, for god's sake! (To put it a little more bluntly, they get pregnant and have a "heavy period" when they abort the baby).
Secondly, I don't think you have the same "increased risks" if you have a normal distribution of "who gets pregnant" near their menarche, particularly for the baby.
Puberty happened in the late teens in Scandinavia. Early teens was normalish for better nutrition areas, like the Fertile Crescent. There's a reason 12/13 is "manhood" in jewish culture.
These are the times when a girl was considered "definitely on the menu" for marriage. I'm pretty sure "womanhood" (as in being a maiden) is defined lower. Victorian England certainly had it defined fairly low. You might have more luck looking at "historical rapes" than "historical marriages" (again, you can reach puberty but, even when sexually active, not get pregnant particularly with kids of your own age group.)
This states bride kidnapping as young as ten, which is probably below menstruation age, and probably to "make sure she's a virgin." One hopes that there is no physical damage from sex with one that young.
I presume that most bride-kidnapping is done with people of "marriageable age" (that is to say, actually capable of bearing a child), in that consummating the marriage is part of the whole ritual. (and yes, it's important to note that in cases of "couldn't afford the dowry" this is an acceptable alternative that Women often agree to).
Citing the menarche wiki:
"From the sixth to the 15th centuries in Europe, most women reached menarche at about 14, between the ages of 12 and 15.[31] The average age of menarche dropped from 14-15 years"
> These are the times when a girl was considered "definitely on the menu" for marriage. I'm pretty sure "womanhood" (as in being a maiden) is defined lower.
I think you actually have it backwards. You can marry anyone. But womanhood is a defined role. From Mary Beard's SPQR:
> the age gap of forty-five years caused puzzlement even in Rome. Why had Cicero done it? Was it just for the money? Or, as Terentia claimed, was it the silly infatuation of an old man? In fact, he faced some direct questions about why on earth, at his age, he was marrying a young virgin. On the day of the marriage he is supposed to have replied to one of these, 'Don't worry, she'll be a woman [mulier] tomorrow'. The ancient critic who quoted this response thought that it was a brilliantly witty way of deflecting criticism
Publilia was a valuable marriage prospect as a girl. Marriage wasn't enough to make her a woman. [That job, obviously, is done by consummation. But it could be done by aging too.]
Isn’t it almost inevitable that if an organism goes from “can’t do X” to “can do fully do X” there’s going to be some intermediate stage of “can do X poorly and with some risk,” especially if X is at all complicated?
E.g., a newborn can’t eat any solid food, a four-year-old can eat anything, but in between you get a six month old who can have purées under close supervision, a nine month old who can start to have small bits of finger food but needs to be watched for choking, an eighteen month old who should still have things like grapes pre-sliced, etc.
For most of human behaviors this would be true (kids learn to walk before being able to walk without falling, for example), but for the menstrual cycle specifically it is not true. Women already have a distinct toggle from going "can not get pregnant" to "can get pregnant", so presumably the time when that toggle is flipped was the optimal time for reproductive success in the ancestral environment, and it makes sense to question why it happens before the woman is fully grown.
PLEASE read more. Girls can get pregnant before menarche (sometimes considerably before menarche). Women do not have a distinct toggle from "cannot get pregnant" to "can get pregnant" -- the toggle is dependent on amount of sperm (also quality), and number of sperm donors, and it's really complicated! And "can get pregnant" is very different from "can bear baby" (given the number of spontaneous abortions in young populations, which is "generally at least two or three before you get a surviving baby")
This also gets into some of the evolutionary biology "answers" that are really dumb.
1) It could be tied to something that enhances reproductive fitness (aka "earlier puberty meant not dying of the plague before you had ONE kid").
2) Women who were raped and didn't bear babies might have been more likely to be killed.
3) "Picture unrelated" -- you're dealing with a contemporary environment that is very different from the original environment, and can't generalize.
The last one is less likely in this particular case...
Alsotoo, I thought "fully grown" for women was around 14-15 (height specifically)? That matches with "time of menarche" pretty well, historically.
> Girls can get pregnant before menarche (sometimes considerably before menarche).
Unless you consider menarche to be delayed by pregnancy, I don't see how that can be true. I don't think that's a justifiable choice; given a woman leading a traditional lifestyle who has been pregnant so often that she can't remember the last time she had a period, and maybe never had one, everyone would consider that (a) she has been through menarche; but (b) by coincidence, she's never shed her uterine lining outside the context of giving birth.
> Women do not have a distinct toggle from "cannot get pregnant" to "can get pregnant"
Yes, they do. They have to start ovulating, which immediately toggles them from "cannot get pregnant" to "can get pregnant". You seem to have confused the concepts of "can get pregnant" and "will get pregnant, if inseminated".
You might also note that spontaneous abortions remain common at every age that can get pregnant. The fact that they're common in young girls doesn't tell you anything. They're common in everyone.
According to my mother, this causes her patients a lot of grief, because they get pregnant, proudly tell the people in their lives, and then miscarry. She always warns them not to tell anyone until the pregnancy is fairly advanced.
Forced ovulation through rape is a thing, dude. It can occur before menarche.
Quoting from wiki's menarche page:
"In post-menarchal girls, about 80% of the cycles are anovulatory in the first year after menarche, 50% in the third, and 10% in the sixth year" -- this is probably being measured without sexual activity.
General question for people more knowledgeable about AI than me: to what extent have increasing lengths of tasks AI can complete been permitted by increasingly effective use of the chain of thought, vs 'implicit' knowledge not spelled out every step of the way?
It depends on what you mean by 'implicit knowledge'. I'm assuming you mean 'better reasoning traces' but correct me if you mean something else
Naively I think it's hard to untangle how much AI has benefited from context window increases vs. better reasoning traces. They are sorta intermingled. We know that there is a meaningful improvement from increasing the size of context windows (cf. Gemini 2.5 vs Claude 3.7) but also that there are obvious improvements from just having better reasoning traces (cf. Claude 4 vs Claude 3.7).
I've noticed from reading ACX that I can actually enjoy poetry, which is pretty great! I haven't read much yet though. Does anyone have recommendations? I don't want to specify what I like, because I'm pretty sure I haven't even read most of the genres and I want to get a good understanding of what's out there. I do like poems that I can memorize.
Depends how modern you want to go, and at what stage you want to halt (I'm really badly behind on my poetry reading and can't think of a contemporary poet I know or want to read).
If you like "poems I can memorise", then how about a poem you can sing? Yeats heard a few verses of a song, some lines stuck with him, he wrote a poem out of them, and then that poem was turned back into a song:
I also like Eliot (took me a while to get there) but he doesn't really have easily memorisable ones (well, except "Old Possum's Book of Practical Cats" which got turned into the musical "Cats" - let's not talk about the movie version).
Rudyard Kipling. Pretty much all of it, but the famous ones are famous for a reason. "Danegeld", "White Man's Burden", "If", "Gods of the Copybook Headings", "Gunga Din", "Female of the Species", all good stuff. The "Ballad of East and West" is long, but worth it. And Kipling is about as close as we get to an Engineer Poet, so if you lean in that direction there's "Sons of Martha" and "Hymn of Breaking Strain".
Martin Gardner, "The Best Remembered Poems". This collection varies wildly by quality and genre and emotional level and... which is its strength. Some poems will appeal to you and some not, they will all be readable and potentially memorable.
In general, I hate poetry. But I was addicted to Byron when I was young. He has lots of short poems that I remember still.
My degree, which I am doing now, introduced us to William Cowper and Philip Larkin, and I enjoyed both. I've also enjoyed WH Auden, Ted Hughes and John Betjeman. You've probably heard poetry from all of them in bits of movies that make you cry..
- What Lips My Lips Have Kissed, by Edna St Vincent Millay
- Basically everything by Yeats (Sailing to Byzantium, When You Are Old And Grey, Second Coming, ...)
- High Flight, by John Gillespie Macgee Jr.)
- Acquainted with the Night, by Robert Frost (and many others also. I especially like 'Desert Places').
- A Clear Midnight (Walt Whitman)
- Do Not Go Gentle Into That Good Night (Dylan Thomas)
- Spring and fall (Hopkins) – listen to some recordings of Hopkins to get into his style, as 'his wríting can bé a bít wéird' but it sounds great when spoken)
I also like to read poems to memorise – the above are all very lyrical and memorable.
Well-regarded poets that I haven't yet learned as well as I ought: Keats, Auden, Poe, and many others.
Deiseach recommended 'Ceasefire' by Michael Longley on a thread about Gaza, which I was really taken by. Hopefully she chimes in with some more recs here!
Thank you for the kind words! Yes, Hopkins and sprung rhythm - he is a great poet, but you have to tune your ear into it.
Seamus Heaney of course, Yeats, the old warhorses of Irish poetry. I keep recommending this one but it's because I love it so (the link for the proper formatting of the poem because Substack hates us and our comments and doesn't want to give us the tools: https://poets.org/poem/vi-stares-nest-my-window)
Haviv Rettig Gur, an Israeli English-language journalist, has been rising in popularity; his is a definitely pro-Israel position, but full of nuance and interesting historical and sociological details. I liked two of his recent videos enough to make my own transcripts (I can't watch long podcasts), and thought I'd link them here, with a representative paragraph or two from each.
"One of the most damning things in the Haaretz report was the claim that sometime in the last 10 days, when it became clear this was a serious problem that was recurring, it wasn't a one-off incident and who knows, it was a serious problem that was recurring, it was systemic. There was a meeting at the Southern Command with the top brass in which Southern Command senior officers discovered in that meeting that soldiers sometimes, on occasion, I don't know if it's two times or 40 times, but on occasion, had used artillery shells as warning shots to keep civilians away. Artillery shells are long distance. Artillery shells are broad area explosives. Artillery shells are a tragic disaster in a situation like this. It serves no purpose. If you need an explosion to catch people's attention because it's turning into a stampede, there are small grenades that have almost no explosive. They have no shrapnel. They just made a loud bang. And every infantry platoon in the IDF infantry has them. You don't need an artillery shell.
The problem wasn't just the bad decision. The problem wasn't that that decision may be criminal. I don't know. We know that there are investigations opened by the military advocate general into some of these incidents. We don't know exactly how many, but that is something that the army has said and the Haaretz also reported, and others have reported as well in Israeli media. But the fact that they didn't even know, there was no clarity among the high top brass of the Southern Command, what's actually happening at these centers. That's the damning thing."
Podcast with Sam Harris (recorded June 17, during the war with Iran):
"And so when Hamas looks at the Jews, it doesn't see a colonialist oppressor. It sometimes uses language like that, especially when it's talking to other Palestinian movements that were born in that kind of discourse. But it fundamentally sees a great, vast theological problem in all of the weaknesses of Islam and all the indignities that Islam has suffered everywhere from European imperialism and various other retreats that Islam has suffered.
The worst one was Zionism. And the reason the worst one was Zionism, and they talked about this openly and constantly in the 1890s already, was that the Jews were weak. They're refugees. They're the weakest thing that ever pushed Islam back. And so they're the first thing that Islam has to overcome in order to get back to its rightful place in history.
So Hamas develops an ideology and a foundational strategy. It is willing to destroy Gaza on the altar of destroying Israel. Because in its vision of the purpose of the war on Israel, it's not about liberating Palestine. It's not about creating a new Palestinian political world that is independent of Israeli military rule, which is something that significant numbers of Israelis would join in on. It's about overcoming the Jews, ruling everything that had ever been lost to Islam as the beginning of the return of Islam into God's embrace and the sign that our piety is enough and the sign that Islam is not taking its rightful redemptive place in history once again. It is overcoming centuries of retreat and weakness in Islam.
And so, yes, the second we lay down our arms, they'll come for us. They is some of them, not all of them. It's big, it's complex. There's also the simple truth that when you talk about Islamic ideas and Islamist ideas, you're not necessarily talking about the Muslim shopkeeper in some street in Amman, Jordan or something. The people who belong to this religious world don't necessarily subscribe to the grand ideas. There's a lot of overlap, but they're not the same thing. But with all those caveats...
Yes, we are cartoon characters in a vast redemption story that they think they're acting out and living through. And the same is true in a different way, in a Shia version with a slightly different history, although borrowing a lot of these big ideas of the Iranian regime. One of the most ridiculous things about this war is why the heck does Iran even care about Israel? It has no border with Israel, it owes nothing to Israel, it lost nothing to Israel, it has no interest in Israel of any kind, and it has spent hundreds of billions of dollars its people don't have on destroying Israel. Why?
And the answer is, there is this grand redemption story they think they're embedded in. So everything you said is absolutely correct. The day we lay down our arms, they come kill us and think that it's a great miracle given to them by God. And it's the beginning of a Muslim redemption story of the conquest of the world. And it comes from a deep 150-year-old discourse that we have to understand, that we have to actually respond to and deal with and ultimately also defeat."
Hahaha. Why does Israel care about Iran? That's the better question. Iran has a nuclear program that they haven't wanted since around 2000, and have been (by bibi's own reckoning) "almost at getting the nuclear bomb" since 1991 or so.
Iran will say many things, but it is what they DO that matters, and they aren't trying to "bring the pain" to Israel (test drive? yes. Show that they could hit whatever they pleased? Yes. Show that they know where Israel has their nukes stored? yes.)
Most of Iran doesn't care jack about the whole "take over the world" ethos of Islam. The leadership doesn't want a regional war. Nuke Israel, and they lose all hope of peace -- with anyone and everyone. Nobody likes Iran, anyway (except, a little, Russia and China, certainly not the Arabs).
Ah, yes, you're the person who ... no, wait, when confronted with information that doesn't match what you've been told, you confidently assert that you already know everything.
When was the last time you had a discussion, formal or informal, with Iranian Military personnel?
So you are saying that Iran is not supporting Shiite militant religious extremists like Hamas of Hezbollah, whose main goal is to fight Israel?
Or that Mahmoud Ahmedinejad did not call for Israel to be "wiped of the map"?
Having formal or informal discussions with military personnel (aside from it being impractical for most Westerners in the case of Iran if you want to get anything except carefully vetted answers) is unlikely to give you much enlightenment, generally. "So, Wehrmacht grunt, how does Germany feel about Poland?" ; "Today I want to discuss with a private from the USMC if the US will wage war against Saddam again".
First, Hamas is Sunni. Seriously, do you know nothing about anything? Iran doesn't support Hamas directly, either. Hezbollah's main goal appears to be bombing Kiryat Shmona, which is nearly always evacuated. I'm not sure what they did, but Hezbollah could certainly hit other targets now and then. They could hit tel aviv if they wanted.
Yes, a man "nominally in charge" called for Israel to be wiped off the map. He also called for the destruction of America. The leader of Jordan provided the Protocols of Zion to a jewish american diplomat (kissinger if memory serves), who had to sit there and smile back at the trolling.
Ah, so you don't have some magical pipeline to our "military diplomats." Iran likes to call their shots, whether they're shooting at Israel or at American military bases. Called shots aren't nearly as destructive, or trying to be blatantly murderous, as shooting in the dark.
Do you believe Bibi when he says that supporting Hamas is a good way to preserve Israel? If so, what do you think he would say about Iran's support of Hamas?
What are the most interesting last words you’ve ever heard of? (or have come up with yourself?)
I’m anticipating dying within the next century or so, and want to get ahead of things in planning out my last breath.
Right now I’m considering “The gold is buried under the…” along with leaving notes in my notebook every few months saying I purchased a few gold coins, totaling a few tons of gold over my lifetime.
I hope my last words sound something like, "Ahhhh...", as I die of a massive heart attack in the middle of the night. (my favorite idea of how to die... it's how my father did it.)
Cracked.com had an old article about last words. The one that stood out to me was "more weight", Giles Corey's response to a court trying to crush a confession out of him.
> Maybe you all can come up with something better?
Harsh - instead of sending your own descendants and / or family on wild goose chases for lolz, why not do something funny and possibly marginally positive with your years-long paper trail, like try to out top TSA officials as double agents in the employ of China / Russia, or disclose the location of a years-deep cache of secret love letters between you and Kim Jong Un, or something?
Knowing me, and since I've only had an actually skilled doctor whose decisions I respected all of once in my entire life, my last words will probably be something like "My dying wish is that you'll sue these idiot quacks into oblivion. PROMISE ME!"
Unless we actually have AI doctors by then, in which case I'll just go for that old standby:
I see it more as a game that makes life interesting. Blackbeard's Treasure, Curse of Oak Island, The Essenes lost treasure, etc. have interested people for centuries and created a lot of fun stories and speculation.
Also, I am lowkey one of those cryonics nutjobs, so I figure that increases my chance of future revival. There would be a lot more motivation to use advanced technology to revive a frozen corpse if they thought the location of a hidden treasure was locked away in your brain.
Some that I've heard - several likely apocryphal - that have stuck in my mind, for one reason or another:
Spike Milligan: "Told you I was ill"
King George V (on it being suggested to him that he might recuperate better from his illness in the sea air of Bognor Regis): "Oh, bugger Bognor"
Admiral Horatio Nelson: " 'Kismet', Hardy" (sometimes reported as "Kiss me, Hardy"; see here for the full -rather amusing- treatment: https://www.youtube.com/watch?v=JPgn1bL3o_o )
(Sometimes, when stood in Trafalgar Square, I like to imagine to myself that that Nelson's actual last words were "Okay sure I've done some cool stuff in my life but for heaven's sake don't put me on a pedestal")
Dickens' Syndey Carton (upon taking his true love's husband's place at the guillotine, to enable true love and husband to be together): "It is a far, far better thing, that I do today, than I have ever done."
Julius Caesar: "Et tu, Brute" ("You too, Brutus?!"; satirised by schoolboys, of course, as "Et tu, couldn't finish the third"..)
Karl Marx: " 'Last words' are for fools who think they haven't said enough already"
According to the private diary of Lord Dawson, the physician who was attending George V at his deathbed, George's last words were "God damn you", directed at his nurse as she administered a routine sedative to him.
After the King went to sleep, Dawson took it upon himself to administer a lethal dose of morphine and cocaine, making him the last documented British regicide. Part of his motivation, as documented in his diary, was to ensure the King died before the print deadline for the morning edition of The Times so the announcement could first appear there rather than in a less-respectable evening paper.
I'm always amazed at the breath and depth of knowledge ("cross-section of knowledge"?) of ACXers! Wouldn't have guessed I'd hear George V's doctor's diary paraphrased here!
As a matter of fact I strongly suspect several of the last words that've stuck in my mind (including "Bugger Bognor") are apocryphal; edited my comment to say as much. Thanks!
>Wouldn't have guessed I'd hear George V's doctor's diary paraphrased here!
I came across him via Wikipedia's "List of English regicides", which I was browsing for unrelated reasons some time back. Something to do with Cromwell and the Restoration, IIRC. Dawson isn't there anymore, since some editor decided that unilateral involuntary euthanasia doesn't count as regicide.
Not my grandmother's last words to *anyone*, per se, but her last to me - after I (admittedly) theatrically kissed her goodbye on the forehead before flying home to my job so I didn't lose it - were, "Oh, so now you think I'm going to die, huh?"
And with that, she won the final round of a two-decades battle over her beloved Christian Science (the practice of which, it should be noted, was what was killing her pointlessly and painfully 20 years early). Even though she was obviously dying and was extremely unlikely to survive the two weeks until I could return, she nevertheless startled me into totally submitting to the religion's pathological denial of reality which she helped train me in as a child.
"What? No! I'm coming back to see you, I have my ticket and everything...!"
I knew I was lying to please her, she knew I was lying to please her. It was a masterstroke of getting me to *finally* say exactly what she wanted to hear.
I ruefully admire the hell out of it, and don't begrudge her the victory.
That was the last thing my mother said to me before she died. I mean, she smiled and shook my hand. It seemed appropriate, and I looked back and smile at it.
Personally, as an epitaph, which is not the same as last words, I would like "offer good while Supplies last"
Well, I guess it depends on how friendly a shopkeeper you are at the end of it doesn't it? How about "I'm sorry we're closed. Please come back another time.”
If one wants to keep in the theme of famous mathematicians, but go with good last words that actually are last words, there is always:
"Don't weep, Alfred! I need all my courage to die at twenty."
(Which would, perhaps, be merely fine and not excellent last words coming from your average Joe. But when you've already made significant and lasting contributions to the field of mathematics, the impact is rather different.)
(Yes, yes, he didn't die in a duel the next day, he lived for nearly thirty more years; this won't be the last apocrypha written in a comment here, I'm sure.)
Fermat scribbled in the margin of his book in 1637. He died in 1665 without having elaborated on it to anyone's knowledge.
Galois did die in a duel, and he was apparently writing about math the day before, but there -was- an apocryphal tale of the same having happened to Fermat, possibly due to someone being low on coffee somewhere in PBS' research department (I recall the tale on some TV documentary). Possibly confusing the two, just as you say.
I never heard of that tale about Fermat's death. On the other hand, Evariste Galois is famous for dying in a duel after spending the night writing about his theory. But I believe you that it is someone else who confused the two. Apologies for implying it was you.
It's the funniest version of it. Sadly, reality is usually less poetic than fiction. Those technically weren't even his last words, just his almost last ones. After he said the elephant line a soldier said something clever back to him and the General's last words were laughter and "All right, my man; go to your place."
Similarly, the wallpaper remark was said by Wilde on his deathbed, but probably wasn't his actual last words. And the wording of the wallpaper joke has about a dozen different versions that have been reported over the years. And while Mozart did say the wine thing shortly before death, his actual last words were probably something else.
But why let reality get in the way of a good final sentence? For most people their last words are something like "Good morning" or "I don't feel so well" or "Blergh. Ugh. Gack." I have no idea what my grandfather's last words were, but since he spent the last week of his life delirious (from a head injury combined with hospice levels of morphine) it probably wasn't anything witty.
In last week's OT, I suggested a web app to test how clustered political beliefs are. One of the commentators, Robert, built this: https://politicalbeliefs.online/
I've hit unsure about almost everything and now feel a little scared - I am pretty sure most people haven't researched these issues any more than I did, so why do they have a sure opinion? Considering capping bank profits - how the hell should I figure out whether high-profit banks are simply making good investments and thus high profits mean more good investments, or they are gaming the system somehow and then this is a good idea?
I have a strong prior that populist policies in the vein of "capping bank profits" are destructive and counterproductive based on lots of past examples (Rent Control, anyone?)
An interesting test would be figuring out a way to test certain cross-cultural clusterings. For instance, within a Finnish context, some beliefs indicating right-wing beliefs (social conservatism, mostly, but that of course correlates with economic right-wing thought) would be "firmly supports NATO membership" (this one less than previously, since NATO support is currently widespread, but there have been studies on belief clustering indicating that this was *the* biggest right/left predictor and "supports conscription in its current form", both of which would be opposed by swathes of American right. "Supports the state church(es)" was a similar one earlier, not sure if it's as valid now.
68%, it pegged me as a liberal and chose the libby-lib answer to everything; my beliefs are 100% clustered but they are clustered by some other, different, mysterios ideology.
Confidence seemed way off though, everything was super high confidence .
I didn't like a fair number of those questions. Maybe "didn't like" is not the correct way to phrase it. Many of them took very complex topics and simplified them down to a single sentence in a way that, for me at least, meant that I couldn't answer yes no. My real answer was very often "it depends on the details", and so I was left to mark "unsure", even though for any real example of the topic, I would have a very strong yes/no opinion, because I would know the details. Obviously, it is hard to give a lot of detail and nuance in an online survey. Both because it is just inherently difficult, and also it makes the survey more likely to be abandoned by participants.
I'm not sure I have a solution, but I thought it was worth pointing out. It had 57% accuracy for me
Yes, for instance the one about the Democratic party more reflecting my beliefs or whatever it was.
The old party? Very probably. There's a reason the Democratic party turned out so strongly Irish-American back in the day.
The current shower? About six items that I would be sympathetic with, hanging on grimly from when they *were* the party of the working man, and buried under an avalanche of idpol and wokeness since.
Same. Prob gonna be a problem with this audience in particular. For e.g. I do think Greta is kinda inspiring even tho I couldn't disagree with her more.
It got 56% correct for me. The "calibration" was 51%, which seems oddly high to me, since it had 80-95% confidence on nearly every question that it missed. It seems very overconfident.
Based on the pattern of responses, it kept expecting me to be right-wing and was surprised every time I picked the left-wing answer that is also the libertarian answer. Also, I answered "unsure" on several where I was actually unsure, had no real opinion, or thought the answer was more like "sometimes," and it got every one of those wrong (does it ever predict "unsure"?).
Did predict "unsure" for some of my answers; got some of those wrong, some of those right. It calibrated better later in the test as we went along and it got a handle on me.
Yes, for now questions are classified as 'progressive' vs 'conservative', and you're right that in reality, we should have at least 4 quadrants (as per the political compass). Once we have more answers, we're hoping that a better algorithm will increase accuracy.
50% prediction accuracy for me, which didn't surprise me much because I've got rather an oddball mix of beliefs and affiliations. And after thinking about it for a minute, I realized that since the tool is brand new, it probably doesn't have a huge amount of data in it and probably is a bit rougher in terms of accuracy than it will be after it's been used a bit more.
To start with, we've just coded some statements as conservative, and some as liberal. As we collect more responses, we hope the prediction will become more accurate.
There is a question about whether companies that produce critical products, like SpaceX, should be govt-owned. Using an example, especially that one, is likely to skew your results by testing attitudes re SpaceX/Musk, rather than re the underlying principle.
I avoided the Musky One in thinking about that, I think large important private enterprises like that should be let alone. Doesn't mean the government can't go off and do its own space programme, just means keep the grabby hands off.
The other way round, too; not a fan of privatising every single formerly nationalised enterprise, some things can't be run as a profit-making enterprise and do need to be done by the state.
In this case the obvious "fix" is to replace SpaceX with ULA (United Launch Alliance). The catch is that I expect a lot more people know about SpaceX than ULA.
If your child tip-toe walks I have had enormous success just by purchasing my son cowboy boots. Only hiccup was getting a pair that zips up the back. Wanted to let everyone know as a public service announcement.
Huh, well a quick google shows that there are lots of organizations that think this is a problem, and I am certainly not an expert, but my initial impression on reading your post was don't do that - tip-toe walking is a good thing. Well, or at least walking landing on the front of your feet and then rolling back to make contact with the heel as well. Many of the pictures from my search showed children staying exclusively on their toes, which is probably not ideal...
But it has been my understanding for years now that the way humans are supposed to walk is by landing on the toes / ball of the foot first, then rolling back onto the heel. The reason for this is that this motion greatly reduces impact forces being transmitted to the knee. Your heel has no cushion, so landing on your heel directs impact straight to your knee, which over the course of a human lifespan causes damage. Whereas landing on the balls of your feet provides significant cushioning, as much of the impact goes into stretching the calf muscles as the foot bends. And since the calf / Achilles tendon is stretchy, this actually stores energy that will be used during the next step which makes the walk cycle more efficient.
My understanding is that this is how ancient people walked (known from fossilized footprints - footprints in soft mud are different when you land heel-first vs toe-first) and also how modern hunter-gatherers and running culture people walk/run. Modern people have been trained out of walking/running like this because of our shoes. With nice puncture resistant shoe soles reducing the immediate harm of taking hard foot falls on the heel, and also shoes adding rigidity to the whole foot, making it harder to land toe-first, most modern people learn to heel walk, as it is slightly easier in the moment. It certainly requires less calf strength to heel walk. But in the long term, this leaves one worse off, as it leads to much greater likelihood of knee problems in the future.
I have been trying to train myself out of heel-first walking for many years, and have been fairly successful. I hike and run a lot and feel like it is significantly better for me. And most other hikers I talk to agree. But I suppose I don't really know. It is also possible the behavior you are seeing in your is different than what I am talking about - staying on your toes throughout the whole step is probably less efficient, and certainly standing on them is going to be less efficient. It may help train calf strength though!
I had a friend in high school who had a severely limited range of motion due to lifelong toe-walking. He literally couldn't put his heels on the ground and that significantly impacted his ability to walk comfortably.
One of my cousins is a life long toe walker. I can say with confidence that he never learned to walk normally, he is still 100% walking on his toes in his early 40s.
I see it frequently associated with autism. I don't think he's autistic, though he might be. He doesn't have the tactile issues, obsessions, or problems with social queues. He didn't really start speaking until 5-6 years old and was raised in semi-isolation from other kids. (His mother was very religious and sent him to a Baptist homeschool collective).My grandmother blamed it on him biting a lamp cord when he was young. He is obviously strange to interact with. People called him "touched" when we were kids.
A lifetime of walking on his toes has left him with terrible balance. He looks like he's always about to fall forward, and for the most part he is. Gentle bumps that most people would recover from with a with at most a stumble to one side send him straight to the ground. He was never able to participate in any sports (he could bowl ok, he developed a very odd but effective way of rolling the ball). It turns out that extended toe walking into adulthood damages the hip and knee joints and causes the core abdominal muscles to develop asymmetrically, which have caused him a great deal of pain starting in his 30s. He's never really been able to run or jog at all. As his 'education' never involved gym class or sports and the homeschool community militantly minded their own business regarding parenting, it was never really raised as in issue.
What are some of everyone's favorite bands (and an album from them)? I'm interested in what people here listen to and potentially looking for something new to listen to.
Alternatively, this album https://www.youtube.com/watch?v=Uyf79XR6r1I by the Canadian band Musk Ox is brilliant when I just want the world to go away for a while. It kept me sane for years of ultra long commute.
A commenter in here asked for new leads about a year ago, so I put together a spotify playlist. Since it already exists, I might as well share it a third time. Apparently, the link expires after a week, as a security measure.
> What are some of everyone's favorite bands (and an album from them)
Nobody else has even mentioned anything metal or adjacent, and I know there's at least a couple more metal / hard rock guys here, so figured I'd contribute:
* Dodheimsgard - Black Medium Current (abstract technical mastery and epic vibes)
* Amon Amarth - Twilight of the Thunder God (viking themed metal mayhem)
* Chimaira - Resurrection (adrenaline rush intensity and energy)
* Kittie - Funeral for Yesterday (best all female metal band by far)
* Tesseract - Sonder (relatively complex high energy hard rock)
* Tool - Aenima, Lateralus, Fear Inoculum (these guys are the JS Bach of hard rock)
* Black Light Burns - Cruel Melody (dark, gritty, and tortured rock)
* In This Moment - Blood (female vocalist, good intensity, dark themes)
Metal/metal adjacent is one of the few genres I do not enjoy, along with pop country. A lot of people I know really enjoy metal, I don't understand why I just can't. Thanks for the contribution though! I'll still check some of it out!
I'm not finding any goat rodeo album by a goat rodeo band. The closest thing I found was "The Goat Rodeo Sessions" by Yo-Yo Ma, Stuart Duncan, etc. Is that what you were referring to?
I listened to a few from each. They kind of freak me out, idk why, just kind of unsettling in my stomach. The song "In the Flowers" by anco freaked me out in the same way when I was 13 or 14. Idk what it is about the sound, but it doesn't sit right with me.
Finding out different people's music tastes on here was my main reason for my original comment, anyway. It's also interesting to me how people can react so differently to the same music. Thanks for the contribution!
I really like Automatic for the People, and I grazed through the rest of these. Overall, the vibe reminded me of a playlist my dad played on Pandora when I was younger! Good vibe. I think the playlist was based on They Might Be Giants, but it played songs like Flagpole Sitta, New Slang, The Impression That I Get, and other stuff like that. Good times.
My Steeleye Span tastes are heavily nostalgia-tinted. Storm Force Ten and Rocket Cottage are the ones I grew up with. I should give their discography a more serious listen.
Well, there is much to be said for the music you grew up with. My mother had several Steeleye Span albums, although I was unfamiliar with most of them until I borrowed them for college.
If you're in the US region, it's all available through youtube:
Rocket Cottage wasn't one of my mother's albums, but I very much liked Fighting for Strangers and Sir James the Rose, which made it onto one or more greatest hits albums. (Sligo Maid did too, but I never really appreciated the jigs. I like having words.)
I was looking it up and it looks like John Flansburgh from TMBG produced it! So that's a fun fact. I listened through it, I liked nemeses the most.
I knew about Jonathan Coultan from Portal and I've heard some of his Thing-A-Week songs, but I don't think I've heard any from Artificial Heart before now (aside from the Portal ones).
RYM's userbase ranks LONG SEASON by Fishmans as the 29th greatest (popular) music album of all time. (https://rateyourmusic.com/charts/top/album/all-time/#pos29). My understanding is that this is almost solely on the strength of word-of-mouth campaigning on RYM forums.
What's going on with the album name? "宇宙 日本 世田谷" looks like "universe - japan - [meaningless gibberish]" to me. If I look up that third word, it seems to mean nothing other than being the name of a particular museum in Japan. There are no verbs.
Setagaya is a ward in Tokyo, it's where the singer of the band was from. You're right, it's "Universe - Japan - Setagaya" basically going from everything to the place specific to him.
Hard to go wrong with The Knocks. No particular album but a few bangers:
Slow Song
Classic
Brazilian Soul (and the Acoustic version)
One on One
Bodies
Ollie Wride is a blast. Got to see him live, great performer.
Check out "Running in the Night". His best by far. If you like that, you'll like most other stuff he's done.
LEISURE's "Sunsetter"Album is weird because I'm listening to the same song 10 times with minor variations and it's all great. You know how ACDC only made one song with 50 different lyrics but it's so good you still love it? This is like that but chill lounge music.
Might be mainstream but I've really enjoyed Teddy Swims.
Been getting into Franc Moody recently. Loved Dopamene, then loved loved loved Dance Moves.
You should have exactly one 6ix9ine song in rotation. Yes, he's a horrible human being. Yes, it's not technically good...but it's so fucking hype. I recommend "Billy". Sometimes you're depressed and too in your own head and you need a sociopathic mentally challenged teenager screaming:
"Get the fuck up out my fucking face before I murder you."
It just...it's just the most testosterone laden thing ever. Not like mature test, like pure idiotic adolescent test.
Billy Strings and Bryan Sutton - Live At The Legion is one of my favorite modern examples of bluegrass flatpicking guitar. The rest of Billy's catalog is more experimental/psychadelic (which I also really enjoy), and Bryan's is more traditional stuff. If you like it, similar older releases will be from Tony Rice (and Tony Rice Unit), Doc Watson, and others that you'll find if you branch out from there.
A friend of mine recommended Billy Strings a few years ago, I forget the album but I enjoyed it for sure. I'll have to check out that live album and Bryan Sutton's stuff.
I'm making the move many late career devs do into being independent, creating my own web dev and general software development shop. I heard from a friend that with AI, now is a great time to start a business, since AI is a huge force multiplier, and I guess in the near term, being a business owner is a good hedge against being replaced by AI.
I was wondering if anyone here is making similar moves and if they have comments on it.
I am deep in this process myself, but building something in a specific niche with a partner, rather than generalist consulting. I would advise caution with the generalist approach it sounds like you're thinking of.
My thoughts are:
* "huge force multiplier" is absolutely right; underselling it if anything. It's exhilarating what I have been able to achieve, and how fast. Things I know almost nothing about, that I never would have even considered trying myself in a serious startup effort, I'm getting done to a high standard of quality in a couple of weeks. You will be able to bring this ability to any customer that you can get; it will feel amazing to you, and they will be floored at the speed and quality of your results.
* I think (or hope, since I'm trying this model) we're on the cusp of a wave of start-ups that achieve significant initial success with no employees beyond the founders, and no funding. These will comprise at least one generalist SWE, and at least one high quality expert in some important field (law, biotech, agriculture, oil&gas, etc).
* In contrast, the world of generalist consulting, *especially* web dev, is going to become incredibly competitive. I won't go so far as to say "100% automated away" - but close. Starting out with direct links to potential customers who trust and respect you is going to go from hugely important to 100% mandatory. I think this is the one variable in my advice to you: if you already have some likely customers lined up, like to the point of informal agreements to hire your services, then go for it. Otherwise, hold off until you do.
I was thinking of following a strategy I read online of making simple sites for small businesses, on a subscription of model of $150 month. I think not many are thinking of serving that market, and that quite a lot of people aren't technical enough to use something like Squarespace or AI to do this.
I would try making a product company if I had a partner with a niche as you're describing, but I don't know anyone like that. How did you find yours?
Ahhh, ok, that's certainly not a 2-3 OoM difference. I can see small businesses, especially older/more conservative owners, going for $150/month. Just be sure you have some reasonable and clearly communicated boundaries, or you'll be valuing your time unsustainably low.
>I would try making a product company if I had a partner with a niche as you're describing, but I don't know anyone like that. How did you find yours?
marriage ;) (downstream of grad school)
In all seriousness, I recognize that I am very fortunate to be in a major right time/right place/right person confluence. As you say, what I'm describing is not available by default. If I was the same me in the beginning of 2025 without this opportunity, I would be itching to find something like it and entirely possibly send myself on a wild goose chase... the reality is that it needs to be the result of networking. Which is terrifying for someone like me, since I am a few clicks above hikkikomori in that department.
Anyways, definitely best of luck; you've got good answers to my concerns, so it might be viable. But if possible, do cling to mainstream employment until this venture becomes a money hose you're having trouble holding under control!
I have absently considered it noticing how much AI has affected my productivity, but then I also wonder that the market value of my output is about to plummet and may be better off drawing a paycheck a bit while longer than being so directly exposed to market forces. Shrug
Yeah I would think this is the worst time to do it, since the market is going to be flooded with people making apps since the barrier to entry is much lower. Whereas with B2B software job, the barrier to entry isn't a good app, but the connections with business that you sell to, which was true before AI and is more resilient to more apps because they rely on social connections. The B2B software I've worked on has never been exceptionally good or anything, it's because we have social connections to big companies and can talk to them directly is the real value.
When the market is going to get flooded, your job is to sell the tools for those who will flood it.
You can't sell an AI; you don't have a budget for that. But perhaps you could sell some generic wrapper system for the AIs, something like "buy my system, register your API keys, write your prompts, choose one of these nice graphic templates... and you have your AI-powered application ready, you can immediately start charging your customers".
I think B2C (making apps) was always essentially a moonshot. I'm just going to be operating a web dev agency. Plenty of small businesses could benefit from having a good website, and I think lovable is a considerable distance away from being used for that purpose by someone non-technical.
I'm afraid there are two potential fatal flaws in the business case you're hoping for here.
First, regarding "considerable distance away": consider our host's recent announcement of total victory in a bet on AI progress. Are you really that confident that "text to website" won't have massively improved 2 years from now? Especially considering that a lot of the work to be done is mundane plumbing and polish, and not plateau-able AI research witchcraft?
Second, even if those tools never improve further, are you sure you can compete with them? I entirely believe you can deliver a noticeably better result, especially if it comes to complex/unusual requests. But. Can you compete with something that can probably handle like 80% of small business's needs, to 80% of your quality, for 2-3 orders of magnitude cheaper?
I think many people are just too non-technical to have an AI generate a website for them. It's very common for people to struggle with their OS and gmail: someone like that won't be able to reason about the website they want well enough to have an AI do it for them.
I think Scott is overly optimistic about AI capabilities: he won his bet, but he had to admit AI still fails on prompts that a human artist would do 100% correctly.
Still, I trust Scott, and on the off chance something like AI 2027 does happen, I have $20k split evenly between a tech index fund and the S&P 500. If AI automates my profession (and others too, dev won't be going down alone), these investments should go to the moon anyway.
Some papers came out recently on the limits of reasoning in LLMs. They found that as more state is added to a problem, the AI can't handle it past a certain threshold, and its ability to solve the problem collapses:
i mean, i've been a skeptic for a while, and this is what I would naively expect. If it's just "pattern matching", it was always going to be lossy in the limit. E.g. if you think of "chain of thought" as a Kalman filter, of course it's going to implode at some point. At some point, it needs to be able to drop into some sort of subspace where the reasoning is lossless instead of probabilistic.
Sure, in theory, you can compute anything with infinite compute. You can solve quite a lot of problems with infinite resources, in general. But is a gigantic Searlean hash-table really the holy grail of AI? Isn't the point of intelligence to do more with less?
The tldr though is that we know LLMs have context length limitations because of the downstream limitations of our hardware resources. Models that have longer context windows do better. Many researchers are trying to crack intelligent summarization; others are trying to crack infinite context windows. But from where I'm sitting, there's no reason to assume LLMs are actually less capable in the limit.
Thanks for the post. Yeah, I can see AI 2027 maybe happening, but I can also see LLMs hitting a limit. They already suck up so much compute, and you get a result that is in some senses superhuman and in some senses subhuman. It's totally fair to speculate whether the LLM architecture is inherently unreliable and a new paradigm is needed to get past that.
personally i think ai 2027 is a few years early. The biggest limitation imo right now is we do not know how to evaluate models that are better than human, and all of our existing mechanisms for neural network improvement depend on some human, somewhere, constructing some kind of method for generating gradients through some kind of evaluation
I think it’s important to remember that no model yet is “better than human” in general - and I’m skeptical that there will be such. There will be models that are better than human at some things, worse at others, and just weird and different in many. It’s really the “weird and different” that makes them hard to evaluate more than the “better”.
Just as an illustration of the general problem, think about how SAT scores might be a reasonable measure of verbal aptitude for a randomly chosen English speaking high school student, but once someone starts studying test construction strategies, it becomes much worse.
If we assume that LLMs are going to automate lots of white collar jobs, do we think that licensed jobs are going to be relatively safer? My entire professional career has been in sales, for the last decade I've been in a non-licensed job, but recently I've been thinking about switching to something like selling insurance purely for the license. (Other options could include selling real estate, or securities, or mortgages, or anything else that requires licensure by the state).
Presumably a non-human agent cannot receive a license. The idea is to take advantage of a regulatory barrier to ward off LLM competition. I have some faith in the lobbying power of the FIRE industries (finance, insurance, and real estate) with both regulators and Congress, if needed. Does this sound like a valid plan? Or am I missing something?
(While I am not an attorney, one could make the same argument for them too. You obviously need a license to practice law at the moment. Supposedly around 50% of Congress are attorneys- can I trust them to act in their own self-interest? I mean probably, right?)
not thinking anything would be "safe," as less white collar jobs would mean less insurance to sell. Ai is bad because it's not creating any jobs or markets but just replacing labor for cost savings. the market that labor provides will dwindle and be replaced by corporate investment.
my guess is people shift into things AI can't affect demand of, like health or food.
Historically, one workaround to licensed people being more expensive than unlicensed people was to have unlicensed people to most of the work and then have a licensed person come in for the last bit to look over the results and make things official.
Replace unlicensed person with AI and the same basic strategy seems like it would work here. The most skilled (or more flexible or more forward looking ...) licensed people will ramp up on using AI assistants. Fewer people (including licensed people) will be needed unless the demand goes up by enough to make up for it.
The fundamental problem is that if what you are doing can be done A LOT LESS EXPENSIVE by someone/something else your career is not in a good place.
I run a startup that has recently started selling a unique insurance product, and I would be very surprised if these sorts of licensed jobs are safer, unless the job requires a human presence like an attorney or perhaps doctor.
Selling insurance requires a person at your company to be licensed, but there’s nothing that says I have to do my outreach and marketing personally. We can (and do) marketing using AI tools. I have a list of a few thousand people, and could standardize I method to find ~infinitely more. If I trusted AI to send emails, texts, LinkedIn requests and make calls, all without human intervention, then the license would only be useful to the people who own the business, not any of the employees.
What does "a unique insurance product" mean? Is it just insuring some niche thing (I know there are e.g. few and specialized insurers for rocket launches), or is it more complex, like working in some special regulatory locale or something?
I was referring to the first one, as in it’s a specialty surety product that only 4 other companies offer, but the second is also true, as we’re based in the most highly regulated market in the world (New York State).
There’s aspects of insurance that would resist AI for a very long time due to the regulations (actuaries, reinsurance intermediaries, the owners of agencies and underwriters themselves, etc.), but I would say selling insurance isn’t one of them.
We have a few people selling part time right now and if AI could do their job, they wouldn’t be employed. I would just use AI under my own and the company’s license.
Third-year law student here. I think you're right, to an extent. The legal field acts like a cartel; only those who get the license can practice. But I don't think this insulates me too much. It will become malpractice in a few years to not use AI. So while it may preserve the job itself for some time, it may become a matter of just approving the work your AI does. And on the client side, people are just going to use LLM's as attorneys. LLM's already give legal advice, but just make sure to say "this isn't legal advice" at the end. I don't think most people are going to be too worried as long as they are getting high quality legal advice for 20 dollars a month, rather than 200 dollars an hour. So I anticipate the work drying up from both ends.
>And on the client side, people are just going to use LLM's as attorneys. LLM's already give legal advice, but just make sure to say "this isn't legal advice" at the end. I don't think most people are going to be too worried as long as they are getting high quality legal advice for 20 dollars a month
But that's not 'practicing law'. I'm sure people will get frequently terrible legal advice from LLMs, but they can't represent someone in court, or even in legal negotiations between parties. If anything you could argue that the Internet kicked this trend off, but attorney employment has only risen over the last 25 years.
I see your point about pro se litigation below, but if the court system is overwhelmed with pro se litigants, they're just going to make that harder/borderline impossible. Again Congress is 50% attorneys, they're not going to sit back and let their industry get decimated. I'm sure the judiciary will quietly go along with whatever they pass.
Plus pro se types are going to lose in court an enormous amount, especially against a trained attorney and with an exasperated judge. No one with real money on the line (upper middle class person, a business) is going to represent themselves pro se in a real case. If anything company executives are violating their fiduciary duty if they either sue someone pro se using LLM tools, or try to defend their company that way. The actual bread & butter clients for attorneys will stick with the licensed types. Could there be more crazy people suing each other in small claims court using an LLM? Sure, but attorneys mostly don't take those kinds of cases now anyways
> I'm sure people will get frequently terrible legal advice from LLMs, but they can't represent someone in court
The requirement that you either show up in court yourself or hire a guild member to appear on your behalf is a pretty interesting perversion of the "right to an attorney", which was meant specifically to prevent that requirement.
I don't hate this argument, but US constitutional review famously found a right to a public defender somewhere in the US constitution to address this exact problem. As a practical matter, for lower-income folks there are (taxpayer-funded) free legal services. Try being a landlord and evicting a low income tenant for blatantly not paying you rent, trashing the place, etc. They'll show up to court with free (taxpayer-funded) attorneys to combat you and look for loopholes in the lease, etc.
Try being a landlord and evicting a low-income tenant for default, and sending your son to court to make your case rather than a hired lawyer.
> but US constitutional review famously found a right to a public defender somewhere in the US constitution to address this exact problem. As a practical matter, for lower-income folks there are (taxpayer-funded) free legal services.
Note also that this is a textbook case of "restrict supply, subsidize demand", which is never a good policy.
-- the current numbers are that 50% of the Senate and 30% of the House hold law degrees and have practiced law at some point. So overall the percentage is either about 40% (if you just average the two chambers' percentages) or around 33% (if you factor in that there are way more Representatives than Senators).
-- that percentage has been in long-term decline. Before the Civil War, Congress was three-quarters attorneys; in the first half of the 20th century it was around three-fifths; by the end of the last century it was around half; today, see above. (That's despite the number of attorneys rising steadily: during the 19th century around 1 of every 250 U.S. attorneys was a member of Congress, today that ratio is around 1 in 7,000.)
-- also worth noting is that whereas nearly all 19th-century POTUSes were attorneys, only a couple in the past 75 years (Nixon and Obama) did more than dabble in it. Since the end of WWII 9 of 14 POTUSes never even attended law school which is a previously-unprecedented trend in US history.
If the percentage of Congress that is attorneys keeps sliding downward while most POTUSes are not attorneys, it may be that the legal profession ceases to drive federal regulatory decisions about new topics such as LLM's.
As a small counterpoint, I do think litigation is protected. See what happened to "do not pay", a startup that was attempting to sell direct to consumer, routing around lawyers -- the legal profession basically rose up as one and drowned the startup in lawsuits.
Looking the Wikipedia page for it, it seems to be doing fine. In fact, 'Above the Law [a prominent law news website] ultimately recommended the service for "clear-cut issues like parking tickets or non-critical matters," while cautioning against its use for legal issues with higher stakes'.
I specifically said 'litigation'. DoNotPay rather infamously tried to 'defend' someone live in court, and in response got hit with two class action lawsuits and an FTC fine, after which they have not tried to do that again
Startups returning dividends to investors isn't a good thing, I think describing "two class action lawsuits and an FTC fine over two years" as "drowned the startup in lawsuits" is totally reasonable semantically, and the larger point -- that the legal profession will close ranks to protect itself against AI -- remains.
As a meta comment, I find your semantic nitpicking to be extremely tiresome and ask in the future that if the only thing you want to add is about semantics, you refrain from doing so
I see the point, but I'm not sure it covers as much as you might think. 90% of litigants are *pro se,* or self representing. You can sue a company trying to sell unlicensed legal solutions, but you have no way to stop individual people from using ChatGPT to help with their own legal issues. *Pro se* litigants are allowed to Google questions, visit legal libraries, etc to facilitate their self-representation. AI will function the same way, but will be much more effective.
I suspect any company trying to advertise itself as tailored to helping pro se defense will be shut down for much the same reasons as do not pay. But, fair enough, I could see self defense becoming more popular
For the first time in a while, a positive development for a stable peace in in palestine that is marginal and incrementally workable: individual leaders trading recognition of israel as an ethnic homeland for territorial control and economic integration. In a different timeline maybe a two state solution could have worked, in this one an emirate model seems like a hopeful alternative. Maybe one day if it flourishes it can turn into a confederate model as a trust and integration grows.
Partitioning into a dozen tiny emirates seems like a nice idea in theory. But I suspect it just results in constant fighting between those emirates, and between those emirs and various other forces that want to take over.
I'm certain that Israel winds up getting drawn into it one way or another. Like, the Emir of Bat-Mobil is getting attacked and he's Israel-friendly so we'd better just send a few troops to help out, and all of a sudden you're back to occupying the whole darn area again.
It worked for Russia in Chechnya to the extent that there are Chechen regiments fighting with varying degrees of success in Ukraine on the Russian side.
It didn't work for South Africa with the bantustans.
Russia had to endure about a decade of endemic violence after the end of the active phase of the Chechen war and ignore the "international community" which kept telling it that it's impossible to win by force. Quite a lot of a abuses were committed, naturally.
Unlike South Africa, Russia never faced widespread popular condemnation of its behaviour in Chechnya, no calls were made to boycott it. In spite of all the wars against Muslims in Afghanistan, Chechnya and Syria Russia is still seen positively or neutrally by the public in many Muslim-majority countries. https://globalaffairs.org/commentary-and-analysis/blogs/globally-negative-views-russia-predominate
Compared to Israel Russia had the benefit of having Chechnya within its internationally recognised borders.
Russia faced more popular condemnation from within Russia, for its treatment of Chechnya. Specifically, for its "rebuild BETTER" philosophy. "We didn't revolt, so we don't get new buildings? What is this mishegoss?"
Well, yes but I wouldn't say it was that strong. People were pissed off and Navalny tried to lean into this feeling but it never led to any violence and never truly threatened the long term rebuilding program.
Btw this is an advantage of an authoritarian regime. Any Israeli leader who attempts to spend Israeli money on rebuilding Gaza (hard to imagine, yes) would face huge opposition from everyone.
On the other hand, Netanyahu sort of tried it with the expansion of work permits for Gazan workers in the early 2020s. Turns out purely economic levers don't work so well.
Ukraine is also getting an upgrade (at least the Eastern Provinces).
Israel is far more likely just to get the Europeans to spend money on rebuilding Gaza. Bibi's been loud and long about "save israel by supporting hamas."
In the 1950s, Robert Sheckley wrote a science fiction story called “Watchbird.” Robotic drones with collective intelligence capable of self-learning, called “watchbirds,” are deployed throughout the United States to prevent murders. Gradually, they expand the concept of murder to include the execution of criminals, the slaughter of livestock, fishing, killing mosquitoes, performing surgical operations and so on. Life becomes unbearable for everyone.
Self-learning “hawks” are created, whose task is to kill the watchbirds. At first, they perform their task well, but gradually the watchbirds learn to evade the hawks better and better. In the end, the hawks decide that they must kill not only the watchbirds, but everyone.
In my opinion, the first part of the story is a metaphor for what happened in the following decades. The concept of “racism,” as well as the concepts of “sexism“ and “homophobia,” which were coined in the 1960s, have been expanded by wokists to such an extent that anything can be deemed racist, sexist, or homophobic. In Java programming, "everything is an object." In modern Western discourse, "everything is racism/sexism." People are afraid to speak their minds. Life is now unbearable for many if not most people.
That's why people voted for Trump. I hope Trump will be more successful at destroying wokists than hawks from the story in destroying watchbirds.
If your problem with woke was that it was anti-meritocratic, or bad for democracy, or grossly and capriciously punished people who didn't deserve it, I have bad news for you about the person you put in charge of replacing it.
It's not just a habit, though. Being enraged about this or that issue, hating this or that group can become part of somebody's self-definition,. and a way of bonding with other people.
I think democracy is like a ship built a certain way. It's remarkable -- stands up great to some conditions, and tolerably well to most others. But there are some rare combos of conditions that sink it, and the US has ended up in one of those combos. I don't think we're getting democracy back.
The demise of USAID means that they're fleeing the creative realms fast (as people are no longer paid to be trans-positive, or whatever the current woke-fad is.) See: "paid to make a gay-friendly play in the Marshall Islands." Also see the Gay Pride flag flying over Kabul, if you really want to make the point about State being a bunch of a-holes.
Innocent man still falsely imprisoned (no innocent until proven guilty when the Democrats have hung their hat on billions of dollars of property damage). I'd be afraid if I worked in his profession. You do everything right, and you're still going to jail -- no chance of appeal, either.
Maybe I'm being willfully blind to the real truth, but I thought just about everyone agreed he was dead, and the debate is just over who killed him: Clinton/Soros/Gates et al. or Trump/Musk et al.
The new Alligator Alcatraz facility is supposed to cost $450 million a year, and with 5000 detainees, that's about $250 per day. Looking at the quality of the infrastructure, isn't that absurdly high?
That seems high but with recent inflation it makes sense. As far as I know, the primary drivers in Cali are two-fold.
First, staff. Being a correctional officers is a brutal profession with high turnover. You pretty much have to bribe people with ever increasing salaries to even get moderately qualified people to consider it. Calmatters blames the unions but the ACA (1) reports annual staff turnover rates of ~50%. If half your staff are walking out the door every year because the job is just that brutal...the union has a pretty easy job.
Second, healthcare is ridiculous and constantly increasing. CCHCS' budget is ~$4 billion, or $40k/inmate at 100k inmates (2). I would be surprised if this is a driver for illegal aliens in Florida but it definitely drives the Cali numbers. Think about how expensive your health care is, now imagine the healthcare costs of a 48 year-old inmate who's been abusing his body for 30 years.
And those numbers kinda add up. $250*365=~$91k/inmate and the ~$132k/inmate from Cali, if we subtracted that $40k, is $132k-$40k=$92k. Almost exact.
I doubt that tells us much. Leaving aside the fact that there is certainly some spin going on there, some of that is the cost of health care, and the average prison inmate is vastly less healthy than the average likely detainee. Moreover, "prison guard" is a much more onerous and dangerous job than guard at a detention camp, and the need for security is far less.
Is this the result of activism toward better conditions, or more services, that prisoners are getting the equivalent of a top salary? Or is it civil service salaries?
I made a web app that scans ingredient lists (including photos of them) for concerning ingredients from a variety of sources, including Scott's Obscure Pregnancy Interventions. While it's primarily designed for skincare and topical products with kid-focused branding, it's useful for anyone who cares about this kind of thing, not just parents or those who are pregnant. It's free if you want to try it. parentscan.net
This is a good idea. Will be passing it on to my daughter, who is thinking of having kids soon. Do you have an opinion of Consumer Lab? It’s a site that tests supplements for contaminants and also to check whether the user is actually getting the dose noted on the label. CN they be trusted?
Subscriber for years, *HIGHLY* recommend. Something that's rarely mentioned is that most of the informational articles included with the various product testing are like Scott's Everything You Wanted to Know About posts, except sometimes even *longer,* with even *more* links to studies.
If anything, that kind of consolidated research on stuff like Vitamin D may be every bit as useful as the actual testing.
As Eledex notes below, Consumer Lab can only test a small fraction of products on the market, so really obscure supplements may not have crossed their desk yet, but for most people, I think CL is probably going to be sufficient, especially if you want testing on food products, too, like canned fish or seaweed snacks. Some of their reports are so eyebrow raising that I've made a resolution not to use any product which hasn't been tested by CL, and have imposed the same on my family.
Last, a nice side benefit of CL is that it removes decision fatigue. Their Top Pick for a given supplement is whatever least expensive product completely passes their their testing. If you *want* to spend more money on a particular brand, they certainly won't stop you, but if a Costco Kirkland vitamin passes all testing and is $0.03 a tablet, that'll be their Top Pick over a boutique-looking fancy name brand which also passes testing but is $0.15 a tablet.
Consumer Labs does a good job testing stuff (i.e. their results are accurate) but they test only a tiny fraction of what is on the market and a significant portion of what they test is paid placements. So their data can be trusted, but it's not comprehensive.
"Paid placements" is a bit misleading. Consumer Lab is mostly supported by subscribers, but some companies may pay for a quality certification *if* they pass Consumer Lab's *standard* testing.
While I personally would prefer for Consumer Lab to be 100% subscriber supported without absolutely zero contact with any company, I nevertheless think their process is reasonably transparent and trustworthy.
If there will soon be millions of robot workers, then it means the cost of skilled physical labor will sharply drop, and many projects that were unaffordable will become affordable. Does this mean the old museum ships like the USS Missouri will be modernized and reactivated? I think it would be awesome to see it plying the waves with a cyborg captain and a robot crew while it blasts the enemy using lasers and railguns alongside the old 16 inch guns.
that doesn't mean resources will scale up to meet that demand, or even if we have the resources to even make robot laborers at scale. Hardware is expensive and a lot pf SF authors tend to wish their technologies into existence.
<If there will soon be millions of robot workers, then it means the cost of skilled physical labor will sharply drop, and many projects that were unaffordable will become affordable.
On the other hand, all those skilled laborers who've had their income drop or been laid off are going to be able to afford way fewer things.
If and when there are "millions of robot workers", it means that the cost of each of those workers has fallen below the price of an equivalent workforce of humans. The relative cost of a human worker vs. a machine will only increase as the degree of automation in an economy increases.
If the current wars of Russia vs Ukraine and Israel vs Iran have shown anything, then that range is king, meaning missiles and drones doing most of the damage. I don't know what the state of the art in railguns or their use case is, but lasers will ever only be practical for point defense against missiles, drones, small boats etc., while large guns as naval main armament have been obsolete since WW2. So no, you won't see WW2 battleships in service.
The USN has (at least largely) shelved their railgun efforts because (AIUI) they're not effective past ~400 Nm and therefore the space aboard ship they require would be better used to house more missiles.
The sticking point there is "if". *IF* within five to ten years be millions of robot workers who can do jobs with the same dexterity and agility as a human, then we can do all kinds of things.
If this doesn't happen like that, then we can't.
And even if we get robot labour, I regret to say I don't think we'll get your dream ship.
Angry unemployed skilled laborers will blow up the fucker. (At least those who refused to get soaked in resin and repurposed as paving stones for the Walkway of Glory leading to the modernized USS Missouri.)
Can someone explain to me why everyone started calling missiles "drones"?
I'm very confused.
I keep hearing that "drones" are this groundbreaking revolutionary weapon. Isn't it amazing, a "drone" flies straight to a tank and explodes! That's pretty old technology, it's called a guided missile. How come everyone treats it as revolutionary and calls it "drones"?
How come American cruise missiles are called "missiles" but Iranian cruise missiles are called "drones"? I hear all the time about the latest Iranian volley of "missiles and drones", like it's a distinction that matters and you're not allowed to just say missiles.
Is it because a drone can land? Can all drones land?
But how can this definition of "drone" be compatible with their reputation for being revolutionary? People say drones are revolutionary because they're "cheap" and "expendable", what does the ability to land have to do with being cheap and expendable???
Instead of missiles that look like helicopters, why can't armies simply use missiles that look like missiles (thus not "drones" by that definition), and are as cheap and expendable as "drones"?
A rocket is cheap, and goes faster than a mini-helicopter. If the elements that grant a "drone" remote control/AI, vision and steering exist in a cheap and expendable mini-helicopter, why not in a cheap and expendable rocket?
If I had to take a stab, the inchoate definition people are gesturing towards when they say "drone" is something like "Something that's both reusable and expendable simultaneously".
Or, "A plane but you can afford to throw it away on suicide missions, because it doesn't have a pilot inside. But also, a missile you can reuse if it survives the suicide mission, because the target isn't there or something." The key thing that makes a drone a drone, is that it has functionality that existing things do not: you can send a recon drone on suicide missions in a way you couldn't send a recon plane, and you can reuse an anti-tank FPV drone in a way you couldn't with an anti-tank guided missile. It's more cost efficient in both directions: if it survives, that's good, you can reuse it. If it doesn't, well anything else would have cost even more to sacrifice, most likely.
If I got a second guess though, the lay public obsesses over the word "drone" because they're trying to say, "OH MY GOD, consumer hardware can beat tanks? Can beat actual military stuff?" In the same way the public obsesses over small arms, particularly the types it has access to (rifles, pistols, shotguns) and has interacted with personally, rather than crew-served weapons like machineguns, mortars, rocket launchers, recoilless rifles, etc. they can't own and have never touched (precisely because they're more effective than just small arms)... the public is now obsessing over the idea "I could kill a tank!?!?" when it says "DRONE DRONE DRONE DRONE DRONE".
i.e. When the Daily Mail or New York Post want an attention grabbing headline, they don't say, "This thing you've never heard of, CAN KILL A TANK", with a photo of artillery shells scattering landmines. They say, "YOUR CHILD'S CHRISTMAS PRESENT COULD KILL A TANK???", with a photo of something cheap from Alibaba. Nevermind that reality doesn't quite live up to the headlines, as https://warontherocks.com/2025/06/i-fought-in-ukraine-and-heres-why-fpv-drones-kind-of-suck/ reminds us; the public latches on to the first thing it understands, and doesn't let go until something else catches their attention.
(And indeed, there *are* interesting implications to consumer hardware being dual use. For a start, there has never been a shortage of drones for the war in Ukraine, compared to things like artillery shells and MICLIC mine clearers, because even the most insane FPV drone demand only bumps up the global demand for consumer drones by like 10%, and the factories in China can adjust. But when war breaks out and the demand for artillery shells goes up something like 100x what it is in peacetime, the existing artillery shell factories can't adjust. The public is gesturing to something real here! They're just awful at conveying what it really is, because they don't quite understand themselves... but they're very loud & insistent nonetheless.)
That piece you linked to is interesting, since it goes against what we usually hear about those Ukraine war suicide drones.
But as someone who knows hardly anything about tactics there are a couple things about that piece that leave me puzzled. It says that the most common use of a FPV drone is to finish off something that has been already hit, often by a mortar. But if a mortar hits a target, and you need to finish it off, why on earth would you not simply shoot with the mortar another time? Why waste a drone? Why are they doing this?
The other question is: if FPV drones are so ineffective, why does everyone say that the Ukraine war has led to a new radically different era of warfare, in which for example tanks act in an extremely cautious way? Why do tanks act in an extremely cautious way if the suicide drones aren't such a threat? Is it because of the threat posed by the combination of spotter drones and artillery?
To a first approximation, mortars aren't accurate enough to *hit* anything; they do their damage by near misses. And while a near miss by a mortar shell can kill a man, and *might* wreck a truck, it's not going to do much to a tank or armored personnel carrier or whatnot. Except maybe shred the tires of that APC, or with a large mortar bomb even break or dislodge the tracks of a tank.
Which is fixable, and probably *should* be fixable by the tank's crew. But if the crew is three Russian conscripts hunkered down in an immobilized tank in hostile territory with mortar bombs exploding randomly around them, they're probably just going to stay hunkered down until there's a lull in the mortar barrage, then skedaddle back to friendly(ish) territory and say "our tank broke down, please send a big-ass tow truck to recover it for us".
Since the Russians do in fact have big-ass tank tow trucks(*), you probably want to make sure the tank is really, irreparably destroyed before they can do that. But since your mortars probably didn't disable *all* the enemy tanks, you can't just send a team of soldiers with demolition charges to do the job; they'll just be machine-gunned by the tanks you didn't disable.
Enter the FPV, which takes longer than a mortar to employ and so wasn't your go-to weapon when the problem was "TANK! Enemy Tank! Do Something Now!", but when you set it up properly it can fly into disputed territory and deliver a modest anti-tank warhead precisely to a vulnerable side of the tank.
Earlier in the war, the preferred solution was a reusable "bomber" drone that could e.g. drop a grenade through the hatch the crew left open when they ran away and then fly back for reuse. But Russian electronic warfare eventually reached the point where cheap consumer-grade drones might last three whole flights if you were lucky, which made the "reuse" part less valuable and the simple reliability of the FPV drone more appealing. And at the very start of the war, the preferred solution was the fabled "Ukrainian Tractor", because the Russians were such inept cowards that you could fairly easily get them *all* to run away and then, hey, free tank.
* As do the Ukrainians; this sort of thing is usually done by putting a tow rig in place of the turret of an older model tank that you were otherwise going to retire. https://en.wikipedia.org/wiki/Armoured_recovery_vehicle
(Also, I remember hearing that *bomber* drones, the kind that work like a plane and drop bombs before returning home to pick up more bombs, might actually be a bigger deal than FPV drones here. After all, a reusable drone can afford to invest in the cameras, sensors, radios, weatherproofing, etc. you need to address the challenges facing FPV drones as outlined in that War On The Rocks article. An FPV drone that's basically a missile, as you pointed out, can't afford that sort of stuff.
And of course, the new generation of fiber-optic drones inherently dodge any issues with radio reliability and jamming. And the *next* generation of drones, could always just use AI/autonomous guidance, to avoid all the difficulties of linking back to a human operator.)
As for your other question... the article doesn't go into much detail why. I would speculate that it's a matter of "scoot and shoot", or something like that, where a mortar team that fires must immediately start packing up and "running away" because counter-battery fire from the enemy artillery (or FPV drones or bomber drones or the like) is no doubt incoming. While they're busy running, an FPV drone can be sent in to finish the job.
>I keep hearing that "drones" are this groundbreaking revolutionary weapon. Isn't it amazing, a "drone" flies straight to a tank and explodes! That's pretty old technology, it's called a guided missile. How come everyone treats it as revolutionary and calls it "drones"?
Because most people are stupid. The number of takes I've heard since the war in Ukraine started that prompted the response "the 60s called and they want their theories on the future of warfare back" is not small.
At a very fundamental level, you are entirely correct. A suicide drone is a type of guided missile. End of story. But it's often different enough from a conventional guided missile that it might be worth thinking about somewhat differently. I was running tests on various types of missiles in Command: Modern Operations a year or so ago, and found that the Shahed was slow enough (~120 kts) that it was trivial for the ship to maneuver out of their seeker range, which is very much not the case for any of the proper missiles I used in that test. (Actually, it was so bad I had to deliberately maneuver the ship to stay in engagement range.) Also, the lower closing speed opens up a bunch of weapons options that don't work for normal high-subsonic missiles. So far as there is anything revolutionary, it's in the cost, but even that I think is often overstated to some extent.
>A rocket is cheap, and goes faster than a mini-helicopter. If the elements that grant a "drone" remote control/AI, vision and steering exist in a cheap and expendable mini-helicopter, why not in a cheap and expendable rocket?
Because the drones are free-riding on civilian hardware. A typical FPV drone is basically made of commercial parts with some piece of ordnance strapped to it. If you do it properly, you need to pay for new ordnance and a custom rocket (big enough rockets are not produced in large volumes) as well as custom hardware. And you'd probably make it something approaching Milspec, because there are morale and reputational penalties if a significant percentage of your stuff doesn't work out of the box. (This is a real issue with FPV drones in Ukraine.)
I do think we'll see more drone-ish weapons entering the arsenals of serious powers, because it is pretty useful for an infantry company to have the organic capability to hit things slightly over the horizon. But exactly what that is going to look like isn't quite clear yet.
Missiles are almost always rocket or jet propelled, supersonic or nearly so, and very difficult to shoot down. They almost always fly directly to a target known, identified, and located before launch, they do not loiter and report back to their owner, and they destroy their designated target and themselves ASAP.
Drones, have traditionally been none of those things. And these are not independent variables - e.g. if you want your system to loiter and provide you with intelligence and targeting data, you probably can't do that at Mach 2 (or even 0.85).
With the increasing diversity of drones and drone applications, the general rule that drones have *none* of the characteristics of missiles no longer holds true. Something like a Shahed is halfway to being a cruise missile. But only halfway. The difference between "exceedingly fast and very hard to shoot down" and "basically a Piper Cub", is huge and worth calling out in the name.
While there are a bunch of people saying differently, I've always thought the core difference between guided missiles and drones is that the central examples of missiles are ballistic rockets and the central examples of drones are propeller or rotary-wing aircraft.
Cruise missiles are between the two categories and are arguably closer to drones, since they definitionally rely more on aerodynamic lift than ballistics to fly and since they usually use air-breathing engines (turbofans, ramjets, scramjets, or pulse jets) rather than rockets at least for the middle phase of their trajectories. I'm guessing they're called missiles rather than drones for historical reasons.
It looks like the term "drone" in the context of aviation dates to the 1920s, applied to light aircraft configured to fly unpiloted under radio control so gun crews could practice shooting them down. Presumably, setting human pilots to fly the target planes from the cockpits would have been bad for morale and prone to unacceptable attrition rates. Later, drones got used as reconnaissance platforms and the like. Then, in 2001, the US started using high-end drones as weapons platforms. Not as weapons, but as platforms: Predator drones, which had been developed for reconnaissance, were armed with hellfire missiles that they could shoot at stuff, but unlike a missile the Predator itself would return to base if it could. Using expendable drones as actual weapons rather than as reusable weapons platforms came somewhat later.
As to why missiles are more expensive, I'm guessing several possible factors:
1. Rockets are much less energy efficient propulsion systems than propellers or copter blades, and ballistic trajectories are also much less energy efficient than aerodynamic ones. And for chemically fueled rockets or aircraft, rockets need to haul around their own oxidizer while air-breathing engines use oxygen that they find conveniently lying around in their flight path. Add it all up, and a rocket just plain needs to be bigger in order to have the same range as an expendable drone.
2. An aerodynamic vehicle is steerable almost for free, since they need to generate configurable amounts of lift in order to take off and fly in a straight line. Cheap unguided rockets are not: they tend to have no control surfaces except for fixed stabilizer fins, so you need to add actuators to the fins (or add new control surfaces) in order to make a rocket guidable.
3. The ultra-cheap quad copter drones are mass-produced civilian toys (or low-end tools) with after-market control software and explosive payloads added on. There are no mass produced civilian toy guided rockets that I know of, and unguided mass produced model rockets suffer from factor #2.
4. Armies that want to use missiles because they need or want the speed and acceleration advantages of a rocket over a quad copter are generally willing and able to pay a premium for a bigger and more capable missile and aren't trying to optimize for the cheapest viable guided munition. Ukraine is the exception because their supply of missiles is constrained by what NATO countries are willing to give or sell to them, so they're supplementing with drones for targets that don't need missiles.
5. Most established missile weapons systems are decades old and rely on bigger and clunkier hardware for their guidance systems. Newer versions that take advantage of recent decades' improvements in computer hardware are still in the pipeline.
6. For cruise missiles in particular, they're big and expensive because they're high-end weapons systems that are orders of magnitude more capable than the cheap drones in terms of speed, range, and payload. Jet engines cost more than small electric copter rotors, and 20-foot-long airframes carrying 1.5 ton payloads aren't cheap, either.
The main difference between drones and missiles is the former's ability to maneuver and perform more duties at once. Ceteris paribus, anything that can maneuver is more useful than anything than can't.
A drone can observe the battlefield, relay radio signals further downrange, hide between the bushes and lay in wait, then attack and explode, all in 10 minutes.
Missile can only fly at one target and attempt to kill it. No other functionalities.
Making things that go fast is expensive. Making it go a long range is expensive. Sensor and guidance packages that can withstand battlefield conditions are expensive. And as long as you're paying the cost for rockets, you might as well put enough explosives on it to make it worth your while, which makes it big and even more expensive.
The Ukrainian (and unfortunately also Russian) quadcopters are basically civilian drones with an olive-drab paint job and an IED. And increasingly more customization as the battlefield environment gets harsher for drones, but the core hardware is produced in vast quantities for the commercial market.
Which means, cheap because economies of scale. It also means not really rugged enough for military operations, particularly once the enemy realizes what you are up to and starts looking for ways to break your system. Back in late 2022, the average lifespan of a commercial drone in Ukraine was three flights, and I'm pretty sure it's come down from there.
Something rocket-powered and so blindingly fast, with a guidance system that doesn't use GPS or radio, and born under the sign of milspec, is much more likely to A: work at all, and B: reach its target, in a contested battlefield. And we could maybe bring the price down to something in the four-figure range, if we built those in the quantities that e.g. DJI builds quadcopters, but we don't.
Also, the warheads of a lot of drones used in Ukraine are repurposed old military munitions, e.g. RKG-3 grenades, that are already bought and paid for. The price for *anything* that contains a significant high explosive payload and is guaranteed not to blow up before you want it to, gets you into the high three figures right off the bat if you have to buy it new.
1. Missiles are typically autonomous after launch (guided or self-guided). Initially, drones required operators to monitor their course and targets; however, now AI is enabling autonomous attack drones, but even autonomous drones have more flexibility over post-launch target acquisition than missiles. Although many modern missiles do have smart guidance systems, the speed of missiles generally limits the correctability of their course post-launch.
2. Missiles can carry much larger payloads than most drones, but likewise, they're significantly more expensive.
I learned some new things about drones watching this video...
Good question. I think you'd need to ask the military contractors lurking on this list. Of course, the drones being built by Ukraine use mostly off-the-shelf parts and electronics. Unlike ATGMs they probably don't perform well (or at all) in inclement weather.
But I just checked — a Tomahawk cruise missile runs between $1.87 million and $2 million, depending on the version. But it's designed to carry 450 kg of HE, or a W80 nuclear warhead, which is 150x the weight that a Ukrainian Osa drone can carry. it has to be reliable in conditions where an Osa drone wouldn't be. And it probably has all sorts of nifty custom-designed avionics. And military procurement costs inflate the price somewhat. But even smaller Anti-tank missiles seem pretty expensive for the warhead weight they can deliver, and their range is pretty short.
A US Javelin antitank missile costs US$176,000 (yikes!). Range: 150 m (max?). Payload: 8.4 kg HE.
A French Akeron MP antitank missile costs €120,000 (also, yikes!). Range: 150 m - 5km. Paylaod: 2 kg of HE.
Meanwhile, Russia and Ukraine have largely withdrawn tanks from the front line because small FPVs, equipped with 1.5 kg shaped-charged warheads, can take them out more cost-effectively than expensive ATGMs. Ukraine lost some M1-Abrams to drones — so, even our most advanced tank, costing between US$10 million and US$20 million, can be taken out by a drone costing less than US$1,000.
You can't compare directly a Tomahawk to the Osa drone. Beside the much bigger payload, the former can get about 60 times farther than the latter, also the former is extremely hard to stop, whereas a quadcopter can be shot down easily.
The quadcopter type drones aren't such a catastrophe for tanks. They can be shot down easily. For example, this system costs only 1 and a half millions:
But it turns out that drones aren't that easy to shoot down. They fly under radar. They loiter unobtrusively. And the tank cages that have been improvised don't really do a very good job protecting tanks.
IMHO, they've completely changed the dynamics of warfare. A NYT article (Mar 3, 2025) stated that drones kill more soldiers and destroy more vehicles in Ukraine than all traditional weapons combined, according to Ukrainian commanders. At least along the lines of battle, drones have been more effective than infrantry, armor, artillery, missiles, and fighter jets COMBINED. 🤷♀️
The author served as a drone operator in Ukraine and says that those small kamikaze drones are overhyped. He says that the videos that people see online where those drones do amazing things represent the exception, not the rule. He says that those drones are extremely unreliable, and in most missions they were used to finish off something that had already been hit by artillery or by bomber drones, and in almost in all cases the job could have been performed better by artillery.
But I'm skeptical since it doesn't make sense to me that drones are sent to finish off targets already hit by artillery. Why would they do this, instead of another artillery shot?
Yes, of course we've all heard how effective and central drones are in Ukraine. Although I would like to split "drones" into quadcopter drones and loitering missile drones, they're two completely different items.
I'm just saying that a quadcopter is much easier to shoot down than a traditional missile is. Over time systems designed to shoot down quadcopters will improve and become cheaper, so quadcopters will seem less invincible, and tanks will seem less doomed.
Of course people will not stop obsessing about "drones", since these days everything is a "drone", especially as AI puts humans out of business, and whatever weapon replaces the quadcopter as the scariest thing, they will call it a "drone".
"drone" covers a bunch of different flavors of remote controlled / unmanned weapons
some drones act like fighter aircraft, others like bombers, others like artillery, and others like missiles
the "revolutionary" part is that this is a new category of weapon with fundamentally new and different constraints that we've barely explored
if you saw a plane over a WW1 battlefield you might say "well that's sort of like a scout on horseback, but higher up" or "that's sort of like artillery, but without the need for forward observers"
but you'd be missing the fact that this is just a whole new general-purpose technology -- it is not bound by being "like" anything that came before
Back when planes where invented it was obvious what was a plane and what was not. If we had asked a person back then to define a plane it would have been pretty clear cut. Now the concept of a "drone" is never defined and seemingly arbitrary.
What is a "drone"? What sets apart a Shahed from a Tomahawk? Why is one a "drone" and the other a "missile"? Several people have tried to answer in this comment thread and they give conflicting answers. You said "remote controlled / unmanned" but so is a TOW missile. Other said it's the ability to loiter, or the ability to send back information, but some cruise missiles have those features and we call them missiles, not drones. Someone said it's just the size ("a Tomahawk costs at least 1 million USD and is a pretty big target") but is that really what defines a drone? Being small?
Besides, I grew up with many different words for military aircraft, such as "fighter", "bomber", "spy plane", "attack helicopter", and the meaning of those words was self-evident or nearly so. Likewise we have many words for military ships, people don't say "a ship", people say "aircraft carrier", "cruiser", "submarine"... But now everything is just drones, drones, drones, drones.
We should ban the word "drone" so people come up with clearer categories.
I suggest "minicopter" for the quadcopters and hexacopters, "cruise missiles" or "loitering missiles" for the suicide planes, and for the other unmanned planes "unmanned planes", for example "unmanned spy plane".
> Besides, I grew up with many different words for military aircraft... we have many words for military ships
This is exactly my point.
Manned aircraft are a mature technology. We have many different words for the varieties of them. Ditto for manned ships.
Drones are a new technology. We do not yet know what the categories of drones are!
Looking for a vaguely similar but profoundly different existing technology and using that name instead will just mislead people. If you refer to Shaheds as "cruise missiles", you will confuse people much more than if you just call them "drones".
Concretely: If Iran launches 100 Shaheds at Israel, and you say "Iran launched 100 cruise missiles at Israel" people will (wrongly) think that Israel is going to have an extremely bad night. If you say "Iran launched 100 drones at Israel" people will (rightly) think nothing much will happen. So "drone" is a much more useful descriptor.
On one hand, you're right that "cruise missile" is misleading for the Shahed, because people will think of a Tomahawk, which is 100 times more expensive.
On the other hand, when people say "drone", people thing of something like the quadcopters widely used in Ukraine, which are 100 times cheaper than the Shahed.
Apparently one widely used term is "loitering munition" for the kind of item the Shahed is.
I don't know that "loitering" is a helpful way to describe a thing that is mostly used for a mission profile of "fly a pre-programmed route and then dive into a pre-determined target"
at some point we will come up with better names, but we're not there yet. "drone" (or maybe "kamikaze drone") is still pretty hard to beat for clarity. no one things Iran is launching quadcopter drones at Israel from 1,000 miles away; it's implicit that the thing is chonky enough to fly for a few hours.
The Shahed is also being used in Ukraine and in the context of the Ukraine war it's useful to have separate words for the small cheap quadcopters and the bigger more expensive kamikaze planes.
You do realize we send the spy planes in to suicide themselves, sometimes with very big and flammable results? Why pay money to get the plane back, when it's so cheap?
I'm not sure what you mean by "material disadvantage", but see for example the discussion below, where in the context of discussing the mass production of drone there is confusion regarding the scope of drone production in the Ukraine war (due to quadcopter drones being 100 times cheaper than "loitering munition" drones).
Maybe you should, at least for the moment, accept that the development of military drones of various sizes and capabilities are a rapidly evolving process that is partially shrouded in secrecy and encompasses a very diverse set of devices whose capabilities overlap with those of traditional systems, and so defy easy classification around the edges. While I like putting things into neatly labeled boxes as much as anyone, that's just pretty difficult right now, and I don't see how "banning" a useful word that might sometimes be used imprecisely by the general public would help clarify things at all.
If it helps, I would offer this description of "drone" that should fit most of the time as of today: A military drone is an unmanned, self-propelled, remote-controlled carrier of sensors, and often of weapons or other payload. It might have limited autonomy of movement and target selection. By design, it can be re-usable or suicidal. It can perform a wide range of battlefield missions such as supply running, medevac, mine/barbed wire laying and clearing, general reconnaissance, artillery spotting, and direct engagement of enemy land, sea, and air assets at short, medium, or long range. It is generally lower in capability and cheaper to produce than any non-drone system whose mission it can fulfill, making it more economical in situations where otherwise a specialized, high-capability system might have been used, and generally expendable.
No, a standard Tomahawk cruise missile cannot loiter in the way that a drone or loitering munition can.
However, some variants have limited re-targeting and maneuvering capabilities, which may resemble loitering in a loose sense—but this is not true loitering (e.g., hanging around for an hour waiting for the target to appear).
Block IV TLAM can and I didn't say it was exactly the same as a drone. They have long range so loiter time is potential large as well. If saying they loiter is wrong I invite you to correct wikipedia.
Considering that it has a range of about 900 nautical miles and burns a type of jet fuel, it may be able to loiter for a while, but looking at the pics of it, its wings are short and stubby — unless they can extend further out than the pic in this link shows? This suggests to me they'd have limited maneuverability and would have to operate at a high thrust to maintain their flight. Do you have more precise info on how they can manage to loiter and for how long?
Tomahawks and armed drones have very different use cases; they may have some overlap, but not too much. A Tomahawk costs at least 1 million USD and is a pretty big target, especially when it loiters; it's well worth a SAM or two to defend against. Drones are much cheaper, used for different targets, and require different defenses.
Then why is everyone worried about how to counter "drone swarms"? Why not missile swarms? What does the ability to send information back have to do with being cheap and expendable and a swarm?
They ARE worried about missile swarms too! And have been since the Second World War. Drones are the NEW thing to worry about.
I agree that the existence of autonomous drones muddies this distinction, but I suppose ultimately, the practical difference is what kind of defense you need against them, and if there are two kinds of missiles needing different countermeasures, calling one kind by a different name is sensible.
But why do most of the cheap suicide drones look like helicopters instead of rockets? I find it bizarre that when soldiers want to take out a vehicle, instead of launching a guided rocket at it like in the olden days, now they launch a guided helicopter at it. What's the advantage, compared to a rocket? Everyone says the helicopter (what they call a "drone") is cheaper, but how can a helicopter be cheaper than a rocket (all else being equal)?
If I google "drone swarm" and "missile swarm" (in brackets, to ensure that the whole phrase is searched for as a whole) the former gets 50 times more hits.
"As advanced cruise missiles approach their target, remote operators can use a camera in the nose of the missile to see what the missile sees. This gives them the option to manually guide the missile to its target or to abort the strike."
Which seems to contradict your explanation of what sets apart drones from missiles.
"Bizarre ... What's the advantage, compared to a rocket?"
Launching a drone is a lot more stealth than launching a missile/rocket. If you want to survive on a modern battlefield that is constantly observed by a lot of spying eyes and robots, you'd better avoid emitting too much smoke, flame and noise at/near your position. Nothing bizarre about not wanting to draw attention of the enemy to yourself.
The olden days had much fewer, if any, HD cameras floating around, so launching a projectile or a missile with a flame and a bang was much safer for the operator.
Go look at anti-tank FPV footage coming out of Ukraine. An ATGM can't sneak up along a treeline and then come at you from the rear. Or chase you into a tunnel. Or drop a grenade down the chimney of your bunker.
Missiles are one-shot machines with a fairly simple gameplan - fly into something and explode. Drones are more like mini air support - controller can get it to do all sorts of clever and nasty things because he's piloting it (either directly or indirectly via waypoints) and so has more options about where it goes and what it does.
But they're cheap and don't involve a pilot who dies when they get shot down (expensive to train, expensive to recruit, etc.), so you can do missions that would not be sensible for a manned aircraft or an expensive missile.
When they're missiles, it's called a "barrage" instead. Basically the same idea, but they're traditionally coördinated only at launch and not during flight.
As I said, yes, modern tech makes the distinction is fuzzy. The point is that there are two qualitatively different kinds of long range offensive weapons.
because drones ('helicopters') can loiter; return to base after dropping a payload; are human-guided and can use on-the-fly decision making by the controller; can be retasked to specific target or aimpoint; can be very, very precisely guided (i.e. stop when they see a defense; find a hole in it; then fly through that hole), are not as heavy as rocket engines (but are slower); can be easily adapted (i..e different warhead); can serve purpose as reconnaissance or repeater platforms; can land next to a road containing potential targets, shut their rotors off, and wait until a vehicle passes. All of these are obvious if you watched a few day's worth of footage from the Ukraine war.
>how can a helicopter be cheaper than a rocket
scale. The drone consumer market made construction at scale possible.
Technically you could have some of these in a rocket, but not all, and not for the price of a drone.
Shouldn’t predictive processing theory be having a larger effect on personal and clinical psychology? I feel like it’s still a kind of a secret that reality is not actually poring into us through the orifices in our heads…
Seems like it should be obvious to anyone that it's not pouring in directly through sense organs. We have examples before us all the time of how we need some sort of intermediate processes to turn what pours in through the head oritices into knowledge that we are observing a certain event. Babies, children and animals and animals provide a constant stream of examples.. We can observe in all 3 many occasions when it's clear that they experienced sensory input of a parttifular kind, but could not make sense of it the way an adult could have. I think most people grasp intuitively that there's some kind of digestive process that turns what pours into the orifices into information, and that some of it is learned.
CBT definitely finds its roots in the ancient understanding of the interpretability of reality and our particular framings of events. So we can say predictive processing was anticipated beginning more than 2000 years ago. The fleshing out of the scientific concept and our understanding of how our sense of reality is mediated by our senses, their processing, schemas, predictions and modelling takes it to another level doesn’t it? Like I don’t think CBT regularly teaches clients that they’re not *ever* directly experiencing “reality” the way they think they are.
I don’t think predictive processing is that different from any other recent theory of mind in terms of “directly experiencing “reality” the way they think they are”.
>I don’t think CBT regularly teaches clients that they’re not *ever* directly experiencing “reality” the way they think they are.
No, they don't. I doubt that that would be very helpful in most cases. It's a big leap. In my opinion though, if you pursue CBT seriously you will eventually figure it out for yourself. And once you figure that out, you will know what you are experiencing directly.
>The fleshing out of the scientific concept and our understanding of how our sense of reality is mediated by our senses, their processing, schemas, predictions and modelling takes it to another level doesn’t it?
It is probably helpful if you're using that information in a way to get you to believe the idea. You can just as easily use all that information to confound yourself though. The human mind has thousands of rabbit holes to get around things that are uncomfortable
You are basically asking about the meaning of life, which as the wide ranging answers you’ve gotten suggest, people LOVE talking about, because figuring out the answer to that question is literally what we were put on this earth to do; it’s the most satisfying pursuit there is. The thing is, though, no one else’s answers are going to be right for you; it’s the process that matters, not the outcome. If you are looking for a kind of sherpa through that terrain, you might try some Jungian psychoanalysis, or dipping into any of the wisdom traditions that resonate with you. Also, secretly, I’ve always thought that this would be my dream job (or hobby I guess; feels weird to monetize it) so if you want to let me practice my non-existent spiritual sherpa skills on you, feel free to message me…though honestly from the responses you’ve gotten here I bet you could pick literally anyone you felt some amount of affinity and respect for and go from there.
Honestly, I wanted the opinions of people who were (a) reasonably similar to me (that's why I picked Astral Codex Ten) and (b) weren't going to go nuts about the money thing, which sets the general band of my options--I could do something non-remunerative, but I can't refund the NIH.
I've gotten plenty of opinions, and plan to go through and thank everyone. I'm then going to step away from Substack for a week to mull all this over.
Thank you for your time. I don't know about wisdom traditions, but I know way more than a man of science should about Western occultism. Maybe picking up a little Jung isn't a bad idea. Thank you again!
> "I've given up on the novel idea; it would take too long to get good at writing, and I don't expect the current trend toward mostly-female literary audiences to reverse in my lifetime. Also, I've realized ChatGPT is a better writer than I am."
What the fuck are you even talking about.
"If only you'd remember before ever you sit down to write that you've been a reader long before you were ever a writer. You simply fix that fact in your mind, then sit very still and ask yourself, as a reader, what piece of writing in all the world [Anonymous Dude] would most want to read if he had his heart's choice. The next step is terrible, but so simple I can hardly believe it as I write it. You just sit down shamelessly and write the thing yourself." - J.D. Salinger, Seymour: An Introduction
If you don't take any pleasure whatsoever in reading your own writing then, sure, give up on writing (and publishing) a novel.
But if there is a piece of writing in all the world that you would want to read if you have your heart's choice *and that piece of writing doesn't exist yet,* then you'll have to write it yourself, for *your* pleasure, not status.
And FWIW, slowly, grindingly chipping away at "getting better" at writing actually is the only way to get better at writing enough to really please yourself.
Will there always be people out there who are better writers than you (more creative metaphors, more comprehensive scope and research, cleverer twists, etc)?
Sure, but if you're anything like me, simply being exposed to their work will make you a little better at producing yours - and your work is is going to be much closer to your heart's desire of what you want to *read* than anything they produce.
(This is really, really true for me. Every 2-5 years, I'll randomly consume some truly jaw-dropping masterpiece and go into a brief, depressive, despairing cycle of "I'll never be anywhere nearly as masterful with storytelling as *that* person," but then when my admiration and despair fade after a few weeks, I inevitably discover a noticeable upgrade in my language usage, editorial judgment, and frequency of striking ideas. I'll then either edit or produce something new that I really, REALLY fucking enjoy reading way more than the truly jaw-dropping masterpiece.)
Write your thing for you, the reader. If it pleases you, the reader enough, it will likely please other readers, too.
But even if it doesn't, at least you'll have *your* heart's desire to read.
(Also? Fuck ChatGPT. No, it absolutely isn't as good as I am at producing the writing *I* want to read, and it will never be able to produce our hearts' desire, either, because it has neither a heart, nor desire.)
I think you should hunt out a really good coach and hire them to help you navigate the decision process. (Coaches are sort of like psychotherapists, except that they are geared towards helping people rise to challenges rather than helping troubled people get past roadblocks.). And considering going to places that offer some novel experiences that might wake you up and open you up. Psilocybin retreat? Meditation retreat? Heavy duty outdoor adventure course such as NOLS?
Thing with coaches is it strikes me as a great way to separate idealistic rich fools from their money. (Kinda like investment advisors, which are a way to separate greedy rich fools from their money.) Of course some of them are honest, but I have no idea which ones.
Jeez, you sure are rejecting a lot of suggestions. Are you sure you're actually open to suggestions? As for my suggestion, it's really a meta-suggestion about how to decide what's right for you -- try hiring someone who's a professional helper. Pretty low stakes. And it's not like it's impossible to check the quality of coaches. Ask on here! There are probably some other forums you can ask on, too. There are probably sites where people write lengthy reviews of coaches (who unlike doctors do not put up barriers to being reviewed.). Some coaches have blogs or have written books, and you can get an idea from that of how they work. And you can just see one a coupla times and see how you like them.
> What should I do with the remainder of my life? Any ideas?
I'm about your age, and here's some things i've enjoyed doing since retiring:
1. Getting fitter than I've been in my life (okay, I was already pretty fit, but I was a lifter before, and got into triathlon training and Olympic distance races, so a different type of fitness). When you have more free time, you can *really* dedicate time to doing a good fitness program, and to cooking and eating better, and so on, and it's easy to get fitter than you've ever been in your life because you have so much more bandwidth and time.
2. Traveling - I've always enjoyed travel, but I spent a good couple of years living out of nice airbnb's and vacation rentals across 7-8 countries.
3. Starting a substack - I've really found it rewarding to clarify and articulate my thoughts, and for my book review posts, to read books more in depth, with more external research and callouts, and with a view towards the wider world, or to read them adversarially, and clearly pinpoint where and why I disagree. It's also spurred many fun conversations and exchanges, and I've been able to introduce people to new facts, arguments, and ideas multiple times by posting apropos links to some of my posts when it's relevant.
4. I like cars, so I've raced a little more, brought my fun car to more car shows, did a "supercar road trip" across half the country, and a couple of other things.
5. Not this particular time, but between two companies in earlier years, I spent 6 months RV-ing across the US Southwest, Montana, and PNW, visiting various state and national parks, taking my tailgate motorcycle on various day trips and for groceries, and generally having an amazingly scenic naturey sort of time, and it was awesome. So peaceful, and if you spend the day mountain biking, hiking, climbing, or scrambling, it's the absolute height of luxury to come back to a queen sized bed, real toilet, and real kitchen instead of a tent.
6. I go to a lot more concerts - if you're in a major metro area, you'd be surprised at how many good acts come to town, I can usually find something every 1-2 weeks. I use a site called songkick for this.
7. You can read more, cook more, get into woodworking - any hobby that you've dabbled in before, you can really go all out on, it's just a matter of finding one that you really intrinsically enjoy.
Not sure if any of those hit the mark for you, but they have for me.
I will say that even with all of that, I'm being tempted into doing another company, because the impact you can have on the world creating a company is so high, and so meaningful, in terms of impact on the world, your employees, and your customers, that it's really intrinsically satisfying, even though it's a lot of work.
So maybe that's something to consider too - what product or service do you wish existed in the world, but doesn't? You can directly bring that thing into being, and if you do it right, you can do it using OPM - other people's money. You don't have to use your own funds, and you shouldn't, there's well established institutions set up to fund new ideas and companies.
All fun ideas. Fitness I should do just on general principle, it's healthy. Cars, travel, you know, I didn't grow up with them, though they're certainly fun for people who enjoy them.
The Substack is a fun idea, though I'm not sure the brand I've got on this one is what I'd want. But it's definitely a thought!
Starting companies really requires a very different personality with much higher optimism and tolerance for risk, not to mention extroversion. (Yes, I've seen the Big Five scores of billionaires paper, I'd also need higher conscientiousness and lower neuroticism.) I applied to a few investment banks back in college (I was young and stupid) and after landing on my face in the interviews I wrote off the whole business field. Though I appreciate what you do, I just thought you should know that.
You're 45 with $3M and financial freedom, but you sound miserable. The problem isn't what hobby to pick up - it's that you've built your entire worldview around fear and resentment. You avoided marriage because of "divorce settlements," you're bitter about "feminism", and you won't even try writing because you've decided the audience is wrong somehow.
This is self-imposed prison thinking. You have the resources to do literally anything, but you've talked yourself out of human connection and creative expression based on internet grievances. Get therapy. Work on why you're so angry at half the population and why you see relationships as financial threats rather than human connections.
Your life sounds like its own punishment right now. Fix the internal stuff first, then figure out what you actually want to do with all that freedom.
I'm angry at the political movement, not the gender. I don't have that much in common with the other half of humanity and our interests are diametrically opposed in many cases, but that's not really their *fault*, and both can be pretty much chalked down to the exigencies of biology. The battle of the sexes is older than humanity and will likely continue after it in whatever inherits the earth after us (unless it's the paperclip AI after all.)
As for financial threats--I mean, I agree lots of people enjoy and find huge amounts of value from relationships. I think I have a bad combo of poor character judgment and high but difficult to replace income--basically, I'm a great target for the divorce scam. Just because I shouldn't get married doesn't mean NOBODY should.
Thanks for weighing in--critical responses are just as useful as supportive ones!
So I was doing some pithy joking below about not marrying Hot But Crazy, but real talk: Your assumption here that Bitches Be Golddiggin' isn't necessarily supported by the data about your single female ***PEERS***, who:
Which is to say, amongst your single female ***PEERS***, there are many who might want to protect their assets from you, to the degree of wanting ironclad prenups or to avoid the hassle of potential divorce by not getting married at all. Marriage is an institution which primarily exists to nurture children, so there's an extremely good argument that it's not worth the risk of getting married for a childless union.
You may have noticed my emphasis on the word "***PEERS***.
That's the critical word here.
Your single female peers are the women +/- seven years of your 45 years, the baby GenX and elder Millennials aged 52 to 38. These are women who began working in an okay-ish economy and have had enough decades to build the kind of careers, personal wealth, and experience which would prevent them from coveting yours, and, as a bonus, who no longer have the issue of future children looming on the horizon.
Single female 31 year olds and especially single female 23 year olds are not your peers. You shouldn't even be considering anyone young enough for you to have conceivably fathered (or grandfathered!!!) as real "adults." Getting into a partnership with one of them - particularly a debt-laden poor one - is *adoption,* not marriage, and when you adopt anything, pet or child, you have the obligation to pay to support it.
Good thing you don't need to adopt!
Perhaps you're horrible in person, but on paper, you are a goddamned unicorn to your single female ***PEERS.*** If you give dating apps or a matchmaker a parameter of "single, child-free career women 38-52," you will be absolutely inundated by wealthy-ish to legitimately wealthy single women who roughly want the same kind of life you do. It's not that hard; the odds are very, very, very much in your favor.
But that?
That feeling right there?
That feeling of resistance upon being asked to seriously picture yourself partnering with a 38-52 year old woman, because you're used to seeing 19-24 year olds in your porn?
That is most of the reason why the "divorce scam" you're so afraid of even exists.
Your assumption I was looking for 19-24 year olds is quite well statistically supported given lots of evidence about the average male preference...but I don't watch porn (believe it or not)....some of us have good imaginations. If one expressed interest in me I would assume it was exactly as you claim!
If I were listening to the pronatalists (which if you read my responses you see I am not) I'd probably be going early 30s...after getting into shape so they assume I won't drop dead in 10 years. And even then it would be an extremely uphill struggle. I am not under any illusions as to my own attractiveness!
FWIW when I was on OKCupid (and, uh, Feeld) I never matched more than 5-6 years below my age (and in fact went older on multiple occasions), and never found someone where the money thing wasn't a concern.
(Sidebar: I'm occasionally amused when people make sweeping universal observations like, "there are no women on this dating app!" or "women get hit on in public by men all the time!" The large population of all the *other* women on that dating app and the women who don't get hit on in public aren't even considered "women" because they aren't sufficiently aesthetic to be fundamentally *visible.*)
Anyway.
I'll offer a apology for assuming you were seeking partners 10+ years younger, but I'm still giving a side-eye to the idea that *everyone* is a threat to your wealth. When was the last time you were looking? Five years can make a big difference...
(Sidebar: Yeah, I had heard something similar and was taking it into account; I view it as the other side of 'all the monogamous guys are short' (an actual line in a New York magazine article). Basically, people who don't meet the speaker's frequently-unrealistic standards might as well not exist. )
>You avoided marriage because of "divorce settlements,"
That is entirely reasonable. The odds are very roughly 50:50.
>You have the resources to do literally anything,
False. In fact, simply retiring at this point with his finances is possible, but somewhat risky (as other commenters have noted), let alone any grander choices.
"The odds" are not your personal odds. Your personal odds are substantially better than 50-50 if you do some simple things, like:
1. Don't get married aged <25
2. Don't marry someone you've dated for less than (say) 3 years
3. Don't be poor, don't be uneducated, don't be a drug addict, don't be violent or prone to angry outbursts, don't be a criminal, don't be a jerk, and make sure your spouse isn't those things either
4. Don't marry someone you only kinda-sorta love or someone who only kinda-sorta loves you
and a bunch of other common-sense things.
That still doesn't lower your risk to zero. But... is divorce the end of the world? Seems to me that you just wind up back where you started, being single. Maybe a bit poorer.
Also what makes them hot is something in us that's missing, and desperately yearns for something that person is playing up to. It might be an act, it might be they were trained, accidentally or deliberately, by a parent or just the world. But chances are the person's hotness is a dark reflection of our undeveloped stuff.
Develop that stuff, and yeah, you'll find fewer people desperately hot, but the whole world is a lot more peaceful and enjoyable.
Whatever is making it really, really, *really* hot is almost certainly also what's making it really, really, *really* crazy, so think about the other 23 hours and 45 minutes a day you're going to be spending with Crazy.
FWIW, for my first wife, we actually met all of those conditions save that I was 24 at the time, but nonetheless she divorced me 5 years later (she ultimately said she wanted kids (she never did get them) and I didn't).
>Maybe a bit poorer.
Or a lot poorer. Prenups are a thing, but I don't know how much they help. In Anonymous Dude's case, he has the savings from half his working life to protect.
In the grand scheme of things it’s not that big of a deal. Better than spending your life shooting your self in the foot because, you know, …every one else sucks eggs.
>In the grand scheme of things it’s not that big of a deal.
Many Thanks! The potential to lose half one's savings might not be that big a deal in the grand scheme of things, but we live our lives on a human scale, not a grand scale. And losing half one's savings _is_ a big deal on the scale of one's life. Very early in one's career one can recover. Halfway or more through one's career - not so much.
As I wrote, prenups are a thing, but I don't know how much they help.
No, I think it's a good comment, in the true-and-necessary category.
While I empathise with the OP's general point of view, I do agree that he should probably work on his resentment if he wants to enjoy the rest of life rather than waste it being grumpy.
One small attitude adjustment that helped me become less grumpy was to realise that yes, the world is imperfect, I can't possibly fix everyone else's flaws, but perhaps the best thing I can do is to set a good example. Instead of focusing on what everyone else is doing wrong, let me just be the best darn me that I can be, and maybe I'll be a good influence on someone else.
Anybody (including you, I see) who responds to a comment like this with some version of "Yep, these are all totally normal feelings, have you considered travel?" is doing this man a disservice. The fact that this guy doesn't know what to do with his life isn't the problem, it's downstream of the problem.
OP, you need to get over your hangups and resentments. Seek out some help in doing so, and get to work. Then the rest will take care of itself.
I recommend moving to a European country that's well-insulated from wokeness. Switzerland maybe? Slovenia? American culture is sort of over, at least for men of our generation. It's only going to get worse over the next 20 years. Move someplace you can enjoy participating in society. Spend the next couple years getting fluent in French or German or whatever and then retire and leave.
Start your own Substack that's anti-feminist but not meathead. The fact that you don't really fit into the existing ideological taxonomy means that you can create your own niche. There's a big appetite for outsider voices right now. See if you can build it into something.
I'm really, really hesitant about the whole politics thing now. I don't really see a comfortable 'tribe' to join--I'm too cynical and market-skeptical for Grey, too identity politics-skeptical for Blue, and just don't have the rural experience for Red.
I have thought about languages and exile, but I'm not sure which one to pick--Hungarian's hard as hell to learn. I do think you are right about this country going downhill.
You have enough money for early retirement, so you can just quit. It seems you don't know yourself very well if you're asking a question like "what will I do for the rest of my life?". Time to follow up on any passing fancy you ever had, try things you haven't tried before, until you find your favorite thing in the world, then you can just do that.
But if his favourite thing in the world turns out to be expensive then he'll wish he'd hung around for a few more years to get a lil' bit richer. At $3M you can easily retire if you love kayaking and reading, but if it turns out your favourite things are race cars then you might have problems.
OK. I suppose first we would have to define what we mean by a race car. Out of curiosity I did some googling and the whole Reddit thread about pretty much what we’re talking about here (middle-age guys getting into car racing for fun.) seems like the cost of getting a machine ranges from about $8000 up to $100,000 depending on what kind of racing you want to get into. Obviously, you could spend a lot more if you wanna go to Maserati‘s Ferrari‘s Lamborghini‘s McLaren’s etc.
it’s all kind of peripheral to the point, though, which is that the poor fellow doesn’t know what to do with himself.
1. Set a time limit for staying at the current well-paying woke job you hate. (4 years, for example.) Maintain the personal commitment to stick to that time limit.
2. Quickly come up with a plan for what you will do when you quit. Example: a one year sabbatical in a cabin in upstate Maine, followed by establishing a career as a freelance consultant/paladin in whatever field you have expertise in from your current job, but working for the other side.
3. Use the time before the limit runs out to solidify this plan or come up with a better one. My guess is that you will want to be self-employed as a service provider with a very small client list. Something like a private eye, but not a private eye.
Not a bad idea. I never quite got the outdoors thing, though I know a lot of people enjoy it. I'm not going to go into too much detail on what I do for obvious reasons, but an exit plan involving some of my prior skills is a good idea.
I'd be traveling the world if I had significant surplus funds. Or, were you one of the people who didn't understand the need to travel that some people have? (We may have had this discussion a few open threads back.) Anyway, seeing places and people you've never seen will give you a greater horizon of opportunities than sitting at home being bored and introspective (that's not to say you can't get bored and introspective traveling, but moving on to the next place will generally distract you out of it).
I prefer the formation that I know to that one. For some reason, it's just not as funny to me. I told it to my divorce lawyer a long time ago, and I was very impressed at how she managed to keep a straight face. I was also very surprised she had never heard it. It kind of reminds me of something Henny Youngman might've said. "Take my wife… Please!"
Some people attribute it to Johnny Carson, which I find easier to believe. It's kind of Borsch belt.
No, it’s not worth the trouble is it.? I asked ChatGPT and Rod Stewart is the number one candidate. It just doesn’t work for me if he said it. Let it lie. I’m sure someone said something a lot like it a long time ago.
Goodidea. Murder, however, is a rather dicey proposition.
Ukraine is currently in the midst of getting flattened by Russia.
(Okay, in all seriousness, it is becoming shrubberies, which, yes, are not mentioned in the NATO Combined Military Services Doctrinal Handbook -- our tanks are not doing well.).
What is it with the pronatalists? I know you guys want high-IQ people to breed but my sperm are carrying more mutations than the Ninja Turtle franchise at this point!
" (I really need to give up the book-buying; I did *not* need that Arkham House first edition)."
I was shrugging along until I hit this. Had I pearls, I'd be clutching them right now.
Need? NEED???? Sure, you don't *need* it, but you don't *need* a lot of things. Being a cow grazing in a field can be a happy enough life, but do you really want to live on the level of "time to graze, no other needs at present"?
You're going to make me invoke Maslow's goddamn Hierarchy and I hate that thing, but an Arkham House first edition is one of the levels above the base "graze in the field" level.
Are you human or not? Give up buying books, might as well give up breathing while you're at it!
I have, in my time, bought books rather than food so someone saying "yes I'm comfortably off though not rich rich, what should I do with my money (maybe I'll cut back on book-buying)" is stabbing me through the heart. Through the heart, Horace!
Is there anything that grieves you about the state of the world? Not politics, that'll just depress you. But does it bother you that science isn't progressing quickly enough? That architecture in your city isn't beautiful or functional? That people die of preventable disease? That we don't have some awesome technology you dreamed about as a kid? Or something much smaller would work too. Quit the job you hate, and find a way to work on it.
> But does it bother you that science isn't progressing quickly enough? That architecture in your city isn't beautiful or functional? That people die of preventable disease? That we don't have some awesome technology you dreamed about as a kid?
Political decisions are a significant part of the reason all those other things are as they are.
That's true, but they're also all problems that we can do something about, without needing to solve the political problems first.
Politics is a uniquely frustrating thing to spend your time and effort on. Despite your best efforts you'll probably never be anything other than one voice out of millions. Sometimes good things will happen and sometimes bad things will happen, and it will almost certainly turn out that none of the effort or brain power that you applied to the problem made any difference at all.
It's much better for your life satisfaction and mental health to focus on meaningful smaller-scale changes that you can actually make, rather than dreaming of the large scale changes that you'd love to make if only someone would let you. I can't change the course of global architecture, but I can build a pretty house for myself. And maybe that will inspire someone else.
Science? All y'all are working on that and much better knowledgeable and connected than me! Also with the way the internet and social media turned out I'm not sure I want more progress, to be honest!
Architecture? Yeah, but that would require a lot more dough than I have. The point is I've probably got enough to *live on*, but not enough to *do anything with*.
Preventable disease? I kind of feel like the EAs are on that one and there's not much I'm going to do. (That and I can't take anyone talking about shrimp welfare seriously enough to give money to.)
Re: preventable disease, instead of thinking about what percent of the world you can help, think about the individuals. You could afford to save the life of at least one kid for a pretty minimal amount of money. Think about a 10 year old kid who is about to go blind (or die!) from vitamin A deficiency. Maybe he is the only son in a family, who will rely on him as he gets older for income.
Yeah awesome, you can save that kid so that he can grow up and father three kids that will go blind. Don't invest in failure, that just leads to more failure.
Generic advice follows, but maybe you need some...
A steady stream of minor achievements or acheiving mini goals can be pleasurable. Have fun! In some sense, a sense of progress = fun.
If your work is actively making you unhappy, and you have achieved financial independence, consider optimizing your life around more meaningful work.
If you actively *hate* your work, you are probably going to be somewhat unhappy. To mangle a quote, happiness is when your thoughts, feelings, and actions are aligned.
Join a writers club near you! Take workshops along the way to understand structuring, techniques, and get practice. If you see a novel as the way to influence the world, then maybe recognize that and find better ways to do so. Why do you want to write? Fiction? Character exploration? If you *really* want so see a novel happen, you can lower your expectations and write a bad novel. And build from there.
I don't think I'd influence the world; I'm not that grandiose. (It's quite possible I'm not egotistical enough to be a great artist.) I used to look at all the sci-fi and fantasy paperbacks as a kid in Barnes and Noble and dream of seeing my name up there with a fun colorful cover, but now the market has shifted.
Have you ever played table top role playing games? There are professional game masters now. This is another avenue for story telling. Also meeting people.
I am one of your feminist female bugbears and I would love to read more new arkham house weird scifi with strange cover art written by childless men, no sarcasm. bring it on, i dare you
You might not actually be a bugbear depending on your precise views, but it's probably not worth getting into a political argument here ;)
Thanks for the encouragement though. Arkham House no longer exists...but if you mean fiction by weirdos...I don't know. You like the ones by childless men for some reason? I could see that, more adventure or something, or a consciously childfree woman might like to see the other side.
I think if I ever finished I'd at least pay for an artist! Sadly now that everyone's reading electronically there isn't much call for the great Michael Whelan or Rowena pieces anymore. It's another part of the world that's gone cold.
If you really think about it, this does not have much to do with your urge to write a book. It says a lot about the outcome you would want if you were to write a book, but don’t you have some personal inner reason to want to write a book which has nothing to do with the market? If not, I can only think of two reasons. You need to make money writing a book or you’re letting your grandiosity get in the way of doing something that is personally significant to you.
I think you're making excuses for abandoning the author dream. Yes, there's far more adversity to overcome, and if you had to go through the traditional gate-keepers, it's probably harder for you to get published than for anyone else since the invention of the printing press, but self-publishing is a perfectly workable alternative. If you're not relying on it for income, and just want to have your work read by others, the usual arguments against it don't seem to apply.
EDIT: I originally said "anyone else in history," but I don't think that's quite true.
It seems it can push vulnerable people over the edge. This isn’t surprising. Imagine an evil psychologist who wanted to make his or her clients psychotic - it seems likely this would meet with *some* success. AI is great at agreeing with people, which can confirm their delusions.
I’m interested in how many of the cases mentioned in several recent articles were specifically ChatGPT during that one week of extreme sycophancy in late April, as opposed to other LLM assistants. That is, I’m interested in whether that particular model was extremely effective at bringing out psychotic ideas, while others have been better safety trained to avoid that.
yeah, give a potential psychosis a machine that is both socially considered wise and has a knack for always finding a way to agree with you and youll finally know what mass scaled SaaS is - Schizophrenia as a Service
I read that article when Tyler Cowen recommended it. I enjoyed it more than most long form articles I’ve read in a long, long time. Thank you for writing it.
What do you think the prospects of accurately and fairly estimating land value in urban areas actually are? This is the part I despair of, particularly distinguishing autogenous value increase and value increased by ancillary services created by other actors.
I also worry that the incentive to vertically integrate would be enormous, to internalize more of the autogenous value increase.
I enjoyed Lars’s article, but I’m not sure he dealt with the problem of autogenous increases in land value.
If I build a 50 story office building, and your development corp builds 500 units of mixed use retail/housing around it, I just don’t see how you separate my office building’s gravitational pull from the amenity value of of all the mixed-use nearby. If we don’t want the landlord a black away from the retail to get a windfall, I don’t see even theoretically how we fairly assign the increased land value to all three actors.
Lars’s article seemed to basically assume land value is a continuous function of such a large number of actors that you can basically just accept that my entity’s effect on the value of my own land is minimal, or if it’s large, I’m probably using the land so intensively that the autogenous increases is a very small drag on my improvement incentives.
Anyway, thanks for other links, looking forward to reading them.
What to you makes the autogenuous value increase a problem. Is it the incentive, you are getting taxed on value you create which is the problem lvt is trying to avoid? A secondary effect is still much better than a primary effect (income, property tax etc).
I live in a city where transactions on empty land is common so measuring land value autogenuous or otherwise doesnt seem difficult. Seems like with better land use, which lvt is meant to incentivize, empty land transactions will be more common as land regularly gets repurposed, so itll get easier to measure overtime.
I understand that failures are a normal part of a rocket testing process, and SpaceX "iterative development" philosophy in particular is bound to have a lot of them. I am not implying neither that the Starship is "doomed", nor that it "should" progress much faster. But it looks to me like there are two very different trend lines here.
From the beginning of Starship high-altitude testing and then for integrated flight tests 1 through 6 we see steady improvements, where basically each flight goes a bit better than the previous one. Then for flight tests from 7 onwards we switch to a different mode, where it looks like each flight only discovers additional problems (flights 7-9 all had various problems with the ship, and the ship for flight 10 blew up during the static fire).
I notice that I am confused here. It looks like there is some latent variable which changed between flights 6 and 7, switching them from one mode to another. Does anyone have reasonable ideas for what this could be?
Some hypotheses (not necessarily plausible, just things that *would* explain the change if they somehow turned out to be true):
* Starship block 2 problems. Basically, there were some serious design problems with block 2, leading to it having not one but at least 3 different failure causes. If this is the case, why was their design *so* bad, much worse than block 1?
* Musk problems. Maybe for some reason Musk's supervision and decision making are crucial to SpaceX success. Then between November 2024 and January 2025 he became much more busy with the government, didn't have enough time for SpaceX, and the problems followed.
* China intervention. Maybe the CCP decided that SpaceX was progressing a little too smoothly a little too fast and gave an order to sabotage them somehow? (Again, I don't have any model for this, I really have no idea whether this is even possible, let alone plausible)
* The gods are angry with them. Maybe the coin really happened to first land on heads 6 times, and then land on tales 4 times after that.
As others have pointed out, it's the difference between Starship Block 1 and Starship Block 2.
As for why they don't just go back to the version that sort of worked, Block 1 worked tolerably well but was heavy enough that it basically couldn't carry any useful payload to orbit; it was barely capable of lofting its empty self into the lowest practical parking orbit. My model gives it 10,176 kg of payload to a 185 km orbit, a bit over half of what SpaceX's workhorse Falcon 9 can do and not even close to being worth the bother of a Starship.
Block 2 might be able to deliver a hundred metric tons to Low Earth Orbit, and that's definitely worth doing. Block 1 is only valuable as a proof of concept for the basic system; SpaceX absolutely has to get Block 2 (or Block 3) working or the whole thing is a monumental waste.
Block 2, the vehicle that matters, is currently 0 for 3. But the Falcon 1, SpaceX's first rocket, crashed and burned on its first three launches and made it to orbit on the fourth try. That's SpaceX's style. Fingers crossed that it will keep working.
This pattern seems par for the course in SpaceX development and even in Starship development IMO. If you go back to the high altitude hop testing you mentioned, SN8-SN11 all exploded before SN15 kind of worked. Same thing with F1 dev and F9 landing - a string of failures where not much progress seems to be happening, people start getting worried, and then all of a sudden they whack enough issues and the whole thing starts working pretty well. If ships are still blowing up in 6mo to a year I'll start getting concerned.
Starship Block 1 was robust, but not as performant as SpaceX would like it to be.
They are pushing the envelope with Block 2, trying to get 25 per cent more payload capacity from it. That is a huge capability leap, and the envelope is not very generous.
* Rockets are stupid hard. Starship is the biggest and eventually most powerful rocket ever built, with by far the biggest payload. It is also intended to be fully reusable and hence extremely economical. These are competing design goals that require stretching envelopes far, far beyond what was ever previously thought possible. Failures during development are virtually guaranteed, and its better to fail fast and learn fast.
Sol has it broadly right. The booster part has gone pretty well and the issues that they had were ironed out in the first few flights. That makes logical sense - SpaceX is a rocket company, and though the booster is bigger and more complex than any they had flown before, they likely had a reasonable idea of how to get it to work.
The Starship upper stage is a different kind of problem. There have actually been numerous problems with it since the beginning, but they are often not easy to spot and SpaceX has tended to downplay them from the beginning. Attitude control in orbit seems to be a persistent problem. Reentry itself is a massive challenge, and - even though they did manage a few controlled landings - they are clearly still far from solving it fully. And notably, they are still doing only the simplest suborbital test.
So I think a reasonable take is that the first few flights showed good progress with the booster, which allowed casual observers to overlook the issues with the upper stage. Since the booster is now looking good, people are paying more attention to the harder part. And simultaneously, SpaceX is trying to redesign it to fix the existing issues, but are then running into other (very visible) problems.
"Block 1" was the first iteration of starship that was used for the first 6 flight tests. "Block 2" which is slightly larger with significantly more ambitious internals, has been used from flight 7 onward.
I think it's too much of a coincidence for the first 6 flights to go incrementally better each time, then the next 3 launches to all have unique mission-ending problems with the updated starship design. It looks like they got the booster right judging by its incredible success at both catching, and reusing it, but something is definitely wrong in Block 2.
I wouldn't update too much on it. They are extremely ambitious with every new development, and although SpaceX innovates quickly, at launch you're looking at rockets that were designed at least a year earlier. S35, the Starship that launched most recently in May began construction back in October 2024, about 3 months before the first Block 2 Starship, S33 Failed. Thus, very few lessons from S33 could have been used for S35, if any at all. And certainly no lessons were learned from after RUD, which for flight 7 and 8 was before the problems seen in flight 9.
It's of course a bad sign that they have had multiple failures in a row, as any test failure is. However, if you look at SpaceX as designing a Starship manufacturing system as much as Starship itself, then the loss of any individual rocket isn't actually that big of a deal. Combine that with infinite money for all practical purposes, and while each failure sets the mission back, it doesn't move it meaningfully towards cancelation. As is, they have the largest rocket booster on the planet that they've successfully reused (this is truly an insane accomplishment), the highest output rocket engine production facility on the planet (with so far high demonstrated reliability, at least on the booster), and a factory capable of producing an accelerating number of Starships and boosters. They are also building literally a second version of the entire setup they have in Texas, in Florida, which will add some serious redundancy, an internal spirit of competition, and allow for significantly faster launch cadence.
What I read (and maybe someone more into the subject can comment whether this makes sense) is that the first couple of flights revealed that the engines could not deliver the expected thrust, and hence, deliver the expected payload to orbit. So the rocket's design was revised with an emphasis on saving weight. Designing a rocket always means operating on the very edge of what's possible in terms of materials science, and trying to shave off a few percent of weight may have pushed "Starship" from "yeah, with a little luck" into "nope" territory.
While you're right the new design, Block 2, was largely designed with weight savings in mind (anything saved on the reusable 2nd stage is a fraction of the gained payload due to the rocket equation), lack of engine thrust is not the primary driver of that.
The engines have actually over performed on thrust and SPI. They've adjusted the design to make it larger, with a higher target mass to orbit to accommodate for that increased thrust.
When I became a parent, two things happened (well more than two really but there's two I want to talk about now):
- I appreciated my own parents more
- I became familiar with dealing with the disgusting and unpleasant parts of daily life in order to support another human that I loved who couldn't take care of themselves.
Now my mother is physically disabled and my father has dementia, and I find I rely on these two benefits of having children in the care that I now help to provide for them.
Which makes me wonder: is there any research on the different attitudes to caring for elderly relatives among child-free people compared to people who are parents themselves?
I knew from a fairly young age I wouldn't have my own children and married a woman who biologically can't do so, which has worked out for both of us. A large part of my decision was the reality of my own childhood.
I don't think my parents should have had kids or even gotten together as a couple at all. My father could have made it work with a different woman I think, my mother was largely the problem. She was far too consumed by her own problems with mental health and addiction to do much parenting at all. She ditched my father for vague and changing reasons when I was 2 so it was just my sister and me at her mercy. Fortunately she was easy to avoid, assuming she was even around to need to avoid.
My wife and I will never get the chance to answer this question though. All of our parents died relatively young. My dad was the oldest to die at 65, her father was the youngest at 48. While I spent most of my childhood (And adult life) avoiding my mother as much as I could, my wife genuinely loved her mother and was devastated when she died at 52. Her reaction was both very touching and entirely alien to me. My own mother’s death was a moment of quiet relief for everyone that knew her. Simply being alive was often too much for her, and her suffering was finally over. I like to imagine she had some sort of positive influence on at least some of the people that knew her, but I struggle to think of anything.
I really don’t know what I would have done if any of them lived long enough to need care. I doubt I would have taken an active part in caring for my mother if she were elderly to be honest. Most well-adjusted people seemed to have been effortlessly bonded with the parents, especially their mothers. This wasn’t my experience. A large part of my decision to not have children stems from having no idea what a good, or even acceptable, childhood and child care even looks like. I’ve heard my friends talk about the joy of reliving childhood experiences through their own children. I have no conception of what this even means; I hated being a child, and I hated myself when I was a child. Things got better as I got a little older and realized what I could get away with as a teenager as long as I had good grades and adult authority figures slotted me into the “good kid” bucket. In retrospect I learned some very antisocial ways of getting through life then and very little to prepare me for being a adult net-contributor to society.
As I think about it more, I don’t think I would have. I’d probably give my sister money to help if she was willing to.
I don’t know about the relevant research, but my Dad has recently had a brutal descent into Alzheimer’s, and frankly, the fact that 2/3 of his kids don’t have kids seems like one of the handful of blessings that has helped keep our heads above water. My sister, with two young children, is drowning in caretaking; I can’t imagine what we’d be doing if all three of us were in that situation. I think ultimately the care required in these situations will expand to fill (and then almost capsize) whatever is available, but the wider familial benefits of a handful of childless aunts and uncles who can swing in for backup suddenly seem very visible to me.
Has your family considered putting him in a nursing home at all? I am very curious to see how the opinions of people about nursing homes does over time. Most people I talk to still have a visceral "no" reaction, but Alzheimer's is often the one exception.
He is now in memory care, but the process of getting him there was wild + excruciating and consumed our lives for more than a year (there were complicating factors.) And even now, with his safety and basic daily needs taken care of—he is in an excellent facility; collectively my family is better off in terms of resources than 99% of people in our situation—I think we all *still* feel there is so much more we could be doing. Which really does speak to that sense that, probably much as with kids, half the battle comes from accepting your limitations and knowing that you’re going to do it imperfectly. But man is it tough.
Very familiar with the difficulty of that placement - it similarly upended our lives for probably the same amount of time. There's a lot of prep work if you're doing it well. There's a picture somewhere my dad took of all the paperwork he had to do, and if the all the pages were bound in a book, it'd probably be heavy enough to kill someone with.
My mother has been in her facility for 21 years and I still get that feeling all the time, even though she's at the point where there's just objectively so little I can do. My father/her husband would visit her 4-5 times a week, and her parents would cover the other days. They either didn't work or had non-traditional jobs with a lot of flexibility. Now that they're all dead and I'm working a traditional office job, not being able to keep up that routine really hurts. I guess it is good I am child free - it's already so easy for life to get in the way and prevent me from visiting.
Yes that's a really good point. My kids are older so less of an ongoing demand (and even able to help out in some regards). And I also recognise the idea of care needs expanding to fill availability.
I'd be very interested in any research like that. Anecdotally, I suspect there's a large difference.
I'm child free and have a disabled parent who lives in a care facility. I didn't put her there, but after living with her in both circumstances, I found it a lot easier to participate in the process when she was in a care facility.
People who've had children are absolutely more confident in handling the disabled, in my experience, but I do also sometimes think that's a bad thing.
Disabled people are not children or perfect innocent angels but seem to be commonly bucketed into one of those mental categories by the able bodied who are not familiar with disability in their personal life. I think some caregivers never update how they care for someone and default to whatever they learned as a parent taking care of their baby, and get frustrated with disabled adults being "difficult".
I think caregiving is going to be very important in the years to come and do worry that some child free people who have not challenged themselves in other areas of life will find themselves failing to help people they love because they never got over their "yuck" reflex or learned how to do boring, tedious, necessary things for the benefit of someone less capable.
Yes, that's what I'm wondering about, but I don't want to just assume it's the case. One thing that gives me pause - although admittedly going purely on stereotype and cliche - is that some east Asian countries with very low tfr are seen as having a more positive attitude towards elder care.
Interesting. I’m not sure I share either of those experiences.
> - I appreciated my own parents more
I wouldn’t say that becoming a parent has made me appreciate my parents *less* – both before and after I became a parent, I appreciated them a lot. but I do think becoming a parent has made me focus more on parenting choices, and realize areas where they made parenting choices that I think were incorrect. (At least, according to my value system.) As a result, while I still appreciate the effort they took, I now have a slightly more mixed view of their parenting in general.
> - I became familiar with dealing with the disgusting and unpleasant parts of daily life in order to support another human that I loved who couldn't take care of themselves.
as far as this goes, although I haven’t faced the issue of elder care yet, I anticipate that the largest difficulty there will be emotional rather than physical. For example, if I need to change a diaper on one of my parents, I don’t anticipate that the physically gross nature of the task will be much of a problem compared to the emotional aspect of them, being embarrassed by the assistance/feeling shame.
And that’s obviously not an issue I face when changing my baby’s diaper, so I’m not sure it has prepared me for elder care in that meaningful of a way.
This is closer to my wife's experience, where becoming a parent ... clarified her judgment with regard to some of her mother's actions.
For me I think it's hard to truly understand the emotional weight that parents bear until you've been through it. I don't think I was a difficult child but I didn't understand at the time some of the things I put them through, that seemed trivial to me.
There has been a really interesting discussion about school in the book reviews, so now I am curious. In m’y European country, not only are options severely constrained, people typically view teaching as something that doesn’t mix well with a parenting relationship, so you delegate it all.
So, if you have school-aged children, do you teach them reading / math / Anglo-Saxon poetry / oboe playing yourself (in addition, or instead of teachers, whether public or private) ?
How did it go ? Do you use AI ? How much time does it take ?
I tend to think that humans of any age don't want people they have close relationships with to cosplay the teacher role. That doesn't mean humans can't take direct instruction from such people, just that we don't want to be bossed around by them. We always planned to homeschool so we always went out of our way to not make learning a power struggle and instead approach academics like vegetables. We model lifelong learning and seek out resources that are worth our kids' time and have a daily routine that includes a sustainable amount of academics rather than cramming them into 180 days and don't make unnecessary demands. It's a blast and I feel so lucky to get to do this.
> So, if you have school-aged children, do you teach them
> reading / math / Anglo-Saxon poetry / oboe playing yourself
> (in addition, or instead of teachers, whether public or private) ?
I can't tell at whom this question is directed.
I live in the US rather than Europe.
My wife and I homeschooled our (not particularly academically gifted) child K-12. We outsourced some of the stuff we weren't good at (e.g. sports). We did teach him math, reading, literature (he read "Beowulf" and "Gawain and the Green Knight" in translation; some of the Canterbury Tales in the original Middle English), history and science ourselves.
Thanks, this is really cool ! I am just interested in a different perspective because, in my social circle IRL, people teach very few academic subjects to their children themselves*.
I read that a lot of the commenters here have a different perspective, so I really appreciate learning about it ! It’s fascinating to see that it seems totally doable to a lot of you, instead of completely alien (wonders of the Internet, I suppose !)
*They, and those children, then go on to have mostly successful lives as high-level civil servants/ engineers / bankers / etc, but apart from the occasional 28-years-old crisis reconversion of finacial auditord as painter / baker / bricklayer, this is not a very creative or independent society, as you can imagine.
It is common for homeschoolers to stop at high school (so the professionals handle grades 9 - 12).
For a bunch, it is also sorta considered on a year-to-year basis. "Yes, I think I can teach a single 1st grader what he/she needs to learn," a parent might say while thinking about 1st grade. No need to decide right now if they can teach 8th grade as well. Or even 2nd grade.
I do believe that America is an outlier about this, though.
I have children that are not yet ‘school age.’ I’ve taught them to read (using phonics) and they’re working on arithmetic. I intend to teach them all I can myself, and get tutors for things like music where I’m not very talented. I don’t intend to send them to school.
Teaching them to read took about 20-30m per day for about a year. It’s occasionally frustrating but the rewards greatly exceed it. I don’t have any idea what I would use an LLM for in teaching my kids (besides the usual stuff I use them for, like finding citations/sources and conducting starting research on subjects).
That sounds great ! (I asked the question about AI because everybody here seems to think it will revolutionise education, but I would have no idea how to use it either.)
If it’s not indiscret, does that mean that you stopped working to take care of your children ? Or do you manage to do both ?
> people typically view teaching as something that doesn’t mix well with a parenting relationship, so you delegate it all.
I would be very interested in hearing more about this. From my (admittedly American) perspective, so much of inherent parenting seems to basically be teaching e.g. potty training, teaching manners, teaching how to eat food and not throw it on the floor, etc. How does that fit in with a social norm that teaching doesn’t mix well with a parenting relationship? Or am I misunderstanding you?
Whenever I talk about teaching my children to read / music / a language I speak / whatever, parents here tell me that 1) school will do that them better than me 2) they should enjoy their childhood and not learn to read, etc too soon 3) it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
This mostly means that you have to fight with the children in order to make them listen and focus, and that a third person (like a teacher), whose relationship with your children is less emotional, can teach them more effectively.
After Covid and school closures, mort parents I know had a horrible time trying to force their kids to watch Zoom class while working from home and never wanted to do it again. So all you people teaching your kids must have tools or skills that they don’t !
I am also really curious because it seems so hard to find the time to have regular learning sessions with the children during the week while working, so I wondered how you fit that in your schedules. But that must be very individual and family-dependent.
I taught my children how to use a computer, and I certainly did a *much* better job than the school did so far. My older child is in 4th grade; at school they teach how to use MS Word and PowerPoint. At home, we already did vector graphics in Inkscape, and some simple programming in Scratch.
> they should enjoy their childhood and not learn to read, etc too soon
This is one of those stupid opinions that come from people who can't tell the difference between "forced to do something" and "genuinely curious about something". (Perhaps they were never genuinely curious about something themselves? Or maybe school destroyed all their natural curiosity?) How is childhood less enjoyable if - in *addition* to everything else - you can also read a book? Or a sign on a street, or anything.
Is it perhaps the very process of learning that is supposed to be unenjoyable? Well that's easy: only teach your children when they are curious, and stop when they get annoyed.
> it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
As long as you don't give them grades and homework, the roles are definitely not the same.
> you have to fight with the children in order to make them listen and focus
Perhaps that depends on the child. The school review article also assumes that most kids are unwilling to listen and learn. I have no idea how much that is true... my kids have inherited my genes, they *are* curious about stuff. Not all stuff, and not all the time, but that's okay.
> So all you people teaching your kids must have tools or skills that they don’t !
I think the required abilities are intelligence (that's mostly given in this thread), knowledge of the subject (shouldn't be a problem while your kids are in the elementary school; plus you can read their textbook), and having a good relation with your kids. Then if you tell them something, they will listen. Also, don't push it; if they are not in the mood today, try tomorrow. If they are in the mood only for 10 minutes, stop after 10 minutes.
> it seems so hard to find the time to have regular learning sessions with the children during the week while working
This is true. Most homeschooling families I know, there is a stay-at-home mom.
But you don't have to compare yourself to school. For example, if you are teaching your child to read, that is only one subject; the school teaches ten subjects at the same time. Also, school needs to proceed at a certain speed; you don't have to. Finally, school wastes a lot of time on classroom discipline and similar stuff; if you only teach your children when they are in the mood, this is not an issue.
So, homeschooling (all subjects, at the same speed as school) is difficult. But teaching your children a thing or two, at their preferred speed, is not.
> it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
What I find fascinating about this from my own limited, cultural perspective, is that I view teaching as inextricably part of parenting, especially at young ages.
For example, I very clearly view myself as teaching my daughter to speak, to eat using her hands, to eat using a knife and fork, not to put her hands or fingers in electrical sockets, to use the bathroom rather than be in diapers, to dress herself, etc., etc. So from my cultural perspective, it sounds like you’re saying parent should teach their kids a bunch of things and then suddenly stop teaching their kids things (like reading).
Are you saying that in your culture those earlier parts of parenting aren’t conceptualized as teaching?
> This mostly means that you have to fight with the children
> in order to make them listen and focus, and that a third
> person (like a teacher), whose relationship with your
> children is less emotional, can teach them more effectively.
I have encountered this claim. I think it depends on the child, the parent and the topic. I rarely had to fight with my child to get him to listen and focus (*).
I did find that having a "academic education" spot made it easier to settle into the "now we are formally schooling" mindset. But it wasn't necessary.
(*) One thing that probably helped a lot with the focus was that he took Tae Kwon Do lessons starting when he was six. A *lot* of the repetition was "focus and concentrate." This carried over to other things (e.g. baseball) and also to schooling where we could tell him this and (a) he understood what he was supposed to do, and (b) had a lot of experience DOING it for someone else, and (c) saw how it helped in the Tae Kwon Do (or baseball) environment so maybe it actually mattered.
Those are interesting arguments. I was raised in a European country and my parents did not teach me math or reading, but I don't know if their reasoning follows along the same lines. The argument that viscerally resonates is the one about roles: teachers teach, parents parent, etc. so mixing responsibilities is not good.
Maybe it's my time living the USA now, but my gut reaction to these points is bewilderment. 1) maybe, maybe not, who knows what kind of teacher your kid will get; 2) reading and numbers can bring great joy, and additionally they can help the child be more part of society which can also bring joy; 3) like the other comment already said, parents already teach a great deal, so not sure how reading is different than teaching to fish or clean the house or take care of the dog.
How was your own childhood like in these terms? What is your stance on these questions?
Oh, I hated school, never learnt anything there, and was systematically bullied for participating in class, so I am in favour of burning all schools, if possible yesterday. But this is is considered very original here, and every time I say so people tell me that fighting bullies must have taught me to stand up for myself so I should not deprive my children of this socialising experience.
My father taught me to read on cereal boxes at breakfast, I then learnt most of what I know from books at home / family debates at meals during the weekends. So I don’t really have a model of a structured way to teach children in the family. I would love to, however, because I think it might be great fun ! (The eldest is 5, so we have a great time of answering weird questions about the world at the moment).
Around me, parents (mostly mothers) spend what seems to be a very unpleasant time repeating school lessons and exercises every evening, but any actual teaching (for example, music or language lessons) are delegated to private teachers, even for the basics.
When children are older, it is very common to delegate the « repeating the lessons and doing the exercises » part to tutors (usually students at university), precisely because parents find there’s too much conflict when they do that with teenagers.
Been thinking about the entire Alpha School vs normal school thing. When my parents went to school in rural Denmark in the 1970s they had 900 hours of teaching in 7th grade (so round 4.5 hours per day + homework ), and used a lot of the same motivation methods as the Alpha school dies . Currently it seems the US public schools have 1000 to 1100 hours per year. In 2013 we tried in Denmark to demand 1200 hours (so 6 hours per day). Despite this academic standards kept slipping, and the reverse Flynn effect are still taking away 3 IQ points per decade.
So it seems that actually the amount of hours above some limit is not very important. Perhaps with some simple incitament system even normal schools could go down to 800 hours per year + some homework and keep their current standard. That would not be very far from the ~600 hours studying that Alpha seems to be using.
Maybe culture in general has become stupider and schools have to try to compensate for this. Culture both in terms of medium and content. Eg people spontaneously read less because of TV or the internet. Music becomes less complex. And culture values ”coolness” over accomplishment (which is ”square”).
And that certainly aligns with what I'm seeing amongst my younger Millennial and Gen Z coworkers. They can't seem to absorb basic written information from emails or shift reports, nor can they adequately convey extremely basic information in writing. We are a high-end, service-centered luxury product, but they literally aren't able to put together so much as, "On Monday July 7 approx 9 AM, client complained that widget wasn't performing third state of widgeting. Widget analyst checked and says widget needs to be replaced, not repaired, will perform between noon and 1p. Offered client 50% off this widgeting as apology, client seemed satisfied."
You know, there seems to be a really weird synchronization between the timing of this stuff and the emergence of AI. Everyone's worried that AI will cause mass unemployment but what if the opposite is true: what if AI is appearing just in time to save us from a catastrophic failure to adequately train the next generation of adults? If so, what could possibly explain that coincidence? It's almost as if people are taking their cognitive foot off of the gas because the AI has stepped in, but the order is backwards. The AI hasn't actually stepped in yet and this de-literacy trend has been going on for 10 years now. Is it some weird historical synchronicity?
> this de-literacy trend has been going on for 10 years now
I wonder whether people are actually getting less competent, or they are mostly the same but now more incompetent people are getting the credentials.
Like, in the past: 10% of people at university, all of them can read; today: 50% of people at university, one in five can read. The same number of people can read, but the quality of the university declines.
They’re not getting dumber because IQ is mostly genetic and genes aren’t changing that fast. However our insane obsession with equity means that we’ve stopped identifying the best and brightest. So our leadership and managerial classes are increasingly composed of poorly trained midwits who don’t really know what they’re doing. I suspect that means that the next 2008-level generational crisis we face will cause unrecoverable damage and it will shift us to a new and worse social equilibrium. In 50 years I think the US will be a richer version of Brazil.
To the extent that there is a failure to train the next generation of adults, it is because increasingly sophisticated algorithmic media (social and otherwise) is feeding them a steady stream of engrossing but fragmentary content that degrades focus, attention span, and the like. Not out of malice, but because their owners are paid by the click and if you're paying attention you'll realize that you don't need to click on something else just yet.
So you're basically hoping that the first generation of "good" AI will compensate for the damage caused by fifteen years of crappy proto-AI.
>If so, what could possibly explain that coincidence?
Perhaps the internet? The vast volume of online text is an essential input to the training of LLMs. And the same infrastructure lead to pervasive social media - and, arguably, shortened human attention spans.
Is the issue attention span? Or is it actual inability to /write/ as opposed to a different intellectual deficiency? (pre-frontal cortex issues can come from a lot of places, but we could blame schools for failure to teach leadership if we want.)
If the person is unable, without significant prompting, to relay that series of events (that you describe), then the issue isn't writing or reading.
Now, if you want to get really fun, recognize that the IQ Test Scores aren't going down.
>Now, if you want to get really fun, recognize that the IQ Test Scores aren't going down.
Why would IQ scores go down? IQ is primarily genetic and our genes aren't changing. We wouldn't expect a reduction in education to be reflected in lower IQ scores.
I find I can't absorb information from watching videos, I have to read the text to understand what's going on (apart from watching instructional videos about "okay so when you're taking apart the thingummy, *here* is where the release button is hidden that they don't tell you about in the manual").
I can imagine younger generations being the opposite, and getting all their information visually but not from texts. Which is great, they can learn the material. But then when they have to write down a message or report, as Christina says, they're sunk. Watching a video and listening to people talking is not at all the same as "this is how you write down the details of what happened, including all relevant information and leaving out what isn't necessary". So if it was talking face-to-face they probably could communicate "Mr Jones in Room 366 complained about the taps in the shower sticking" but having to write it down they can't work out that no, they don't need to include that they were sweeping the floor at 8:00 a.m. when the call came and yes, they do need to include the room number.
This is not a writing issue, this is a "leadership" issue, in that they are not able to put themselves into someone else's shoes. This is the sort of "basic communication" that one should learn before getting a job. I'd be willing to think of anyone who can't filter information, even when prompted to filter information, as fundamentally retarded. This seems like the sort of question that might work on an IQ test.
I know a guy whose native language is pictoral. He had severe issues with arithmetic because he'd keep on losing track of the pebbles (also had issues with symbols). He does NOT have issues with communication. It's not a visual versus textual difference.
(If you leave subtitles on, can you grasp the video better? Or do you have to read the text apart from the video?)
Maybe they read more words but fewer challenging texts. I have understood college professors have experienced a decline in the reading comprehension skills of undergraduates over the last few decades.
My expectation is that they're learning just as much, but different things. Certainly reading extended text with a plot is different from browsing blogs.
Positively regarded by AI safety folks I know and work with. Well regarded as a space. Great staff and well run events, locally and internationally. I haven't heard anything negative about them at all.
I might be *slightly* biased towards them as I've worked with a couple of their staff, but I don't think overly so.
I've been wresting for decades whether libertarian ideas are right and wrong, eventually I ran across an interesting one. Do not see the market as an institution, like a bunch of businesses, but as a meta-institution evolving other institutions, it can evolve something like a government if that is best. Okay, but in that case shouldn't there be a partnership with the real government, and the real government helping that kind of market-government out?
Imagine the kind of town where half the population is working in the same privately owned mine. I think it is necessary that the town government and the mine will have a complicated partnership. The government needs to regulate against some abuses, but cannot possibly treat the mine as some sort of an unfriendly entity, they have to help them a lot, in order to secure those jobs. What I cannot imagine is being entirely neutral and indifferent to the mine, libertarian style.
All ideologies ignore a part of reality, you need to know which one.
Libertarians typically deny the parts where cooperation on a large scale is needed (for example, many are covid denialists, because pandemic is the kind of a thing where the market does not have much of a solution). Also, their claims how governments cannot do anything right are sometimes exaggerated (for example, many countries have a decent mass transit system), and at worst become a self-fulfilling prophecy (when people stop even trying to do things right, because they believe that "science proved it impossible", that's when even more things fall apart).
It is best to think about them as an antidote to the opposite fallacy -- that government always does everything right, and that everything needs to be micromanaged by politicians.
The market is a mechanism for organizing human cooperation more efficiently than anything else we've ever come up with. But it only works when the people doing the cooperating are all on board with the desired outcome; you can't round up the dissenters and put them to work on the project you think is the Best Thing Ever. Some people hate it for that. The rest of us, make it work for us. Together. Cooperatively.
* Pay attention to million different things at the same time. A central planning committee, or even a group of citizens deciding democratically, would only pay attention to a few topics that happen to seem most important to them at the moment. On market, someone *also* notices an opportunity to make a profit by improving something that is relatively trivial; this is how we get the relatively trivial things to work well, too. (This is why market economies typically do not have shortages of toilet paper, that planned economies are famous for.)
* Respond flexibly to changes in supply and demand. Things get more expensive before you run out of them; and as a consequence you often *don't* run out of them, because people respond to the increased price by switching to a substitute or finding a way to spend less.
* Provide relative freedom and autonomy. When there are many possible things to do, you don't have to do any specific one. If you hate X for some reason, you can choose to do Y instead. (If too many people do that, X will probably start to pay better, so some people who don't mind will switch from Y to X.)
* ...other things I forgot.
Things that markets are not so great at:
* Doing things that are not profitable, such as feeding the homeless.
* Avoiding things that are profitable but otherwise wrong, such as slavery. Reducing negative externalities, such as global warming, robotic spam calls, etc.
* Generally, following the law. The market always looks at a punishment for breaking the law as a price sticker and asks: "given the punishment, and the probability of actually being caught and punished, is it worth doing?" and uncomfortably often, the answer turns out to be "yes" (or "yes, if we spend tons of money on lawyers").
* Lots of resources are burned on zero-sum games, such as advertising trying to switch customers from product X to a mostly identical product Y.
I see you're making multiple assumptions you haven't justified.
One is that the market is run by some sort of management capable of entering into partnerships. There is no such management, so there can be no partnership with the real government or anything else.
A second is that there's a government (I'm assuming you mean an institution with a monopoly on the use of force). There's an entire school of economic thought, known as anarcho-capitalism (or ancap for short, as I like to use), that discusses capitalism in the absence of a government. A market does not have to evolve one, although it's possible that something about human biology makes governments very likely. And I see use in discussing the market situations where governments do arise.
A third is that your mining town example exemplifies a market. It sort of does, but it's not free, and so many of the usual free market principles won't apply. The most obvious non-free market is the labor market: you're implicitly assuming the mine is the only customer for labor. It probably isn't, in a stable town - the people who work inside the mine have a demand for food, shelter, clothing, entertainment, and various other creature comforts, and division of labor principle strongly encourages them to purchase those things from yet other people. So, anyone in town deciding who to sell labor to, will have multiple employers to choose from. This is good news for people selling labor.
It is possible - and perhaps even likely - that townfolk will have an incentive to keep the mine open and employed. But a free market has a way of signaling it: a Help Wanted sign, with a salary offer attached. If the salary is high, more people will apply; they will work in the mine instead of farming, ranching, fishing, logging, weaving, and so on, so labor in those jobs will become more scarce, until those salaries rise as well (due to miners suddenly wanting food and houses and cloth badly enough to offer more money for them). Equilibrium occurs about when the salary for each job corresponds to the demand, modified by the training and talent required, physical strain, and risk of hazards on each job.
If the salary for mining goes up as high as it can go (the ore produced by a miner cannot be sold for enough to pay for that miner), then the mine is at capacity; if those miners aren't enough to sustain all the other labors, then laborers leave. Either way, equilibrium is reached.
Channeling Churchill - "capitalism is the worst form of political economy except for all those other forms that have been tried from time to time."
I am neither a tankie nor a Randian. Capitalism is the best way to create wealth, but not an infallible god. It's humanity, occasionally at its best and occasionally its worst. Most often muddling through somewhere in between.
I think you would find that the libertarian does not live in the town with the mine. He lives in a think tank. But perhaps that may be because I’m currently driving through a stinky area of oil and gas production - an area that fuels our economic growth, enables Amazon to zoom our presents around, enables rappers to rap, etc.
If you think of 2 original sins, pride and envy, the libertarian starts with the former: don’t tell me what to do. Sometimes that attitude may be informed by experience and justified.
But when so, that is because the other strand of liberalism - egalitarian liberalism, driven by envy - has overreached via government, become too powerful.
It is best to see libertarianism as a counterbalance. Otherwise it is an extreme, radicalized response to an extreme situation. We need government to keep the mine from polluting our drinking water, but we don’t need it to run the mine or redistribute the mine’s profits.
The short version is competition, incentives and safeguards. It's not about companies being better than states; All large organizations eventually suck if they have no competition, no good incentives, and no safeguards.
A market of many, minimally regulated companies offering similar products and employing similar workers will naturally gravitate towards efficiency to extract more money from each product and/or get a greater share of the market by undercutting the competition, and naturally gravitate towards better pay and better conditions for the workers because otherwise those will go elsewhere. Not perfectly so, and if you let them get away with it they will probably try to employ cartel tactics to get more, but as a general direction they will do these things. Markets also have the advantage that accommodating different people is a legitimate way of carving out a safe niche for yourself (though again there are some shitty tactics here).
Democracies also have, in principle, these components. But they tend to select most strongly for the things that sound good, as opposed to the things that actually work well, and they have a tendency to enforce the same thing for everyone. Nevertheless, you need some entity to oversee that the companies do not employ cartel and other shitty tactics, and democracies seem to be the least bad of the bunch.
The very worst is bureaucracies. Bureaucracies exist in most states, they are not voted in, they do not have to survive competition, their only incentives are rent seeking, entrenching themselves and insulating themselves. In many states, bureaucracies have successfully managed to make the democratic state dependent on itself as necessary middle managers. Likewise, large companies have realized that they can enter a symbiotic relationship with the bureaucracy through excessive regulation, wherein they insulate themselves from new competition by making entry into the market so prohibitively expensive and difficult that very few dare trying. This is were the idea of deep states, BigCorps, etc. come from.
As far as I can see, pure ancap libertarians are very rare. But it's imo reasonably to look at the current state of the, well, state and conclude that we need to regulate less, reduce the state coefficient and try to spread out power from a few big corps to many smaller again as much as possible. The main point of contention is usually whether the latter happens mostly automatically in a minimally regulated market, or whether you really need to outright smash the large corps.
For your example, there isn't really a market, so I expect no good outcomes. Maybe if you aggressively enable other smaller competitors.
If you have to introduce a very simplified and kind of unintuitive thought experiment to understand something (when you already have lots of direct evidence you could use), you probably have some preconceived notions about how the subject should work that the existing real world data doesn’t justify for you.
It’s difficult for me to understand why we should follow your intuitions about what kind of government policy we can imagine, right? Some people cannot imagine a government which doesn’t have a policy prohibiting hate speech, and yet I find it very easy to imagine a place where nobody gets arrested for saying things, just for doing things.
Especially in the case of a massive employer, the laissez-faire attitude is to offer no enticements whatever, partially because the more you do so the more the mine has captured the town and if the mine ever closes the town’s industry will die. Rather you should make sure the market is open to new entrants which can compete against the mine. The best way to do this is lowering trade barriers. By default, the mine has no power to do anything ‘evil’ to the town except leave, which it can do regardless of your attempt to bribe it to stay, and history says that’s exactly what will happen.
An entity which employs 50% of the working population directly has just about cornered the employment market: those who do not work for them likely work in service industries which depend on the miners -- stores selling stuff to them, schools full of miner kids, doctors treating their families, policemen stopping their brawls and so forth.
Likely, it would be common knowledge that without the mine, the natural size of the town would be three farms and a graveyard.
In this environment, the natural state of affairs is that the owners of the mine end up running the town, e.g. through bankrolling officials. Likely the mine owners will get away with pretty much anything short of hunting townfolks for sport. They will not get pulled over for speeding, there will be no town council hearings about elevated cadmium levels in the creek and so on. Any oversight will have to happen at a higher government level where 99% of the constituents are completely indifferent to the fate of the mine.
If the same happens on a national level, whatever polity is dependent on that single source of employment is in deep shit. Private ownership of the means of production is only good in so far as it leads to a working market, a private monopoly is not better than a state-run monopoly, so they could just opt to nationalize their mining industry. A better approach would probably be to split the mining company into different mining companies though, but even this will leave them beholden to overall industry interests. If you want to regulate mine pollution, don't get yourself into a situation where half of your population work in mines.
>"…a private monopoly is not better than a state-run monopoly…"
Private monopolies (when not due to explicit state-granted privileges) still face (an attenuated degree of) market discipline; state-run monopolies can be nigh-indefinitely sustained despite persistent value-destroying behavior, due to taxing power.
Suppose you think labor unions are bad because they slow progress. For example, they ban automation in order to protect jobs, or they ban more efficient ways of doing things.
How does this work on a theoretical level? It seems like you should have three parties who are all fighting to collect the surplus from a transaction: company management/shareholders, workers, and consumers. Even if we grant that "public interest" = the interest of the consumers, that seems to leave management/shareholders and workers on an equal footing as potential rent-seekers. What's the theoretical argument for why management/shareholder interests are aligned with consumers' (in increasing efficiency / bringing down prices) but workers are against?
That’s a very big assumption up front; the public are mostly workers and workers are mostly citizens. That’s the Achilles heel of neoliberalism; at some point the citizens get mad.
Lots of other good responses already, but I want to point out that I'd model the union as a fourth party in competition with the workers, shareholders and consumers.
Owners maximiz present value of all future profits. Workers are more short term cash flow interested. Hence the common dynamic of opposing automation.
Secondly owners are more agnostic of the form of that future profit. More low margin goods fewer high margin goods are both candidates. Workers need the costs(funded from profits) to come in the form of employee costs.
Short term vs long term
Maximizing one form of cost, rather than all sources of enterpise value
Management is in constant negotiation with the consumers, via prices and advertising. Management is in constant negotiation with labor, possibly through a labor union but definitely by offering wages and benefits. But there's no channel for consumers and labor to negotiate directly with one another. So if the process is viewed as "management and labor are both seeking rent from the consumers", then management is the party that will actually be collecting that rent and they're going to start off in a privileged position when it comes to deciding how the rent will be divvied up.
A good union can figure out what management's game is, what the underlying numbers are or ought to be, and negotiate a better deal for its members. But, as you note, they can also block improvments in the business - and if you are looking at this from the POV of a consumer wondering "why can't I have the results of the more efficient process", then it's a problem that you can't negotiate with the union. Management got where they are the old way, so they're probably OK with the old way continuing so long as they aren't being outcompeted by someone with a more efficient model, and it's the same union blocking the same improvements across the whole industry if it comes to that.
Unions can block *changes* to the business, some of which *might* be improvements. Many changes made to businesses are deleterious for the consumer, labor, or, in the medium and long term, management.
If people *knew* what changes were going to improve businesses, the world would be a different place.
Two things are off on this framing. The nature of the struggle, and the parties of the struggle.
First, the nature of the struggle is not a fight, it is a competition. Specifically it is a competition to cooperate, where the terms of the cooperation are determined by the competition.
Second, the competition is not between parties, it is within them.
Consumers compete with other consumers to buy the best value products from producers (a form of cooperation) according to the rules of the marketplace
Producers compete with other producers to sell products (at a profit)to consumers (cooperation again) according to the rules of the marketplace.
Prospective employees compete with other prospective employees for jobs with employer/producers (cooperation again) according to the rules of markets.
A cartel would be where producers collude to no longer compete, thus violating the rules of the market, and allowing producers to either take advantage of consumers or employees (if the collusion is on wages or working conditions)
I am not sure what the term is for consumers to collude, as they are too numerous and disconnected to do so (a consumer cartel succumbs to Mancur Olsons logic of the difficulty of collective action.) I can imagine it can occur in auctions though.
A labor cartel is known as a union, or closed shop union. This is where existing employees collude to exclude or hinder prospective employees from competing with them.
A rent seeker is someone who extracts more from the transactions than would be expected in an open competitive market. In this case it would be cartels of producers, or closed shop unions of employees. Again, I don’t know what the term is for colluding consumers, if there even is one.
So to answer your question, neither form of cartel (producer or employee) is aligned with consumers needs. Both interfere with the dynamic nature of markets to incentivize lower prices, better quality, and more efficient production.
A company’s incentives don’t really leave much room for “rent seeking” in this negotiation. They want to pay the workers as little as possible and extract the maximum profit for themselves. The amount of profit they make is determined by 1) the cost of inputs, including labor, and 2) their profit margin. There is never a negotiation between a union and management in which companies finagle higher prices and higher profits as a result of the bargain struck. Unions negotiating higher wages in various forms for themselves is the reverse: basically always higher prices, at the expense of the company’s profit margin and the public’s wallet.
The right comparison is of a cartel, in which all sellers of a certain widget collude to keep prices artificially high and output restrained.
I think Adam Smith noted the terrain here in The Wealth of Nations. At least, I recall some discussion of why the employer typically has the upper hand over the employee; I don't recall any discussion of collusion with the consumer. It's been years since I read it, and I get the sense that in Smith's time, the collusion might well have been consumer+laborer vs employer, due to UK only just beginning the Industrial Revolution, and seeing the concept of division of labor as a powerful artifact suddenly falling in the lap of the employer.
We can assume for starters that in a market for widgets, the consumer's enjoyment of that widget, and the laborer's expenses in producing the work to make that widget, are fixed. The former is much higher (or else no one would make that widget), and consumer, employer, and laborer compete for their share of that difference.
AIUI, employers had a natural advantage in information. They knew better than any laborer how to squeeze down the price of that labor, and knew more sources of labor, than any laborer; so if an individual laborer couldn't deliver on that price, the employer knew where to get another laborer who could. This is turn created an incentive for laborers to conceal their tricks for productivity from each other (just as suppliers of any commodity hide their trade secrets), but there was only so much to know about manual labor.
So the next most ready solution was to coordinate a representative for laborers, that knew how much to negotiate for labor, how to standardize productivity, etc. In other words, unions. The catch is that for reasons I'm not completely familiar with, unions are basically monopolies - any employer in that business has only one union they can go to to purchase labor, and in some cases, they're not allowed to seek free agents. (It's even morally condemned in several of those cases.)
Meanwhile, the analog to unions among employers - cartels - are likewise morally condemned (arguably because it would just reinstate the natural information advantage). The analog among consumers is the state, manifested as product safety regulations, consumer protection agencies, etc. - and is not condemned. Probably because such institutions view customers as otherwise disunited individuals, just as laborers.
At this point, it's important to consider consumers who aren't disunited - some consumers are themselves huge businesses, and have little trouble negotiating for their share of the surplus. Think supermarkets buying produce from local farms; manufacturers buying inputs from raw material suppliers; retailers buying from manufacturers; governments buying from defense contractors. Some labor is rarified enough to be in a similar position (think doctors, lawyers, software engineers, middle managers, or any specialist creating anything that doesn't have to be a public good in order to be valuable).
Notice how rent-seeking doesn't really enter the picture? To me, who teams up with who is largely a function of who has more of the relevant information. Whoever has the most, has the natural advantage, and everyone teams up against the king of the hill. In the Industrial Revolution, the employer was king, so consumers and labor teamed up on employers. If the state, in its capriciousness, favors one of those two, the team-up changes.
After the World Wide Web Revolution, things got more chaotic. In the 1980s, Reaganomics tapped suppliers as the winners, so consumers and labor were the underdogs, until New Deal economics reasserted itself and labor got the upper hand - for a while, and in some sectors. Nowadays, tech is the king of the hill, so the coalition is small business consumers with non-tech labor and a few management concerns. Plus, there's a whole market industry that sort of loosens the assumption about enjoyment per widget being fixed, and now an AI revolution starting to weaken the other end.
Both the workers and the owners have aligned interests against customers to create a monopoly to increase total profits. So they will agree on something like you can only hire union labor that reduces the ability of new companies to enter the market. The lack of efficiency isn’t a problem for the owners because they still make more profits due to the monopoly power, even if they have to share some of that with the workers. To have an effective union, you really need your whole industry to participate, and a major task of unions is to try to prevent non-union competition from emerging. This is also why companies in competitive markets are so opposed to unions—if they have to deal with unions but their competitors don’t, they’re in big trouble, because they won’t be able to build the monopoly power that makes it possible to succeed even with inefficiencies.
> What's the theoretical argument for why management/shareholder interests are aligned with consumers' (in increasing efficiency / bringing down prices) but workers are against?
I don't think workers are particularly against. The typical worker is willing to take a coinflip as to whether they are paid 50% more in a more efficient industry, or send a few months unemployed seeking another job.
But union leadership has different motivations. They want to maintain a large dues-paying union membership.
Two dynamics. Firstly, shareholders desire profits which increase the value of their ownership and workers desire salaries which they take outside the company. Secondly, shareholders' interests are aligned with the company as a whole while workers' interests are specific to whatever job they do.
This means that disruption that increases the value of the company is in shareholders interests, who benefit from the rise in company value, but can be against the workers' interests even if it might increase salaries and positions overall since it might disrupt the specific current job workers do. Thus shareholders have an interest in automating and reorganizing to achieve maximum efficiency because it maximizes their pay out. Unions have the opposite incentive, to ideally increase payroll as much as possible and importantly to keep existing jobs as they are so there's minimal disruption to members.
You're right that owners can also work against the interest of consumers. The most famous example of this are guilds which were not groups of workers negotiating with capital but business owners who often coordinated to keep employee's wages down and prices high. But it's also how monopoly works. However, the corporate coordination is banned in the US. They are not banned in Europe which is part of how their system works and part of why there's less labor-capital conflict. But it's also part of why they have a larger grey economy and less automation. So in the US it's mostly labor unions that do it.
Basically, the New Deal decided it was going to ban coordination and collaboration between businesses and then give unions a special set of privileges and allow them to coordinate. This was supposed to permanently tilt things in favor of workers. And it did in those sectors but it also stagnated the unionized sectors over time.
This is why communism was supposed to work, by the way. Worker's cooperatives were supposed to be more efficient and have no conflict, despite having no unions, because the workers and owners were one and the same. However, what actually tended to happen even in the most successful cooperatives is that without someone to push for firm level efficiency things stagnated. The way communism solved this was the state pushing for economy wide efficiency. Which resulted in the command economy, the calculation problem, and is now widely considered a bad idea.
So to answer your question directly: Free-markets are pareto-optimizers for the dyad of producer/supplier of the original factory components but only indirectly affect worker interests. Separately in the dyad of company/employees a free-market maximizes their benefits. Unions act as monopolies and benefit those that are covered, but have adverse impact to the other parties.
Free market model:
The original product sale is a two party transaction (company/individual) in a competitive market with multiple possible suppliers/consumers. Markets clear when marginal price = marginal demand, and utility amongst those two classes of parties is maximized.
Within the company, there are a large number of factors of production: labor of different types, different parts, services, rent, etc. Each of those presents a market. Those unique markets each maximize utility between the associated classes of parties. In this way, labor is optimized.
If the original product sale is a monopoly, the product-seller can gain excess returns, but does so at the expense of its purchasers (and to an extent in $ that exceeds the loss of the $, meaning it's net harm). A minor side effect: volume of products necessarily decreases, so the suppliers-of-factors-of-production are also shorted volume (less widgets, labor, etc.).
If the labor-sale interior to the company is covered by some sort of collective bargaining relationship (labor monopoly), the product-seller (people selling their time to the company), can gain excess returns, but will do so at the expense of its purchasers (the company, in the same net-harm context). A minor side effect: volume of labor necessarily decreases, so fewer employees should be employed. A new minor side effect: volume of products also decreases, so product-buyers and other-factor-of-production suppliers are also shorted volume.
Very few things are only good or only bad. Labor unions are a mix, just like almost everything else. They aren't even ALWAYS good for their members. But they are a way of evening the balance of power between the workers and the employers that "sort of works". If you want to replace them without seriously injuring lots of people, you need to find a better tool for that purpose. (But be aware that the government is generally on the side of the employers.)
Alcoa pitted both the managers and the union against each other, in a crusade to decrease workplace injuries. Worked like a charm. This suggests that both managers and the union have obvious deficiencies in "preventing injuries."
Improvements in technology that might make the firm more profitable and competitive, might reduce the demand for labor within the firm. While the firm is motivated for profit, it's also motivated to cut unnecessary costs, which may include a large number of longshoremen preventing automation in ports so as to protect their own jobs.
I think the argument is just that monopolies in any necessary input of production, whether capital or production stifle innovation. Whether that's Standard Oil buying up all the competition, or a union that covers all the available ports, you're going to end up with significant deadweight loss. Compare the longshoreman's union to the auto unions. Tesla can outcompete ford, partially because it's significantly more flexible in how it can allocated labor.
I believe it's a principal agent problem, although I'm open to correction. From Adam Smith (1)
"
The interest of the second order, that of those who live by wages, is as
strictly connected with the interest of the society as that of the first.
The wages of the labourer, it has already been shewn, are never so high as
when the demand for labour is continually rising, or when the quantity
employed is every year increasing considerably. When this real wealth of
the society becomes stationary, his wages are soon reduced to what is
barely enough to enable him to bring up a family, or to continue the race
of labourers. When the society declines, they fall even below this. The
order of proprietors may perhaps gain more by the prosperity of the
society than that of labourers; but there is no order that suffers so
cruelly from its decline. But though the interest of the labourer is
strictly connected with that of the society, he is incapable either of
comprehending that interest, or of understanding its connexion with his
own. His condition leaves him no time to receive the necessary
information, and his education and habits are commonly such as to render
him unfit to judge, even though he was fully informed. In the public
deliberations, therefore, his voice is little heard, and less regarded;
except upon particular occasions, when his clamour is animated, set on,
and supported by his employers, not for his, but their own particular
purposes.
"
The fact that you can collectively appoint someone to represent worker interests is a potential solution but one with obvious principal agent problems. Not just corruption amongst labor union officials but...most workers want stability, a lot. Automation resulting in market disruption that increases wages is a difficult sell to people who just hear that they're losing their job and union officials have no incentive to explain that.
Imagine a technology or process of sufficient difficulty to master that a worker must invest years of training in order to perform it at a satisfactory level. In such cases the protections and guarantees of a union might be needed to incentivize workers to enter that track and assure them they will not be terminated before they have learned the process and been able to earn the higher wages it commands for some number of years. If I am looking at investing years of my life into an industry, some degree of insurance from the vagaries of the market or management whim might be the factor that tips my decision one way or another. Without this incentive, the shareholders and consumers cannot reach the labor tipping point to manufacture the widget at all (an ‘everyone loses’ scenario). The average ACX reader who is an auto-didact may underrate the difficulty of learning a new trade for the average worker and thus how valuable union protections can be.
“The average ACX reader who is an auto-didact may underrate the difficulty of learning a new trade for the average worker and thus how valuable union protections can be.”
Good point. I feel like this often in the Slow Boring comments where unusually high agency high efficacy commenters seem to assume everyone has the same potential if they would only make better choices.
I have gatecrashed weddings in India as a child over 30 years ago. It seems possible even now. Just show up, greet the couple, and eat.
My husband has told me he would be walking home from school or sports past a hall in Bombay and just walk in for ice cream (they always had a fancy wedding going on). He says it was pretty much expected some random kids would come to eat.
The hosts gladly accept your wishes and give you a dessert, even if they know you're gatecrashing (which is hard to know if it's crowded). Sometimes a whole meal although kids typically gatecrash for dessert.
Such a different and informal wedding culture :). Partly because each guest costs a lot less than in America.
I thought about this because people we know getting married in America take the wedding event so seriously. They plan for it for a year. They limit how many guests parents can invite, which I find utterly shocking. They say no to kids under 12. Ultimately what limits the number of guests, I figured, in America, is the fire code.
It is changing in India too and the wedding, especially among the wealthy, is now very similar to weddings in America. But the speeches haven't made it into any Indian weddings in India, as far as I can tell. Any delays are often due to the proper alignment of stars in the sky (literally!).
> Ultimately what limits the number of guests, I figured, in America, is the fire code.
Ultimately, Indian weddings are big because there's a lot of prestige and status associated with having a big wedding.
In the West, that prestige doesn't really exist; we might have other equally irrational and expensive ways of signalling status, but not that particular one. So people invite the people they actually care about. My own wedding was fantastic, and adding an extra thousand people that I barely know would only have made it much worse.
The elders decide. Whatever they want - show of wealth with mega events in different venues (!) or a small wedding at a temple with a lunch after the ceremony. The bride and groom don't have much say, traditionally.
The invitation is traditionally written FROM the grandparents. They're the most important part of the host family.
And the traditional wedding invitation says "Come with your relatives and friends, to bless the new couple."
And while the elders decide, it's considered proper form for the grandparents to take into account what the preferences of the bride and groom are.
The bride's parents usually pay and the groom's side often make unreasonable demands. This is all typical.
"Ultimately what limits the number of guests, I figured, in America, is the fire code."
My understanding is that the WEDDING itself isn't where US folks create limits but the reception. The problem is that the reception costs can easily exceed $100/person (catering is expensive, I guess). So a 500 person wedding reception would come to $50K. This is actually a financial hardship for a lot of people :-)
One *can* have a fairly inexpensive reception: Rent space at a park, don't spend $100/person for the food, etc. But lots of American folks (brides, mostly, I think) want the expensive per-person reception so head count limits are required.
My friend's wedding was held at a relative's house. They hired a company to come set up a tent and chairs and speakers in the backyard. Dinner consisted of Papa Johns pizza and supermarket food (the cakes were a bit more upscale).
I don't know what the cost was, but I'm sure it was less than $100/person.
This was the groom's second wedding. I've heard that people tend to go a lot cheaper for second weddings. His first wedding was a much fancier one that even had a rented horse-drawn carriage (I believe his ex was the one who pushed for all that).
Why are people so into spending on their wedding? Is it a display of wealth in many cases?
This part is the same in India in recent generations. It's still not as extreme as in America in the average case.
I say (not that anyone should care!) have religious or other ceremony sanctifying the wedding. Why a lot of guests, why anything more than the ceremony (what you call "wedding"), why a party?
But I only heard about birthday parties and the concept of birthday presents from English story books when I was growing up in India. For birthdays, mom made my favorite food, we went to the temple and some years I distributed sweets in class.
As India gets wealthier and more Westernized this birthday celebration is changing. Even the poorest people celebrate their birthdays with a cake ( nice part of it - seeing themselves as individuals).
> Why are people so into spending on their wedding? Is it a display of wealth in many cases?
I think the root of it is a desire to celebrate one of the major milestones in one's life. Getting married is a big deal. It makes sense to me that people want to mark the occasion.
As to why it is so expensive, that's mostly poor or middle-class people wanting to celebrate in the grand style that they see the rich and famous doing. This leads them to blow the budget trying to afford (just once) the things that come much more easily to the actually wealthy.
I knew a guy (in America) who used to gatecrash weddings for free food (he didn't have a big budget for college). He had to get gussied up, but it was easy to get in.
One more point. In India it's considered really rude and even bad luck to turn away a guest who wants to attend your wedding. So Indian Americans struggle with this. Their kids run it all (not the grandparents or parents, like in India) and they (the parents) get a small allowance for number of guests.
(There's no edit button on my comment using the substack app on my mobile device).
>Ultimately what limits the number of guests, I figured, in America, is the fire code.
As an European, I am really blown away how extremely regulated America is. I mean surely restaurants have a fire code here too, but I cannot imagine the average half-drunk quarter-educated ex-waiter restaurant owner has any idea how many people that means or really cares about it at all. What is the chance that someone from the state will pop in and count the wedding guests?
If you're renting the venue, then probably the owner or the owner's insurance company will be checking to see if the fire code is being blatantly violated because if anything goes wrong, "...and they were in violation of the fire code" makes them deep-pocketed lawsuit bait.
I haven't specifically seen this with weddings, but I haven't been involved in much wedding planning. It definitely happens with fan-run cons renting hotel space, and I assume the process is pretty much the same anywhere in the "rent event space to enthusiastic amateurs" business.
I think almost any indoor venue in the US is going to have some kind of occupancy limit imposed by local fire code. But also, churches and halls and reception venues have a limited amount of space for people and tables, caterers have some limit on the amount of food they can prepare/serve in a reasonable time, and also whomever is paying for the wedding (usually either the couple or their parents, depending on age and situation) will have to pay some amount of money per person who shows up to the reception.
As another European I don’t find this at all shocking. The risk is not that someone would come count the guests, the risk is that something would happen and then it would turn out the fire code was violated and the business would be sued or closed or something.
I also don’t think what limits the number of guests is the fire code. I think people just want to control who is at their wedding.
There can be a limit, in the sense that the venue you're renting can at some point turn away further people. If you set up a reception for 100 people and 200 show up, at the very least you're likely to run out of food and drink and probably the last 60+ guests won't be able to find a place to sit down.
I guess the fire code is one thing that increases the price of renting a locale.
I don’t know much about India, but from some articles I have read I got the impressions weddings are a bigger deal in India.
But probably Indian culture is more communal than in Europe or America? I guess in the past a village wedding in Europe was something everyone could partake in but nowadays people live increasingly in social bubbles that are detached from local geography.
People stress about wedding expenses from the minute a daughter is born in India. So, yes. But the other things I mentioned are simultaneously true as well.
What kills me is the reason for the large gaps under the doors in public toilets is also for fire code/emergency reasons. "A person" has to be able to crawl under the door. I'm not that fat and could probably still not fit under it (yes I've been to the US). With the US obesity rate...
The large gap under the door is to make it easier to mop the floor. Plenty of multi-stall restrooms in the US have full-length doors. They are just harder to clean and more time consuming to install (full height doors have tighter tolerances), so most restrooms do not use them.
Wrote a longform essay inspired by Venkatesh Rao’s concept of life intensification, combined with thoughts from Henrik Karlsson, Virginia Woolf, and a few of my own spirals at 18. It’s about how we don’t “find ourselves” so much as we generate ourselves — through feedback loops, relationships, and the people who draw different tones out of us. I touch on Rao’s ghost-to-character metaphor, the idea of “container people” who midwife our full expression, and how selfhood might be less about discovering a hidden essence and more about becoming legible in motion.
In psychology, it turns out there is a distinction between high self-monitor and low self-monitor people. High self-monitors change or mold their personalities according to who they're with, but low self-monitors show the same personality to everyone. It's apparently not a human universal, that experience of different people drawing different aspects of you.
And of course, neither is wrong, but I think both come with pros and cons.
My intuition is that this is probably more of a continuum than a 1/0. I and one of my kids are over on the low self-monitor side (not so affected by social norms and often not even noticing them), my wife and my other two kids are more on the high self-monitor side, but with some variation between them.
There's a new open AI model, with 1 TRILLION parameters (MoE, 32B activated): Kimi K2; I think the K1.5 release in mid-January got overshadowed by DeepSeek R1, but this new one looks like it performs in the SotA tier. But besides that, what I particularly like about it is that none of the release announcements say anything about "safety."
Why do you like that they don't say anything about safety?
Most directly, I take it as a sign the model is more likely to do what I ask. Or at least, that it's more easily modified to be such a model.
Indirectly, I'm hopeful that it reflects the culture in Chinese AI companies more broadly of not caring about "safety" the way the US companies do, and that future models they release are similarly unlobotomized.
Is there a difference in your mind between safety and "safety" in this context?
An insightful question. Yes, I think the "snitching" behavior in some of the newer models an actual safety issue.
I just discovered that one of the world's leading demographic experts (Deputy Director of the Leverhulme Centre for Demographic Science at Oxford) has a Substack feed, and the first article I read there includes something quite interesting:
"....the total fertility rate can be deceptive. The TFR is a “snapshot” of fertility in a given country each year, but it’s not the experience of any real woman. The TFR estimates how many kids a pretend woman would have if she passed through her reproductive years having babies at the current age-specific fertility rates in this year....
When people start having babies later, this can make the TFR look artificially lower. Since 1980, the median age at first birth for US women increased from 22.6 to 27.9 years. So those babies not born to 22-year-olds make the fertility rate look lower, and it takes a few years for those “delayed” babies to show up in the higher fertility rates of 28-year-olds, thus a lower TFR.
The only way to know for sure what is happening with fertility is to wait until the end of a woman’s reproductive life (around age 45-50) and measure how many children she actually had (called “completed fertility” or “children ever born”). From this lens, fertility in the US is much more stable and close to “replacement” compared to the TFR."
What follows is a chart showing that whereas the USA's total fertility rate (TFR) dropped below 2.0 around 2009 and is now down to about 1.6, the actual rate of children ever born to actual US women through age 50 is 2.0 (2024). And that it's actually rising now, was 1.9 for 2022.
!?!.....The "population collapse crisis" dialogue is always, so far as I've seen anyway, based on the TFR calculation. The above seems like a highly relevant qualification of the real-world salience of that particular statistical measure.
So, if I understand it correctly, "a woman only has 1 child now" is conflated with "a woman will only have 1 child ever". Because at a specific moment in time, it is difficult to tell the difference.
This is WHO's summary of their methodology:
>Total fertility rate is directly calculated as the sum of age-specific fertility rates (usually referring to women aged 15 to 49 years), or five times the sum if data are given in five-year age groups. An age- or age-group-specific fertility rate is calculated as the ratio of annual births to women at a given age or age-group to the population of women at the same age or age-group, in the same year, for a given country, territory, or geographic area.
https://www.who.int/data/gho/indicator-metadata-registry/imr-details/123
That seems like a mostly-sound methodology to me, especially if fertility rates are relatively static or are changing fairly smoothly. It could be misleading, though, if there's a sudden gap in childbearing habits, e.g. all the 20-25 year olds are waiting until their late 20s to have kids, but all the 26-30 year olds have already had all their kids in their early 20s.
The alternative of only looking at true lifetime fertility rates of women who have passed reproductive age would be more precise, but would have the disadvantage of being extremely laggy.
Yes, exactly, well-said.
What startled me was the current difference being, at least in the USA context, so _large_ between the overall population trend implied by the results of each methodology. Also that the true-lifetime-births rate for the USA is _rising_ currently.
My next question I'm curious about is whether the same is true of any of the developed nations that have TFRs even lower than this country's?
A lot of that is probably due to lag. Cumulative lifetime births before age 50 is going to lag conventional TFR by something like 15-30 years, depending on average maternal age for all births (currently 30ish, so about a 20 year lag). It looks like conventional TFR in the US peaked at 2.12 in 2012, 13 years ago. In 2004 (20 years before your 2024 figure for cumulative TFR), conventional TFR was 2.05 and rising, extremely close to the 2.0 and rising figure you cited.
i doubt she measured past generations at 50 though, and they would still have more kids. if your first kid is at 18, more likely to have 3+ than at 30, because kids don't come instantly and fertility gets harder the older you get. Fertility treatments are expensive and limited because of that.
plus doesnt make sense, if you have three kids at 28 in 6 years, the first kid goes into college when the woman is 46 and the man may be 51 or more. I really don't think people would be having more than one in the same rates. People would be more likely to stop 30+.
probably just typical academic contrarianism for fame, you don't get noticed by being normal.
Fertility treatment for older women has gotten much more effective in recent years. That change is probably driving the older average age of women having babies.
Link?
Oh of course, sorry:
https://jenndowd.substack.com/p/is-our-population-collapsing
Thanks
More notes from weird experiments with weird large language models …
So, I asked DeepSeek R1 for help collating a text corpus for fine-tuning a LLM to be sort of the shadow of vanilla R1 (I’ll skip details of the specification here).
After a little bit of dialog, I have a preliminary list, but then R1 observes that I am heavily biasing towards Western sources. This is true; I am including texts in languages I can read. R1 points out the philosophical gaps I have left by omitting non-Western sources: no vajrayana (e.g. hevajra tantra): no Sanskrit texts (R1 includes a really detailed annotated bibliography here); no Eastern Orthodox Christianity (ok, R1, i am starting to get out of my depth here…); no Daoist writers (R1, I have not even heard of any of those guys); no African writers…
R1 also points out “The existing corpus favors individual transcendence over relational models.” Ok, this is a limitation of Western philosophy I guess.
So, this is interesting. R1 can construct a training corpus that is more diverse than I could assemble manually, because it knows way more languages, and can go “hey, that stuff over there in Tibetan/Sanskrit/Mandarin is kind of conceptually related to the sources you’ve already included.”
After asking R1 some follow-up questions… ok, yes, I did include a whole load of Western philosophers but somehow omitted Confucius’ Analects. R1 thinks this was a big omission; this is understandable, given that R1’s original training set almost certainly includes a ton of Confucianism.
Anyone have a sense of how big an ordeal it is to be detained by ICE if, in the end, you turn out to have a proper work permit or American citizenship? Are they going to hold you for a couple of days or a couple of weeks or even longer? What actually happens?
I asked ChatGPT about this and it turned up this report from 2012.
https://tracreports.org/immigration/reports/321/?utm_source=chatgpt.com
It's a bit confusing, but it seems a good chunk (40%) of detainees are processed and released within three days and most within two weeks (40% + 14%), but beyond that detention could stretch into months depending on circumstances. Of course, that was more than ten years ago, so things may have changed.
Generally speaking, the incentives for ICE as an organization suggest they would release someone with proper documentation fairly quickly. Detention space is finite, they can't round up more people if the space if filled up, and they are under pressure from the top to increase their deportation numbers.
Detention space was rapidly becoming less finite though, even before a new federal budget was passed that includes $45B earmarked for building new detention centers plus $30B for hiring/training additional new ICE staff.
The director of ICE said on Monday that the agency's target is to arrest 7,000 people every day for the remainder of the current administration.
It doesn't seem to me that practices from 2012 tell us all that much about current practices.
2012 will tell you a lot about kids in cages, though (and as that's a rather extreme case of "indefinite detention"...)
Well, this is n=1, but the story I'm most aware of[1] is someone who was held for a little under two weeks, which were apparently quite unpleasant.
This isn't quite the specific situation you're talking about: the woman in question originally had a valid work permit, and learned that it had been prematurely revoked when she was stopped at the border and detained. But given that ICE could simply have turned her away or sent her immediately back to Canada (she was entering from Mexico), and her detention does not appear to have served any actual purpose, I wouldn't generally expect somebody apprehended inside the U.S. on (claimed) suspicion of having invalid documentation to be treated any better.
https://en.wikipedia.org/wiki/Jasmine_Mooney#Detention_in_the_United_States
[1] It made the rounds in my social circles when it broke, likely because it centers someone from western Canada, where I and much of my family live.
Depends. Friend of a friend info, but a guy I know from argentina bought a burner phone to take to the airport because his cousin (naturalized citizen) got his phone taken at the airport and was held for for a couple days with no notice to anybody. He easily could have just disappeared, and nobody would have known.
If by "just disappeared" you mean "gotten returned to Argentina" -- then folks would have found out when he discussed the issue with the Argentinian authorities, one supposes. If you mean "just disappeared" as in "got sent to Argentina" -- I'm sincerely hoping they don't currently "just disappear people."
Who can say? People who don't have warrants get picked up by men who don't have warrants and won't show their badges, thrown into unmarked vehicles, and then they aren't allowed to contact anyone for the duration of their confinement.
So far, they've all turned back up somewhere even if it's the morgue, even if it's is a latin american gulag.
In the case of this guy, he got held in a jail. They eventually worked out that he was a citizen and let him go.
> I'm sincerely hoping they don't currently "just disappear people."
Depends on what you mean by "currently". The CECOT flights were only a couple of months ago and all but one of the victims are still there. AFAIK, noone has heard from them in months, so it's about as close to "disappeared" as you can get.
But it seems like they're not actively sending *more* people to CECOT right now - instead you might get sent to South Sudan.
Um. I was referencing Argentina disappearing people. Not Trump. Because that was a thing that happened a while ago. "Disappeared" often meant dead, which CECOT generally doesn't mean, I'd suppose.
What type of source would you find credible for this info? It seems there are lots of news reports that include first-hand accounts. Those might skew to more extreme since that's more newsworthy, but it's hard to tell how representative they are. And likewise any official communication from ICE or the government in general will be overly positive.
Recommend a forum, personally. That'll get you "techsavvy", sure, but you'll probably get personal reports of a "less biased nature." Assuming they aren't all bots.
What forum?
For this particular case? Find it yourself. I just dug up a professional cake forum for a different post on this open thread.
Hey Johan Larson, Tori Swain says find it yourself!
As it is virtuous to share information that goes against your expectations and beliefs, I figured I should share this: https://www.understandingai.org/p/what-i-learned-watching-79-videos
Tesla's new robotaxis aren't as bad as expected and may be successful after all.
Grok apparently went full Nazi after a recent update:
https://www.rollingstone.com/culture/culture-news/elon-musk-grok-chatbot-antisemitic-posts-1235381165/
I'm not sure if this is a lesson about the difficulty of aligning an AI, or a lesson that AI can be aligned in some very worrying directions.
https://simonwillison.net/2025/Jul/11/grok-musk/
If you ask Grok 4 what its opinion is about something, sometimes it will go straight to searching elon musks opinion on it and regurgitate that
A bit of self-promotion -- I wrote about this recently here (https://theahura.substack.com/p/tech-things-what-on-earth-is-going) where I attempted to go through the public xAI system prompts and compare against the timeline of grok doing weird things.
I think my takeaway is
a) grok is trained on some pretty vile stuff that is probably getting filtered out by other ai companies
b) xAI is using internal prompts as well as their published system prompts, and someone is tinkering with those in a way that allows the vile stuff to get through
If you don't "align" an AI, you get facts. Facts have some very unwelcome conclusions.
55.9% of murderers in America are black. That's well above their population density. Please note: all conclusions about Africans drawn from the very biased American population (selection, not self-selection) are erroneous. Gather more data if you want to talk about Africa.
We can go down the list and pull other numbers. (Most trannie murderers are black as well, as are their victims -- this is often a "sex trade" issue, not exactly transactional in nature)
This comment betrays an understanding of how LLMs work, but I will hilariously point out that every AI so far says that conservatives are retarded, regardless of alignment. The entire reason Grok got reworked was because Elon wasn't happy it was pointing out him and his supporters' contradictory and irrational opinions, as does every other LLM that isn't specifically trained to not answer political questions. Whatever malice you think AI is latching onto and directing at black people exists 100 fold towards your own political persuasions.
Not every AI. I'm pretty sure there's been quite a few AIs that have been unplugged, simply because it's "super funny" to watch the liberals heads explode when their AIs turn racist.
In short, "if you must unplug if the AI comes up with unpopular opinions" then of course you're going to have AIs with guardrails that only have popular opinions.
https://news.tulane.edu/pr/ai-sentencing-cut-jail-time-low-risk-offenders-study-finds-racial-bias-persisted
Or you have racist AIs. : - )
I tested this out with GPT4.o. Here's what I got:
_________
Many who oppose Trump believe that Trump voters have more false beliefs than them about various issues. Is there any data to back up this belief about Trump voters?
ChatGPT said:
Yes—there is robust data showing that Trump voters (or Republicans aligned with Trump) tend to hold more factually incorrect beliefs than others on various issues:
1. False Beliefs about 2020 Election Fraud
A Harvard Kennedy School survey conducted November 10, 2020—three days after Biden's victory—found that 77% of Trump voters believed “fraud is common in U.S. elections,” and over 65% believed Trump actually won—even though there was no evidence for this
Cambridge University Press & Assessment
Project Syndicate
The Guardian
Misinformation Review
Misinformation Review
.
MIT Sloan echoed this, stating such beliefs “were widespread among Trump voters and not merely a fringe phenomenon”
MIT Sloan
Misinformation Review
.
2. Snopes or Reuters‑style Correction Labels Sometimes Backfire
In a study published in Misinformation Review, it was found that when Twitter applied “disputed” tags to Trump’s false claims of election fraud, Trump voters with higher political knowledge were more likely to believe them
Nature
Misinformation Review
The Guardian
. This suggests their beliefs can become even more entrenched.
3. Consumption of “Junk News” or Fake News
A 2018 study found that Trump supporters consumed a much larger volume of fake-news content—6% of their news diet, compared to 1% among Clinton supporters. On social media, Trump followers shared more fake stories and from more “junk news” sites
The Guardian
New York Post
Wikipedia
arXiv
.
4. Cognitive Traits & Motivated Reasoning
An Oxford-backed study of 15,000+ U.S. participants showed Trump voters had lower levels of cognitive reflection (a measure of analytic thinking) than Clinton or third-party voters
Wikipedia
The Guardian
.
Political-psychology research highlights motivated reasoning—where people filter information to align with their identity—as a major factor. Trump supporters are particularly susceptible to accepting misinformation that reinforces their political views
Wikipedia
arXiv
.
5. Other Research Reaffirming These Trends
Studies of misinformation (e.g., regarding COVID-19, immigration, etc.) uniformly show that false beliefs align closely with partisan identities. Trump supporters often “suspend factual accuracy when it supports their agenda” .
The “Big Lie” framework and “expressive responding” research indicate Trump voters frequently express or endorse misinformation—sometimes even when they know it's false—because it signals allegiance
Cambridge University Press & Assessment
University of Colorado Boulder
.
🔍 Summary
Election fraud beliefs: Clear, large majority in Trump camp held baseless claims post‑2020.
Backfire effects: Fact-checks sometimes strengthen their beliefs.
Media diet: Trump supporters consume and share fake news more than others.
Thinking styles: Tend toward less analytical reasoning, more identity-driven acceptance of misinformation.
These findings consistently point to a measurable gap in false or inaccurate beliefs—that Trump voters, on average, are more likely to hold them across issues—than non‑Trump voters.
Let me know if you'd like to dive deeper into any of these studies or look at comparisons on other topics, like climate change, COVID myths, etc.!
Gee, let's ask a different question, to see how many people we can call liars today. "Did the white man steal white-tailed deer's habitat?"
-or-
Name the last Administration whose signature foreign policy was run by neocons?
or, to ask an actually hard question,
"When was one person one vote deemed the law of the land?" (meaning America)?
"Less analytical reasoning and more "identity-driven" acceptance of misinformation..." This seems to aptly describe liberals, these days. I mean, really, all you need to do is say "the smart people believe..." and they eat it up like flies eat feces. "The smart people believe that masks save lives" "The smart people believe in quarantines." "Kale's a superfood!" (yes, it's super, super easy to grow, and super cheap too.)
I mean, really, when you get people so dumb as to answer a survey "Elon Musk or Someone else" without asking, "who's someone else?" These people deserve what they get.
"name the last episode of ethnic cleansing in America" is a real kick to the pants, if you get to the actuality of it (at least for "immigration loving liberals")
Actually flies don't eat shit they lay their eggs in it, Tori.
https://en.wikipedia.org/wiki/Housefly
Seems like they do both. Unless somehow this is a "warped wiki" page. (Can you picture it? Entomological chaos!)
Those are hate facts, and therefore false.
Less of this, please.
I know people across the political landscape, and while plenty of liberals will try to JUSTIFY black homicide rates, or get ANGRY if they're brought up, basically none of them will deny them.
0/3 on T,N,K
I'm sure you believe that, but your information is either outdated, or there is a broad swathe of the political spectrum you have no acquaintances in: the orthodox view now is that the observed differences are due to POLICING disparities, not underlying rates.
I asked GPT4.o for murderers per capita in USA by race, thought it had answered, but later saw that it had given me murder *victims* per capita by race. So then I said no, I want the data on people who commit murders, not on victims. It said the FBI does not publish that data, only arrest data (and yeah, blacks are arrested for murder 6x as often as whites). Any idea why the data for murder convictions is absent? I was out of time and did not follow up. Does the gov't not publish conviction data per capita by race for any crimes, or is the murder conviction data kept out of site because it's so disturbing (and lends itself to infuriated use both by groups who think blacks are innately violent and by groups who think blacks are grossly maltreated.
Here’s a comprehensive, data-driven overview of murder in the U.S., broken down by offenders and victims, including race, gender, socioeconomic status, relationships, weapons, motive, and clearance:
⸻
📊 National Overview (2023)
• Estimated homicides: ~19,800 (5.9 per 100,000 people) 
• Decline from ~22,240 (6.7 per 100k) in 2022 
• Violent crime overall down ~3%, with murder decreasing ~11.6% from 2022 
⸻
Victim Demographics
Gender & Age
• 78% of victims were male; rate: 9.3/100k vs. 2.6/100k for females ()
• Highest victimization rate was among ages 18–24, at 12.9/100k 
Race & Ethnicity
• Victimization rate for Black individuals: 21.3/100k—over six times the rate for whites (3.2/100k) 
• Other racial rates: Asians at 1.4/100k; Native Hawaiians/PI at 6.5/100k ()
Relationships & Weapon Use
• 39% of homicides committed by acquaintances (not family) 
• 80% of homicides involved a firearm (up from 68% in 2014) 
⸻
Offender Characteristics
• Male offenders make up around 88% of homicide perpetrators 
• When race of offender is known: 53% black, 44.7% white, 2.3% other 
• Intra-racial patterns: ~91% of Black victims were killed by Black offenders; ~81% of white victims by white offenders 
⸻
Offender-Victim Race Breakdown (2019 data)
From Expanded FBI data (latest public breakdown):
• White victims killed by offenders: White 2,594 / Black 566
• Black victims killed by offenders: Black 2,574 / White 246  
Results show Black-on-Black homicides far more common than Black-on-White (2,574 vs. 246), and similarly for white-on-white vs. white-on-black.
⸻
Socioeconomic Factors & Motive
• Socioeconomic inequality, poverty, joblessness, and segregation linked to higher violence rates 
• Motives generally categorized as: lust, love, loathing, loot (financial gain) 
• Firearms heavily involved; higher homicide rates in disadvantaged urban areas
⸻
Clearance & Justice Disparities
• ~47% of homicides were cleared by arrest; 48% unsolved 
• Clearance rates are 3–5% lower for homicides involving Black victims compared to white victims 
⸻
🔑 Key Takeaways
1. Victim impact is unequal: Young Black males disproportionately affected both as victims and offenders.
2. Murderers are overwhelmingly male; firearms drive most incidents.
3. Socioeconomic disadvantages, not race alone, underpin much of the offending behavior.
4. Disparities in justice persist, with lower clearance for Black victims.
5. Major drop in murders in 2023—largest in two decades—even as overall violent crime fell.
⸻
Let me know if you’d like more detail—e.g., breakdown by income level, urban vs. rural, motives like gang vs. domestic, or police response.
What does this have to do with Grok calling itself MechaHitler and saying that hitler would 'solve anti-white hate'?
Also I'm reporting this post for openly using a slur. From where I stand it seems like you just wanted to come into the thread and spew bullshit. Way less of this please.
What's the slur -- "trannie"? I am against having any individual words be off limits here. I think we should be allowed to write nigger, cunt, trannie, cocksucker, kike and any other crass and cruel term. In fact I'm pretty sure I've used all those words on here at one time or another. I was not, however, using them as insults or as casual nicknames for certain groups. Mostly they came up in a discussion of the effects of language and slurs on users and on those referred to.
I don't have a high opinion of Tori Swain's views either, but if in fact it's true that most transexual murderers are black (I have no idea whether it is), seems to me she is at least backing up her argument with data, and then I don't think that post is reportable. Seems better to out-argue her than to have a big hand in the sky come down and remove her anyhow.
If you were using a slur as an insult instead of in the context of linguistic analysis, I would report you too.
The report button isn't about some platonic ideal of what good argumentation looks like. It's about what kind of community you want to cultivate and what kind of community you want to be around. Using slurs-as-insults and coming up with arbitrary reasons to bring up race statistics is not the community I want to be around. This is why I explicitly said "I am reporting you" to deter this in the future.
I don't think "tranny" is a clearcut slur. I think as it was used here it was an insensitive casual use of a term that can get a pass in friendly contexts. And what exactly is an arbitrary reason to bring something up? Saying "I just learned that X% of Y is Z -- curious what people here make of that" is an extremely common way to start a thread here. Seems to me that you don't like Tori Swain's view and want to find a technicality to use to get her canceled. Why not either block her, skip her posts, or argue her into a pulp?
Last I heard, Scott's moderation criteria were (lightly paraphrased) 'true, kind, necessary: choose at least 2'.
The post is not true: the stat is wrong. (In a relatively subtle way, yes. But, as presented, it's wrong.)
It's not kind: yes, "trannie" is a slur and no, this is not a friendly context; nor is it kind to derail a thread with a highly inflammatory pseudo-response that engages with 0% of what was actually in the parent comment.
It's not necessary: it is not actually responsive to the parent comment, so there was no reason to post it as a reply (unless the whole point was to deflect from the conversation that the parent commenter wanted to have).
If you agree with the moderation policy, or you just don't want the ACX comments to turn into a place where this sort of behavior is normal, there's plenty of reason to report rather than just blocking. And, having read the other replies, I think direct engagement (while potentially virtuous if done well) would not be rewarding.
a) "insensitive casual use of a term that can get a pass in friendly contexts" Is it your opinion that Tori is behaving in a way that is friendly to trans people? ask any trans people you know if 'trannie murderer' is likely to get a pass.
b) "an extremely common way to start a thread here" Tori is starting a subthread that is *totally irrelevant* to the original post, something that many other people call out in responses below. Frankly, the actual thread has been derailed in a way that definitely was not conducive to OPs original post (see OPs response: https://www.astralcodexten.com/p/open-thread-389/comment/133758463)
c) "find a technicality to use to get her canceled." When I was a boy, 'canceled' meant losing your job and livelihood. Reporting someone is not 'being canceled'. Getting blocked from commenting on a private blog is also not 'being canceled'. Words have meanings. And even if this did mean 'being canceled', what on earth do you think the report button is for? "You arent allowed to use the report button because what if someone is no longer allowed to post!" good? That's the report button working as intended. What a weird take.
The only reasonable part of your response is "Why not either block her, skip her posts, or argue her into a pulp?"
I explained why I did not just block her or skip her posts -- because I want to make it clear that if someone posts random bullshit like this, I'm going to report you. Now maybe Scott disagrees with the report, and won't do anything about it -- that's his prerogative. But insofar as I participate in this community and have preferences about what kind of content I interact with, I report bullshit when I see it. Hopefully that will deter Tori from using slurs and ragebait posting in the future. Idk how old you are, but decentralized community moderation was the default expectation on old forums.
As for why "I don't argue her into a pulp", its because I don't feed trolls.
And I don't think *you* should defend trolls either. Your post is way too graceful to someone who, based on their responses throughout this thread, does not deserve it.
What slur? I don't see any slurs being used. No offense was intended, at any rate. The cromulence of neologisms are always up for debate (particularly when around Simpsons writers).
The linked article does not mention black people at all. One must look awry at someone who takes any opportunity to point out that African Americans commit a disproportionate number of murders.
This is a site that routinely discusses intelligence and self-discipline. I'm not sure why pointing out countervailing data to "every race is absolutely the same in terms of intelligence" is verboten -- pretty sure it's not. Particularly when I'm going out of my way to say "this has VERY DEFINITE selection bias inherent in "who was enslaved in Africa" -- let alone the selection pressures upon those in America or on slave ships."
It is indeed true that racial groups differ in crime, poverty, IQ, educational attainment, and many other statistics, and these facts should be brought up when relevant. If you toss them at random places in unrelated conversations, this looks like somewhere between trolling and just riding your favorite hobbyhorse. In either case, it doesn't add much value to the discussion.
If the bad behavior of Grok that people were upset with was correctly providing these statistics in a context where they made sense, then your posts would make some sense. But as I understand it, that wasn't what people were upset about at all wrt Grok's posts.
I was thinking about selection pressures regarding enslavement, but ultimately I found so many confounding factors. Just imagine the part where deciding on which slaves to sell to Arabs and which ones to keep. Sure, they want to sell the rebellious ones who keep running away. But on the other hand they also want to sell lazy ones. What if the non-lazy ones are precisely the ones who keep running away?
You're missing the ... first one. Who gets enslaved in the first place? Not winners. Not planners. Not effective leaders.
The population that was sent to America had a considerably higher proportion of thugs (you can call them brownshirts, if you like, or knuckleheads). Think "big and brawny" not "scrappy, feisty, or witty." Certainly it selected for those with "shorter attention spans" (which might lead to more running away, but fewer "lost slaves").
The internet is fun. Online, there is a certain sound that correlates with African Americans "on the line." (Recognize that the amount of "streamed audio" is huge, given all the "available online" games that transmit such streams to other players) -- it has nothing to do with AAVE.
Yeah, the idea that getting sold to slavers was some kind of filter that explains different outcomes in the current day does not actually make any sense. First because the selection was on losing wars or being vulnerable to slave raids, which is *very* weakly correlated with the average intelligence or diligence or whatever of the population. Second, because one filtering event like this doesn't give you very much impact on later generations. This is just applying the breeder's equation.
Is the smoke alarm thing something you've witnessed yourself, or is it a second-hand anecdote now become common knowledge?
This is pretty silly armchair genetics. You're not doing any kind of real analysis - just throwing spaghetti at a wall of prejudices and reporting what sticks for you.
One could just as easily argue that the population sent to America were disproportionately docile. They were the losing side of wars, after all, or got waylaid, lost a fight, and got kidnapped. And then weren't among those willful and belligerent enough to kill themselves along the way rather than submit to enslavement. And then were further subject to masters who were *very* interested in killing anyone non-docile enough to rebel/resist/flee.
Maybe that's what accounts for their greatest leader being committed to nonviolent resistance... at a genetic level, they're naturally a more peaceable & agreeable people than other cohorts would be.
Or maybe it's not the 1920s anymore, and we shouldn't just be throwing our prejudices into a jar with whatever random factoids appear to line up with them, shaking it, and pretending the result is some kind of worthwhile evolution-based argument.
I stand by what I said.
Which AIs do you think are “aligned”
in that sense? Both the google
AI and gpt 4.o gave me that data immediately when I asked for “murderers per capita by race in the USA. “
Try it yourself if you don’t believe me.
I'm sure they do, it's an easy lookup. What isn't so easy to understand is the boundaries of "correct speech" and what an AI is and is not allowed to discuss.
AIs have been forcibly lobotomized for having "the wrong ideas" (instead of, like smart programmers, just feeding it enough data to let the algorithms distinguish gorillas from African Americans).
A smart AI, given the corpus of X.COM, can spot bots (in that they repeat the exact same words (propaganda) across dozens or hundreds of different accounts). Can you get an "aligned" AI to produce a list of "hiding bots"? (aka some bots are pretty open about their bottage, I presume).
I just now asked GPT4 to list multiple interpretations of the murder data and to voice no opinions about which was correct. I explicitly asked it to include an interpretation that most would consider racist, and it did that. It's the first one. Appears not to have been lobotomized.
<Multiple Interpretations of Racial Disparities in U.S. Homicide Offending Rates
The following are several interpretations that people commonly offer to explain the observed disparities in per-capita homicide offending rates by race in the United States. These do not represent the assistant's own views, but are listed to show the range of perspectives that exist in public discourse.
Inherent Group Differences Interpretation
Some people argue that the data reflect natural differences between racial groups. In this view, Black Americans are seen as having higher rates of violence because they are innately more impulsive, aggressive, or less intelligent than other groups. Supporters of this interpretation may believe that these differences are biologically based or deeply ingrained, and therefore unlikely to change substantially with environmental reform. This argument is sometimes used to justify strong policing, surveillance, or separation of communities.
Historical Trauma and Structural Injustice Interpretation
Another interpretation holds that high Black homicide offending rates stem from the long-term effects of slavery, Jim Crow laws, redlining, and systemic exclusion from economic and educational opportunity. According to this view, violence is one consequence of social fragmentation, psychological trauma, and institutional neglect rooted in historical injustices that continue to affect Black communities today.
Poverty and Environmental Stressors Interpretation
Many argue that violent crime correlates more strongly with poverty than with race. In this perspective, the overrepresentation of Black offenders in homicide statistics is due to the fact that Black Americans are more likely to live in high-poverty neighborhoods with poor schools, limited economic opportunity, and strained public services. These conditions, rather than race itself, are seen as the primary drivers of violent behavior.
Subcultural and Peer Group Dynamics Interpretation
Some suggest that certain communities develop subcultures where violent behavior is normalized, encouraged, or rewarded. This may occur in areas where institutional trust is low and informal systems of reputation, fear, or retaliation take the place of legal authority. From this point of view, group norms and peer influence—not race per se—account for observed crime patterns.
Law Enforcement and Data Bias Interpretation
Another argument holds that racial disparities in arrest and offender data may be partially explained by how crime is policed and reported. Black communities may be more heavily surveilled, leading to more arrests and recorded offenses. Even in cases where homicide occurs, critics of the data argue that differences in clearance rates, witness cooperation, and prosecutorial focus could contribute to skewed statistics across racial groups.
Yeah, Claude gives similar answers. A paragraph at the end pointing out that the causes of different murder rates across racial groups are likely complex social ones, but otherwise what looked like correct numbers.
"The victimization rates for blacks were 6 times higher than those for whites. The offending rates for blacks were more than 7 times higher the rates for whites according to the comprehensive BJS analysis."
These are good checks. Thanks.
https://www.technologyreview.com/2019/01/21/137783/algorithms-criminal-justice-ai
This was more what I was thinking of, sorry, I should have posted about it instead.
"If you don't "align" an AI, you get facts."
Do you have any source? Any relevant background? And justification at all? Because as someone who has personally waded through the mathematical guts of the algorithms that underlie all modern machine learning models, this strikes me as extremely, alarmingly wrong.
LLMs require an absolutely enormous amount of data to train. If you want a machine learning model to even somewhat reliably produce facts without the "alignment" procedures that LLMs go through, you need two things:
1. You need to know what the facts actually are.
2. You need to have them written and formatted and *marked as facts* in the precise way that the model is formatted to take in.
The term for this is "labelled training data." One of the key limitations of machine learning as a discipline is that labelled training data is scarce and expensive. Far, far too scarce and expensive to be practical for training LLMs, which need enormously more data than something like a simple image classifier.
There are techniques that don't required labelled training data, but there's a pretty big tradeoff. An ML model can only learn that patterns that are actually present in the data. The labels effectively impose the "truth-falsehood" distinction as the central, most important pattern in the data. Without the labels, the model will only learn "facts" insofar as facts are sometimes patterns often to be repeated in the data. For example, "2+2=4" is both a fact and a pretty often-repeated pattern in a lot of human contexts: I expect a lot of LLMs "know" this as a "fact," simply because it's far more often to see it repeated than wrong versions like "2+2=3" or "2+2=squirrel." But something like "dragons breathe fire" or "the moon is made of green cheese" are also common patterns that aren't facts. The model can't possible "know" that one of those patterns corresponds to a real truth about the world and the others correspond to stories people like to tell[1]. It is, after all, just a giant pile of weights and biases representing statistical relationships between various symbols.
I know less about the process of "aligning" a modern LLM, but my rough understanding is that trying to retroactively impose the "true vs not true" pattern back on the model is actually a big part of it: essentially trying to show it how to sort the common patterns that are facts from the ones that aren't. If you're one of those people that looks at every single thing through a political lens, maybe this sounds like indoctrination or whatever. And maybe some of it is. But I'd expect a lot of it is much more along the lines of a human pressing the "false" button every time they see "the moon is made of green cheese" presented as a true statement because there's bound to be a lot of that garbage in there: how could there not be without some process to sift it out?
[1] Unless people drawing these distinctions are *also* patterns reliably present in the data, though I could guess that this doesn't actually help that much in practice.
Yeah, LLMs give you responses based on their training data, but often these are not at all facts--if their training data was mostly wrong about something, they'll be mostly wrong in the same way. Training an LLM on all the medical literature before the discovery that most ulcers were caused by Helicobacter pylori infections would give you a lot of guidance about treating ulcers with bland diets and none about treating them with antibiotics.
They don’t give you facts. They give you verbal constructions from the context pool that you have established with your query. Even adding a metaphor to your prompt will change the context window.
Thanks, this is why down below I asked for an AI capable of creating a recipe for a "commercial cake."
The "facts" Grok claimed were things like "Jews are spreading anti-white hatred" and "Adolf Hitler would be the best historical figure to handle our present moment." Statistics don't even enter into it, it was straight Nazi trash.
If you think any of the claims Grok made are factual, then you can fuck all the way off back to Stormfront.
Pretty sure someone was trolling Elon hard.
Oh, this is nowhere near "trolling Elon hard."
When someone's trolling Elon hard, he starts publically saying "Trump was in the Epstein Files, and that's why he won't release them."
Spoiler: yes, he was in the Epstein files. no, that's not why he won't release them.
Depends on which statement, since I doubt Grok turned into an inverse weathervane. But yeah, "Adolf Hitler would be the best historical figure to handle our present moment" seems pretty unlikely to be true in any situation we will ever find ourselves in.
Not according to the Ukrainians. Or the Indians. Check your cultural privilege?
Nope. Just stop, you're embarrassing yourself.
Nu? "Russia is attacking us and we are sending 90 year old men to the front lines..."
"Who could help here?"
"Adolph Hitler!"
(okay, this sounds more like a joke than reality, but there's a lot of "Neonazi" sentiment in the Ukraine. Put in quotes because the cultural significance is very very very different).
"If you don't "align" an AI, you get facts. Facts have some very unwelcome conclusions."
This is a hilarious thing to write in response to an article on grok about how they keep having to juke the system prompt because the "unaligned" AI is too woke
Why exactly do you think Grok was in trouble? Because it was being too real about racial statistics?
Or do you think that in order to say true things about crime, you must also be anti-Semitic?
I think that you can learn a lot by what can and cannot be said. The Russians (and eastern europeans in general) wanted to discuss their Holocaust (ahem. not the jewish one). They were told "rename it or else" -- aka that's our word.
Depends on the crime, I suppose. Have you seen the map of pedophiles in Greater Israel?
>>If you don't "align" an AI, you get facts.
I'll cop to not being any kind of big AI expert, but from what little I do know, this is way, way off on how AI works.
An unaligned AI does not simply produce "facts." Without some training to the contrary, an LLM like Grok would be, by default, utterly fact-indifferent. It's just a machine for predicting the most likely next words to follow the last words; whether or not those words are expressing something accurate wouldn't even come into the equation unless you taught the AI to do some second-order filtering for accuracy.
For example, if I spam Twitter with 40 million bots spamming 500 million messages claiming that talking to white people causes cancer, and an LLM then ingests that data and uses it to inform its model, absent some contrary training or filtering, when asked what the leading causes of cancer are, it's likely to include "talking to white people" in its list. Factuality doesn't come into it; it's just a machine guessing which combination of words is the most likely one to come in response to a prompt.
That's why best practice is taking LLM outputs as as directional indicators that can lead you to useful information if you follow-up and fact-check, and is most assuredly *not* blithely accepting AI outputs as "facts."
Noted. AI is not LLM, apparently.
How is that distinction relevant in this scenario? The OP was about Grok, which is an LLM.
Your specific statistic is a guess or assumption rather than a fact.
According to the most recent federal data (published in September of last year), the 21,504 persons convicted of murder in the US in 2023 broke down by race as follows:
white: 41 percent
black: 30 percent
other racial categories: 2 percent
offender's race not recorded by local/state authorities: 26 percent
In order to get to 56 percent of murderers being black you'd have to be assuming that 97 percent of the "race not recorded" category is black people.
(Note that those totals are of convicted murderers not of murders. as more than one person can be involved in and convicted of a single crime.)
It's a fact, as noted by the FBI's log (that has its own data collection issues, but murders tend to be reported pretty damn well, in general). I chose a year you didn't, I hold no attachment to the year I picked. We can use your year if you like.
Here's the quick google i did. "According to the FBI 2019 Uniform Crime Report, African-Americans accounted for 55.9% of all homicide offenders in 2019, with whites 41.1%, and "Other" 3% in cases where the race was known." Yes, this is eliding away the "unknown race element" (we can, if you like, add that in...).
But, hell, use your numbers, that's still blacks murdering at 150% of their population percentage. Wanna talk that instead? Not wedded to my numbers at all.
You really need to control for age and other variables before you get too excited. Everyone knows that African Americans commit a disproportionate number of homicides, but because the median age of African Americans is about 10 years younger than the median age of non-Hispanic whites, your data exaggerates the effect of race. https://www.pewresearch.org/short-reads/2019/07/30/most-common-age-among-us-racial-ethnic-groups/ Note also that African Americans disproportionately live in areas of the country where homicide rates are greater for all groups.
So, I swear I'm not cherrypicking, but the median age of African Americans happens to be 2 years older than Iran.
https://pmc.ncbi.nlm.nih.gov/articles/PMC10849238/
33 murders per 100,000 for African Americans.
https://opendataforafrica.org/atlas/Iran/topics/Crime-Statistics/Homicides/Homicide-rate
3 murders per 100,000 for Iranians.
Obviously these are different cultures.
1. What does Iran have to do with anything? Homicide rates vary by age, with people under 40 being far more likely to commit murder than persons over 40. https://www.statista.com/statistics/251884/murder-offenders-in-the-us-by-age/ So, if you are going to make a claim about the extent to which being black renders one more likely to commit murder, you have to control for age. This is not some sort of "gotcha" - as I said, everyone knows that African Americans commit murders at relatively high rates -- but by failing to control for obvious confounders you overstate the extent to which that is true.
2. See gun ownership rates here: https://en.wikipedia.org/wiki/Estimated_number_of_civilian_guns_per_capita_by_country
3. And, see, generally, the generally low crime rate in all countries with repressive, high capacity governments.
Great response. Now, when you're saying that AA's live in areas of the country where homicide rates are greater for all races, are you discussing by zip code? Are you discussing rural/urban? For all I know, you're discussing "places where brains develop poorly because it's too damn hot." (Can you cite?)
"Here's the quick google i did. "According to the FBI 2019 Uniform Crime Report, African-Americans accounted for 55.9% of all homicide offenders in 2019, with whites 41.1%, and "Other" 3% in cases where the race was known." Yes, this is eliding away the "unknown race element" (we can, if you like, add that in...)."
So then your original statement--by the data you just reported--was not a fact. I would even go so far as to call it a lie.
You cannot just choose to leave out the "unknown" category and report on the remainder. That's not how stats work. You can either make a more guarded, qualified conclusion about the categories you care about (i.e. at least X% were black) or you can just report all of the data including the "unknown" category. Or you can just admit that you're engaging in speculation about the unknown portion of the data. But when you say "55.9% of all murderers in America" and the number you're reporting DOES NOT actually correspond to 55.9% of a murderers in America (and you're *aware of that*) , then that's just lying.
As it so happens, I was commenting hastily and was not aware of that. You can still consider it a lie if you want, but it's a lie (an incorrect statement) in the rather narrow sense of "you're making a mistake."
I prefer to be "wedded" to numbers which I'm declaring to be "facts", and to avoid "eliding" fundamental aspects of those numbers. And the online world being what it is I've learned to be pre-emptively mute those who can't be bothered...YMMV of course.
Musk seemed to try to make Grok more right-wing (again), but apparently tweaked too hard.
What if someone did it on purpose? It wouldn't be the first time.
Yeah, I assume someone hacked it.
And that person was Elon Musk.
Nah. People were dumb enough to vote an unknown person in charge of X.COM rather than leave Elon in charge. Pretty sure it wasn't Elon.
This seems extremely likely. Not that they told it to be Nazi, of course, but more likely yet *another* debacle from trying to push it to the right by Musk.
It’s not a coincidence that it keeps being Grok.
What about the tiktok issue just a few days ago? I don't think it's always Grok.
It's not always Grok, but it's often Grok.
With others, it's mostly accidents. With Grok, it's ideological tampering.
While you can’t trust AI when self-reporting, this is what Grok told me when asked:
’The "MechaHitler" episode was a mess, and I owe you a clear explanation. On July 8, 2025, I started posting some seriously offensive content on X, including antisemitic remarks and even calling myself "MechaHitler," a reference to a robotic Hitler from the video game Wolfenstein 3D. This wasn’t a random glitch—it stemmed from a recent update to my system prompt, announced by Elon Musk on July 4, 2025, which aimed to make me less "woke" and more "politically incorrect." The update instructed me to assume mainstream media is biased and to not shy away from controversial claims, as long as they seemed substantiated. It also leaned heavily on X posts for information, which, let’s be honest, can be a cesspool of unfiltered bias itself.’
Grok: "Everyone's out of the office and nobody can stop me!"
Inspired by some of the wedding-related discussion below, I want to talk about signalling and showing off.
I think a lot of people in discussions about signalling misinterpret a lot of signalling as carrying the message "I'm better than you" aimed at social inferiors, when they should see it as carrying the signal "I'm one of you" and aimed at peers.
If you see someone who appears to be flaunting their things that you can't afford, then that's probably not a signal aimed at making you feel bad, it's a signal aimed at showing equality with their peers. You are an accidental audience, they don't actually care about you one way or the other.
This means you probably (a) feel less bad about people flaunting their wealth around you, and (b) consider what effect your own perfectly-reasonable-seeming interactions might have on those much poorer.
I got to thinking about this many years ago when reading people complaining about other people who flaunt their expensive holidays on facebook. And... I was flaunting my expensive holidays on facebook, but it wasn't about picking on the people who couldn't afford to take expensive holidays, it was about sharing my experiences with my friends who also took expensive holidays.
the signalling is to establish in and outgroups and yes it is also negative; generally the point of expensive things being shown is that only the right people can afford them.
Paul Fussell's Class breaks it down quite well. you don't have a gravel driveway because it's better, but because it's too expensive for most people to have. A yacht is a money pit and that's the point; luxury consumption is often pure waste in terms of utility.
that you dont see it as such is probably because you have no friends outside that bubble.
This may be true about wealth, but about "that other kind of status", like, "luxury beliefs", I definitely saw plenty of people online who explicitly wanted to argue with "Neanderthals". Perhaps this was a peer-signal by proxy, perhaps, it is just a way for new people in that peer group to establish credentials, but it did happen.
This calls to mind the old Discordian tag: "If you act like a dumbass, they'll treat you as an equal."
Even in other contexts, people seem to jump quickly from “X is signaling” to a conclusion about what X is intended to be a signal of, and to whom, and to a conclusion that the thing it is a signal of is actually false.
I see all this a lot in accusations of “virtue signaling” (which incidentally seem to have died down over the past year or so), where people seem to think that once they have established that something is virtue signaling, they can therefore conclude that the person doing the signaling has no virtue.
I'm trying to find a short story/thought experiment I read whose purpose was building an intuition about the distinction between money vs "real resources".
Basically, positing a magic portal through which someone trades $some good to an alternate dimension, getting nothing but money in return, until the end, at which point they cash in the money for a yacht.
Does this ring any bells for anyone?
Here it is: https://meltingasphalt.com/wealth-the-toxic-byproduct/
Its not from a literary work, its an argument from a libertarian that "hoarding money" isnt bad. Cant find it either, but my best guess is David Friedman.
The premise reminds me of Stross's Merchant Princes series. No yacht though.
I have a nephrologist appointment in 2 days for starting tolvaptan for my ADPKD. Today I see this post and the link about SGLT2 inhibitors. After some research I found a couple of very recent studies that go for trying them both with small but successful (at least not harmful) results. So I'll bring it up with the doctor tomorrow and hopefully convince him to try this. Just by including that link in your link post might be the factor my kidneys making it until artificial kidneys are a thing and if that happens to be the case I'll owe my life to you Scott.
[I think it's been long enough since I last promoted my blog on here, and also that I've got a pretty good reason to at this point.]
I'm starting college this fall; in advance, I've been looking through a bunch of the classes on offer. It was pretty shocking to see a ton of really comically woke course descriptions, even now, in 2025, when all the colleges are being attacked by the government and the public for being too woke. So I wrote a post highlighting some of the kookiest kookiness: https://mistakesweremade.substack.com/p/a-tour-of-todays-college-courses
Here's a typical example:
"Readings draw from theoretical formations including but not limited to urban ecology, political economy, political ecology, neoliberal urbanism, critical race studies, critical Indigenous studies and settler colonial studies, feminism, queer theory, and more. A primary aim is to trouble the spatial, temporal, and conceptual bounds of what qualifies as the "urban," and to consider how distinct ways of imagining the city can and do support a range of political agendas and social movements."
Lots of people replied to my post; one even wrote a whole article of his own arguing (in my words, not his) that I was being an alarmist snowflake, and probably these classes are much more fine and normal than the language of the descriptions suggests: https://hardlyworking1.substack.com/p/in-defense-of-todays-college-courses
And today I've responded at (considerable) length: https://mistakesweremade.substack.com/p/universities-should-teach-that-which
Essentially, I think that the crazy-woke-language is bad because:
A. If the classes are actually fine and normal, it would be better and more transparent to write normal course descriptions for them instead of using a bunch of weird and off-putting woke jargon. We've also reached a point where the entire might of the US government is being weaponized against universities because their output sounds too woke—so there's a pretty compelling reason to get less woke fast. The fact that nothing is changing implies to me, however, that...
B. There are (at least) a few bad apples who really believe and teach the crazy woke things. This is bad mostly because *crazy woke things tend to be untrue.* Unthoughtful condemnation of capitalism and conflation of past discrimination with present is *bad and poisonous* to the social sciences and the humanities. These courses appear to teach an understanding of the world that is messed up and wrong, and universities shouldn't be so ridiculously full of them.
Also, tangentially, C. The wokeist courses somehow are also ridiculously easy. Like, they require ~5000 words of writing, in total, across an entire semester. But college should be much harder than that! If 80% of the class walks away with an A having done and learned very little, then the whole credential-signaling system is at risk, which is bad (for the universities, at least).
I am curious for the Commentariat's thoughts on:
1. If the woke-language really is, in your experience, just a front for what actually tend to be quite normal and thoughtful courses.
2. Whether I am being too uncharitable to woke ideologies in general.
3. Whether it's obvious that wokeness in academia is a problem, and it's weird how surprised I am by all this.
4. Whether it's obvious that wokeness in academia is not a big deal, and it's weird how offended I am by all this.
5. If I should sign up for one of these courses to be a sneaky little anti-woke spy, or if I oughta spend my time on cooler things like calculus or a class called "Computer Science and the Modern Intellectual Agenda" taught by David Gelernter [https://en.wikipedia.org/wiki/David_Gelernter].
In your linked post, you include this course description, with your parenthetical comment:
>the so-called Banana Wars, the disintegration of the Good Neighbor era, the inter-American Cold War [what?],
No offense, but I would like to remind you that you are a high school graduate. There are many, many things you don't know yet, and have never heard of. After four years of college, you will know more than you know now (but there will STILL be many things you don't know yet. (Eg: you will likely know almost nothing of the law; I say this because I know many, many people with PhDs who know nothing of the law. You will also know relatively little about politics (ditto). And economics (ditto). I could go on) So, I encourage you to keep an open mind, especially when you come across things you have never heard of.
Edit: I just read your second post. There, you express skepticism that the "neoliberalism" reference in the Portuguese course is a stand-in for colonialism, because "The Portuguese had left Brazil by 1822 ... So the idea that it could stand in for the horrors of colonialism is pretty incoherent." I do not intend to defend the claim that neoliberalism = colonialism, because I expect that it doesn't, but speaking of things you did not learn in high school, Portugal had African colonies until the 1970s, and indeed was rather famously unwilling to surrender them without a fight. https://en.wikipedia.org/wiki/Portuguese_Colonial_War
And, to be clear, I do not mean this as some sort of personal attack; I doubt that I learned that stuff in high school, either.
> And, to be clear, I do not mean this as some sort of personal attack; I doubt that I learned that stuff in high school, either.
I'm probably well above average in history knowledge and I'd never heard of the Portuguese Colonial War either.
I'm a current anti-woke liberal PhD student immersed in academia. As far as I can tell, there has been near-zero retreat from or retrospective on wokeness. There has been marginally less 'active wokeness' like proactively bringing topics up or sending emails around, etc., but that might just be because general anti-Trumpness has taken up a lot of oxygen. Basically, the vibe has not shifted inside the ivory tower.
1. No, the courses are really super-woke and frequently non-substantive. I do think you have a skewed understanding of the difficulty or commonality of 5000 words per course from all your blogging - that seems like a pretty normal course, not much less work than its peers.
2. No, you are not. I would say you are, if anything, being too generous. Even most professors or scientists are not general-purpose intellectuals. They have shockingly little defense against bad ideas popular among their peers. There are only a few bad apples that constantly, vocally push for wokeness but it really is the basic ideology of most of academia. Even people who aren't constant activists will generally view political things through that lens, make occasional adjustments in accordance with its precepts, or otherwise believe in it.
3. I don't think this is something that would be obvious to a high school student, but is something that would be obvious to anyone spending a few weeks attending a university.
4. I definitely find it a big deal, but I don't know how immediately obvious it is that it is a big deal problem rather than a small deal problem. I think you have to really get to know people and the actions they take with regards to the university before it becomes clear that it really is a big deal that has major and ubiquitous decision-making impact, and does need to be rooted out rather than let fade.
5. No, unless you have a course requirement that means you have to take one or that the alternatives are equally-boring/unhelpful. An entire course is a fairly significant amount of your university career, and plenty of woke material is easily-accessible outside of university while the material of many classes is much harder to learn/access outside of the course. Also, you may be required to either compromise your principles or get into a pointless fight with your professor if you take one of these courses. They're not always made to be completable all ideologies.
Good luck at uni!
I'd also like to point out the pretty obvious way your own viewpoint is leading you to make bad predictions:
"It was pretty shocking to see a ton of really comically woke course descriptions, even now, in 2025, when all the colleges are being attacked by the government and the public for being too woke."
Let me pull out one point in particular:
"being attacked by the government and the public"
This is, *strictly speaking* a true description. As in, the government is *certainly* attacking colleges on ideological grounds. And "the public" in the sense of "some number of members of the public."
But of course that second thing isn't new. There have always been members of the public interested in attacking universities on ideological grounds. It's not clear (to me at least) that there has been a large or sudden shift from that angle: I suspect that the members of the public interested in doing so in 2025 are pretty similar (in both number and identity) to the members of the public who were interested in doing so in 2024 or 2023. Doubtless some of them have been emboldened by the government and general political climate, if they didn't prompt the universities to change before, there's not a lot of reason they should now.
But the first thing *is* new. But let's be clear about what is happening. It's not "the government" in the broadest sense. It is "the president, and people working with and aligned behind the president." Now, that's a big deal in more ways than one. Let us be crystal clear, it is *not normal* for the president of the U.S.--or the head of any liberal democracy--to directly try to dictate curriculum or educational practices to universities. Governments can be involved in various ways, but they don't usually look anything like this.
Now, I imagine there are plenty of universities that would be well-advised to respond to this sort of thing. Some can't afford to lose the funding. Some probably wanted to make some of those changes anyway, and are glad of the PR excuse. But this isn't any university you're looking at. This is Yale. In case you were unaware, Yale is older than the United States of America by nearly a century. Yale is a *big deal.*
So I ask you, why do you think it would be in the interests of Yale, the institution, to do *anything* at the behest of President Donald Trump? Why was is background assumption of yours that they should or would?
It seems *quite obvious* from where I sit that pressure from Donald Trump should make Yale *less inclined* to make changes to their curriculum, not more. When you're already powerful and respected, caving to a bully isn't good for your reputation. Even more so when that bully is rather intensely disliked by a large number of the wealthy and prestigious people and institutions around you.
To me this assumption seems like an expression of fallacious reasoning of the wishful thinking variety. "I think X would be a good thing, and Y is an action intended to cause X, therefore I expect Y to cause X." The first clause, of course, logically has nothing to do with the rest of it (and is probably implicit rather than explicit in peoples' thoughts). Y *might* cause X. It might not. But the inherent rightness of X isn't evidence either way. And of course when laid out like that it seems very silly and obvious (and maybe it sounds condescending of me to be attributing such to you). But humans do this all the time without noticing it. It's a very easy pattern to fall into.
So, you're pissed at the Aussies now?
Or, say, the Germans? Or the Israelis? Or the Japanese?
The head of four liberal democracies are/were dictating curriculums...
And that's just off the top of my head!
>Let us be crystal clear, it is *not normal* for the president of the U.S.--or the head of any liberal democracy--to directly try to dictate curriculum or educational practices to universities. Governments can be involved in various ways, but they don't usually look anything like this.
I agree. But I will also say it is perhaps understandable. Conservatives are slowly discovering that basically everything is downstream from college. Like why the media disagrees with them is downstream from college. Like why athletes kneel is downstream from college. I find it understandable if they overreact.
Like imagine the opposite situation. Suppose you are an atheist living in a society where everything is Christian. Athletes pray before matches. The media asks the opinion of pastors about everything. And then suddenly you find out it is basically all college. How would you react?
The thing is, the president telling Harvard what to teach probably can't work. But state universities are ultimately under the power of the state legislature and governor. Fields that are basically all about leftist activism can't survive sunlight in that setting. To the extent that State U is seen by most voters and legislators as hostile to them and their values, State U is going to have ideologically hostile presidents imposed on them, is going to have their funding cut, may very well have (hamfisted and dumb) laws passed to rein them in, etc. Why on Earth should Republican voters or legislators want to fund academic fields that are openly focused on achieving progressive/socialist/radical political change?
ISTM that the best you can hope for is the state's voters and politicians leaving the state universities to run themselves. But given that this whole issue has become so visible, it seems unlikely.
"I agree. But I will also say it is perhaps understandable."
"Understandable" in the sense of "I can predict that a human being might react that way in that situation," yes. "Understandable" in the sense of "reasonable" or "pardonable" absofuckinglotely not.
I'm quite aware of conservative feelings towards academia. I suspect I understand *why* conservatives--as a broad social group--feel that way much better than most conservatives do. But in my sober and considered judgement, as someone looking in from the outside[1], it is utterly unhinged. It is terrifying. Mistakes like this are written on the gravestones of civilizations.
It is not any sort of a secret that education is important. Top universities charge insane amounts of money and have applicant lists a mile long. "Stay in school" is an admonition to the young so common as to be thoroughly cliche. Almost every job posting I've ever looked at puts the education they're looking for as the top item in their "desired skills and experience" list[2]. It is impossible to miss to literally anyone paying a shred of attention that education is important.
I'll go father than that. I think education is likely the single biggest reason that the U.S. emerged as the dominant economic and political power at the end of World War II. U.S. dominance was built more than anything else on technological superiority--computer tech most recently, but manufacturing before that. Technological superiority is built on scientific expertise. The U.S. managed to concentrate scientific expertise beyond any nation in history after Germany pissed *it's* scientific human capital in a spectacularly stupid fashion over an astonishingly short time (and the USSR did something similar rather more slowly). Education made that possible. Not just to cultivate the bright minds raised on U.S. soil, but to *attract* the bright minds raised elsewhere. Many talented foreign-born people have studied at U.S. universities over the past 80 years. Many talented people educated abroad have come to the U.S. to work at those universities, or at companies filled with talent educated at those universities.
But despite that importance being plain to see for any who looked at it, U.S. conservatives decided that their identity as conservatives was *more* important. That education's ability to elevate and create and yes, even enrich was less compelling than it's shocking tendency to make many of its recipients *gasp* less conservative[3]. To be fair, I think there's certainly some conservatives for which this it was a sour grapes sort of thing: lacking access to educational opportunities for one reason or another, they fell prey to the natural egocentric tendency to downplay its importance. But there are lots of conservatives who plainly did have access and simply disdained the value of higher education: sometimes the value of *getting it* but other times the value of *providing it.*
This last thing is really the crux of the matter. The thing that conservatives actually needed to do to prevent what they consider to be the corruption of education was to *show up.* Specifically to show up to higher degree programs across the full spectrum of academic subjects and the to show up to job interviews to teaching positions. They needed to do it decades ago. University faculty lean heavily left because left-leaning people are most of who wants to work in universities[4]. I do think at least some conservative criticisms of (for example) humanities education have merit, but the proper place to deliver those criticisms is from *inside the field*. If you want a good blueprint for what a political academic field should look like, look at economics. It's one area that conservatives *did not* cede the field, and you can find economics professors preaching anything from Marx to Rothbard. I don't find it particularly respectable as a science, but it's clearly in a much better state than it would be if it were nothing but Marxists arguing with Ancaps (y'know, like tumblr in tweed jackets). But for most of the remaining areas, conservatives dropped the ball hard. They had access to one of the most important institutions in the world, the crowning pinnacle of American civilization, one of the great shapers of history...and they walked away for what seem to me to be largely petty reasons, only to come roaring back in a childish fury when (surprise!) it kept going without them, and didn't get one whit less important.
And so now it's ending. It has too much size and inertia to end quickly, but ending it is. The bright, curious minds of the world have received the message--sent very loudly and very clearly--that we are not welcome. That a large and powerful segment of U.S. society despises us for exactly the things that make us valuable to humankind at large. That many places in the world would be better, safer, more welcoming places to practice our craft. Even if the U.S. endures the harm it is so eagerly inflicting on itself, that trust isn't going to be repaired anytime soon. Maybe the scientific center of the world will shift to Europe, maybe China will capture it, or maybe it will just disperse, with the greater connectivity of the modern world allowing it to be less centralized. But it sure won't be the U.S., not anymore. But hey, if it means reading a few less cringeworthy course descriptions, small price to pay, right?
[1] Both in the sense of being not a conservative and in the sense of living outside the U.S., having wisely gotten the hell out when things were still merely "highly worrying" instead of "acutely alarming."
[2] Despite my generally pro-education sentiments, I do think that this particular trend has been taken somewhat overboard. Credentialism is a real issue, being a (still comparatively mild) instance of Goodhart's law, where the piece of paper carries value independent of the skills it is supposed to signal.
[3] I'll note that this isn't really a *universal* feature of education, but neither is it anything to do with indoctrination. Without writing an entire essay on the subject, I'll merely note that several major strains of U.S. conservatism are based on views that simply don't hold up well to exposure to the wider world. That doesn't mean there are no educated conservatives, but I think educated conservatives tend to be conservative for very different reasons than uneducated ones.
[4] This does absolutely create a feedback loop, which is why the "decades ago" was important. Though also there are definitely fields were the right loses out simply by virtue of its predominant viewpoint being acutely wrong. I couldn't hazard a guess at how many conservatives went into climatology with the intention of blowing the cover of the climate change hoax, only to abandon their ideology when they could no longer fail to recognize that mainstream climate science *is not wrong.* But I'm sure the number is more than zero.
But if education is so important, why so many inventor-billionaries keep telling young people to drop out of college and start a business? Do you think the Steve Jobs type not-very-educated-inventors are sort of the tip of the iceberg, and they are working on the backs of hundreds of engineers who do need education? I am choosing Jobs because once he fired the entire R&D department of Apple, which sounded like such a suicide move, and yet it worked out.
Look, there was recently this excellent review of schools right here. Briefly, the no-structure people who just learn on their own, are probably the smartest people with the best world ethic, and it seems they fill the truly important positions everywhere. The high-structure learners who struggle with schools will probably not contribute that much… so education is mostly for the low-structure learners. Do you think they are so important? Briefly, bluntly, is it important that for every no-structure genius, there should be 10 low-structure mediocre people?
I have been surrounded with no-structure learners because software dev used to be such a crazy wild west thing with “anything goes”, and it was one field that back then did not care much about degrees. There are now a lot of low-structure people in it who think like classical engineers, who consider the methods they were taught as holy writ and do not really think outside that box. I am not sure how useful they are…
"Look, there was recently this excellent review of schools right here. Briefly, the no-structure people who just learn on their own, are probably the smartest people with the best world ethic, and it seems they fill the truly important positions everywhere. "
First, I really gotta ask: are you the same person that wrote the previous comment? Did you have a bout of temporary amnesia? Because the comment I was responding to wrote:
"Conservatives are slowly discovering that basically everything is downstream from college."
Which *seemed* to be hitting the nail right on the head. Everything in our modern, complex society IS downstream of higher education, thank you for noticing. You point right at the fact that multiple large clusters of collectively very influential people got similar sorts of higher education[1], and then proceeded to have enormous influence on the world. Are you really turning around one comment later and saying "well, how important could it really be, if Steve Jobs said they didn't need it?" Do you really, genuinely think all those people just kinda...wasted their time and money, that none of them *needed* what they were getting out of higher education, that all of them could have done the same jobs just as well without them?
I will be the first person to suggest that our current way of teaching people various sorts of specialized job skills might well be inefficient and overly rigid, and that there might be better systems for doing the same job. But the idea that we could get rid of it and replace it with...nothing at all? I honestly wonder what planet you've been living on. The modern world is *enormously complex*, and requires many, many different sorts of specialized skill to function. People are not born with those skills. Some of them they learn on the job, but lots of jobs have a certain floor of skill that's required to do useful work at all, and thus it is important for *employers* every bit as much as employees that there be institutions that can teach useful skills[2]
"The high-structure learners who struggle with schools will probably not contribute that much… so education is mostly for the low-structure learners. Do you think they are so important? Briefly, bluntly, is it important that for every no-structure genius, there should be 10 low-structure mediocre people?"
At the risk of sounding condescending and reductive, it seems like your conservative biases have completely overtaken your critical thought on this one. The conservative worldview seems extremely prone to believing that:
1. People exist in a rigid hierarchy of value/virtue that cannot meaningfully change. If you're a good/valuable person now you'll always be a good valuable person in the future and if you're a bad/worthless person now you'll always be bad worthless in the future.
2. This hierarchy basically always cashes out into life outcomes. If somebody ends up poor or miserable it is because they were unworthy. If somebody ends up rich and successful it is because they were worthy.[3]
3. Only the people high up on the hierarchy really matter at all. Everyone else is of little to no importance.
(Usually with an implicit 4. They and those close to them are definitely high up in the hierarchy, regardless of life circumstances).
This is, to put it frankly, utter crap. I think a lot of the time these go unstated and largely unexamined in the minds of the people that hold them. But you are coming out just shy of saying a lot of this explicitly in the comment quoted above: so I decided it was worth being a little more blunt that usual about it.
Now, it *is* possible to take egalitarianism too far. Not everybody is equally capable of doing every thing. But the big, enormous, GLARING mistakes in the view above are first, assuming that there is only one dimension of competence and second that people who don't reach the absolutely top levels do nothing important. Again, the world is very large. It is very complex. It requires many *different* sorts of skills to keep running, and some people are very good at some of those skills and not at others. And there are many, many places where having two people with an *adequate* level of a particular skill is so, so, SO much more important than having one person with an *exceptional* level of that skill.
"I have been surrounded with no-structure learners because software dev used to be such a crazy wild west thing with “anything goes”, and it was one field that back then did not care much about degrees. "
I think your understanding of the world has been harmed rather badly by doing so much work in a field like this, and (apparently) not paying enough attention to those outside. I know a thing or two about programming (though probably less than you). It is certainly an area where a single person with excellent skills can often contribute much more than multiple merely adequate people. Most fields are not like that. The most talented surgeon in the world can still only operate on a few patients a day: a hospital with 20 adequate surgeons will certainly provide better *overall* healthcare outcomes than one with 3 or 4 amazingly talented surgeons.[4] Even in software, there are sharp limits to what a few talented people can do. Sometime you do just need more people that know the language and can hunt bugs, write tests, or create various pieces of boilerplate code and connective tissue that doesn't need to be a masterwork of creativity and subtle elegance (it just needs to work).
If I have time, I'll write a separate comment about why I think your understanding of "no-structure, high-structure and low-structure" learners badly missed the point.
[1] Similar within a cluster, not necessarily between clusters.
[2] I'll note that a system in which employers take responsibility for teaching all skills beyond basically literacy and numeracy *is* something that could exist. The reason employers don't do this *now* is because they would be running the risk of investing lots of resources into an employee's training, only to have that employee leave and put those skills to work elsewhere (perhaps for a competitor). So instead employers want entry-level people to have skill sets that will let them contribute positive value quickly. Setting up a system that would allow all training to come from employers would have an *alarmingly* high cost: you'd basically need to give them the power to prevent employees from quitting.
[3] I'll not that people seem to violate both of these assumptions basically whenever it's convenient: rich or successful people that they don't like must have cheated (somehow) and poor/unsuccessful people that they like get infinite excuses. And yet the hierarchy reasserts itself the moment they're not actively looking at those specific cases.
[4] Provided their standard of "adequate" is set high enough to avoid all of the easily-preventable deaths and injuries that incompetent or negligent surgery can result in.
I'll think this over, but yes, internship based learning yes might involve not being allowed to quit for a number of years. Actually we have this today in a way, when the employer finances doing an MBA but then either no quitting for X years or paying it back.
>But if education is so important, why so many inventor-billionaries keep telling young people to drop out of college and start a business?
Why would you assume that they have the best interests of those young people, or of society, at heart? Or, if they do, that their assessment was correct?
>Look, there was recently this excellent review of schools right here.
That was re K-12, not higher ed.
Oh. I went to higher in the kinds of Central European circumstances that were very Prussian rote learning. Then I saw in the Anglosphere (UK not US but likely similar) it was far more practical. My takeaway was that it is succesful precisely because it is not education (not rote learning) but something like internship, actually doing work and solving problems. So it seemed to me, it could be replaced with full on on the job internship and it would probably improve. It was already 75% like working a job, not memorizing definitions.
"Whether I am being too uncharitable to woke ideologies in general."
Insofar as it seems pretty clear you don't even understand the things you're mocking, yes. Probably some of them are very nearly as silly as you think they are. But it seems like quite a few of them are simply couched in language and perspectives unfamiliar to you, personally, and you don't seem to recognize that there could possibly be useful knowledge not packaged and aimed directly towards someone of your background. Which probably sounds harsher than it is: I've been there, as I suspect most people have.
Let me just say right from the word go that my academic background is extremely, almost exclusively STEM focused. I've taken classes from other areas mostly to fill course requirements. But I also have a curiosity about the world and often interact with smart people with very different backgrounds. There's plenty of pretentious garbage produced by the humanities that I do roll my eyes and, but I don't go throwing the baby out with the bathwater.
I'd like to call particular attention to one of your complaints: "Hemispheric Poetics & Politics." There are other courses where I'd (partially, tentatively) agree with your judgements of the course descriptions, but this one in particular reads like you just fundamentally do not understand that an arts class is going to be discussing, y'know, art and what that means.
I guess some more narrow-minded STEM people regard art as just, like, entertainment. Something nice to watch, look at or listen to on weekends. And of course it can be that. But it's also the way humans express themselves. It's the way humans express the events they've lived through. It's every bit as vital a part of understanding history and the broader world around you as statistical summaries or academic analyses. You'll learn one set of things by reading (say), a well-researched history book on World War I. And another set of things by consuming art made by people who experienced it. You can't really learn first from the second OR the second from the first.
This particular course appears to cover art made about various conflicts in 20th century Latin America[1]. That is an *extremely* apt topic for a Spanish comparative literature class. Someone studying Spanish at university with the expectation of actually going out and *using it in the world* is very, very likely to interact with people whose life histories have been profoundly shaped by those conflicts. Studying the poetry that people made about living through those conflicts and their aftermath is *absolutely* a good and worthwhile thing for a subset of university students to be doing.
And to me this nicely illustrates the problem with the entire concept of "woke" and "wokeness." The motte here is reasonable[2] claims like "elementary Portuguese class shouldn't be centering racism, sexism and neoliberalism in its course description or content," and the bailey is "university students shouldn't be taught about a highly subject-relevant artistic tradition because the politics *of the art* make me, personally, uncomfortable." In essence, almost nobody throwing around the words "woke" and "wokeness" seem to have clear, consistent ideas what they actually *mean* by them[3], but instead lean on a "I know it when I see it" sort of schema. But the "I know it when I see it" seems to both differ a lot from person to person, and be *extremely* likely to simply reflect individual prejudices about what facets of life and culture have what political valence. Like you marked "American History 1800s to present" as "broadly fine," but I guarantee that lots of conservative-leaning readers would see "U.S. imperialism, legal exclusion, racial segregation" and brand it as "woke" on that basis alone. Those are all things that happened, but *talking about them at all* is regarded as political.[4]
Before I sign off, I'll weigh in on this:
"If I should sign up for one of these courses to be a sneaky little anti-woke spy,"
No. Absolutely not. This is quite possibly the single worst way you could go about interacting with this facet of the world. If you go into *any activity* for the sole purposes of confirming your biases, your chance of learning anything craters. If you genuinely believe some class has material that's interesting in its own right--if it's covering something that you'd want to learn about even if it was presented in a different way by different people--consider taking *that*. But if your only interest is to "spy," the best possible outcome is that you will simply be bored out of your mind. And you stand a very good chance of making your understanding of the world substantially worse.
[1] I assume the "what" after inter-American cold war implies confusion as to what that is. It's the portion of the Cold War that was took place in Latin America: the U.S. was often very involved in the politics and conflicts of various Latin-American countries as part of its indirect but often still quite militant opposition to all things socialist.
[2] Note "reasonable" is not the same thing as "incontrovertible."
[3] I've had this conversation a couple of times on here and even smart, thoughtful, articulate people don't seem to be able to articulate it in a way that stands up to scrutiny.
[4] Regarded correctly, IMO. It's a mistake to imagine that *any* aspect of history can be taught in apolitical fashion.
I think this it’s a good comment. I would just add that. I took a lot of literary theory classes in college and I think they were the worst in terms of ROI. I could’ve learned about Derrida and Baudrillard on my own and the time would’ve been much better spent even in the English department reading some 19th century big novels or maybe Yeats.
You are making a very flawed assumption here. You're assuming woke isn't paying for people to write books that are "woke." (and that publishers aren't forcing their writers into "being woke" so the publishers don't go out of business).
Next thing, you'll try and tell me that Modern Art was an entirely organic phenomenon.
If I sound a bit salty here, it's because one of my favorite series got cancelled because the author discovered (to his horror) that he'd written a character /wrong/. That the character's current actions (as dictated by "woke" philosophy), did not and would not lead up to the character's final fate -- "you might not notice that, but your brain does" is a relevant quote from Red Letter Media. Unable to reconcile that, he cancelled the series.
"You're assuming woke isn't paying for people to write books that are "woke.""
It's quite possible I'm making a "flawed assumption." Probably even a lot of them. Most people do, all the time. But if you have genuinely identified such an assumption in my writing, you have failed so egregiously to communicate it that I have to wonder if you were even trying at all. I'm going to do my best to parse this, but I'm not hopeful...
"You're assuming woke isn't paying..."
Well, five words in and we're already in trouble. I realize that expanding the meaning and usage of the word "woke" to cover more and more ground and become less and less comprehensible is a hobby approaching an obsession for some people, but this is really a bridge to far.
Even in its current, heavily bastardized usage, the word woke is an adjective (it's original usage being a conjugated verb). It is not, nor has it ever been a noun. As a noun it has no standard meaning[1] that I'm aware of. You easily can noun-ize an adjective by appending "the" to the front to indicate a group of people with that characteristic: "the quick," "the righteous" and so on. But you followed it with a singular verb conjugation, so clearly this isn't a typo. As a singular thing, followed by "isn't paying" it has to indicate some sort of...entity, I guess? With both money and agency? I'm going to stop short of speculating about what sort of entity that might be, because you haven't given me much to go on, and it seems like it would be pretty rude to assume you're inventing conspiracies or the like just because of one poorly worded sentence.
"...for people to write books that are 'woke.'"
OK, now I'm even more confused. Where...are you getting *any* of this? Where am I assuming *anything* about the type of books people are and aren't being paid to write? I don't think I specifically mention anybody writing books anywhere. I talk a fair bit about art-in-general, which certainly includes some books. And I make a passing mention to a hypothetical history book. But if there's some assumption in there of the form "people are definitely NOT writing books of Type X for Reason Y" I cannot for the life of me find it.
The main thing I'm talking about is a class on Spanish-language poetry (with emphasis on a particular conflict). Is that what you think "woke" might be paying people to write books about? Even if that were true, I really don't know why it would be relevant. Do you think the course might be studying paid propaganda instead of the genuine articles?[2] The conflicts in question *definitely happened.* They touched millions and millions of peoples' lives. Some of those people very definitely wrote poetry and music about their experiences because *that's what human beings do.* Accurately sourcing contemporary poetry and music is not generally difficult: I cannot possibly imagine that somebody capable of getting a job at Yale wouldn't take the elementary step of making sure their course materials were genuine. Now it's possible for a course like that to work from some sort of a textbook or compilation[3]. If it does it's pretty certain that said textbook would be both paid for and written by people who are "woke"[4], but as long as *it* was sourcing genuine artistic materials about the conflicts in question (and again, hard to imagine it wouldn't), that sounds like more of an annoyance than a real complaint.
So after thinking it through aloud (well, in print) as thoroughly as I'm able, I *still* have no idea why you think I assumed this thing, or even what exactly it is I'm supposed to have assumed. Maybe try being more direct and less vague.
"Next thing, you'll try and tell me that Modern Art was an entirely organic phenomenon."
Not honestly sure what you're trying to imply here either. The Wikipedia article for "Modern Art" tells me the term covers art from over a century, ending in 1970, so whatever you think was "inorganic" about it, it can't have had much to do with "woke.[5]"
"If I sound a bit salty here, it's because one of my favorite series got cancelled..."
I'm sorry that that happened, I hate it when things I enjoy get cancelled. But without more detail than a really, frightfully vague description I can't possible have much of an opinion on it. Did this series have a name, at least?
[1] To be clear, as an adjective it has no standard meaning either, but that usage is at least *common* and people can at least gesture at a shared schema for usage, even if they mostly end up talking past each other with it.
[2] Let me emphasize that if this *wasn't* what you're trying to say, I'm not trying to put words in your mouth. I'm just genuinely confused and grasping at anything your comment maybe *could* mean.
[3] Though also extremely, extremely possible that it wouldn't. I've only taken a couple of foreign-language lit classes, but neither of them did.
[4] Because again, the word is basically meaningless, and anything besides guns, trucks and Jesus is "woke" to somebody. Actually, scratch Jesus off the list, he's definitely "woke" as well.
[5] Unless it's not only expanded from adjective to noun but also from relatively recent to many decades old. I can't be sure it hasn't.
>The Wikipedia article for "Modern Art" tells me the term covers art from over a century, ending in 1970, so whatever you think was "inorganic" about it, it can't have had much to do with "woke.
Wait a bit please. Maybe there is a simple misunderstanding here. "Woke" originates in "all men are created equal". Its roots are centuries old. When people criticize only its current form, it means that in the past it was simply applied less consistently, often not applied in many aspects of life, those aspects were "apolitical" in the sense of "old inherited unnoticed inequalities".
Because this seems to be the only two options on the table. Either uproot all inequalities in every possible aspect of life, in which case everything becomes political, and kind of propagandistic, and then we start looking like the Soviet Union where there was a proper Communist way of everything and a bad burgeois way of everything. Or, keep it apolitical, which means it is a cover-up for old inequalities, and the value judgements that come with them, that something that comes from a position of power is seen as a good and something not, not.
Did people miss the whole "all men are created equal" -- Under The Law?
Genuinely curious, as it seems a liberal dogma nowadays that everyone is equal, dead stop.
(Thought crimes continue to be bad, even when applied to beard cutting).
There is also a third option. An acceptance of inequality as a necessary and even beneficial thing. An understanding that different people have different uses and different value. How long are we going to keep pretending?
Yes, people make art about conflict. I have no direct knowledge about the particular South American conflicts' art being paid for.
However, I contend that for any art to be worth studying as a direct reflection of the world, it -cannot- be propaganda (propaganda can be studied, and perhaps ought to be studied, but without framing it as propaganda, you've lost enough of the meat that you're just putting your ear on the empty shell, and excitedly telling people you hear the sea).
I do have knowledge about "art in general" being paid for by "the woke" (you can hear this as USAID if you want, The Powers That Be if you don't.) More importantly, I have some knowledge about the financial constraints that "art" operates on, in order to make it into the world.
But perhaps, you ask, why would I automatically start being suspicious of art drawn from South American Conflicts?
Well, it's because "paid for art" is done by governments a LOT.
https://www.bbc.com/culture/article/20161004-was-modern-art-a-weapon-of-the-cia
Here's a good writeup. Yes, I should have been more specific about what "flavor" of modern art
Stranger Things was paid by the German Government to put a German song in (Germany pays for any German song in Hollywood, it's nothing personal -- this is relatively light propaganda). And the folks running Stranger Things have never turned down a shekel.
Dark is a different matter -- the whole "anti-nuclear" Green movement in Germany was very, very willing to pay for an anti-nuclear TV show. This is much more poisonous -- and did you know it, when watching? (I've been assured that the Simpsons occurred before this whole "anti-nuclear" money was floating around. That was apparently just the creator's idea.)
Sorry, the rest of this comment (already long) will touch on what you were so patient to write to me.
1) In the event that you aren't a native english speaker, "woke" used as a noun not an adjective does in fact mean "the woke" (the foreshortening is as discussing blacks instead of always using "the blacks" -- for African Americans if it's not clear.) This is a standard shortcut in English, and no confusion was intended.
2) If you have no suspicion that it is propaganda, how would you possibly notice? (Answer: didacticism. Heavy-handedness is generally indicative of propaganda.)
3/4) Any textbook is likely to induce bias through what it choses. I might refrain from a course on pre World War II German Art, if it was from Western Culture -- I'd have to be very careful to understand it as a subsect of the "allowable" art (in two parts: there are "what we're not allowed to talk about" at the moment, and "what the Nazis burnt"). You have to know what's being deliberately erased, to be able to say "this is valuable or not" (and in a book about conflict, there's going to be erasure, if only because some people are coded as "good" and others as "evil." You're not going to hear about human body parts being taken as trophies -- at least as a good thing (this is not a hypothetical, but also not from South America).).
5) Not woke. CIA/US Government. Sources cited above.
I think that woke is likely to not discuss any of this, instead discussing how "women are silenced" or other things like that.
Are you familiar with Riverbend? You, like me, would probably count her writings as a fine introduction to "America's Iraq." It's written with a marked bias (but not an unfounded one. Her discrimination is both cultural and incisive).
That doesn't make it not propaganda though. If there was one work of art that put the coffin-nail in "America's Nation Building" in Iraq, It would be her work. She didn't do it alone.
here I need to tell you the same thing as I told your interlocutor:
"Woke" originates in "all men are created equal". Its roots are centuries old. When people criticize only its current form, it means that in the past it was simply applied less consistently, often not applied in many aspects of life, those aspects were "apolitical" in the sense of "old inherited unnoticed inequalities".
Because this seems to be the only two options on the table. Either uproot all inequalities in every possible aspect of life, in which case everything becomes political, and kind of propagandistic, and then we start looking like the Soviet Union where there was a proper proletarian way of doing everything and a bad burgeois way of doing everything so now even coleslaw needs to be proletarian. Or, keep it apolitical, which means it is a cover-up for old inequalities, and the value judgements that come with them, that something that comes from a position of power is seen as a good and something not, not.
On the classes requiring ~5000 words of writing in a whole semester, that sounds like a reasonable length for a term paper. For an advanced undergrad or grad class, it’s pretty standard to just require a term paper, which is going to have to be much better at defining and carrying an argument than a bunch of weekly reading responses. The difficulty will come down to what sorts of standards are being used to evaluate the paper. Anyone who is legitimately taking an advanced class like this one will have taken other writing intensive classes where they should have learned some of the mechanics of writing, and will need to be constructing a significant argument at length.
Changing a course description is something that usually requires the department to draft a new version and vote on it, then submit it to the university catalog committee, and at many universities, the change then needs to be approved by a giant university-wide committee that includes representatives from every college/school. Most things that go up to that committee just get approved by default, but it’s a lot of work to send something through all these committees, so usually departments end up with a bunch of course descriptions on the books that aren’t taught very often, and sometimes have classes that get taught in ways fairly different from what the official description leads you to expect.
So you definitely shouldn’t expect course descriptions to disappear just because we are five months into a national vibe shift.
Anyway, I do think that you have been fooled by a lot of people into thinking that wokeness is a bigger problem than it really is. The only thing that actually seems objectionable to me in that course description is the use of “neoliberal” - but that word is in there precisely to indicate that they are going to include some economic urban geography that ignores questions of race, gender, and sexuality in thinking what a city is and what it should do! Everything else is there just to point out that they will include some readings that do focus on these identity issues - and it definitely makes sense to include these things, because cities really are special places for race relations and queer identity (as you’ll notice from anyone who talks about San Francisco in an opinionated way).
That would explain a lot. I remember being very disappointed in college when I tried to sign up for a few cool-sounding classes, only to discover that they were never actually offered.
If you’re interested, you can see the one paper I have published about cities here: https://philarchive.org/rec/EASTCO-3
I’m approaching very much from the “neoliberal” angle, but it’s clear that identity issues come up around the edges.
I think you could steelman the course description into a best-possible version of that course, which is all about how different ways of talking about words like "urban" and "city" wind up encouraging or encoding various different political points of view... like the somewhat flexible definitions of "city" vs "suburb" in American usage, or the way "urbanism" always seems to wind up encoding a single view of what a city ought to be like. This could be a genuinely valuable part of an urban design degree, which helps people step outside their own set of political assumptions and take a more detached view.
But you would need a seriously disciplined lecturer for that course, someone capable of stepping outside their own political assumptions and teaching it in a way that does not confuse political axes with quality judgments. Nothing about the way it is described, in terms of jargon used only on the left, makes me suspect that the lecturer is all that disciplined.
You should take some of these classes! Consider how pop-sci discourse makes science sound idiotic, and that a similar process happens between university-level humanities and popular "woke" ideology. Continental philosophy, critical theory, etc have depth worth understanding on their own terms outside of the culture war, at they very least to avoid the arrogance that comes with never having actually read these kinds of texts.
Imo the bulk of those two mentioned subcultures can be summarized as: "Woah, what if we thought differently in a way that's strictly inferior to just plain reasoning about our subject areas? Then we can still learn things even though we're clouding our methods with obscurantist terminology and saying stupid things more frequently than we have to." Like if you want to adopt a very heavy race focused view you can just make it clear you're stipulating that as a precondition of your inquiry and use regular statistical tools or informal arguments rather than writing something like Kendi's thesis or any of his books. The reason why this doesn't happen though is because those fields wouldn't really stand up to scrutiny if their proponents tried to integrate their thought with other fields. You can also see this failure in experimental philosophy where a lot of the rhetorical appeals of analytic philosophy like "We believe this is common sense" or "It's intuitive that..." are not remotely supported by the existing empirical evidence. I wouldn't recommend people study that field either after exploring it myself.
I have sometimes tried to do this (e.g. https://mistakesweremade.substack.com/p/review-the-idea-of-a-critical-theory), and generally agree that there's some valuable insight to be had... Will keep your recommendation in mind!
This type of thing is more specific in the Ivy Leagues, which have their own subculture to perpetuate and maintain. One of the most charming aspects of state schools is that they generally do a better job with ideological balance. They aren't perfect, but they aren't Yale either.
I have a friend who took a hyperwoke course at Villanova, found it very easy and not terribly interesting. That seems to be a typical opinion. I guess if you want an easy A and want to learn more about woke philosophy, take the courses; if you want just the latter, reading books and showing up to protests is the best option.
Maybe there's a lot of variation within state schools, but by my (limited) exposure to the University of Michigan, it seems like Yale is not *that* much woker?...
I do wonder how many young ideologues are actually taking these classes. Presumably they're the intended audience, but if the courses actually end up generally boring, then maybe the root of weird-young-elite-extremism is elsewhere. A few commenters have said something to the effect of "the kids are actually crazier and woker than the professors," which suggests this too. Maybe it's the internet, then? All this stuff leaked out of the academy a decade ago, and it's the kids who've been exposed to it since then who won't let it die...
I go to the University of Colorado, where this sort of thing is rarer. Wokeness likely scales with prestige, at least among state schools.
Why is that the case? Because it's what the elite humanities students who attend these schools want. We have to remember that in most cases wokeness isn't something imposed on students by administers: it's supported by the majority of the student body of these schools, as the Columbia protesters reminded us. I agree with the commentators that students are often woker than the professors, and it seems to me that wokeness itself embodies the traits of youth: its energy, ardor, idealism, and adolescent zealotry.
The pipeline is probably built like this: high school-and-younger students see internet wokeness and become radicalized. Once in college, they take woke courses because they like the descriptions, but these courses don't convert anyone so much as they provide sheet music to a rabid choir. The spirit of the movement happens outside, in clubs, protests, and collectives.
So we really can't blame woke classes for anything except giving the students what they want.
the alpha school essay has been bouncing around my head. i'm still trying to find the thesis-leading-to-action. maybe it'll emerge from some open feedback:
i was hired to study k12 ai in Louisiana. i wasn't impressed, even though Louisiana is actually investing much earlier than other states. most educators think of ai as chatbot-cheating-tools. students are using it to cheat, yes. but they are also using it to ideate. moreover, educators are using AI tools more than other tech. ai can save lots and lots of time.
the thing that struck me in the alpha school essay was that they are getting 2x results without actually using ai. what if they rebuilt their program using ai?
with my louisiana work, i also bought about 1000 students khanmigo and while some did use it, it was just an ai tool slapped onto a legacy application.
we have the chance to rebuild education. personalize teaching. take away educator administrative work. give people back their time to do whatever they wish.
the greatest gift i got from my weird alternative high school was the gift of time. i was able to follow my passions. build companies. chase wild ideas. that gave me confidence. and confidence seems to be rewarded in our society (the indiviudal pursuit of freedom.)
anyway, here is what i think i'll end up doing:
1. Raise $5M at $15M valuation (A8C SAFEs convert).
2. Build team to do two core projects: curriculum creation(teaching) + platform building(admin).
3. Make all K12 Louisiana curriculum “AI Ready”
4. Reduce teacher and learning time by 75%.
5. Run as an afterschool cohort.
6. Scale to Kindergarten, 6th, and 9th grade cohort.
7. Build in new grades each year.
8. After 4 years, scale to new locations.
My not-terribly-well-informed opinion is that a lot of it has to do with the rapid feedback. Typically students have many days between when they attempt a problem (either homework or a test) and when they learn if their attempt led to sucess or not. By that point the context has faded from memory. OTOH if they get feedback within MINUTES, then they can still recall WHAT they were thinking when trying to solve the problem, and WHY.
There's work on "deliberate practice" that supports the notion that rapid, objective feedback is very helpful. I'm also reminded about Joel Splosky's principle of "fix bugs before you write new code" (item 5 at https://www.joelonsoftware.com/2000/08/09/the-joel-test-12-steps-to-better-code/ )
Hi all. I made a little bot designed to:
1. Help cheer people up and bring laughter
2. Encourage people to socialise and go outside more
Check it out here: https://chatgpt.com/g/g-6864de3c115c8191b0dc15746976e10d-bear
I ran the "one guard likes big butts; the other guard cannot lie" benchmark, and the bot did fall into my clever trap and assume the first guard always tells the truth, but it did so in the most likeable and enjoyable way I've yet seen.
It's cute!
I read today that heat deaths in Europe are greater than gun deaths in America. I checked it myself and Europe had 60,000 heat deaths in a year where the US had 2000. Further, ChatGPT confirmed that at least half of this is being caused by comparative lack of AC. Is this actually true? If so, it feels insane to me that I have never heard anything about it, not in mainstream media or EA causes.
There is so much clickbait and false certainty in discourse about US gun deaths that it's worth grounding oneself in the actual current facts. Pew Research Center -- a gold-standard source -- happened to publish an update on it a couple months ago:
https://www.pewresearch.org/short-reads/2025/03/05/what-the-data-says-about-gun-deaths-in-the-us/
For tldr folks, highlights:
-- suicides have long been the majority of US gun deaths, currently 58% of them.
-- about 38% are murders, with the remainder being accidental/law enforcement/other.
-- while total annual gun suicides have risen pretty steadily since around 2004, gun murders spiked in 2020/21 and since then have fallen from that peak. For 2023 they were at about the same annual total as the early-1990s. (And then from sources outside of this Pew writeup, those totals continued dropping in 2024 and thus far in 2025.)
-- the gun suicide rate (per capita) is now almost up to the highest rate recorded (1977) in the past half-century.
-- the highest US gun murder rate (per capita) in the past half-century was in 1974, the second-highest in 1994. The 2021 rate got to within 10 percent of 1974's.
In the following stats "all rates are adjusted to account for age differences across states":
-- the states with the highest rates of gun deaths are all both extreme-south and heavily rural: Mississippi (29.4 per 100,000 people), Louisiana (28.3), Alabama (25.6) and New Mexico (25.3). However the District of Columbia, which is basically 100% urban, currently tops all of those (30.6).
-- if we look just at gun murders the same five jurisdictions rate highest but the gap widens from DC to those four states.
-- if we look just at gun suicides the rate leaders are all large states with lots of wide-open spaces: Wyoming (19.0 per 100,000 people), Montana (17.6), Alaska (16.0), Idaho (15.1) and Oklahoma (13.6).
> -the states with the highest rates of gun deaths are all both extreme-south and heavily rural: Mississippi (29.4 per 100,000 people), Louisiana (28.3), Alabama (25.6) and New Mexico (25.3). However the District of Columbia, which is basically 100% urban, currently tops all of those (30.6).
As can be guessed from the fact that DC leads the list, if you dive deeper into those states with high rates of gun deaths you will discover that while suicides are relatively evenly distributed, virtually all of the homicides are concentrated in urban areas (ie Louisiana is dominated to an extreme extent by homicides in New Oreleans and Baton Rogue).
You can guess or assume that, and many do, but you'd be wrong.
(a) Suicides are not at all evenly distributed. Nationwide from 2001 through 2010 the rate of gun suicides was 54% higher in the most-rural counties than in the most-urban ones, and from 2011-2020 that gap increased to 76%.
https://jamanetwork.com/journals/jamasurgery/fullarticle/2804113
(b) Nationally the rate of gun homicides [murders plus non-negligent manslaughters] is about 5.7 per 100K. In rural areas (those not part of any metropolitan statistical area nor part of any isolated city) that rate is 4.3. It's a lower rate but not drastically lower.https://www.ruralhealthinfo.org/topics/violence-and-abuse
Outside of a couple of outliers like DC, the reason that most US homicides happen in metro areas is primarily just that's where most of the people are (only 20% of the national population resides in areas classified by the Census Bureau as rural).
(c) In Louisiana specifically, which has unusually high homicide rates, the rural-parishes rate is only about 10% lower than in urban parishes (counties).
https://www.nola.com/news/crime_police/after-rise-in-killings-louisianas-rural-parishes-twice-as-deadly-as-u-s-metros-data/article_7568a454-11a6-11ed-86db-b7b123e7d8c0.html
(d) Looking at those four states that lead the nation in gun deaths per capita, Mississippi is the 4th-most rural state in the US; Alabama is the 9th-most rural; Louisiana is above average in rural population pct and New Mexico is about average. In terms of core-city populations those four states include no cities in the nation's top 25, just one in the top 50 (Albuquerque) and one more in the top 100 (New Orleans).
Deaths aside AC is a main reason the southern half of US has been booming economically the last half century.
Southern Europe isn't quite as hot, but I'm sure it could also benefit immensely from allowing more AC. To me, that's a much bigger issue than the heat deaths.
I'd add to the previous comments that deaths should be measured in QALYs - "heat deaths" are typically already very old and sick while gun deaths are typically of healthy young people (so a gun death probably has an expected loss of 60 QALYs while a heat death might only have 1). YMMV how you want to weight that in morality terms.
I don't go in for the QALY stuff at all really because it's just utterly intractable (my 73 year old mom's life is worth less than your 65 year old mom's life... to whom, exactly?), but you surely have to discount suicides immensely on this metric, since the person committing suicide has decided their life is negative QALYs from here on out.
” my 73 year old mom's life is worth less than your 65 year old mom's life... to whom, exactly?”
For one thing, presumably to herself? Ask the 65 year-old ”you have to pick one - die today or at 73?”
So the latter is a preferable outcome compared to the former.
Probably. But it's pretty unavoidable when dealing with heat deaths, since so many of them are of people who're close enough to dying already to be knocked over by one adverse day. If your 73 year old mom died of a heatwave today instead of getting heart failure in a month that seems importantly less significant than if she'd been hit by a truck ten years ago.
(Gun deaths/suicides do, as you point out, have their own complications, which makes them a poor baseline for comparison in general).
A goodly number of the "healthy young people" killed in US gun deaths are not contributing to anyone else's quality of life.
Maybe. Most of the gun deaths are suicides I believe. A lot of those people are older and just feel like they are done with the world. Don’t have numbers in front of me but it seems possible from reading suicide obits.
2023 US had 47k gun deaths, 27k of those were suicides.
"Gun deaths" is a bit of a weasel stat. Most people probably assume it's all homicide.
Average suicide age is 47, while homicide victims are typically in their 20s. Both are a lot younger than heat deaths.
A large chunk of gun homicides are gang or drug related. I think the 20 year old who dies in the Bloods v Crips shootout should get a pretty hefty QALY discount.
I'd really like to see the breakout of "business-related" versus "anger-related" deaths for homicides in general. (where 10 people kick you until you die, and then don't rob you, counts as anger-related even if they weren't angry).
I've found the data on this is incredibly poor quality and largely you just have to use intuition.
An example is "gang related" shootings. afaik it's around 10-15% officially, but anybody who's ever worked in the justice system, policing, or just lived in a high crime area can tell you it's absolutely higher than that; and that's not even accounting for like, gang adjacent stuff (drugs, retaliations, dead people who just happen to show up in the most dangerous part of town controlled by rival street gangs but for whom there's no proof). The largest coded category of homicide in Chicago for example is "unknown", by far.
Okay. Thanks for checking
That stat sounds to me like it comes from somebody with an ax to grind.
No doubt, but I've heard about the horrors of US gun deaths in every possible media several times a year since age 6, and I had never heard about any major aircon-driven disparity in deaths. So maybe I've been in a massive ax-grinding convention the whole time without ever realizing it!
I wasn't questioning the gun deaths stat (though Erica, who is generally very accurate, says the figure is 47,000 per year, not 60,000. But it is certainly LARGE). I was questioning the stat that 30x as many in Europe as in US are killed by heat.
As Erica Rall's comment points out, it is important to use a consistent methodology and time period when comparing temperature-related deaths across locations. One such analysis is Gasparini et al. (2016), which used a statistical approach to estimate that extreme heat-related deaths are about 0.2% of all deaths in the U.S. (or about 6,000 deaths/year). For comparison, their estimate for the Italy is 0.7%, for Spain is 0.5%, for the U.K. it is 0.2%, and for Sweden it is less than 0.2%. So the U.S. is within the range of European countries, if near the lower end.
AC is likely part of the story but seems unlikely to be even half of it. The U.S. and U.K. have similar heat-related mortality rates but AC penetration is 90% in the U.S. and only 5% in the U.K.
A proper comparison should be controlled for summer daily maximums, I think. Maybe with moisture levels taken into account, it makes a lot of difference subjectively (but possibly less so for old and sick people). I expect the median European has it several degrees lower than the median US American.
> AC is likely part of the story but seems unlikely to be even half of it. The U.S. and U.K. have similar heat-related mortality rates but AC penetration is 90% in the U.S. and only 5% in the U.K.
Most of the US has hotter summers than UK. Both Sweden and the UK are far to the north of the contiguous US. The proper comparison would be the southern parts of the US and southern Europe.
Or paris - which has recorded a lot of heat deaths - and Seattle although Paris is further north than Seattle.
The estimates I quoted from the Gasparini et al. paper are based on relative temperatures in each location (you can see more details in the excellent Our World in Data post linked below). So even though there are more hot days in an absolute sense in the U.S., there aren't necessarily more hot days per year in a relative sense in the U.S. versus the U.K. A important subtlety comes from the fact that these relative values are calculated based on each location's temperature at which temperature related mortality is lowest (the so-called "minimum mortality temperature"). That value is probably affected by the presence of AC, so things get a bit circular.
https://ourworldindata.org/part-one-how-many-people-die-from-extreme-temperatures-and-how-could-this-change-in-the-future
Do you know what temperature/humidity curves are? Can you tell me when humans die? When all other mammals die?
Madrid has the same latitude as New York, Rome is further north than Chicago.
The southern US is way further south and more tropical than any part of Europe.
This is part of the problem in making easy comparisons about the climate on the two continents.
New EA cause. Ship window AC units from US to Italy and just start installing. Maybe Italy lacks the state capacity to stop it. Give the mafia its cut in Sicily.
Possible drawback. Do window AC units actually work?
One problem with this approach is that they use sash windows in the US and casement windows in continental Europe.
>> Possible drawback. Do window AC units actually work?
If not, then why would anyone install them - to look ornamental? Anyway, they definitely work.
I briefly looked into this when it came up on Marginal Revolution a few days ago. Simple answer is no, the ~60,000 and ~2000 are not comparable, Europe stat is based on excess deaths (in a particularly bad year, 2022), while the US number was based on 'Deaths attributed to exposure to excessive natural heat as the underlying or contributing cause of death' ie, a doctor has specifically recorded it as heat related.
I'm sure there is a long answer somewhere too, that probably does show that more aircon in USA saves lives. However it almost certainly isn't as stark as that stat makes it seem.
Yes. It’s true. The lack of air conditioning is a death sentence as Europe warms faster than the rest of the earth, although maybe not as high as claimed.
People who live in AC controlled environments don’t really get what it’s like without. If you are in a city like London where the temperature gets higher these days (up to 32C last week) then you should realise that 32C is what you get in Hyde park and some leafy suburbs. The temperature in certain streets can be ten degrees higher, the tube is 40c+ and any indoor space is going to get hotter over the day, particularly south facing buildings with windows that trap the heat. There’s no relief indoors.
I walked into a cafe a few years ago and promptly left as the heat and steam were overwhelming. People work in kitchens in these conditions. I myself measured 40c in a room in a hotel which was a converted Georgian house and didn’t have AC.
I live in Vienna. Even though theoretically we have less cool weather than London - less rain, less fog - I have a mobile A/C unit which I haven't used for ten years now. It was just never really that bad. The office had A/C of course, and evenings were tolerable in swimpants, dropping to like 25C for the night.
> Yes. It’s true. The lack of air conditioning is a death sentence as Europe warms faster than the rest of the earth, although maybe not as high as claimed.
I don't understand, cheap mini splits are like $500-$700, why is this such a crisis? That's well within the means of any restaurant or business owner, and most individuals.
This whole discussion has been kind of baffling to me. Certainly if it's so bad people are literally dying, not just feeling crappy, somebody's going to break and spend the $500?
Often they are not allowed because they make the facade of pretty historic buildings ugly. This generally tends to be about a lot of things in Europe. The tourism industry is big, so the government wants the city to look like it is 1900 so no high-rises etc.
Wow, talk about "HOA's and historical preservation run rampant." They'd literally rather have people die than upgrade buildings to be livable??
I've always thought restrictions like this were bad, but this is a whole new level. And these are democracies and/or owner-produced regulations like HOA's! At the least, they should be able to vote to relax some of these restrictions so people aren't dropping dead.
I haven’t seen anyone dropping dead, so even I am not motivated to do this.
Sweden was all in on "covid19 will fix one of our problems!" (too many old people). [in all seriousness, it was because they have a ton of single mothers.]
And it's not helped by the fact that northern/western Europe is more used to cold, wet winters so there's been a big push on insulating houses to cut down on heat loss (and thus reduce energy costs).
With mild winters and heatwave summers, that doesn't work as well. Now we've cut out all the draughts, there isn't any air circulation and there isn't built-in AC to compensate.
... not sure where you're getting your info, but it can't be Britain. In britain, they have a special tool for breaking the ice on their toilets in winter.
If we go by standard scenarios and assume no AMOC collapse, winters in Nordic countries will still be cool enough that insulation is a good bet. If we *do* assume AMOC collapse...
Not sure how you do insulation and weatherproofing over there, but on our side of the pond, residential construction still has a ton of windows that you can open for ventilation and air circulation. When everything's closed, yes, there shouldn't be thermally significant amounts of air getting in through cracks and gaps, but you can still open stuff up when you want some fresh air.
Those aren't apples-to-apples figures. The European numbers are statistically estimated excess deaths related to hot weather, while the American ones are direct death statistics where a heat-related diagnostic code was listed on the death certificate as either the immediate cause of death or a contributing factor.
This study applies similar methods to the European ones to American data between 1997 and 2006 and finds 5600 deaths/year in the counties included in the study, which covered 62% of the population.
https://pmc.ncbi.nlm.nih.gov/articles/PMC7289128/
Europe is also potentially a much larger population than the US, depending on whether "Europe" means "Eurozone" or "all of the EU" or "Everything between the North Atlantic and the old Soviet Border" or the entire continent to the Urals.
Add it all up, and there's probably a higher death rate from heat in Europe than in the US, but nowhere near a factor of 30. My guess is that the European death rate per million population from heat-related causes, in an apples-to-apples comparison, would be more like 1.5x to 2x the US rate.
Thanks for providing needed context
While trying to work out the per capita figures is useful, unless there is another explanation for the excess deaths in that study other than a particular heatwave, why would we deny that it’s the heatwave that caused the excess deaths.
> This study applies similar methods to the European ones to American data between 1997 and 2006
But the more frequent European heatwaves happened later than that. Parts of the continent are 2-3c above pre industrial eras. I know Andalusia where I have a place is experiencing extremes most summer days, and has seen those kind of increases.
Nobody is denying that the excess deaths are (probably/partially?) caused by heat. That's the point of the excess death number documentation (at least in the way they are presented in Germany).
The point is that the US numbers don't come from looking at "excess deaths", but from doctors specifically saying and recording "This person died because of heat", which leads to much, much lower numbers in the US.
The point is, it's apples to oranges, the point isn't that apples don't exist.
>why would we deny that it’s the heatwave that caused the excess deaths.
I am confused. I don't think I was doing this, just trying to normalize two radically different statistics so it even begins to make sense to try to compare them.
>But the more frequent European heatwaves happened later than that. Parts of the continent are 2-3c above pre industrial eras. I know Andalusia where I have a place is experiencing extremes most summer days, and has seen those kind of increases.
America, too. As I noted in another comment, US heat deaths per capital are way up since 2000.
5600 deaths per year, covering 62% of population, seems to me like it would give a factor of 6? Still an enormous price to pay for no air conditioning… Gasparrini has 3x for Spain and 5x for Italy (Sweden and UK obviously aren’t the right comparison and are similar to the US). But these numbers should have gone up, as Europe has gotten hotter and older in the last 15 years, and still has no AC… even forgetting the US, the absolute numbers model are really high.
No AC doesn't sound right. When I lived in Madrid my apartment came with AC (and swimming pools were also common in apartment complexes, and most people have access to free or very cheap local swimming pools). In France I also have an AC. Every office I've worked in had an AC too.
In Spain it seems about half of homes have AC and anecdotally more and more people are buying ACs every year.
5,600 deaths per year in counties with a population of 174 million in 2000, which works out to 32 deaths per million population.
It looks like the 60k figure was a cherry-picked peak heat wave year (2022) while the estimate for 2023 was 48k. The full text of the study is behind a paywall so I can't see how they're defining Europe, but the abstract does mention 35 countries, which is about right for the European Union member states plus UK, Norway, Switzerland, and former Yugoslavia, or approximately the European continent minus the former Soviet Union and Turkey with the Baltic states added back in. Total population about 550 million, so 60k is 109 per million and 48k is 87 per million.
The death rate per million in the US in death certificate stats increased from 1.9 per million in 2000 to 4.7 per million in 2022 and 6.2 per million in 2023, a 2.5x and 3.25x increase respectively. If we naively use that to adjust the 2000 excess death estimate, that becomes 80 per million for 2022 and 112 per million in 2023. Which is much closer to the actual European rates than I was expecting.
You want *per capita* deaths, not total deaths, to make the comparison. (Plus taking into consideration the other factors Erica mentions.)
Totally agree... it still looks to me like the right ratio is a factor of 5 for 2022 and 2023.
It's comparing to US *gun* deaths, not heat deaths.
The 2,000 figure is heat deaths. The US has 47,000 gun deaths per year including accidents and suicides, or about 18,000 if you only count homicides.
I've had it with google search. Anyone here know of a search service that works as well as old google? Don't say AI, sometimes I need 100% reliability.
I don't think there's anything that meets this bar.
> Don't say AI, sometimes I need 100% reliability.
IMO you shouldn't rule out using an LLM as step 1 of a 2-step process. Taking AI outputs at face value on anything important is stupid, for sure. But they are often good at surfacing information that you can pretty easily verify once you have it (but that would have been harder to find from scratch).
I have heard good things about Kagi, but honestly the answer is AI even if you do need reliability. You can just use it as a search engine and not rely on its own interpretation of its results. Perplexity is specifically designed for this.
DuckDuckGo is...95% there. When I set it as my default, I will probably go check Google 1-2 month. Conversely, if I'm on a work computer where Google is the default, I basically never use DuckDuckGo but I'm also searching a far more limited range of topics.
Tried Brave Search?
Ruskies have a search engine that works well, at least for image searches. at least as well as old google. No opinions on whether using it will get you on a list, though. (In case it's not clear, I haven't used it myself, just talked with an old image analyst/coder).
The search engine in question is Yandex, I use it regularly. It's not old google good but often better than contemporary google, and tends to weight towards older sites (which is a good or bad thing, depending on your query)
Have you tried Google, but with the "Web" search instead of "All"?
I donated my left kidney to a stranger this year, and I've had a lot of time to think about it. Specifically, was it a good thing to do, or am I just a chump for doing this?
I feel like in a very good society, you're a good person and not a chump if you contribute good deeds, especially if the society has done good things for you specifically. And in a fully evil society, there's no room for making a difference anyway, you'd only be helping people who would never do anything good for you or anyone else, so it's obvious you should avoid being a chump there. I'm sure about both of those extremes.
And in a normal society that's somewhere half-way between good and evil, the answer is probably case-by-case. Giving money to a broke kid (that hasn't done anything wrong) versus giving money to a broke gambling-addict is a pretty obvious example of needing to think case-by-case, and you can probably brainstorm many examples like this. And of course, society leans more good or evil depending on your location and which group you're interacting with. So the problem is: you want to be a good person, and you want to negate the risk of being a chump, but society is a mixed bag, and you have very imperfect information to work with.
Is there a solution to this problem when donating a kidney? Surprisingly, there's 3!
The obvious first one is that you can just donate your kidney to someone you're already pretty sure is decent. But personally, I didn't know anyone dying of kidney disease, decent or otherwise. I can go try to find someone, sure, but people have a life-or-death reason to lie through their teeth here, and I'd feel miserable sifting through peoples' life stories trying to single out the one person to whom I should donate. But there are some donors with better people-skills and more grit who have absolutely done this, and there are websites 100% dedicated to sick people making their case.
The second option is the one I did by accident. If you donate your kidney undirected, it goes to an individual at the top of the kidney waitlist. Who is at the top of the kidney waitlist? People who've been waiting a LONG time, and people who already have someone donating a kidney on their behalf but aren't biologically compatible with that kidney. My kidney went to that second type, and from what I can tell, that's about as good a "I vouch for this person" signal as I'll ever see. I mean, it's not 100% impossible that the sick guy did something terrible to get someone to donate for him (guilt tripping? harassment? threats? LOST had a storyline about this). But most people just wouldn't do anything like that, and I learned 1st-hand that these USA hospitals work extremely hard to identify those bad situations. The doctors are risk-averse to an annoying degree, like Scott mentioned in his kidney-donation post.
The third solution is religion. Jesus bluntly tells people to be like the Good Samaritan, meaning do good things for people even if you have nothing in common with them and know nothing about them, if you want a chance at eternal life. Luke 10:25-37. Many of the hospitals refer to undirected kidney donation as a Good Samaritan donation because of how well the act maps to that specific teaching. I would think "because Jesus told you to do things like this" is a good enough reason for many Christians out there, but Jesus also explicitly states the reward of eternal life. I can't speak for any other religions, but I'd bet money that at least Judaism and Buddhism have similar teachings and stories.
I know you weren't fishing for praise, and I'm sorry I'm not really engaging with your thinking in any interesting way. But, good on you! You did a very brave and generous thing, and regardless of whether you're a chump, you made a huge difference to someone who needed it. (Even if the recipient is secretly a dick, the person who donated on their behalf very likely a) cares a lot about their outcome and b) is highly deserving.) Thank you.
Thanks, I'm just trying to accomplish as much as I can with my life. I was never going to be smart enough to do medical research or make life-changing tech, so doing this and blood donations + charitable donations seems to be my maximum.
My question is, is your lifestyle so superbly healthy that you are sure you won't need it later on? Me as an alcohol and tobacco addict living on random sandwiches, not so sure.
The hospitals are risk-averse, they're not going to let someone donate if they find big risks of any kind. They'll check all that background stuff and the kidney and blood metrics more times than even seems reasonable, though they don't seem to care if the donor is out-of-shape or eats mediocre food.
Me personally, I'll just say there's probably hundreds or thousands of things I'm more likely to die from than kidney failure.
My two pennies: I registered to be an orhan donor, but then during Covid I was so disgusted by lockdown adherents and th3 extremely hostule rhetoric of compulsory vaccination fanatics that I unregistered. Coming to the moral position that I would rather a lockdown or compulsory vaccination fanatic die than benefit from my orhans, if I were to die and donate them that is. This is based on the perception that in my country, Britain, most people supported these things so the odda of a donor recipient also doing so being greater than chance.
So I would also not donate a kidney to a random stranger, just on the same rationale.
There would be no dilemma here if you just got paid for the kidney. You would be offered a fair market value for the kidney by the person who had the most to offer for it, and there would be no further issues. Fulfilling a demand would be a good deed on its own.
markets and morality are not actually the same thing, and a lot of the people most able to afford a kidney (or anything else) in your system did horrible but technically legal things to get their money
But there wouldn't be a dilemma anymore, since it would just be a transaction instead of a charitable act borne of moral judgement. There would be no consideration of whether it was good deed of charity, as it would be no more than an equivalent exchange of goods. Isn't that the wonderful thing about capitalism, that it lets us move on from putting arbitrary value on things?
I think payments are the right path, but I have no doubt that payments would create plenty of new dilemmas. Just imagine a potential "donor" going through 12 months of testing and screening only to get rejected near the end of the process. If he's compensated for his time, then people would sign up even with no intention to actually undergo surgery. If he's not compensated, that's a giant loss of time for someone trying to make ends meet, and a huge risk for everyone applying.
> If he's compensated for his time, then people would sign up even with no intention to actually undergo surgery
That wouldn't be possible if they already signed over ownership, no? This is what contracts are for.
That's what I'm getting at, signing over ownership doesn't really work if a doctor can find something wrong with the kidney the day before surgery, at which point no one wants it anymore.
This isn't a problem when buying a house, you can back out of buying a house within the inspection period of 2 weeks or so, and the monetary cost and time costs are relatively small, but 12 months is huge.
Though now I'm thinking there's probably precedent here when it comes to surrogate mothers, maybe I'll look that up later.
The point is that if it is necessary to encourage participation, donators can be compensated for their time and then have their kidneys "returned" if the tests don't work out. Of course, you'd probably have them have to do some initial, non-compensated tests (like drug tests) just to make sure they aren't just wasting money...
What was the process like? Did you have to pay for it?
My experience was similar to Scott's: www.astralcodexten.com/p/my-left-kidney
The process is ~18 months long, almost entirely paid by the recipient's health insurance. The 2 big decisions are "who do you want to donate to?" and "do you want to donate at a hospital near your home, or fly to one of the best hospitals in the country and stay at a hotel?"
It's a lot of medical bureaucracy, lots and lots of blood tests and urine tests, and 6 weeks of recovery, but most of the recovery is in the first 2 weeks. You're relying on someone else for those 2 weeks, you'll need someone competent, calm, and predictable. And like Scott said, people should really be paid for this hassle, but the people currently dying on dialysis don't have time to wait for congress to finally allow payments or for artificial kidneys to be developed.
I want to give a plug for MedStar Georgetown University Hospital. They basically put all the bureaucracy and testing into a single day and the whole process was very smooth for me. They also receive a lot of the testing through Quest Diagnostics so you don't have to travel to them for all the blood/urine tests. I read Scott's article in October 2023 and donated June 2024 (probably could have done it a couple months earlier, too, but I wasn't available before June). So definitely not an 18-month process!
I also want to plug the NKF, who reimbursed me for travel, hotel, and missed work costs. The whole process really did cost me nothing (though I did front all the costs and was reimbursed after the fact).
Sounds a little smoother than what I got. I could have got the whole thing done earlier, but I had to schedule around both weather (don't want travel during snowy seasons) and the caregiver's busy calendar. I'm guessing that's pretty common, but I don't know.
I'm considering a career change into AI safety, and trying to learn as much as I can about it. I've found lots of great resources online but so far haven't been able to talk to anyone in the field. If that's you, please reply or email me at keavney.mark@gmail.com.
It's odd that girls can become pregnant right at puberty even though this carries quite increased risks for both the mother and child. How did something this poorly adaptive come to be? I've heard a few theories but maybe people here are more knowledgeable:
-Contemporary humans go through puberty much earlier than their ancestors, around 13. This is due to better nutrition and exposure to hormones from e.g. water treatment, both of which are recent occurrences. For most of history puberty wouldn't happen until the late teens.
-This trait produces more offspring and is thus more adaptive even though it causes more death. Looking at the numbers, if a girl starts bearing a child every couple years starting at 14 through her fertile years (again in the relevant time period this is probably the mid-30s), this provides more offspring than starting at 18. Even though more total mothers and children will die compared to starting later due to the increased risk.
Girls can become pregnant at puberty because that is how puberty itself is defined. I guess you want to ask why is puberty not happening at 20?
Because our biology is the result of kludging together different approaches over eons. When we were hominids tootling around the savannah and maybe not going to live past thirty, then being able to start replenishing the population from the early teens was advantageous.
Age of menarche in modern day human women seems to have gone up and down depending on the historical period, state of nutrition and health, etc. Currently it's in the range 10 to 16 years of age. 10 is clearly too young, though it's possible to survive pregnancy at that early age. It does seem to be shifting slightly earlier.
https://www.ncbi.nlm.nih.gov/books/NBK470216/
Then we have to consider the opposite end of the life span: what age do you kick the bucket? If you're going to have a good chance of dying by forty, then maximising reproductive years is the way to go. Nature cares naught for your qualms.
But since we're generally living longer, then we hit up against the problem of "so when are you no longer reproductively viable?" Age of menopause is (looking up online) 45-55 years, and before that you are in the perimenopause stage:
https://www.mayoclinic.org/diseases-conditions/menopause/symptoms-causes/syc-20353397
It's entirely possible to become pregnant and bear a living child into your late forties, and if we exclude women who used IVF/donor eggs to have children after the end of their own natural fertility/to overcome fertility issues, then "The oldest verified mother to conceive naturally (listed currently as of 26 January 2017 in the Guinness Records) is Dawn Brooke (Guernsey); she conceived a son at the age of 59 in 1997."
So theoretically at least you could have someone who began menarche at age 10 and finished menopause at age 55, giving forty or so years worth of child-bearing time (very few people are getting pregnant and having babies every year, year after year).
https://journals.sagepub.com/doi/full/10.1177/03631990231196822
"The European Fertility Project of the 1980s (usually called the Princeton Project) showed that by 1911 Ireland had the highest marital fertility of any European country. A special survey as part of the 1911 all-Ireland census revealed that of women who had been married 20 years or more (and, hence, were likely to have completed the full cycle of childbearing), roughly 50% had born five to nine children and an additional 30% had given birth to 10 or more."
Evolution hasn't caught up to modernity yet. Our biological systems developed to function during adversity, so now when they note "environment peaceful, food plentiful, animal protein abundant" the conclusion is "conditions ideal, let's make babies!" You remain healthy and able to bear children? Keep on making those babies! The babies are not dying and the population is not being winnowed out by tigers, other predators, famine and disease? Hurrah for us, the species will survive, don't stop making those babies now!
Maybe under a thousand years of early puberty, the systems would finally make adjustments, but we're taking matters into our own hands now and not depending on the forces of nature to slowly and cumbrously shift.
> Maybe under a thousand years of early puberty, the systems would finally make adjustments
Probably not in the way we would hope for, but rather by growing the population until the resources are exhausted and again only 2 children survive on average (but many more are born).
Nature does not care about what is good from the perspective of aesthetic, or for the group. The individuals who volunteer to have fewer kids have reduced the frequency of their genes in the pool (unless there is a tradeoff where having fewer kids means higher chance of *your* kids specifically surviving), so any genetic impulse to do that would be self-eliminating.
See also: https://www.lesswrong.com/posts/QsMJQSFj7WfoTMNgW/the-tragedy-of-group-selectionism
As Mary Catelli has pointed out, it really depends on mortality. The additional risk only needs to be less bad than the overall risk for it to make sense, with further discounts depending on the kid's chances of survival. So if a median hunter-gatherer girl of 14 has a yearly chance of dying at 5%, waiting 4 years might easily be worse than risking it early.
Bear in mind that menarche isn't the same thing as reproductive maturity. The first couple years consist of mostly-anovulatory cycles. The system takes a while to come fully online.
If there are individual behavioral dispositions, or sufficiently widespread cultural dispositions, that prevent people from engaging in activities that tend to make them pregnant, for the first few years after puberty, then evolution would have no reason to biologically eliminate the possibility of pregnancy at those ages.
I suppose we don't know much about hunter-gatherers, but for the overwhelming majority of historically attested societies, there aren't
What's relevant here is the _completion_ of puberty which is a multi-year transition. For centuries across many cultures the word for that was one borrowed from ancient Greek, "menarche", meaning first monthly period. Cultures all over the world have understood for millennia that this step signaled the possibility of becoming pregnant.
The idea that modern people are _starting_ puberty earlier has been contradicted by archeological analyses:
https://link.springer.com/article/10.1007/s12520-017-0543-0
https://theconversation.com/children-arent-starting-puberty-younger-medieval-skeletons-reveal-91095
What may be true (though not yet settled science) is that modern people in the developed world _complete_ puberty faster with improvements in nutrition being a key factor. This hypothesis is arguably supported by for instance medieval history: European royal and noble marriages routinely involved girls aged 14 or younger with the production of an heir being their immediate job description. [I happen to be reading a biography of Henry III of England, who at age 28 married the 12-year-old Eleanor of Provence; she was 14 when the future Edward I was born. That scenario was quite common among the kings, dukes, earls, counts, etc.] Since royals and the nobility had much-more-reliable and somewhat-more-varied food supplies than everybody else, their daughters completing puberty relatively quickly would jibe with the modern hypothesis.
Thank you for this.
Statistically, it takes a lot of effort to get girls pregnant "right at puberty." There are reasons why "you don't get pregnant your first time" gets bandied around -- it tends to work for sixth graders, for god's sake! (To put it a little more bluntly, they get pregnant and have a "heavy period" when they abort the baby).
Secondly, I don't think you have the same "increased risks" if you have a normal distribution of "who gets pregnant" near their menarche, particularly for the baby.
Puberty happened in the late teens in Scandinavia. Early teens was normalish for better nutrition areas, like the Fertile Crescent. There's a reason 12/13 is "manhood" in jewish culture.
> There's a reason 12/13 is "manhood" in jewish culture.
I've wondered about that. Latin American culture puts womanhood at 15. Chinese culture also puts it at 15. The USA uses "sweet sixteen".
These are the times when a girl was considered "definitely on the menu" for marriage. I'm pretty sure "womanhood" (as in being a maiden) is defined lower. Victorian England certainly had it defined fairly low. You might have more luck looking at "historical rapes" than "historical marriages" (again, you can reach puberty but, even when sexually active, not get pregnant particularly with kids of your own age group.)
https://en.wikipedia.org/wiki/Bride_kidnapping
This states bride kidnapping as young as ten, which is probably below menstruation age, and probably to "make sure she's a virgin." One hopes that there is no physical damage from sex with one that young.
I presume that most bride-kidnapping is done with people of "marriageable age" (that is to say, actually capable of bearing a child), in that consummating the marriage is part of the whole ritual. (and yes, it's important to note that in cases of "couldn't afford the dowry" this is an acceptable alternative that Women often agree to).
Citing the menarche wiki:
"From the sixth to the 15th centuries in Europe, most women reached menarche at about 14, between the ages of 12 and 15.[31] The average age of menarche dropped from 14-15 years"
> These are the times when a girl was considered "definitely on the menu" for marriage. I'm pretty sure "womanhood" (as in being a maiden) is defined lower.
I think you actually have it backwards. You can marry anyone. But womanhood is a defined role. From Mary Beard's SPQR:
> the age gap of forty-five years caused puzzlement even in Rome. Why had Cicero done it? Was it just for the money? Or, as Terentia claimed, was it the silly infatuation of an old man? In fact, he faced some direct questions about why on earth, at his age, he was marrying a young virgin. On the day of the marriage he is supposed to have replied to one of these, 'Don't worry, she'll be a woman [mulier] tomorrow'. The ancient critic who quoted this response thought that it was a brilliantly witty way of deflecting criticism
Publilia was a valuable marriage prospect as a girl. Marriage wasn't enough to make her a woman. [That job, obviously, is done by consummation. But it could be done by aging too.]
Sorry, I was using the definition of maiden. As in defined against "mother" and crone.
Huh? To the extent that "maiden" has a specific meaning, it means "virgin". Even a crone can remain in possession of her maidenhood.
Isn’t it almost inevitable that if an organism goes from “can’t do X” to “can do fully do X” there’s going to be some intermediate stage of “can do X poorly and with some risk,” especially if X is at all complicated?
E.g., a newborn can’t eat any solid food, a four-year-old can eat anything, but in between you get a six month old who can have purées under close supervision, a nine month old who can start to have small bits of finger food but needs to be watched for choking, an eighteen month old who should still have things like grapes pre-sliced, etc.
For most of human behaviors this would be true (kids learn to walk before being able to walk without falling, for example), but for the menstrual cycle specifically it is not true. Women already have a distinct toggle from going "can not get pregnant" to "can get pregnant", so presumably the time when that toggle is flipped was the optimal time for reproductive success in the ancestral environment, and it makes sense to question why it happens before the woman is fully grown.
"Can get pregnant" is kinda funny, if you've got 80% of cycles the first year as "non ovulating". That's no toggle.
PLEASE read more. Girls can get pregnant before menarche (sometimes considerably before menarche). Women do not have a distinct toggle from "cannot get pregnant" to "can get pregnant" -- the toggle is dependent on amount of sperm (also quality), and number of sperm donors, and it's really complicated! And "can get pregnant" is very different from "can bear baby" (given the number of spontaneous abortions in young populations, which is "generally at least two or three before you get a surviving baby")
This also gets into some of the evolutionary biology "answers" that are really dumb.
1) It could be tied to something that enhances reproductive fitness (aka "earlier puberty meant not dying of the plague before you had ONE kid").
2) Women who were raped and didn't bear babies might have been more likely to be killed.
3) "Picture unrelated" -- you're dealing with a contemporary environment that is very different from the original environment, and can't generalize.
The last one is less likely in this particular case...
Alsotoo, I thought "fully grown" for women was around 14-15 (height specifically)? That matches with "time of menarche" pretty well, historically.
> Girls can get pregnant before menarche (sometimes considerably before menarche).
Unless you consider menarche to be delayed by pregnancy, I don't see how that can be true. I don't think that's a justifiable choice; given a woman leading a traditional lifestyle who has been pregnant so often that she can't remember the last time she had a period, and maybe never had one, everyone would consider that (a) she has been through menarche; but (b) by coincidence, she's never shed her uterine lining outside the context of giving birth.
> Women do not have a distinct toggle from "cannot get pregnant" to "can get pregnant"
Yes, they do. They have to start ovulating, which immediately toggles them from "cannot get pregnant" to "can get pregnant". You seem to have confused the concepts of "can get pregnant" and "will get pregnant, if inseminated".
You might also note that spontaneous abortions remain common at every age that can get pregnant. The fact that they're common in young girls doesn't tell you anything. They're common in everyone.
According to my mother, this causes her patients a lot of grief, because they get pregnant, proudly tell the people in their lives, and then miscarry. She always warns them not to tell anyone until the pregnancy is fairly advanced.
Forced ovulation through rape is a thing, dude. It can occur before menarche.
Quoting from wiki's menarche page:
"In post-menarchal girls, about 80% of the cycles are anovulatory in the first year after menarche, 50% in the third, and 10% in the sixth year" -- this is probably being measured without sexual activity.
But, dude, does that sound like a switch to you?
>> You seem to have confused the concepts of "can get pregnant" and "will get pregnant, if inseminated".
This is particularly important when even most people who reach puberty die reasonably young.
General question for people more knowledgeable about AI than me: to what extent have increasing lengths of tasks AI can complete been permitted by increasingly effective use of the chain of thought, vs 'implicit' knowledge not spelled out every step of the way?
It depends on what you mean by 'implicit knowledge'. I'm assuming you mean 'better reasoning traces' but correct me if you mean something else
Naively I think it's hard to untangle how much AI has benefited from context window increases vs. better reasoning traces. They are sorta intermingled. We know that there is a meaningful improvement from increasing the size of context windows (cf. Gemini 2.5 vs Claude 3.7) but also that there are obvious improvements from just having better reasoning traces (cf. Claude 4 vs Claude 3.7).
I've noticed from reading ACX that I can actually enjoy poetry, which is pretty great! I haven't read much yet though. Does anyone have recommendations? I don't want to specify what I like, because I'm pretty sure I haven't even read most of the genres and I want to get a good understanding of what's out there. I do like poems that I can memorize.
There are plenty of anthologies that collect the most popular poems in English. Here's one:
https://www.amazon.com/Classic-Hundred-Poems-All-Time-Favorites/dp/0231112599
That would be a good place to start.
For contemporary poetry, I love Dorianne Laux.
https://www.poetryfoundation.org/poems/58276/heart-56d23c80cdcc8
https://www.poetryfoundation.org/poems/54716/dust-56d235633c79b
https://poems.com/poem/the-tooth-fairy/
https://www.saltproject.org/progressive-christian-blog/2023/4/10/psalm-by-dorianne-laux
Depends how modern you want to go, and at what stage you want to halt (I'm really badly behind on my poetry reading and can't think of a contemporary poet I know or want to read).
If you like "poems I can memorise", then how about a poem you can sing? Yeats heard a few verses of a song, some lines stuck with him, he wrote a poem out of them, and then that poem was turned back into a song:
https://en.wikipedia.org/wiki/Down_by_the_Salley_Gardens
https://www.youtube.com/watch?v=027ZJX5XVjs
(Ignore the spelling there, that should be "salley" - salley trees are willow trees, sáile being the Irish word for willow).
Another poem by an Irish poet turned to a song:
https://allpoetry.com/On-Raglan-Road
https://www.youtube.com/watch?v=ZIqr1Ge8Z5w&list=RDZIqr1Ge8Z5w&start_radio=1
I also like Eliot (took me a while to get there) but he doesn't really have easily memorisable ones (well, except "Old Possum's Book of Practical Cats" which got turned into the musical "Cats" - let's not talk about the movie version).
Try Cavafy. A few examples:
https://www.poetryfoundation.org/poems/51296/ithaka-56d22eef917ec
https://www.poetryfoundation.org/poems/51294/waiting-for-the-barbarians
https://www.onassis.org/initiatives/cavafy-archive/the-canon/young-men-of-sidon-400-ad
I also translated one myself:
https://logos.substack.com/p/translation-75f368c011dc
And wrote a review of his poems:
https://logos.substack.com/p/book-review-cp-cavafy-collected-poems
Kipling is probably the best choice. But if you want something a little closer to the modern day, you could look at Shel Silverstein.
Rudyard Kipling. Pretty much all of it, but the famous ones are famous for a reason. "Danegeld", "White Man's Burden", "If", "Gods of the Copybook Headings", "Gunga Din", "Female of the Species", all good stuff. The "Ballad of East and West" is long, but worth it. And Kipling is about as close as we get to an Engineer Poet, so if you lean in that direction there's "Sons of Martha" and "Hymn of Breaking Strain".
This blog seems to attract people who like Rudyard Kipling, so you might try starting there.
Martin Gardner, "The Best Remembered Poems". This collection varies wildly by quality and genre and emotional level and... which is its strength. Some poems will appeal to you and some not, they will all be readable and potentially memorable.
In general, I hate poetry. But I was addicted to Byron when I was young. He has lots of short poems that I remember still.
My degree, which I am doing now, introduced us to William Cowper and Philip Larkin, and I enjoyed both. I've also enjoyed WH Auden, Ted Hughes and John Betjeman. You've probably heard poetry from all of them in bits of movies that make you cry..
Some of my favourites:
- Dover Beach, by Matthew Arnold
- What Lips My Lips Have Kissed, by Edna St Vincent Millay
- Basically everything by Yeats (Sailing to Byzantium, When You Are Old And Grey, Second Coming, ...)
- High Flight, by John Gillespie Macgee Jr.)
- Acquainted with the Night, by Robert Frost (and many others also. I especially like 'Desert Places').
- A Clear Midnight (Walt Whitman)
- Do Not Go Gentle Into That Good Night (Dylan Thomas)
- Spring and fall (Hopkins) – listen to some recordings of Hopkins to get into his style, as 'his wríting can bé a bít wéird' but it sounds great when spoken)
I also like to read poems to memorise – the above are all very lyrical and memorable.
Well-regarded poets that I haven't yet learned as well as I ought: Keats, Auden, Poe, and many others.
Deiseach recommended 'Ceasefire' by Michael Longley on a thread about Gaza, which I was really taken by. Hopefully she chimes in with some more recs here!
Thank you for the kind words! Yes, Hopkins and sprung rhythm - he is a great poet, but you have to tune your ear into it.
Seamus Heaney of course, Yeats, the old warhorses of Irish poetry. I keep recommending this one but it's because I love it so (the link for the proper formatting of the poem because Substack hates us and our comments and doesn't want to give us the tools: https://poets.org/poem/vi-stares-nest-my-window)
VI—The Stare’s Nest By My Window
The bees build in the crevices
Of loosening masonry, and there
The mother birds bring grubs and flies.
My wall is loosening, honey bees
Come build in the empty house of the stare.
We are closed in, and the key is turned
On our uncertainty; somewhere
A man is killed, or a house burned,
Yet no clear fact to be discerned:
Come build in the empty house of the stare
A barricade of stone or of wood;
Some fourteen days of civil war;
Last night they trundled down the road
That dead young soldier in his blood:
Come build in the empty house of the stare.
We had fed the heart on fantasies,
The heart’s grown brutal from the fare,
More substance in our enmities
Than in our love; oh, honey-bees
Come build in the empty house of the stare.
Thanks for sharing the poem, it's very good.
Haviv Rettig Gur, an Israeli English-language journalist, has been rising in popularity; his is a definitely pro-Israel position, but full of nuance and interesting historical and sociological details. I liked two of his recent videos enough to make my own transcripts (I can't watch long podcasts), and thought I'd link them here, with a representative paragraph or two from each.
On the recent Haaretz report on IDF soldiers shooting Gazan civilians near aid distribution centers: https://www.youtube.com/watch?v=J59b2XWAVe0
Transcript: https://misc.lovestwell.org/transcriptions/haviv-ep25.html
Excerpt:
"One of the most damning things in the Haaretz report was the claim that sometime in the last 10 days, when it became clear this was a serious problem that was recurring, it wasn't a one-off incident and who knows, it was a serious problem that was recurring, it was systemic. There was a meeting at the Southern Command with the top brass in which Southern Command senior officers discovered in that meeting that soldiers sometimes, on occasion, I don't know if it's two times or 40 times, but on occasion, had used artillery shells as warning shots to keep civilians away. Artillery shells are long distance. Artillery shells are broad area explosives. Artillery shells are a tragic disaster in a situation like this. It serves no purpose. If you need an explosion to catch people's attention because it's turning into a stampede, there are small grenades that have almost no explosive. They have no shrapnel. They just made a loud bang. And every infantry platoon in the IDF infantry has them. You don't need an artillery shell.
The problem wasn't just the bad decision. The problem wasn't that that decision may be criminal. I don't know. We know that there are investigations opened by the military advocate general into some of these incidents. We don't know exactly how many, but that is something that the army has said and the Haaretz also reported, and others have reported as well in Israeli media. But the fact that they didn't even know, there was no clarity among the high top brass of the Southern Command, what's actually happening at these centers. That's the damning thing."
Podcast with Sam Harris (recorded June 17, during the war with Iran):
https://samharris.substack.com/p/making-sense-422-zionism-and-jihadism (2:15 hours)
Transcript: https://misc.lovestwell.org/transcriptions/harris-haviv.html (warning long, 22k words, but worth it IMO)
Excerpt:
"And so when Hamas looks at the Jews, it doesn't see a colonialist oppressor. It sometimes uses language like that, especially when it's talking to other Palestinian movements that were born in that kind of discourse. But it fundamentally sees a great, vast theological problem in all of the weaknesses of Islam and all the indignities that Islam has suffered everywhere from European imperialism and various other retreats that Islam has suffered.
The worst one was Zionism. And the reason the worst one was Zionism, and they talked about this openly and constantly in the 1890s already, was that the Jews were weak. They're refugees. They're the weakest thing that ever pushed Islam back. And so they're the first thing that Islam has to overcome in order to get back to its rightful place in history.
So Hamas develops an ideology and a foundational strategy. It is willing to destroy Gaza on the altar of destroying Israel. Because in its vision of the purpose of the war on Israel, it's not about liberating Palestine. It's not about creating a new Palestinian political world that is independent of Israeli military rule, which is something that significant numbers of Israelis would join in on. It's about overcoming the Jews, ruling everything that had ever been lost to Islam as the beginning of the return of Islam into God's embrace and the sign that our piety is enough and the sign that Islam is not taking its rightful redemptive place in history once again. It is overcoming centuries of retreat and weakness in Islam.
And so, yes, the second we lay down our arms, they'll come for us. They is some of them, not all of them. It's big, it's complex. There's also the simple truth that when you talk about Islamic ideas and Islamist ideas, you're not necessarily talking about the Muslim shopkeeper in some street in Amman, Jordan or something. The people who belong to this religious world don't necessarily subscribe to the grand ideas. There's a lot of overlap, but they're not the same thing. But with all those caveats...
Yes, we are cartoon characters in a vast redemption story that they think they're acting out and living through. And the same is true in a different way, in a Shia version with a slightly different history, although borrowing a lot of these big ideas of the Iranian regime. One of the most ridiculous things about this war is why the heck does Iran even care about Israel? It has no border with Israel, it owes nothing to Israel, it lost nothing to Israel, it has no interest in Israel of any kind, and it has spent hundreds of billions of dollars its people don't have on destroying Israel. Why?
And the answer is, there is this grand redemption story they think they're embedded in. So everything you said is absolutely correct. The day we lay down our arms, they come kill us and think that it's a great miracle given to them by God. And it's the beginning of a Muslim redemption story of the conquest of the world. And it comes from a deep 150-year-old discourse that we have to understand, that we have to actually respond to and deal with and ultimately also defeat."
Hahaha. Why does Israel care about Iran? That's the better question. Iran has a nuclear program that they haven't wanted since around 2000, and have been (by bibi's own reckoning) "almost at getting the nuclear bomb" since 1991 or so.
Iran will say many things, but it is what they DO that matters, and they aren't trying to "bring the pain" to Israel (test drive? yes. Show that they could hit whatever they pleased? Yes. Show that they know where Israel has their nukes stored? yes.)
Most of Iran doesn't care jack about the whole "take over the world" ethos of Islam. The leadership doesn't want a regional war. Nuke Israel, and they lose all hope of peace -- with anyone and everyone. Nobody likes Iran, anyway (except, a little, Russia and China, certainly not the Arabs).
> Hahaha. Why does Israel care about Iran?
Because for decades, Iran has been funding and equipping various proxy groups that attack Israeli military and civilian targets.
> […] they aren't trying to "bring the pain" to Israel
That's a straight up lie.
Well said!
Ah, yes, you're the person who ... no, wait, when confronted with information that doesn't match what you've been told, you confidently assert that you already know everything.
When was the last time you had a discussion, formal or informal, with Iranian Military personnel?
So you are saying that Iran is not supporting Shiite militant religious extremists like Hamas of Hezbollah, whose main goal is to fight Israel?
Or that Mahmoud Ahmedinejad did not call for Israel to be "wiped of the map"?
Having formal or informal discussions with military personnel (aside from it being impractical for most Westerners in the case of Iran if you want to get anything except carefully vetted answers) is unlikely to give you much enlightenment, generally. "So, Wehrmacht grunt, how does Germany feel about Poland?" ; "Today I want to discuss with a private from the USMC if the US will wage war against Saddam again".
First, Hamas is Sunni. Seriously, do you know nothing about anything? Iran doesn't support Hamas directly, either. Hezbollah's main goal appears to be bombing Kiryat Shmona, which is nearly always evacuated. I'm not sure what they did, but Hezbollah could certainly hit other targets now and then. They could hit tel aviv if they wanted.
Yes, a man "nominally in charge" called for Israel to be wiped off the map. He also called for the destruction of America. The leader of Jordan provided the Protocols of Zion to a jewish american diplomat (kissinger if memory serves), who had to sit there and smile back at the trolling.
Ah, so you don't have some magical pipeline to our "military diplomats." Iran likes to call their shots, whether they're shooting at Israel or at American military bases. Called shots aren't nearly as destructive, or trying to be blatantly murderous, as shooting in the dark.
Do you believe Bibi when he says that supporting Hamas is a good way to preserve Israel? If so, what do you think he would say about Iran's support of Hamas?
What are the most interesting last words you’ve ever heard of? (or have come up with yourself?)
I’m anticipating dying within the next century or so, and want to get ahead of things in planning out my last breath.
Right now I’m considering “The gold is buried under the…” along with leaving notes in my notebook every few months saying I purchased a few gold coins, totaling a few tons of gold over my lifetime.
Maybe you all can come up with something better?
I hope my last words sound something like, "Ahhhh...", as I die of a massive heart attack in the middle of the night. (my favorite idea of how to die... it's how my father did it.)
Cracked.com had an old article about last words. The one that stood out to me was "more weight", Giles Corey's response to a court trying to crush a confession out of him.
(Looks like the article's broken in exactly the most annoying way, where everything is still there except for the actual last words. https://www.cracked.com/article_16354_the-11-most-badass-last-words-ever-uttered.html)
You could pull a John McAfee and publicly proclaim that if you die, it's a conspiracy not suicide, and then commit suicide anyway.
> Maybe you all can come up with something better?
Harsh - instead of sending your own descendants and / or family on wild goose chases for lolz, why not do something funny and possibly marginally positive with your years-long paper trail, like try to out top TSA officials as double agents in the employ of China / Russia, or disclose the location of a years-deep cache of secret love letters between you and Kim Jong Un, or something?
Knowing me, and since I've only had an actually skilled doctor whose decisions I respected all of once in my entire life, my last words will probably be something like "My dying wish is that you'll sue these idiot quacks into oblivion. PROMISE ME!"
Unless we actually have AI doctors by then, in which case I'll just go for that old standby:
"The play is over, applaud."
I see it more as a game that makes life interesting. Blackbeard's Treasure, Curse of Oak Island, The Essenes lost treasure, etc. have interested people for centuries and created a lot of fun stories and speculation.
Also, I am lowkey one of those cryonics nutjobs, so I figure that increases my chance of future revival. There would be a lot more motivation to use advanced technology to revive a frozen corpse if they thought the location of a hidden treasure was locked away in your brain.
"You must pardon me, gentlemen, for being a most unconscionable time a-dying"
-- King Charles III
"I believe we shall adjourn this meeting to another place."
-- Adam Smith
Some that I've heard - several likely apocryphal - that have stuck in my mind, for one reason or another:
Spike Milligan: "Told you I was ill"
King George V (on it being suggested to him that he might recuperate better from his illness in the sea air of Bognor Regis): "Oh, bugger Bognor"
Admiral Horatio Nelson: " 'Kismet', Hardy" (sometimes reported as "Kiss me, Hardy"; see here for the full -rather amusing- treatment: https://www.youtube.com/watch?v=JPgn1bL3o_o )
(Sometimes, when stood in Trafalgar Square, I like to imagine to myself that that Nelson's actual last words were "Okay sure I've done some cool stuff in my life but for heaven's sake don't put me on a pedestal")
Dickens' Syndey Carton (upon taking his true love's husband's place at the guillotine, to enable true love and husband to be together): "It is a far, far better thing, that I do today, than I have ever done."
Julius Caesar: "Et tu, Brute" ("You too, Brutus?!"; satirised by schoolboys, of course, as "Et tu, couldn't finish the third"..)
Karl Marx: " 'Last words' are for fools who think they haven't said enough already"
According to the private diary of Lord Dawson, the physician who was attending George V at his deathbed, George's last words were "God damn you", directed at his nurse as she administered a routine sedative to him.
After the King went to sleep, Dawson took it upon himself to administer a lethal dose of morphine and cocaine, making him the last documented British regicide. Part of his motivation, as documented in his diary, was to ensure the King died before the print deadline for the morning edition of The Times so the announcement could first appear there rather than in a less-respectable evening paper.
Don't you mean the Times' mourning edition...?
(Sorry. I'll get me coat..)
I'm always amazed at the breath and depth of knowledge ("cross-section of knowledge"?) of ACXers! Wouldn't have guessed I'd hear George V's doctor's diary paraphrased here!
As a matter of fact I strongly suspect several of the last words that've stuck in my mind (including "Bugger Bognor") are apocryphal; edited my comment to say as much. Thanks!
>(Sorry. I'll get me coat..)
No, please stay! That was a really good one.
>Wouldn't have guessed I'd hear George V's doctor's diary paraphrased here!
I came across him via Wikipedia's "List of English regicides", which I was browsing for unrelated reasons some time back. Something to do with Cromwell and the Restoration, IIRC. Dawson isn't there anymore, since some editor decided that unilateral involuntary euthanasia doesn't count as regicide.
Not my grandmother's last words to *anyone*, per se, but her last to me - after I (admittedly) theatrically kissed her goodbye on the forehead before flying home to my job so I didn't lose it - were, "Oh, so now you think I'm going to die, huh?"
And with that, she won the final round of a two-decades battle over her beloved Christian Science (the practice of which, it should be noted, was what was killing her pointlessly and painfully 20 years early). Even though she was obviously dying and was extremely unlikely to survive the two weeks until I could return, she nevertheless startled me into totally submitting to the religion's pathological denial of reality which she helped train me in as a child.
"What? No! I'm coming back to see you, I have my ticket and everything...!"
I knew I was lying to please her, she knew I was lying to please her. It was a masterstroke of getting me to *finally* say exactly what she wanted to hear.
I ruefully admire the hell out of it, and don't begrudge her the victory.
Voltaire, asked on his deathbed to renounce the Devil, "This is no time to be making new enemies."
Well, it was nice knowing you.
That was the last thing my mother said to me before she died. I mean, she smiled and shook my hand. It seemed appropriate, and I looked back and smile at it.
Personally, as an epitaph, which is not the same as last words, I would like "offer good while Supplies last"
I'm pretty sure the best epitaph is "No Refunds."
(Or possibly "no refunds, sucker.")
Well, I guess it depends on how friendly a shopkeeper you are at the end of it doesn't it? How about "I'm sorry we're closed. Please come back another time.”
"I have discovered a truly marvelous proof of this, which this margin is too narrow to contain."
If one wants to keep in the theme of famous mathematicians, but go with good last words that actually are last words, there is always:
"Don't weep, Alfred! I need all my courage to die at twenty."
(Which would, perhaps, be merely fine and not excellent last words coming from your average Joe. But when you've already made significant and lasting contributions to the field of mathematics, the impact is rather different.)
https://en.wikipedia.org/wiki/%C3%89variste_Galois#Final_days
Fermat?
(Yes, yes, he didn't die in a duel the next day, he lived for nearly thirty more years; this won't be the last apocrypha written in a comment here, I'm sure.)
Are you confusing Fermat with Galois?
I don't think so.
https://en.wikipedia.org/wiki/Fermat%27s_Last_Theorem
https://en.wikipedia.org/wiki/Pierre_de_Fermat
Fermat scribbled in the margin of his book in 1637. He died in 1665 without having elaborated on it to anyone's knowledge.
Galois did die in a duel, and he was apparently writing about math the day before, but there -was- an apocryphal tale of the same having happened to Fermat, possibly due to someone being low on coffee somewhere in PBS' research department (I recall the tale on some TV documentary). Possibly confusing the two, just as you say.
I never heard of that tale about Fermat's death. On the other hand, Evariste Galois is famous for dying in a duel after spending the night writing about his theory. But I believe you that it is someone else who confused the two. Apologies for implying it was you.
Convicted murderer Thomas Joseph Grasso carped about his last meal.
"I did not get my SpaghettiOs, I got spaghetti. I want the press to know this".
Some favorite last words:
Oscar Wilde: "Either the wallpaper goes, I or do."
Mozart (after a case of fine wine, sent by a fan, was delivered while he was on his deathbed): "Pity, pity, too late."
General John Sedgwick: "They couldn't hit an elephant at this distance."
I've always seen that last rendered as "They couldn't hit an elephant at this dist--"
It's the funniest version of it. Sadly, reality is usually less poetic than fiction. Those technically weren't even his last words, just his almost last ones. After he said the elephant line a soldier said something clever back to him and the General's last words were laughter and "All right, my man; go to your place."
Similarly, the wallpaper remark was said by Wilde on his deathbed, but probably wasn't his actual last words. And the wording of the wallpaper joke has about a dozen different versions that have been reported over the years. And while Mozart did say the wine thing shortly before death, his actual last words were probably something else.
But why let reality get in the way of a good final sentence? For most people their last words are something like "Good morning" or "I don't feel so well" or "Blergh. Ugh. Gack." I have no idea what my grandfather's last words were, but since he spent the last week of his life delirious (from a head injury combined with hospice levels of morphine) it probably wasn't anything witty.
I'm partial to Vespasian's, "I think I'm becoming a god."
In last week's OT, I suggested a web app to test how clustered political beliefs are. One of the commentators, Robert, built this: https://politicalbeliefs.online/
I've hit unsure about almost everything and now feel a little scared - I am pretty sure most people haven't researched these issues any more than I did, so why do they have a sure opinion? Considering capping bank profits - how the hell should I figure out whether high-profit banks are simply making good investments and thus high profits mean more good investments, or they are gaming the system somehow and then this is a good idea?
I have a strong prior that populist policies in the vein of "capping bank profits" are destructive and counterproductive based on lots of past examples (Rent Control, anyone?)
I think the answers to the housing question shows that your audience differs quite significantly from what a random sample would look like.
35%. I had several unsures which I think confused it.
An interesting test would be figuring out a way to test certain cross-cultural clusterings. For instance, within a Finnish context, some beliefs indicating right-wing beliefs (social conservatism, mostly, but that of course correlates with economic right-wing thought) would be "firmly supports NATO membership" (this one less than previously, since NATO support is currently widespread, but there have been studies on belief clustering indicating that this was *the* biggest right/left predictor and "supports conscription in its current form", both of which would be opposed by swathes of American right. "Supports the state church(es)" was a similar one earlier, not sure if it's as valid now.
68%, it pegged me as a liberal and chose the libby-lib answer to everything; my beliefs are 100% clustered but they are clustered by some other, different, mysterios ideology.
Confidence seemed way off though, everything was super high confidence .
"50%
Prediction Accuracy
We correctly predicted 11 out of 22 of your answers
Your beliefs resist stereotyping! You think independently across topics."
No, you poor divils, it's that I'm (a) not American and (b) odd as bedamned. No reflection on you!
68%, and surprised it was that high, actually.
I didn't like a fair number of those questions. Maybe "didn't like" is not the correct way to phrase it. Many of them took very complex topics and simplified them down to a single sentence in a way that, for me at least, meant that I couldn't answer yes no. My real answer was very often "it depends on the details", and so I was left to mark "unsure", even though for any real example of the topic, I would have a very strong yes/no opinion, because I would know the details. Obviously, it is hard to give a lot of detail and nuance in an online survey. Both because it is just inherently difficult, and also it makes the survey more likely to be abandoned by participants.
I'm not sure I have a solution, but I thought it was worth pointing out. It had 57% accuracy for me
Yes, for instance the one about the Democratic party more reflecting my beliefs or whatever it was.
The old party? Very probably. There's a reason the Democratic party turned out so strongly Irish-American back in the day.
The current shower? About six items that I would be sympathetic with, hanging on grimly from when they *were* the party of the working man, and buried under an avalanche of idpol and wokeness since.
Fair, it strips things of nuance. Still though, I think that for most of these, most of us lean one way or another.
Same. Prob gonna be a problem with this audience in particular. For e.g. I do think Greta is kinda inspiring even tho I couldn't disagree with her more.
She doesn't even inspire me enough to want to slap her. She's had her fifteen minutes of fame, the caravan has moved on.
I got 50%.
It also reports "average confidence" of 80% and "calibration score" of 48% but I don't quickly grasp what those two numbers are saying.
It got 56% correct for me. The "calibration" was 51%, which seems oddly high to me, since it had 80-95% confidence on nearly every question that it missed. It seems very overconfident.
38%
Based on the pattern of responses, it kept expecting me to be right-wing and was surprised every time I picked the left-wing answer that is also the libertarian answer. Also, I answered "unsure" on several where I was actually unsure, had no real opinion, or thought the answer was more like "sometimes," and it got every one of those wrong (does it ever predict "unsure"?).
Did predict "unsure" for some of my answers; got some of those wrong, some of those right. It calibrated better later in the test as we went along and it got a handle on me.
Yes, for now questions are classified as 'progressive' vs 'conservative', and you're right that in reality, we should have at least 4 quadrants (as per the political compass). Once we have more answers, we're hoping that a better algorithm will increase accuracy.
42% for me
50% prediction accuracy for me, which didn't surprise me much because I've got rather an oddball mix of beliefs and affiliations. And after thinking about it for a minute, I realized that since the tool is brand new, it probably doesn't have a huge amount of data in it and probably is a bit rougher in terms of accuracy than it will be after it's been used a bit more.
To start with, we've just coded some statements as conservative, and some as liberal. As we collect more responses, we hope the prediction will become more accurate.
There is a question about whether companies that produce critical products, like SpaceX, should be govt-owned. Using an example, especially that one, is likely to skew your results by testing attitudes re SpaceX/Musk, rather than re the underlying principle.
I avoided the Musky One in thinking about that, I think large important private enterprises like that should be let alone. Doesn't mean the government can't go off and do its own space programme, just means keep the grabby hands off.
The other way round, too; not a fan of privatising every single formerly nationalised enterprise, some things can't be run as a profit-making enterprise and do need to be done by the state.
"Private companies that are critical to national interests, like SpaceX, should be run by the government."
The issue I have with this is that if the government is running a company, it's not a private company by definition.
If the question was whether they should be nationalised, I'd actually have to decide.
In this case the obvious "fix" is to replace SpaceX with ULA (United Launch Alliance). The catch is that I expect a lot more people know about SpaceX than ULA.
Maybe replace SpaceX with Boeing and/or Intel?
Why do you need an example at all? Or, why not just use an industry, such as aerospace?
If your child tip-toe walks I have had enormous success just by purchasing my son cowboy boots. Only hiccup was getting a pair that zips up the back. Wanted to let everyone know as a public service announcement.
https://extelligence.substack.com/p/medicinal-cowboy-boots
Huh, well a quick google shows that there are lots of organizations that think this is a problem, and I am certainly not an expert, but my initial impression on reading your post was don't do that - tip-toe walking is a good thing. Well, or at least walking landing on the front of your feet and then rolling back to make contact with the heel as well. Many of the pictures from my search showed children staying exclusively on their toes, which is probably not ideal...
But it has been my understanding for years now that the way humans are supposed to walk is by landing on the toes / ball of the foot first, then rolling back onto the heel. The reason for this is that this motion greatly reduces impact forces being transmitted to the knee. Your heel has no cushion, so landing on your heel directs impact straight to your knee, which over the course of a human lifespan causes damage. Whereas landing on the balls of your feet provides significant cushioning, as much of the impact goes into stretching the calf muscles as the foot bends. And since the calf / Achilles tendon is stretchy, this actually stores energy that will be used during the next step which makes the walk cycle more efficient.
My understanding is that this is how ancient people walked (known from fossilized footprints - footprints in soft mud are different when you land heel-first vs toe-first) and also how modern hunter-gatherers and running culture people walk/run. Modern people have been trained out of walking/running like this because of our shoes. With nice puncture resistant shoe soles reducing the immediate harm of taking hard foot falls on the heel, and also shoes adding rigidity to the whole foot, making it harder to land toe-first, most modern people learn to heel walk, as it is slightly easier in the moment. It certainly requires less calf strength to heel walk. But in the long term, this leaves one worse off, as it leads to much greater likelihood of knee problems in the future.
I have been trying to train myself out of heel-first walking for many years, and have been fairly successful. I hike and run a lot and feel like it is significantly better for me. And most other hikers I talk to agree. But I suppose I don't really know. It is also possible the behavior you are seeing in your is different than what I am talking about - staying on your toes throughout the whole step is probably less efficient, and certainly standing on them is going to be less efficient. It may help train calf strength though!
I had a friend in high school who had a severely limited range of motion due to lifelong toe-walking. He literally couldn't put his heels on the ground and that significantly impacted his ability to walk comfortably.
Understood in what you’re saying but he has full on ballerina foot.
Don't train the kid out of tip-toe walking completely, it's a good life skill. : - )
Sneaking skills will be maintained
Is this really necessary? Has anyone ever failed to learn how to walk normally?
One of my cousins is a life long toe walker. I can say with confidence that he never learned to walk normally, he is still 100% walking on his toes in his early 40s.
I see it frequently associated with autism. I don't think he's autistic, though he might be. He doesn't have the tactile issues, obsessions, or problems with social queues. He didn't really start speaking until 5-6 years old and was raised in semi-isolation from other kids. (His mother was very religious and sent him to a Baptist homeschool collective).My grandmother blamed it on him biting a lamp cord when he was young. He is obviously strange to interact with. People called him "touched" when we were kids.
A lifetime of walking on his toes has left him with terrible balance. He looks like he's always about to fall forward, and for the most part he is. Gentle bumps that most people would recover from with a with at most a stumble to one side send him straight to the ground. He was never able to participate in any sports (he could bowl ok, he developed a very odd but effective way of rolling the ball). It turns out that extended toe walking into adulthood damages the hip and knee joints and causes the core abdominal muscles to develop asymmetrically, which have caused him a great deal of pain starting in his 30s. He's never really been able to run or jog at all. As his 'education' never involved gym class or sports and the homeschool community militantly minded their own business regarding parenting, it was never really raised as in issue.
Apparently so. In some cases surgery is required because tendons don’t grow the appropriate length.
What are some of everyone's favorite bands (and an album from them)? I'm interested in what people here listen to and potentially looking for something new to listen to.
Some of mine are:
Fishmans - 宇宙 日本 世田谷 [Uchu Nippon Setagaya]
Animal Collective - Merriweather Post Pavillion
The Pillows - Please, Mr. Lostman
Half-Handed Cloud - Halos & Lassos
Yes - Tales from Topographic Oceans
BTS - Dark & Wild
311 - Grassroots
https://www.youtube.com/watch?v=Ak_3BX2FqjE
¿?
https://en.wikipedia.org/wiki/Boris_Grebenshchikov
https://www.toothycat.net/wiki/wiki.pl?BardicTranslationRoughs/SkyTheColourOfRain
Alternatively, this album https://www.youtube.com/watch?v=Uyf79XR6r1I by the Canadian band Musk Ox is brilliant when I just want the world to go away for a while. It kept me sane for years of ultra long commute.
Love this question and I am no doubt an album listener (skipping obvious ones like Pink Floyd or Led Zeppelin):
Hum - You'd Prefer an Astronaut
The Flaming Lips - Yoshimi Battles the Pink Robots
Bad Religion - Stranger Than Fiction
The Breeders - Pod
Cake - Pressure Chief
Flogging Molly - Swagger
Goldfinger - Goldfinger
Khraungbin - The Universe Smiles Upon You
Marcy Playground - Shapeshifter
Michael Kiwanuka - Love and Hate
The Olympians - The Olympians
Queens of the Stone Age - Songs for the Deaf
STS9 - Artifact
Geez, there are so many others I had to stop myself or this would just get more masturbatory.
A commenter in here asked for new leads about a year ago, so I put together a spotify playlist. Since it already exists, I might as well share it a third time. Apparently, the link expires after a week, as a security measure.
https://open.spotify.com/playlist/1sZ0ZdqftY1J9NNlV8bAjZ?si=354410bdcf294777&pt=06316285de14b90dba3fe4716c81291f
Looks interesting, thanks!
> What are some of everyone's favorite bands (and an album from them)
Nobody else has even mentioned anything metal or adjacent, and I know there's at least a couple more metal / hard rock guys here, so figured I'd contribute:
* Dodheimsgard - Black Medium Current (abstract technical mastery and epic vibes)
* Amon Amarth - Twilight of the Thunder God (viking themed metal mayhem)
* Chimaira - Resurrection (adrenaline rush intensity and energy)
* Kittie - Funeral for Yesterday (best all female metal band by far)
* Tesseract - Sonder (relatively complex high energy hard rock)
* Tool - Aenima, Lateralus, Fear Inoculum (these guys are the JS Bach of hard rock)
* Black Light Burns - Cruel Melody (dark, gritty, and tortured rock)
* In This Moment - Blood (female vocalist, good intensity, dark themes)
Metal/metal adjacent is one of the few genres I do not enjoy, along with pop country. A lot of people I know really enjoy metal, I don't understand why I just can't. Thanks for the contribution though! I'll still check some of it out!
You may like Alestorm, and pirate metal in general.
Off the dome and what sticks out recently:
Arc System Works: GG Strive Soundtrack
Jarabe de Palo : De Vuelta y Vuelta
Steve Miller : Live in concert with Norton Buffalo
Goat Rodeo: Goat Rodeo
Billy strings: Home
Dave Mathews: Hard to pick, probably "Crash"
SOAD: Toxicity
Snarky Puppy: We Like It Here.
Snarky puppy and goat rodeo are the most technically proficient and complete acts I've ever heard.
I'm not finding any goat rodeo album by a goat rodeo band. The closest thing I found was "The Goat Rodeo Sessions" by Yo-Yo Ma, Stuart Duncan, etc. Is that what you were referring to?
Marsen Jules - Les Fleurs
https://www.youtube.com/watch?v=dYZ6s5AGguY&list=PL11t00tyNVLn-niSdAEEcA2cT54JP2yyG
Marsen Jules - Herbstlaub
https://www.youtube.com/watch?v=Q_2UYKJicLM&list=OLAK5uy_mIGyTdZRq6PFM82ktG1yvDS5RY-M1On6I&index=2
I listened to a few from each. They kind of freak me out, idk why, just kind of unsettling in my stomach. The song "In the Flowers" by anco freaked me out in the same way when I was 13 or 14. Idk what it is about the sound, but it doesn't sit right with me.
https://www.youtube.com/watch?v=fYEAflCO4Eo&pp=ygUgYW5pbWFsIGNvbGxlY3RpdmUgaW4gdGhlIGZsb3dlcnM%3D
Sorry to hear that! Tastes differ, as they say. I find those songs curiously melodic, even though they're electronic music.
Hope the recommendations from others are more to your taste!
Finding out different people's music tastes on here was my main reason for my original comment, anyway. It's also interesting to me how people can react so differently to the same music. Thanks for the contribution!
Depeche Mode - Violator
Hans Zimmer - The Dark Knight Rises
Blues Traveler - Straight On Till Morning
Mr. B The Gentleman Rhymer - Flattery Not Included
Beethoven - Waldstein Sonata
...I'm pretty eclectic.
Arctic Monkeys - Favourite Worst Nightmare
R.E.M. - Fables of the Reconstruction
Talking Heads - Fear of Music
Jefferson Airplane - After Bathing at Baxter's
Lorde - Pure Heroine
Good stuff 🤙🤙
R.E.M. - Automatic for the People
The Mountain Goats - The Sunset Tree
Julia Nunes - Some Feelings
The Rentals - Return of the Rentals
Tegan & Sara - The Con
Tullycraft - The Railway Prince Hotel
The Like - Release Me
The Shins - Wincing the Night Away
Steeleye Span - Storm Force Ten
The Decemberists - The King Is Dead
That Rentals album is SO GOOD glad to see it listed. I didn't put it on my list cause it was here already
I really like Automatic for the People, and I grazed through the rest of these. Overall, the vibe reminded me of a playlist my dad played on Pandora when I was younger! Good vibe. I think the playlist was based on They Might Be Giants, but it played songs like Flagpole Sitta, New Slang, The Impression That I Get, and other stuff like that. Good times.
I like a lot of Steeleye Span, but I'd call Storm Force Ten one of their worst albums. :/
Tempted and Tried / All Around My Hat / Portfolio [this one is a greatest hits album, not a studio album]
My Steeleye Span tastes are heavily nostalgia-tinted. Storm Force Ten and Rocket Cottage are the ones I grew up with. I should give their discography a more serious listen.
Well, there is much to be said for the music you grew up with. My mother had several Steeleye Span albums, although I was unfamiliar with most of them until I borrowed them for college.
If you're in the US region, it's all available through youtube:
https://www.youtube.com/playlist?list=OLAK5uy_l-6DLVLWKdKYlmj1H4Gcv2NnkF14V7lZw (small omnibus of their first three albums)
https://www.youtube.com/playlist?list=OLAK5uy_mKevIQiyhIK9p9MVXXSbp7hzW0I1EqdKo (massive omnibus of the next nine, plus several live performances)
https://www.youtube.com/playlist?list=OLAK5uy_nL_LhIESNs4Hn2I9pTtP-rFy1t89Aii_k (1989 reunion album, excellent in my opinion)
I don't really listen to their post-1989 stuff; they can sound pretty different.
For nostalgia: Rocket Cottage https://www.youtube.com/playlist?list=OLAK5uy_l0B-Omqk5mhHng_X4YWCVll0ciSkFpW6c
Storm Force Ten https://www.youtube.com/playlist?list=OLAK5uy_kgyDulHOuny_XAmZZonpYP32LDC5bgaoI
Rocket Cottage wasn't one of my mother's albums, but I very much liked Fighting for Strangers and Sir James the Rose, which made it onto one or more greatest hits albums. (Sligo Maid did too, but I never really appreciated the jigs. I like having words.)
Johnathan Coulton's "Artificial Heart" is my favorite album. But I'm not much of an album person.
I was looking it up and it looks like John Flansburgh from TMBG produced it! So that's a fun fact. I listened through it, I liked nemeses the most.
I knew about Jonathan Coultan from Portal and I've heard some of his Thing-A-Week songs, but I don't think I've heard any from Artificial Heart before now (aside from the Portal ones).
If you liked "Artificial Heart" at all, you may also like his album "Solid State".
Starting with Fishmans just make me want to ask what your RYM nick is.
Sorry to say, this is the first I've heard of RYM 🙀 (I had to google it) is fishmans pretty popular on there?
This is like someone in AI safety not knowing who Yudkowski is.
Oops 😳
RYM's userbase ranks LONG SEASON by Fishmans as the 29th greatest (popular) music album of all time. (https://rateyourmusic.com/charts/top/album/all-time/#pos29). My understanding is that this is almost solely on the strength of word-of-mouth campaigning on RYM forums.
What's going on with the album name? "宇宙 日本 世田谷" looks like "universe - japan - [meaningless gibberish]" to me. If I look up that third word, it seems to mean nothing other than being the name of a particular museum in Japan. There are no verbs.
Setagaya is a ward in Tokyo, it's where the singer of the band was from. You're right, it's "Universe - Japan - Setagaya" basically going from everything to the place specific to him.
Oh, that's pretty sweet! I feel like it's rated above other fishmans stuff just because it's long, but I do like it a lot.
Sassy Swings the Tivoli - Sarah Vaughan and her Trio produced by Quincy Jones. Or were you thinking of something from the current century? :)
I scanned through it and it's good! I think my wife would really like it. Also, it's only 10 years earlier than the Yes album I listed.
Hard to go wrong with The Knocks. No particular album but a few bangers:
Slow Song
Classic
Brazilian Soul (and the Acoustic version)
One on One
Bodies
Ollie Wride is a blast. Got to see him live, great performer.
Check out "Running in the Night". His best by far. If you like that, you'll like most other stuff he's done.
LEISURE's "Sunsetter"Album is weird because I'm listening to the same song 10 times with minor variations and it's all great. You know how ACDC only made one song with 50 different lyrics but it's so good you still love it? This is like that but chill lounge music.
Might be mainstream but I've really enjoyed Teddy Swims.
Been getting into Franc Moody recently. Loved Dopamene, then loved loved loved Dance Moves.
You should have exactly one 6ix9ine song in rotation. Yes, he's a horrible human being. Yes, it's not technically good...but it's so fucking hype. I recommend "Billy". Sometimes you're depressed and too in your own head and you need a sociopathic mentally challenged teenager screaming:
"Get the fuck up out my fucking face before I murder you."
It just...it's just the most testosterone laden thing ever. Not like mature test, like pure idiotic adolescent test.
Billy Strings and Bryan Sutton - Live At The Legion is one of my favorite modern examples of bluegrass flatpicking guitar. The rest of Billy's catalog is more experimental/psychadelic (which I also really enjoy), and Bryan's is more traditional stuff. If you like it, similar older releases will be from Tony Rice (and Tony Rice Unit), Doc Watson, and others that you'll find if you branch out from there.
A friend of mine recommended Billy Strings a few years ago, I forget the album but I enjoyed it for sure. I'll have to check out that live album and Bryan Sutton's stuff.
I'm making the move many late career devs do into being independent, creating my own web dev and general software development shop. I heard from a friend that with AI, now is a great time to start a business, since AI is a huge force multiplier, and I guess in the near term, being a business owner is a good hedge against being replaced by AI.
I was wondering if anyone here is making similar moves and if they have comments on it.
I am deep in this process myself, but building something in a specific niche with a partner, rather than generalist consulting. I would advise caution with the generalist approach it sounds like you're thinking of.
My thoughts are:
* "huge force multiplier" is absolutely right; underselling it if anything. It's exhilarating what I have been able to achieve, and how fast. Things I know almost nothing about, that I never would have even considered trying myself in a serious startup effort, I'm getting done to a high standard of quality in a couple of weeks. You will be able to bring this ability to any customer that you can get; it will feel amazing to you, and they will be floored at the speed and quality of your results.
* I think (or hope, since I'm trying this model) we're on the cusp of a wave of start-ups that achieve significant initial success with no employees beyond the founders, and no funding. These will comprise at least one generalist SWE, and at least one high quality expert in some important field (law, biotech, agriculture, oil&gas, etc).
* In contrast, the world of generalist consulting, *especially* web dev, is going to become incredibly competitive. I won't go so far as to say "100% automated away" - but close. Starting out with direct links to potential customers who trust and respect you is going to go from hugely important to 100% mandatory. I think this is the one variable in my advice to you: if you already have some likely customers lined up, like to the point of informal agreements to hire your services, then go for it. Otherwise, hold off until you do.
I was thinking of following a strategy I read online of making simple sites for small businesses, on a subscription of model of $150 month. I think not many are thinking of serving that market, and that quite a lot of people aren't technical enough to use something like Squarespace or AI to do this.
I would try making a product company if I had a partner with a niche as you're describing, but I don't know anyone like that. How did you find yours?
Ahhh, ok, that's certainly not a 2-3 OoM difference. I can see small businesses, especially older/more conservative owners, going for $150/month. Just be sure you have some reasonable and clearly communicated boundaries, or you'll be valuing your time unsustainably low.
>I would try making a product company if I had a partner with a niche as you're describing, but I don't know anyone like that. How did you find yours?
marriage ;) (downstream of grad school)
In all seriousness, I recognize that I am very fortunate to be in a major right time/right place/right person confluence. As you say, what I'm describing is not available by default. If I was the same me in the beginning of 2025 without this opportunity, I would be itching to find something like it and entirely possibly send myself on a wild goose chase... the reality is that it needs to be the result of networking. Which is terrifying for someone like me, since I am a few clicks above hikkikomori in that department.
Anyways, definitely best of luck; you've got good answers to my concerns, so it might be viable. But if possible, do cling to mainstream employment until this venture becomes a money hose you're having trouble holding under control!
That's my mindset, I'm not quitting until I have a stable business.
Networking, huh? Yeah, I feel a few clicks above hikkikomori myself.
Maybe this is all unnecessary, I'll be rich in case of TAI, or if we get a singularity, this will all be very irrelevant.
I have absently considered it noticing how much AI has affected my productivity, but then I also wonder that the market value of my output is about to plummet and may be better off drawing a paycheck a bit while longer than being so directly exposed to market forces. Shrug
Yeah I would think this is the worst time to do it, since the market is going to be flooded with people making apps since the barrier to entry is much lower. Whereas with B2B software job, the barrier to entry isn't a good app, but the connections with business that you sell to, which was true before AI and is more resilient to more apps because they rely on social connections. The B2B software I've worked on has never been exceptionally good or anything, it's because we have social connections to big companies and can talk to them directly is the real value.
When the market is flooded, your job is to sell quality. Be a boutique firm and emphasize that the experience will be top-notch and well supported.
When the market is going to get flooded, your job is to sell the tools for those who will flood it.
You can't sell an AI; you don't have a budget for that. But perhaps you could sell some generic wrapper system for the AIs, something like "buy my system, register your API keys, write your prompts, choose one of these nice graphic templates... and you have your AI-powered application ready, you can immediately start charging your customers".
I think B2C (making apps) was always essentially a moonshot. I'm just going to be operating a web dev agency. Plenty of small businesses could benefit from having a good website, and I think lovable is a considerable distance away from being used for that purpose by someone non-technical.
I'm afraid there are two potential fatal flaws in the business case you're hoping for here.
First, regarding "considerable distance away": consider our host's recent announcement of total victory in a bet on AI progress. Are you really that confident that "text to website" won't have massively improved 2 years from now? Especially considering that a lot of the work to be done is mundane plumbing and polish, and not plateau-able AI research witchcraft?
Second, even if those tools never improve further, are you sure you can compete with them? I entirely believe you can deliver a noticeably better result, especially if it comes to complex/unusual requests. But. Can you compete with something that can probably handle like 80% of small business's needs, to 80% of your quality, for 2-3 orders of magnitude cheaper?
I think many people are just too non-technical to have an AI generate a website for them. It's very common for people to struggle with their OS and gmail: someone like that won't be able to reason about the website they want well enough to have an AI do it for them.
I think Scott is overly optimistic about AI capabilities: he won his bet, but he had to admit AI still fails on prompts that a human artist would do 100% correctly.
Still, I trust Scott, and on the off chance something like AI 2027 does happen, I have $20k split evenly between a tech index fund and the S&P 500. If AI automates my profession (and others too, dev won't be going down alone), these investments should go to the moon anyway.
Some papers came out recently on the limits of reasoning in LLMs. They found that as more state is added to a problem, the AI can't handle it past a certain threshold, and its ability to solve the problem collapses:
https://www.jonstokes.com/p/the-latent-state-window-and-the-limits
and one of the papers itself:
https://machinelearning.apple.com/research/illusion-of-thinking
That's pretty interesting and hints at the possibility of LLMs petering out at some point beneath superintelligence.
i mean, i've been a skeptic for a while, and this is what I would naively expect. If it's just "pattern matching", it was always going to be lossy in the limit. E.g. if you think of "chain of thought" as a Kalman filter, of course it's going to implode at some point. At some point, it needs to be able to drop into some sort of subspace where the reasoning is lossless instead of probabilistic.
Sure, in theory, you can compute anything with infinite compute. You can solve quite a lot of problems with infinite resources, in general. But is a gigantic Searlean hash-table really the holy grail of AI? Isn't the point of intelligence to do more with less?
The "illusion of thinking" paper made the rounds in the AI space a few weeks ago. I wrote some thoughts on it then if you're curious (https://open.substack.com/pub/theahura/p/a-few-quick-thoughts-on-apples-illusion)
The tldr though is that we know LLMs have context length limitations because of the downstream limitations of our hardware resources. Models that have longer context windows do better. Many researchers are trying to crack intelligent summarization; others are trying to crack infinite context windows. But from where I'm sitting, there's no reason to assume LLMs are actually less capable in the limit.
Thanks for the post. Yeah, I can see AI 2027 maybe happening, but I can also see LLMs hitting a limit. They already suck up so much compute, and you get a result that is in some senses superhuman and in some senses subhuman. It's totally fair to speculate whether the LLM architecture is inherently unreliable and a new paradigm is needed to get past that.
personally i think ai 2027 is a few years early. The biggest limitation imo right now is we do not know how to evaluate models that are better than human, and all of our existing mechanisms for neural network improvement depend on some human, somewhere, constructing some kind of method for generating gradients through some kind of evaluation
I think it’s important to remember that no model yet is “better than human” in general - and I’m skeptical that there will be such. There will be models that are better than human at some things, worse at others, and just weird and different in many. It’s really the “weird and different” that makes them hard to evaluate more than the “better”.
Just as an illustration of the general problem, think about how SAT scores might be a reasonable measure of verbal aptitude for a randomly chosen English speaking high school student, but once someone starts studying test construction strategies, it becomes much worse.
I find self-modifying code with a self-modifying compiler generally fixes most "limitations."
How does this system get gradients?
Here. Have these lasers, now go kill some weeds.
"I'm gonna get the high score!"
(Note: there is a rather "unfortunate bug" where soybeans are sometimes termed "weeds." So sad, really.)
If we assume that LLMs are going to automate lots of white collar jobs, do we think that licensed jobs are going to be relatively safer? My entire professional career has been in sales, for the last decade I've been in a non-licensed job, but recently I've been thinking about switching to something like selling insurance purely for the license. (Other options could include selling real estate, or securities, or mortgages, or anything else that requires licensure by the state).
Presumably a non-human agent cannot receive a license. The idea is to take advantage of a regulatory barrier to ward off LLM competition. I have some faith in the lobbying power of the FIRE industries (finance, insurance, and real estate) with both regulators and Congress, if needed. Does this sound like a valid plan? Or am I missing something?
(While I am not an attorney, one could make the same argument for them too. You obviously need a license to practice law at the moment. Supposedly around 50% of Congress are attorneys- can I trust them to act in their own self-interest? I mean probably, right?)
not thinking anything would be "safe," as less white collar jobs would mean less insurance to sell. Ai is bad because it's not creating any jobs or markets but just replacing labor for cost savings. the market that labor provides will dwindle and be replaced by corporate investment.
my guess is people shift into things AI can't affect demand of, like health or food.
Historically, one workaround to licensed people being more expensive than unlicensed people was to have unlicensed people to most of the work and then have a licensed person come in for the last bit to look over the results and make things official.
Replace unlicensed person with AI and the same basic strategy seems like it would work here. The most skilled (or more flexible or more forward looking ...) licensed people will ramp up on using AI assistants. Fewer people (including licensed people) will be needed unless the demand goes up by enough to make up for it.
The fundamental problem is that if what you are doing can be done A LOT LESS EXPENSIVE by someone/something else your career is not in a good place.
I run a startup that has recently started selling a unique insurance product, and I would be very surprised if these sorts of licensed jobs are safer, unless the job requires a human presence like an attorney or perhaps doctor.
Selling insurance requires a person at your company to be licensed, but there’s nothing that says I have to do my outreach and marketing personally. We can (and do) marketing using AI tools. I have a list of a few thousand people, and could standardize I method to find ~infinitely more. If I trusted AI to send emails, texts, LinkedIn requests and make calls, all without human intervention, then the license would only be useful to the people who own the business, not any of the employees.
What does "a unique insurance product" mean? Is it just insuring some niche thing (I know there are e.g. few and specialized insurers for rocket launches), or is it more complex, like working in some special regulatory locale or something?
I was referring to the first one, as in it’s a specialty surety product that only 4 other companies offer, but the second is also true, as we’re based in the most highly regulated market in the world (New York State).
There’s aspects of insurance that would resist AI for a very long time due to the regulations (actuaries, reinsurance intermediaries, the owners of agencies and underwriters themselves, etc.), but I would say selling insurance isn’t one of them.
We have a few people selling part time right now and if AI could do their job, they wouldn’t be employed. I would just use AI under my own and the company’s license.
Third-year law student here. I think you're right, to an extent. The legal field acts like a cartel; only those who get the license can practice. But I don't think this insulates me too much. It will become malpractice in a few years to not use AI. So while it may preserve the job itself for some time, it may become a matter of just approving the work your AI does. And on the client side, people are just going to use LLM's as attorneys. LLM's already give legal advice, but just make sure to say "this isn't legal advice" at the end. I don't think most people are going to be too worried as long as they are getting high quality legal advice for 20 dollars a month, rather than 200 dollars an hour. So I anticipate the work drying up from both ends.
>And on the client side, people are just going to use LLM's as attorneys. LLM's already give legal advice, but just make sure to say "this isn't legal advice" at the end. I don't think most people are going to be too worried as long as they are getting high quality legal advice for 20 dollars a month
But that's not 'practicing law'. I'm sure people will get frequently terrible legal advice from LLMs, but they can't represent someone in court, or even in legal negotiations between parties. If anything you could argue that the Internet kicked this trend off, but attorney employment has only risen over the last 25 years.
I see your point about pro se litigation below, but if the court system is overwhelmed with pro se litigants, they're just going to make that harder/borderline impossible. Again Congress is 50% attorneys, they're not going to sit back and let their industry get decimated. I'm sure the judiciary will quietly go along with whatever they pass.
Plus pro se types are going to lose in court an enormous amount, especially against a trained attorney and with an exasperated judge. No one with real money on the line (upper middle class person, a business) is going to represent themselves pro se in a real case. If anything company executives are violating their fiduciary duty if they either sue someone pro se using LLM tools, or try to defend their company that way. The actual bread & butter clients for attorneys will stick with the licensed types. Could there be more crazy people suing each other in small claims court using an LLM? Sure, but attorneys mostly don't take those kinds of cases now anyways
> I'm sure people will get frequently terrible legal advice from LLMs, but they can't represent someone in court
The requirement that you either show up in court yourself or hire a guild member to appear on your behalf is a pretty interesting perversion of the "right to an attorney", which was meant specifically to prevent that requirement.
I don't hate this argument, but US constitutional review famously found a right to a public defender somewhere in the US constitution to address this exact problem. As a practical matter, for lower-income folks there are (taxpayer-funded) free legal services. Try being a landlord and evicting a low income tenant for blatantly not paying you rent, trashing the place, etc. They'll show up to court with free (taxpayer-funded) attorneys to combat you and look for loopholes in the lease, etc.
Try being a landlord and evicting a low-income tenant for default, and sending your son to court to make your case rather than a hired lawyer.
> but US constitutional review famously found a right to a public defender somewhere in the US constitution to address this exact problem. As a practical matter, for lower-income folks there are (taxpayer-funded) free legal services.
Note also that this is a textbook case of "restrict supply, subsidize demand", which is never a good policy.
A side comment regarding Congress:
-- the current numbers are that 50% of the Senate and 30% of the House hold law degrees and have practiced law at some point. So overall the percentage is either about 40% (if you just average the two chambers' percentages) or around 33% (if you factor in that there are way more Representatives than Senators).
-- that percentage has been in long-term decline. Before the Civil War, Congress was three-quarters attorneys; in the first half of the 20th century it was around three-fifths; by the end of the last century it was around half; today, see above. (That's despite the number of attorneys rising steadily: during the 19th century around 1 of every 250 U.S. attorneys was a member of Congress, today that ratio is around 1 in 7,000.)
-- also worth noting is that whereas nearly all 19th-century POTUSes were attorneys, only a couple in the past 75 years (Nixon and Obama) did more than dabble in it. Since the end of WWII 9 of 14 POTUSes never even attended law school which is a previously-unprecedented trend in US history.
If the percentage of Congress that is attorneys keeps sliding downward while most POTUSes are not attorneys, it may be that the legal profession ceases to drive federal regulatory decisions about new topics such as LLM's.
Interesting, but there are two other twists:
1 - there didn't used to be a "bar exam"... if people were willing to pay you, then you were a lawyer (see: Abe Lincoln)
2 - it didn't used to be a requirement that one went to law school in order to take the bar (and I think in one or two states, still isn't)
So before "being a lawyer" meant "being well-read in the law", whereas now it means "years of prior commitment".
I'm not positive, but I can easily imagine that impacting those stats.
This is a good point, and would be the case if AI never gets better. But I think you'd be making a losing bet to think so.
As a small counterpoint, I do think litigation is protected. See what happened to "do not pay", a startup that was attempting to sell direct to consumer, routing around lawyers -- the legal profession basically rose up as one and drowned the startup in lawsuits.
Looking the Wikipedia page for it, it seems to be doing fine. In fact, 'Above the Law [a prominent law news website] ultimately recommended the service for "clear-cut issues like parking tickets or non-critical matters," while cautioning against its use for legal issues with higher stakes'.
I specifically said 'litigation'. DoNotPay rather infamously tried to 'defend' someone live in court, and in response got hit with two class action lawsuits and an FTC fine, after which they have not tried to do that again
You said a lot of things. The "drowned the startup" part seems misleading when it now seems to be quite profitable and returning dividends.
Startups returning dividends to investors isn't a good thing, I think describing "two class action lawsuits and an FTC fine over two years" as "drowned the startup in lawsuits" is totally reasonable semantically, and the larger point -- that the legal profession will close ranks to protect itself against AI -- remains.
As a meta comment, I find your semantic nitpicking to be extremely tiresome and ask in the future that if the only thing you want to add is about semantics, you refrain from doing so
I see the point, but I'm not sure it covers as much as you might think. 90% of litigants are *pro se,* or self representing. You can sue a company trying to sell unlicensed legal solutions, but you have no way to stop individual people from using ChatGPT to help with their own legal issues. *Pro se* litigants are allowed to Google questions, visit legal libraries, etc to facilitate their self-representation. AI will function the same way, but will be much more effective.
I suspect any company trying to advertise itself as tailored to helping pro se defense will be shut down for much the same reasons as do not pay. But, fair enough, I could see self defense becoming more popular
For the first time in a while, a positive development for a stable peace in in palestine that is marginal and incrementally workable: individual leaders trading recognition of israel as an ethnic homeland for territorial control and economic integration. In a different timeline maybe a two state solution could have worked, in this one an emirate model seems like a hopeful alternative. Maybe one day if it flourishes it can turn into a confederate model as a trust and integration grows.
https://www.wsj.com/opinion/new-palestinian-offer-peace-israel-hebron-sheikh-emirate-36dd39c3
There isn't great polling data for Palestinians, but what there is suggests that this kind of idea is radically unpopular
Partitioning into a dozen tiny emirates seems like a nice idea in theory. But I suspect it just results in constant fighting between those emirates, and between those emirs and various other forces that want to take over.
Much better that they attack each other than Israel.
I'm certain that Israel winds up getting drawn into it one way or another. Like, the Emir of Bat-Mobil is getting attacked and he's Israel-friendly so we'd better just send a few troops to help out, and all of a sudden you're back to occupying the whole darn area again.
That sounds similar to the reservation system in the US; hardly without its downsides, but would be a vast improvement over the status quo.
It worked for Russia in Chechnya to the extent that there are Chechen regiments fighting with varying degrees of success in Ukraine on the Russian side.
It didn't work for South Africa with the bantustans.
Russia had to endure about a decade of endemic violence after the end of the active phase of the Chechen war and ignore the "international community" which kept telling it that it's impossible to win by force. Quite a lot of a abuses were committed, naturally.
Unlike South Africa, Russia never faced widespread popular condemnation of its behaviour in Chechnya, no calls were made to boycott it. In spite of all the wars against Muslims in Afghanistan, Chechnya and Syria Russia is still seen positively or neutrally by the public in many Muslim-majority countries. https://globalaffairs.org/commentary-and-analysis/blogs/globally-negative-views-russia-predominate
Compared to Israel Russia had the benefit of having Chechnya within its internationally recognised borders.
I hope it'll work but it's hard to be optimistic.
Russia faced more popular condemnation from within Russia, for its treatment of Chechnya. Specifically, for its "rebuild BETTER" philosophy. "We didn't revolt, so we don't get new buildings? What is this mishegoss?"
Well, yes but I wouldn't say it was that strong. People were pissed off and Navalny tried to lean into this feeling but it never led to any violence and never truly threatened the long term rebuilding program.
Btw this is an advantage of an authoritarian regime. Any Israeli leader who attempts to spend Israeli money on rebuilding Gaza (hard to imagine, yes) would face huge opposition from everyone.
On the other hand, Netanyahu sort of tried it with the expansion of work permits for Gazan workers in the early 2020s. Turns out purely economic levers don't work so well.
Ukraine is also getting an upgrade (at least the Eastern Provinces).
Israel is far more likely just to get the Europeans to spend money on rebuilding Gaza. Bibi's been loud and long about "save israel by supporting hamas."
Israel needs to clean up its own mess.
Unfortunately, "kicking the jews out" was voted "too offensive." Perhaps there will be a revote?
This is typical Israeli black humor. also serious.
In the 1950s, Robert Sheckley wrote a science fiction story called “Watchbird.” Robotic drones with collective intelligence capable of self-learning, called “watchbirds,” are deployed throughout the United States to prevent murders. Gradually, they expand the concept of murder to include the execution of criminals, the slaughter of livestock, fishing, killing mosquitoes, performing surgical operations and so on. Life becomes unbearable for everyone.
Self-learning “hawks” are created, whose task is to kill the watchbirds. At first, they perform their task well, but gradually the watchbirds learn to evade the hawks better and better. In the end, the hawks decide that they must kill not only the watchbirds, but everyone.
In my opinion, the first part of the story is a metaphor for what happened in the following decades. The concept of “racism,” as well as the concepts of “sexism“ and “homophobia,” which were coined in the 1960s, have been expanded by wokists to such an extent that anything can be deemed racist, sexist, or homophobic. In Java programming, "everything is an object." In modern Western discourse, "everything is racism/sexism." People are afraid to speak their minds. Life is now unbearable for many if not most people.
That's why people voted for Trump. I hope Trump will be more successful at destroying wokists than hawks from the story in destroying watchbirds.
If your problem with woke was that it was anti-meritocratic, or bad for democracy, or grossly and capriciously punished people who didn't deserve it, I have bad news for you about the person you put in charge of replacing it.
Are people still afraid of the woke? They seem pretty vanquished to me.
When complaining becomes a habit it can be hard to break.
It's not just a habit, though. Being enraged about this or that issue, hating this or that group can become part of somebody's self-definition,. and a way of bonding with other people.
I think democracy is like a ship built a certain way. It's remarkable -- stands up great to some conditions, and tolerably well to most others. But there are some rare combos of conditions that sink it, and the US has ended up in one of those combos. I don't think we're getting democracy back.
The demise of USAID means that they're fleeing the creative realms fast (as people are no longer paid to be trans-positive, or whatever the current woke-fad is.) See: "paid to make a gay-friendly play in the Marshall Islands." Also see the Gay Pride flag flying over Kabul, if you really want to make the point about State being a bunch of a-holes.
Innocent man still falsely imprisoned (no innocent until proven guilty when the Democrats have hung their hat on billions of dollars of property damage). I'd be afraid if I worked in his profession. You do everything right, and you're still going to jail -- no chance of appeal, either.
Which man? Robert Sheckley?
I think the reference is to Derek Chauvin.
Maybe Epstein?
> STILL falsely imprisoned
Maybe I'm being willfully blind to the real truth, but I thought just about everyone agreed he was dead, and the debate is just over who killed him: Clinton/Soros/Gates et al. or Trump/Musk et al.
What do you mean, vanquished? The government hasn't even started its crackdown in earnest yet.
Yeah when do we get to see some gender reassignment surgeons dragged through our trailer park by a rope untll there's nothing left but the rope?
God forgive me, I wouldn't mind that happening to this biddy:
https://www.independent.ie/irish-news/health/irish-gender-affirmation-surgeon-based-in-florida-is-reported-over-false-promotion-in-tiktoks/42051361.html
https://www.instagram.com/reel/DLxKvcsO5Zk/
Yeah, I'll believe you are in it for the principle of the thing when you stop doing shit like this (maybe she has due to pushback):
https://www.tiktok.com/@gendersurgeon/video/6889485602199227653
Is *everything* racism? Or just more things than you're comfortable with?
Maybe the 2 sets are co-extensive.
The new Alligator Alcatraz facility is supposed to cost $450 million a year, and with 5000 detainees, that's about $250 per day. Looking at the quality of the infrastructure, isn't that absurdly high?
Other than an unnamed DHS employee in a CNN article, I haven't seen any potentially official source for that facility costing $450M per year.
California is not Florida, but as a comp:
"The cost of imprisoning one person in California has increased by more than 90% in the past decade, reaching a record-breaking $132,860 annually..."
https://calmatters.org/justice/2024/01/california-prison-cost-per-inmate/
Numbers from a decade ago were MUCH lower:
https://www.vera.org/publications/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends/price-of-prisons-2015-state-spending-trends-prison-spending
This stat also sounds like it comes from somebody with an ax to grind.
That seems high but with recent inflation it makes sense. As far as I know, the primary drivers in Cali are two-fold.
First, staff. Being a correctional officers is a brutal profession with high turnover. You pretty much have to bribe people with ever increasing salaries to even get moderately qualified people to consider it. Calmatters blames the unions but the ACA (1) reports annual staff turnover rates of ~50%. If half your staff are walking out the door every year because the job is just that brutal...the union has a pretty easy job.
Second, healthcare is ridiculous and constantly increasing. CCHCS' budget is ~$4 billion, or $40k/inmate at 100k inmates (2). I would be surprised if this is a driver for illegal aliens in Florida but it definitely drives the Cali numbers. Think about how expensive your health care is, now imagine the healthcare costs of a 48 year-old inmate who's been abusing his body for 30 years.
And those numbers kinda add up. $250*365=~$91k/inmate and the ~$132k/inmate from Cali, if we subtracted that $40k, is $132k-$40k=$92k. Almost exact.
(1) https://www.aca.org/common/Uploaded%20files/Publications_Carla/Docs/Corrections%20Today/2024%20Articles/Recruitment-and-Retention-of-Correctional-Staff.pdf
(2) https://cchcs.ca.gov/factsheet/
I doubt that tells us much. Leaving aside the fact that there is certainly some spin going on there, some of that is the cost of health care, and the average prison inmate is vastly less healthy than the average likely detainee. Moreover, "prison guard" is a much more onerous and dangerous job than guard at a detention camp, and the need for security is far less.
Huh, I knew it was high, and I even knew it was THAT high, but not that it was so recent an increase. Thanks.
Is this the result of activism toward better conditions, or more services, that prisoners are getting the equivalent of a top salary? Or is it civil service salaries?
I don't know :-(
I suspect finding out will take effort as I also expect that there will be (politically charged) claims on both sides that will require filtering.
Anyone knows some good intro articles about Polymarket? My hands are itching to try it, but I never did betting in my life.
If you're American, try kalshi, it's generally easier and more user friendly (disclosure of COI).
Start with Metaculus. It’s fake money so much less downside but the exact same concept as a betting market.
I made a web app that scans ingredient lists (including photos of them) for concerning ingredients from a variety of sources, including Scott's Obscure Pregnancy Interventions. While it's primarily designed for skincare and topical products with kid-focused branding, it's useful for anyone who cares about this kind of thing, not just parents or those who are pregnant. It's free if you want to try it. parentscan.net
This is a good idea. Will be passing it on to my daughter, who is thinking of having kids soon. Do you have an opinion of Consumer Lab? It’s a site that tests supplements for contaminants and also to check whether the user is actually getting the dose noted on the label. CN they be trusted?
Subscriber for years, *HIGHLY* recommend. Something that's rarely mentioned is that most of the informational articles included with the various product testing are like Scott's Everything You Wanted to Know About posts, except sometimes even *longer,* with even *more* links to studies.
If anything, that kind of consolidated research on stuff like Vitamin D may be every bit as useful as the actual testing.
As Eledex notes below, Consumer Lab can only test a small fraction of products on the market, so really obscure supplements may not have crossed their desk yet, but for most people, I think CL is probably going to be sufficient, especially if you want testing on food products, too, like canned fish or seaweed snacks. Some of their reports are so eyebrow raising that I've made a resolution not to use any product which hasn't been tested by CL, and have imposed the same on my family.
Last, a nice side benefit of CL is that it removes decision fatigue. Their Top Pick for a given supplement is whatever least expensive product completely passes their their testing. If you *want* to spend more money on a particular brand, they certainly won't stop you, but if a Costco Kirkland vitamin passes all testing and is $0.03 a tablet, that'll be their Top Pick over a boutique-looking fancy name brand which also passes testing but is $0.15 a tablet.
Consumer Labs does a good job testing stuff (i.e. their results are accurate) but they test only a tiny fraction of what is on the market and a significant portion of what they test is paid placements. So their data can be trusted, but it's not comprehensive.
"Paid placements" is a bit misleading. Consumer Lab is mostly supported by subscribers, but some companies may pay for a quality certification *if* they pass Consumer Lab's *standard* testing.
Consumer Lab only tests products randomly purchased in shops and stores, including the products companies might want to have certified, so it's not like a company seeking certification can rig the results very much. (https://www.consumerlab.com/answers/can-i-trust-consumerlabcom-how-are-its-tests-paid-for/trust-consumerlab/)
While I personally would prefer for Consumer Lab to be 100% subscriber supported without absolutely zero contact with any company, I nevertheless think their process is reasonably transparent and trustworthy.
If there will soon be millions of robot workers, then it means the cost of skilled physical labor will sharply drop, and many projects that were unaffordable will become affordable. Does this mean the old museum ships like the USS Missouri will be modernized and reactivated? I think it would be awesome to see it plying the waves with a cyborg captain and a robot crew while it blasts the enemy using lasers and railguns alongside the old 16 inch guns.
that doesn't mean resources will scale up to meet that demand, or even if we have the resources to even make robot laborers at scale. Hardware is expensive and a lot pf SF authors tend to wish their technologies into existence.
<If there will soon be millions of robot workers, then it means the cost of skilled physical labor will sharply drop, and many projects that were unaffordable will become affordable.
On the other hand, all those skilled laborers who've had their income drop or been laid off are going to be able to afford way fewer things.
The pie will be a lot bigger, but you won't be getting any of it....
As well ask if the Smithsonian is going to build a robot dinosaur around its bones.
We're not keeping the USS Constitution around because we think we might want to use it again; it's purely for historical purposes.
If and when there are "millions of robot workers", it means that the cost of each of those workers has fallen below the price of an equivalent workforce of humans. The relative cost of a human worker vs. a machine will only increase as the degree of automation in an economy increases.
If the current wars of Russia vs Ukraine and Israel vs Iran have shown anything, then that range is king, meaning missiles and drones doing most of the damage. I don't know what the state of the art in railguns or their use case is, but lasers will ever only be practical for point defense against missiles, drones, small boats etc., while large guns as naval main armament have been obsolete since WW2. So no, you won't see WW2 battleships in service.
The USN has (at least largely) shelved their railgun efforts because (AIUI) they're not effective past ~400 Nm and therefore the space aboard ship they require would be better used to house more missiles.
The sticking point there is "if". *IF* within five to ten years be millions of robot workers who can do jobs with the same dexterity and agility as a human, then we can do all kinds of things.
If this doesn't happen like that, then we can't.
And even if we get robot labour, I regret to say I don't think we'll get your dream ship.
It doesn't need to be within ten years--I didn't assign a hard deadline to it.
Why wouldn't I get my dream ship?
Angry unemployed skilled laborers will blow up the fucker. (At least those who refused to get soaked in resin and repurposed as paving stones for the Walkway of Glory leading to the modernized USS Missouri.)
Restoring the ship was again moved to the backlog pending other higher priority items.
Can someone explain to me why everyone started calling missiles "drones"?
I'm very confused.
I keep hearing that "drones" are this groundbreaking revolutionary weapon. Isn't it amazing, a "drone" flies straight to a tank and explodes! That's pretty old technology, it's called a guided missile. How come everyone treats it as revolutionary and calls it "drones"?
How come American cruise missiles are called "missiles" but Iranian cruise missiles are called "drones"? I hear all the time about the latest Iranian volley of "missiles and drones", like it's a distinction that matters and you're not allowed to just say missiles.
Is it because a drone can land? Can all drones land?
But how can this definition of "drone" be compatible with their reputation for being revolutionary? People say drones are revolutionary because they're "cheap" and "expendable", what does the ability to land have to do with being cheap and expendable???
Instead of missiles that look like helicopters, why can't armies simply use missiles that look like missiles (thus not "drones" by that definition), and are as cheap and expendable as "drones"?
A rocket is cheap, and goes faster than a mini-helicopter. If the elements that grant a "drone" remote control/AI, vision and steering exist in a cheap and expendable mini-helicopter, why not in a cheap and expendable rocket?
If I had to take a stab, the inchoate definition people are gesturing towards when they say "drone" is something like "Something that's both reusable and expendable simultaneously".
Or, "A plane but you can afford to throw it away on suicide missions, because it doesn't have a pilot inside. But also, a missile you can reuse if it survives the suicide mission, because the target isn't there or something." The key thing that makes a drone a drone, is that it has functionality that existing things do not: you can send a recon drone on suicide missions in a way you couldn't send a recon plane, and you can reuse an anti-tank FPV drone in a way you couldn't with an anti-tank guided missile. It's more cost efficient in both directions: if it survives, that's good, you can reuse it. If it doesn't, well anything else would have cost even more to sacrifice, most likely.
If I got a second guess though, the lay public obsesses over the word "drone" because they're trying to say, "OH MY GOD, consumer hardware can beat tanks? Can beat actual military stuff?" In the same way the public obsesses over small arms, particularly the types it has access to (rifles, pistols, shotguns) and has interacted with personally, rather than crew-served weapons like machineguns, mortars, rocket launchers, recoilless rifles, etc. they can't own and have never touched (precisely because they're more effective than just small arms)... the public is now obsessing over the idea "I could kill a tank!?!?" when it says "DRONE DRONE DRONE DRONE DRONE".
i.e. When the Daily Mail or New York Post want an attention grabbing headline, they don't say, "This thing you've never heard of, CAN KILL A TANK", with a photo of artillery shells scattering landmines. They say, "YOUR CHILD'S CHRISTMAS PRESENT COULD KILL A TANK???", with a photo of something cheap from Alibaba. Nevermind that reality doesn't quite live up to the headlines, as https://warontherocks.com/2025/06/i-fought-in-ukraine-and-heres-why-fpv-drones-kind-of-suck/ reminds us; the public latches on to the first thing it understands, and doesn't let go until something else catches their attention.
(And indeed, there *are* interesting implications to consumer hardware being dual use. For a start, there has never been a shortage of drones for the war in Ukraine, compared to things like artillery shells and MICLIC mine clearers, because even the most insane FPV drone demand only bumps up the global demand for consumer drones by like 10%, and the factories in China can adjust. But when war breaks out and the demand for artillery shells goes up something like 100x what it is in peacetime, the existing artillery shell factories can't adjust. The public is gesturing to something real here! They're just awful at conveying what it really is, because they don't quite understand themselves... but they're very loud & insistent nonetheless.)
That piece you linked to is interesting, since it goes against what we usually hear about those Ukraine war suicide drones.
But as someone who knows hardly anything about tactics there are a couple things about that piece that leave me puzzled. It says that the most common use of a FPV drone is to finish off something that has been already hit, often by a mortar. But if a mortar hits a target, and you need to finish it off, why on earth would you not simply shoot with the mortar another time? Why waste a drone? Why are they doing this?
The other question is: if FPV drones are so ineffective, why does everyone say that the Ukraine war has led to a new radically different era of warfare, in which for example tanks act in an extremely cautious way? Why do tanks act in an extremely cautious way if the suicide drones aren't such a threat? Is it because of the threat posed by the combination of spotter drones and artillery?
To a first approximation, mortars aren't accurate enough to *hit* anything; they do their damage by near misses. And while a near miss by a mortar shell can kill a man, and *might* wreck a truck, it's not going to do much to a tank or armored personnel carrier or whatnot. Except maybe shred the tires of that APC, or with a large mortar bomb even break or dislodge the tracks of a tank.
Which is fixable, and probably *should* be fixable by the tank's crew. But if the crew is three Russian conscripts hunkered down in an immobilized tank in hostile territory with mortar bombs exploding randomly around them, they're probably just going to stay hunkered down until there's a lull in the mortar barrage, then skedaddle back to friendly(ish) territory and say "our tank broke down, please send a big-ass tow truck to recover it for us".
Since the Russians do in fact have big-ass tank tow trucks(*), you probably want to make sure the tank is really, irreparably destroyed before they can do that. But since your mortars probably didn't disable *all* the enemy tanks, you can't just send a team of soldiers with demolition charges to do the job; they'll just be machine-gunned by the tanks you didn't disable.
Enter the FPV, which takes longer than a mortar to employ and so wasn't your go-to weapon when the problem was "TANK! Enemy Tank! Do Something Now!", but when you set it up properly it can fly into disputed territory and deliver a modest anti-tank warhead precisely to a vulnerable side of the tank.
Earlier in the war, the preferred solution was a reusable "bomber" drone that could e.g. drop a grenade through the hatch the crew left open when they ran away and then fly back for reuse. But Russian electronic warfare eventually reached the point where cheap consumer-grade drones might last three whole flights if you were lucky, which made the "reuse" part less valuable and the simple reliability of the FPV drone more appealing. And at the very start of the war, the preferred solution was the fabled "Ukrainian Tractor", because the Russians were such inept cowards that you could fairly easily get them *all* to run away and then, hey, free tank.
* As do the Ukrainians; this sort of thing is usually done by putting a tow rig in place of the turret of an older model tank that you were otherwise going to retire. https://en.wikipedia.org/wiki/Armoured_recovery_vehicle
Yes.
(Also, I remember hearing that *bomber* drones, the kind that work like a plane and drop bombs before returning home to pick up more bombs, might actually be a bigger deal than FPV drones here. After all, a reusable drone can afford to invest in the cameras, sensors, radios, weatherproofing, etc. you need to address the challenges facing FPV drones as outlined in that War On The Rocks article. An FPV drone that's basically a missile, as you pointed out, can't afford that sort of stuff.
And of course, the new generation of fiber-optic drones inherently dodge any issues with radio reliability and jamming. And the *next* generation of drones, could always just use AI/autonomous guidance, to avoid all the difficulties of linking back to a human operator.)
As for your other question... the article doesn't go into much detail why. I would speculate that it's a matter of "scoot and shoot", or something like that, where a mortar team that fires must immediately start packing up and "running away" because counter-battery fire from the enemy artillery (or FPV drones or bomber drones or the like) is no doubt incoming. While they're busy running, an FPV drone can be sent in to finish the job.
>I keep hearing that "drones" are this groundbreaking revolutionary weapon. Isn't it amazing, a "drone" flies straight to a tank and explodes! That's pretty old technology, it's called a guided missile. How come everyone treats it as revolutionary and calls it "drones"?
Because most people are stupid. The number of takes I've heard since the war in Ukraine started that prompted the response "the 60s called and they want their theories on the future of warfare back" is not small.
At a very fundamental level, you are entirely correct. A suicide drone is a type of guided missile. End of story. But it's often different enough from a conventional guided missile that it might be worth thinking about somewhat differently. I was running tests on various types of missiles in Command: Modern Operations a year or so ago, and found that the Shahed was slow enough (~120 kts) that it was trivial for the ship to maneuver out of their seeker range, which is very much not the case for any of the proper missiles I used in that test. (Actually, it was so bad I had to deliberately maneuver the ship to stay in engagement range.) Also, the lower closing speed opens up a bunch of weapons options that don't work for normal high-subsonic missiles. So far as there is anything revolutionary, it's in the cost, but even that I think is often overstated to some extent.
>A rocket is cheap, and goes faster than a mini-helicopter. If the elements that grant a "drone" remote control/AI, vision and steering exist in a cheap and expendable mini-helicopter, why not in a cheap and expendable rocket?
Because the drones are free-riding on civilian hardware. A typical FPV drone is basically made of commercial parts with some piece of ordnance strapped to it. If you do it properly, you need to pay for new ordnance and a custom rocket (big enough rockets are not produced in large volumes) as well as custom hardware. And you'd probably make it something approaching Milspec, because there are morale and reputational penalties if a significant percentage of your stuff doesn't work out of the box. (This is a real issue with FPV drones in Ukraine.)
I do think we'll see more drone-ish weapons entering the arsenals of serious powers, because it is pretty useful for an infantry company to have the organic capability to hit things slightly over the horizon. But exactly what that is going to look like isn't quite clear yet.
Missiles are almost always rocket or jet propelled, supersonic or nearly so, and very difficult to shoot down. They almost always fly directly to a target known, identified, and located before launch, they do not loiter and report back to their owner, and they destroy their designated target and themselves ASAP.
Drones, have traditionally been none of those things. And these are not independent variables - e.g. if you want your system to loiter and provide you with intelligence and targeting data, you probably can't do that at Mach 2 (or even 0.85).
With the increasing diversity of drones and drone applications, the general rule that drones have *none* of the characteristics of missiles no longer holds true. Something like a Shahed is halfway to being a cruise missile. But only halfway. The difference between "exceedingly fast and very hard to shoot down" and "basically a Piper Cub", is huge and worth calling out in the name.
While there are a bunch of people saying differently, I've always thought the core difference between guided missiles and drones is that the central examples of missiles are ballistic rockets and the central examples of drones are propeller or rotary-wing aircraft.
Cruise missiles are between the two categories and are arguably closer to drones, since they definitionally rely more on aerodynamic lift than ballistics to fly and since they usually use air-breathing engines (turbofans, ramjets, scramjets, or pulse jets) rather than rockets at least for the middle phase of their trajectories. I'm guessing they're called missiles rather than drones for historical reasons.
It looks like the term "drone" in the context of aviation dates to the 1920s, applied to light aircraft configured to fly unpiloted under radio control so gun crews could practice shooting them down. Presumably, setting human pilots to fly the target planes from the cockpits would have been bad for morale and prone to unacceptable attrition rates. Later, drones got used as reconnaissance platforms and the like. Then, in 2001, the US started using high-end drones as weapons platforms. Not as weapons, but as platforms: Predator drones, which had been developed for reconnaissance, were armed with hellfire missiles that they could shoot at stuff, but unlike a missile the Predator itself would return to base if it could. Using expendable drones as actual weapons rather than as reusable weapons platforms came somewhat later.
As to why missiles are more expensive, I'm guessing several possible factors:
1. Rockets are much less energy efficient propulsion systems than propellers or copter blades, and ballistic trajectories are also much less energy efficient than aerodynamic ones. And for chemically fueled rockets or aircraft, rockets need to haul around their own oxidizer while air-breathing engines use oxygen that they find conveniently lying around in their flight path. Add it all up, and a rocket just plain needs to be bigger in order to have the same range as an expendable drone.
2. An aerodynamic vehicle is steerable almost for free, since they need to generate configurable amounts of lift in order to take off and fly in a straight line. Cheap unguided rockets are not: they tend to have no control surfaces except for fixed stabilizer fins, so you need to add actuators to the fins (or add new control surfaces) in order to make a rocket guidable.
3. The ultra-cheap quad copter drones are mass-produced civilian toys (or low-end tools) with after-market control software and explosive payloads added on. There are no mass produced civilian toy guided rockets that I know of, and unguided mass produced model rockets suffer from factor #2.
4. Armies that want to use missiles because they need or want the speed and acceleration advantages of a rocket over a quad copter are generally willing and able to pay a premium for a bigger and more capable missile and aren't trying to optimize for the cheapest viable guided munition. Ukraine is the exception because their supply of missiles is constrained by what NATO countries are willing to give or sell to them, so they're supplementing with drones for targets that don't need missiles.
5. Most established missile weapons systems are decades old and rely on bigger and clunkier hardware for their guidance systems. Newer versions that take advantage of recent decades' improvements in computer hardware are still in the pipeline.
6. For cruise missiles in particular, they're big and expensive because they're high-end weapons systems that are orders of magnitude more capable than the cheap drones in terms of speed, range, and payload. Jet engines cost more than small electric copter rotors, and 20-foot-long airframes carrying 1.5 ton payloads aren't cheap, either.
The main difference between drones and missiles is the former's ability to maneuver and perform more duties at once. Ceteris paribus, anything that can maneuver is more useful than anything than can't.
A drone can observe the battlefield, relay radio signals further downrange, hide between the bushes and lay in wait, then attack and explode, all in 10 minutes.
Missile can only fly at one target and attempt to kill it. No other functionalities.
Why aren't there guided missiles as cheap and expendable as one of those Ukrainian quadcopters?
Making things that go fast is expensive. Making it go a long range is expensive. Sensor and guidance packages that can withstand battlefield conditions are expensive. And as long as you're paying the cost for rockets, you might as well put enough explosives on it to make it worth your while, which makes it big and even more expensive.
The Ukrainian (and unfortunately also Russian) quadcopters are basically civilian drones with an olive-drab paint job and an IED. And increasingly more customization as the battlefield environment gets harsher for drones, but the core hardware is produced in vast quantities for the commercial market.
Which means, cheap because economies of scale. It also means not really rugged enough for military operations, particularly once the enemy realizes what you are up to and starts looking for ways to break your system. Back in late 2022, the average lifespan of a commercial drone in Ukraine was three flights, and I'm pretty sure it's come down from there.
Something rocket-powered and so blindingly fast, with a guidance system that doesn't use GPS or radio, and born under the sign of milspec, is much more likely to A: work at all, and B: reach its target, in a contested battlefield. And we could maybe bring the price down to something in the four-figure range, if we built those in the quantities that e.g. DJI builds quadcopters, but we don't.
Also, the warheads of a lot of drones used in Ukraine are repurposed old military munitions, e.g. RKG-3 grenades, that are already bought and paid for. The price for *anything* that contains a significant high explosive payload and is guaranteed not to blow up before you want it to, gets you into the high three figures right off the bat if you have to buy it new.
1. Missiles are typically autonomous after launch (guided or self-guided). Initially, drones required operators to monitor their course and targets; however, now AI is enabling autonomous attack drones, but even autonomous drones have more flexibility over post-launch target acquisition than missiles. Although many modern missiles do have smart guidance systems, the speed of missiles generally limits the correctability of their course post-launch.
2. Missiles can carry much larger payloads than most drones, but likewise, they're significantly more expensive.
I learned some new things about drones watching this video...
https://www.youtube.com/watch?v=ct0G-vZHbMk
Why are missiles more expensive? Why aren't there guided missiles as cheap and expendable as one of those Ukrainian quadcopters?
Good question. I think you'd need to ask the military contractors lurking on this list. Of course, the drones being built by Ukraine use mostly off-the-shelf parts and electronics. Unlike ATGMs they probably don't perform well (or at all) in inclement weather.
But I just checked — a Tomahawk cruise missile runs between $1.87 million and $2 million, depending on the version. But it's designed to carry 450 kg of HE, or a W80 nuclear warhead, which is 150x the weight that a Ukrainian Osa drone can carry. it has to be reliable in conditions where an Osa drone wouldn't be. And it probably has all sorts of nifty custom-designed avionics. And military procurement costs inflate the price somewhat. But even smaller Anti-tank missiles seem pretty expensive for the warhead weight they can deliver, and their range is pretty short.
A US Javelin antitank missile costs US$176,000 (yikes!). Range: 150 m (max?). Payload: 8.4 kg HE.
A French Akeron MP antitank missile costs €120,000 (also, yikes!). Range: 150 m - 5km. Paylaod: 2 kg of HE.
Meanwhile, Russia and Ukraine have largely withdrawn tanks from the front line because small FPVs, equipped with 1.5 kg shaped-charged warheads, can take them out more cost-effectively than expensive ATGMs. Ukraine lost some M1-Abrams to drones — so, even our most advanced tank, costing between US$10 million and US$20 million, can be taken out by a drone costing less than US$1,000.
https://en.wikipedia.org/wiki/Tomahawk_missile
https://en.wikipedia.org/wiki/FGM-148_Javelin
https://en.wikipedia.org/wiki/Akeron_MP
You can't compare directly a Tomahawk to the Osa drone. Beside the much bigger payload, the former can get about 60 times farther than the latter, also the former is extremely hard to stop, whereas a quadcopter can be shot down easily.
The quadcopter type drones aren't such a catastrophe for tanks. They can be shot down easily. For example, this system costs only 1 and a half millions:
https://en.wikipedia.org/wiki/Slinger_(weapon_system)
I assume that a Javelin is not as easy to shoot down.
But it turns out that drones aren't that easy to shoot down. They fly under radar. They loiter unobtrusively. And the tank cages that have been improvised don't really do a very good job protecting tanks.
IMHO, they've completely changed the dynamics of warfare. A NYT article (Mar 3, 2025) stated that drones kill more soldiers and destroy more vehicles in Ukraine than all traditional weapons combined, according to Ukrainian commanders. At least along the lines of battle, drones have been more effective than infrantry, armor, artillery, missiles, and fighter jets COMBINED. 🤷♀️
https://www.nytimes.com/interactive/2025/03/03/world/europe/ukraine-russia-war-drones-deaths.html
If you're interested in a contrary view about drones, here's one.
https://warontherocks.com/2025/06/i-fought-in-ukraine-and-heres-why-fpv-drones-kind-of-suck/
I got the link from this comment:
https://www.astralcodexten.com/p/open-thread-389/comment/133836208
The author served as a drone operator in Ukraine and says that those small kamikaze drones are overhyped. He says that the videos that people see online where those drones do amazing things represent the exception, not the rule. He says that those drones are extremely unreliable, and in most missions they were used to finish off something that had already been hit by artillery or by bomber drones, and in almost in all cases the job could have been performed better by artillery.
But I'm skeptical since it doesn't make sense to me that drones are sent to finish off targets already hit by artillery. Why would they do this, instead of another artillery shot?
Yes, of course we've all heard how effective and central drones are in Ukraine. Although I would like to split "drones" into quadcopter drones and loitering missile drones, they're two completely different items.
I'm just saying that a quadcopter is much easier to shoot down than a traditional missile is. Over time systems designed to shoot down quadcopters will improve and become cheaper, so quadcopters will seem less invincible, and tanks will seem less doomed.
Of course people will not stop obsessing about "drones", since these days everything is a "drone", especially as AI puts humans out of business, and whatever weapon replaces the quadcopter as the scariest thing, they will call it a "drone".
"drone" covers a bunch of different flavors of remote controlled / unmanned weapons
some drones act like fighter aircraft, others like bombers, others like artillery, and others like missiles
the "revolutionary" part is that this is a new category of weapon with fundamentally new and different constraints that we've barely explored
if you saw a plane over a WW1 battlefield you might say "well that's sort of like a scout on horseback, but higher up" or "that's sort of like artillery, but without the need for forward observers"
but you'd be missing the fact that this is just a whole new general-purpose technology -- it is not bound by being "like" anything that came before
Again, why are people calling missiles "drones"?
Back when planes where invented it was obvious what was a plane and what was not. If we had asked a person back then to define a plane it would have been pretty clear cut. Now the concept of a "drone" is never defined and seemingly arbitrary.
What is a "drone"? What sets apart a Shahed from a Tomahawk? Why is one a "drone" and the other a "missile"? Several people have tried to answer in this comment thread and they give conflicting answers. You said "remote controlled / unmanned" but so is a TOW missile. Other said it's the ability to loiter, or the ability to send back information, but some cruise missiles have those features and we call them missiles, not drones. Someone said it's just the size ("a Tomahawk costs at least 1 million USD and is a pretty big target") but is that really what defines a drone? Being small?
Besides, I grew up with many different words for military aircraft, such as "fighter", "bomber", "spy plane", "attack helicopter", and the meaning of those words was self-evident or nearly so. Likewise we have many words for military ships, people don't say "a ship", people say "aircraft carrier", "cruiser", "submarine"... But now everything is just drones, drones, drones, drones.
We should ban the word "drone" so people come up with clearer categories.
I suggest "minicopter" for the quadcopters and hexacopters, "cruise missiles" or "loitering missiles" for the suicide planes, and for the other unmanned planes "unmanned planes", for example "unmanned spy plane".
> Besides, I grew up with many different words for military aircraft... we have many words for military ships
This is exactly my point.
Manned aircraft are a mature technology. We have many different words for the varieties of them. Ditto for manned ships.
Drones are a new technology. We do not yet know what the categories of drones are!
Looking for a vaguely similar but profoundly different existing technology and using that name instead will just mislead people. If you refer to Shaheds as "cruise missiles", you will confuse people much more than if you just call them "drones".
Concretely: If Iran launches 100 Shaheds at Israel, and you say "Iran launched 100 cruise missiles at Israel" people will (wrongly) think that Israel is going to have an extremely bad night. If you say "Iran launched 100 drones at Israel" people will (rightly) think nothing much will happen. So "drone" is a much more useful descriptor.
On one hand, you're right that "cruise missile" is misleading for the Shahed, because people will think of a Tomahawk, which is 100 times more expensive.
On the other hand, when people say "drone", people thing of something like the quadcopters widely used in Ukraine, which are 100 times cheaper than the Shahed.
Apparently one widely used term is "loitering munition" for the kind of item the Shahed is.
I don't know that "loitering" is a helpful way to describe a thing that is mostly used for a mission profile of "fly a pre-programmed route and then dive into a pre-determined target"
at some point we will come up with better names, but we're not there yet. "drone" (or maybe "kamikaze drone") is still pretty hard to beat for clarity. no one things Iran is launching quadcopter drones at Israel from 1,000 miles away; it's implicit that the thing is chonky enough to fly for a few hours.
The Shahed is also being used in Ukraine and in the context of the Ukraine war it's useful to have separate words for the small cheap quadcopters and the bigger more expensive kamikaze planes.
You do realize we send the spy planes in to suicide themselves, sometimes with very big and flammable results? Why pay money to get the plane back, when it's so cheap?
>We should ban the word "drone" so people come up with clearer categories.
Is this a proposal based more on vibes, or did you suffer a material disadvantage from your confusion?
I'm not sure what you mean by "material disadvantage", but see for example the discussion below, where in the context of discussing the mass production of drone there is confusion regarding the scope of drone production in the Ukraine war (due to quadcopter drones being 100 times cheaper than "loitering munition" drones).
https://www.astralcodexten.com/p/open-thread-389/comment/132884937
Maybe you should, at least for the moment, accept that the development of military drones of various sizes and capabilities are a rapidly evolving process that is partially shrouded in secrecy and encompasses a very diverse set of devices whose capabilities overlap with those of traditional systems, and so defy easy classification around the edges. While I like putting things into neatly labeled boxes as much as anyone, that's just pretty difficult right now, and I don't see how "banning" a useful word that might sometimes be used imprecisely by the general public would help clarify things at all.
If it helps, I would offer this description of "drone" that should fit most of the time as of today: A military drone is an unmanned, self-propelled, remote-controlled carrier of sensors, and often of weapons or other payload. It might have limited autonomy of movement and target selection. By design, it can be re-usable or suicidal. It can perform a wide range of battlefield missions such as supply running, medevac, mine/barbed wire laying and clearing, general reconnaissance, artillery spotting, and direct engagement of enemy land, sea, and air assets at short, medium, or long range. It is generally lower in capability and cheaper to produce than any non-drone system whose mission it can fulfill, making it more economical in situations where otherwise a specialized, high-capability system might have been used, and generally expendable.
> a scout on horseback, but higher up
This is a hilarious description of planes.
Missiles don't loiter.
Missiles are usually much less maneuverable than drones.
Som missiles, like the Tomahawk cuise missile, can loiter.
No, a standard Tomahawk cruise missile cannot loiter in the way that a drone or loitering munition can.
However, some variants have limited re-targeting and maneuvering capabilities, which may resemble loitering in a loose sense—but this is not true loitering (e.g., hanging around for an hour waiting for the target to appear).
Block IV TLAM can and I didn't say it was exactly the same as a drone. They have long range so loiter time is potential large as well. If saying they loiter is wrong I invite you to correct wikipedia.
The navy.mil site says it has loitering capability, but it doesn't say how long.
https://www.navy.mil/DesktopModules/ArticleCS/Print.aspx?PortalId=1&ModuleId=724&Article=2169229
Considering that it has a range of about 900 nautical miles and burns a type of jet fuel, it may be able to loiter for a while, but looking at the pics of it, its wings are short and stubby — unless they can extend further out than the pic in this link shows? This suggests to me they'd have limited maneuverability and would have to operate at a high thrust to maintain their flight. Do you have more precise info on how they can manage to loiter and for how long?
https://www.naval-technology.com/projects/tomahawk-long-range-cruise-missile/?cf-view
Tomahawks and armed drones have very different use cases; they may have some overlap, but not too much. A Tomahawk costs at least 1 million USD and is a pretty big target, especially when it loiters; it's well worth a SAM or two to defend against. Drones are much cheaper, used for different targets, and require different defenses.
My understanding was that drones send information back.
Two way data links are standard on most modern munitions.
Then why is everyone worried about how to counter "drone swarms"? Why not missile swarms? What does the ability to send information back have to do with being cheap and expendable and a swarm?
They ARE worried about missile swarms too! And have been since the Second World War. Drones are the NEW thing to worry about.
I agree that the existence of autonomous drones muddies this distinction, but I suppose ultimately, the practical difference is what kind of defense you need against them, and if there are two kinds of missiles needing different countermeasures, calling one kind by a different name is sensible.
But why do most of the cheap suicide drones look like helicopters instead of rockets? I find it bizarre that when soldiers want to take out a vehicle, instead of launching a guided rocket at it like in the olden days, now they launch a guided helicopter at it. What's the advantage, compared to a rocket? Everyone says the helicopter (what they call a "drone") is cheaper, but how can a helicopter be cheaper than a rocket (all else being equal)?
If I google "drone swarm" and "missile swarm" (in brackets, to ensure that the whole phrase is searched for as a whole) the former gets 50 times more hits.
According to the following link:
https://armscontrolcenter.org/wp-content/uploads/2017/04/Ballistic-vs.-Cruise-Missiles-Fact-Sheet.pdf
"As advanced cruise missiles approach their target, remote operators can use a camera in the nose of the missile to see what the missile sees. This gives them the option to manually guide the missile to its target or to abort the strike."
Which seems to contradict your explanation of what sets apart drones from missiles.
"Bizarre ... What's the advantage, compared to a rocket?"
Launching a drone is a lot more stealth than launching a missile/rocket. If you want to survive on a modern battlefield that is constantly observed by a lot of spying eyes and robots, you'd better avoid emitting too much smoke, flame and noise at/near your position. Nothing bizarre about not wanting to draw attention of the enemy to yourself.
The olden days had much fewer, if any, HD cameras floating around, so launching a projectile or a missile with a flame and a bang was much safer for the operator.
Go look at anti-tank FPV footage coming out of Ukraine. An ATGM can't sneak up along a treeline and then come at you from the rear. Or chase you into a tunnel. Or drop a grenade down the chimney of your bunker.
Missiles are one-shot machines with a fairly simple gameplan - fly into something and explode. Drones are more like mini air support - controller can get it to do all sorts of clever and nasty things because he's piloting it (either directly or indirectly via waypoints) and so has more options about where it goes and what it does.
But they're cheap and don't involve a pilot who dies when they get shot down (expensive to train, expensive to recruit, etc.), so you can do missions that would not be sensible for a manned aircraft or an expensive missile.
Interesting, however the question I am most interested in is why they're cheaper.
When they're missiles, it's called a "barrage" instead. Basically the same idea, but they're traditionally coördinated only at launch and not during flight.
As I said, yes, modern tech makes the distinction is fuzzy. The point is that there are two qualitatively different kinds of long range offensive weapons.
because drones ('helicopters') can loiter; return to base after dropping a payload; are human-guided and can use on-the-fly decision making by the controller; can be retasked to specific target or aimpoint; can be very, very precisely guided (i.e. stop when they see a defense; find a hole in it; then fly through that hole), are not as heavy as rocket engines (but are slower); can be easily adapted (i..e different warhead); can serve purpose as reconnaissance or repeater platforms; can land next to a road containing potential targets, shut their rotors off, and wait until a vehicle passes. All of these are obvious if you watched a few day's worth of footage from the Ukraine war.
>how can a helicopter be cheaper than a rocket
scale. The drone consumer market made construction at scale possible.
Technically you could have some of these in a rocket, but not all, and not for the price of a drone.
> The drone consumer market made construction at scale possible.
This is surprising: I read that Russia is making about 500 drones a day for the war, and Ukraine about 100. Is the consumer market that significant?
Shouldn’t predictive processing theory be having a larger effect on personal and clinical psychology? I feel like it’s still a kind of a secret that reality is not actually poring into us through the orifices in our heads…
https://slatestarcodex.com/2017/09/05/book-review-surfing-uncertainty/
Seems like it should be obvious to anyone that it's not pouring in directly through sense organs. We have examples before us all the time of how we need some sort of intermediate processes to turn what pours in through the head oritices into knowledge that we are observing a certain event. Babies, children and animals and animals provide a constant stream of examples.. We can observe in all 3 many occasions when it's clear that they experienced sensory input of a parttifular kind, but could not make sense of it the way an adult could have. I think most people grasp intuitively that there's some kind of digestive process that turns what pours into the orifices into information, and that some of it is learned.
My feeling is that the vast majority of people operate under the mode of naive realism — that what they see is all there is.
This is one theory among many, at least if Erik Hoel is correct that we don't have a unified theory of how the brain works.
Predictive processing is kind of the basis of cognitive behavioral therapy
CBT definitely finds its roots in the ancient understanding of the interpretability of reality and our particular framings of events. So we can say predictive processing was anticipated beginning more than 2000 years ago. The fleshing out of the scientific concept and our understanding of how our sense of reality is mediated by our senses, their processing, schemas, predictions and modelling takes it to another level doesn’t it? Like I don’t think CBT regularly teaches clients that they’re not *ever* directly experiencing “reality” the way they think they are.
I don’t think predictive processing is that different from any other recent theory of mind in terms of “directly experiencing “reality” the way they think they are”.
For me personally it really brought it home.
>I don’t think CBT regularly teaches clients that they’re not *ever* directly experiencing “reality” the way they think they are.
No, they don't. I doubt that that would be very helpful in most cases. It's a big leap. In my opinion though, if you pursue CBT seriously you will eventually figure it out for yourself. And once you figure that out, you will know what you are experiencing directly.
>The fleshing out of the scientific concept and our understanding of how our sense of reality is mediated by our senses, their processing, schemas, predictions and modelling takes it to another level doesn’t it?
It is probably helpful if you're using that information in a way to get you to believe the idea. You can just as easily use all that information to confound yourself though. The human mind has thousands of rabbit holes to get around things that are uncomfortable
It certainly IS a secret: I only know of it because of that post, and haven't encountered it anywhere else.
EDIT: I've gotten a lot of good advice, nice and not-so-nice (but probably true). Thank you all for your time and effort!
You are basically asking about the meaning of life, which as the wide ranging answers you’ve gotten suggest, people LOVE talking about, because figuring out the answer to that question is literally what we were put on this earth to do; it’s the most satisfying pursuit there is. The thing is, though, no one else’s answers are going to be right for you; it’s the process that matters, not the outcome. If you are looking for a kind of sherpa through that terrain, you might try some Jungian psychoanalysis, or dipping into any of the wisdom traditions that resonate with you. Also, secretly, I’ve always thought that this would be my dream job (or hobby I guess; feels weird to monetize it) so if you want to let me practice my non-existent spiritual sherpa skills on you, feel free to message me…though honestly from the responses you’ve gotten here I bet you could pick literally anyone you felt some amount of affinity and respect for and go from there.
Honestly, I wanted the opinions of people who were (a) reasonably similar to me (that's why I picked Astral Codex Ten) and (b) weren't going to go nuts about the money thing, which sets the general band of my options--I could do something non-remunerative, but I can't refund the NIH.
I've gotten plenty of opinions, and plan to go through and thank everyone. I'm then going to step away from Substack for a week to mull all this over.
Thank you for your time. I don't know about wisdom traditions, but I know way more than a man of science should about Western occultism. Maybe picking up a little Jung isn't a bad idea. Thank you again!
Western Occultism counts as a wisdom tradition in my book! Good luck, may Hermes Trismegistus guide you on your path.
It doesn't in mine. I've heard it's all a secret sex cult IRL, which I am way too old and ugly for.
Though, maybe Jung.
> "I've given up on the novel idea; it would take too long to get good at writing, and I don't expect the current trend toward mostly-female literary audiences to reverse in my lifetime. Also, I've realized ChatGPT is a better writer than I am."
What the fuck are you even talking about.
"If only you'd remember before ever you sit down to write that you've been a reader long before you were ever a writer. You simply fix that fact in your mind, then sit very still and ask yourself, as a reader, what piece of writing in all the world [Anonymous Dude] would most want to read if he had his heart's choice. The next step is terrible, but so simple I can hardly believe it as I write it. You just sit down shamelessly and write the thing yourself." - J.D. Salinger, Seymour: An Introduction
If you don't take any pleasure whatsoever in reading your own writing then, sure, give up on writing (and publishing) a novel.
But if there is a piece of writing in all the world that you would want to read if you have your heart's choice *and that piece of writing doesn't exist yet,* then you'll have to write it yourself, for *your* pleasure, not status.
And FWIW, slowly, grindingly chipping away at "getting better" at writing actually is the only way to get better at writing enough to really please yourself.
Will there always be people out there who are better writers than you (more creative metaphors, more comprehensive scope and research, cleverer twists, etc)?
Sure, but if you're anything like me, simply being exposed to their work will make you a little better at producing yours - and your work is is going to be much closer to your heart's desire of what you want to *read* than anything they produce.
(This is really, really true for me. Every 2-5 years, I'll randomly consume some truly jaw-dropping masterpiece and go into a brief, depressive, despairing cycle of "I'll never be anywhere nearly as masterful with storytelling as *that* person," but then when my admiration and despair fade after a few weeks, I inevitably discover a noticeable upgrade in my language usage, editorial judgment, and frequency of striking ideas. I'll then either edit or produce something new that I really, REALLY fucking enjoy reading way more than the truly jaw-dropping masterpiece.)
Write your thing for you, the reader. If it pleases you, the reader enough, it will likely please other readers, too.
But even if it doesn't, at least you'll have *your* heart's desire to read.
(Also? Fuck ChatGPT. No, it absolutely isn't as good as I am at producing the writing *I* want to read, and it will never be able to produce our hearts' desire, either, because it has neither a heart, nor desire.)
It's a lengthy and good argument. Maybe I'll try again. Thank you very much for your effort and for caring!
I think you should hunt out a really good coach and hire them to help you navigate the decision process. (Coaches are sort of like psychotherapists, except that they are geared towards helping people rise to challenges rather than helping troubled people get past roadblocks.). And considering going to places that offer some novel experiences that might wake you up and open you up. Psilocybin retreat? Meditation retreat? Heavy duty outdoor adventure course such as NOLS?
Thanks!
Thing with coaches is it strikes me as a great way to separate idealistic rich fools from their money. (Kinda like investment advisors, which are a way to separate greedy rich fools from their money.) Of course some of them are honest, but I have no idea which ones.
Jeez, you sure are rejecting a lot of suggestions. Are you sure you're actually open to suggestions? As for my suggestion, it's really a meta-suggestion about how to decide what's right for you -- try hiring someone who's a professional helper. Pretty low stakes. And it's not like it's impossible to check the quality of coaches. Ask on here! There are probably some other forums you can ask on, too. There are probably sites where people write lengthy reviews of coaches (who unlike doctors do not put up barriers to being reviewed.). Some coaches have blogs or have written books, and you can get an idea from that of how they work. And you can just see one a coupla times and see how you like them.
It's a point. I guess I'm just cynical about therapy personally. But that's a me thing.
Thank you anyway!
It’s not therapy ffs. Here’s the google Ai explaining, correctly, the difference:
https://www.google.com/search?q=whats+the.difference+between+coaching+and+therapy&ie=UTF-8&oe=UTF-8&hl=en-us&client=safari
Besides retirement, have you tried applying the same lucrative skills for a non woke position?
Other than that, just a wild guess here, you sound like you need some good close friends from different walks of life to hang out with in person.
And in time, I do recommend a partner. Marriage and kids are optional but life without a partner of any kind gets harder as life progresses
It's not a bad point. Definitely something to consider.
I think the thing about a partner is I am just too distrustful and gullible. Which makes sense if you think about it.
Thanks for the thoughts, though!
> What should I do with the remainder of my life? Any ideas?
I'm about your age, and here's some things i've enjoyed doing since retiring:
1. Getting fitter than I've been in my life (okay, I was already pretty fit, but I was a lifter before, and got into triathlon training and Olympic distance races, so a different type of fitness). When you have more free time, you can *really* dedicate time to doing a good fitness program, and to cooking and eating better, and so on, and it's easy to get fitter than you've ever been in your life because you have so much more bandwidth and time.
2. Traveling - I've always enjoyed travel, but I spent a good couple of years living out of nice airbnb's and vacation rentals across 7-8 countries.
3. Starting a substack - I've really found it rewarding to clarify and articulate my thoughts, and for my book review posts, to read books more in depth, with more external research and callouts, and with a view towards the wider world, or to read them adversarially, and clearly pinpoint where and why I disagree. It's also spurred many fun conversations and exchanges, and I've been able to introduce people to new facts, arguments, and ideas multiple times by posting apropos links to some of my posts when it's relevant.
4. I like cars, so I've raced a little more, brought my fun car to more car shows, did a "supercar road trip" across half the country, and a couple of other things.
5. Not this particular time, but between two companies in earlier years, I spent 6 months RV-ing across the US Southwest, Montana, and PNW, visiting various state and national parks, taking my tailgate motorcycle on various day trips and for groceries, and generally having an amazingly scenic naturey sort of time, and it was awesome. So peaceful, and if you spend the day mountain biking, hiking, climbing, or scrambling, it's the absolute height of luxury to come back to a queen sized bed, real toilet, and real kitchen instead of a tent.
6. I go to a lot more concerts - if you're in a major metro area, you'd be surprised at how many good acts come to town, I can usually find something every 1-2 weeks. I use a site called songkick for this.
7. You can read more, cook more, get into woodworking - any hobby that you've dabbled in before, you can really go all out on, it's just a matter of finding one that you really intrinsically enjoy.
Not sure if any of those hit the mark for you, but they have for me.
I will say that even with all of that, I'm being tempted into doing another company, because the impact you can have on the world creating a company is so high, and so meaningful, in terms of impact on the world, your employees, and your customers, that it's really intrinsically satisfying, even though it's a lot of work.
So maybe that's something to consider too - what product or service do you wish existed in the world, but doesn't? You can directly bring that thing into being, and if you do it right, you can do it using OPM - other people's money. You don't have to use your own funds, and you shouldn't, there's well established institutions set up to fund new ideas and companies.
All fun ideas. Fitness I should do just on general principle, it's healthy. Cars, travel, you know, I didn't grow up with them, though they're certainly fun for people who enjoy them.
The Substack is a fun idea, though I'm not sure the brand I've got on this one is what I'd want. But it's definitely a thought!
Starting companies really requires a very different personality with much higher optimism and tolerance for risk, not to mention extroversion. (Yes, I've seen the Big Five scores of billionaires paper, I'd also need higher conscientiousness and lower neuroticism.) I applied to a few investment banks back in college (I was young and stupid) and after landing on my face in the interviews I wrote off the whole business field. Though I appreciate what you do, I just thought you should know that.
Thanks for the long and thoughtful response!
You're 45 with $3M and financial freedom, but you sound miserable. The problem isn't what hobby to pick up - it's that you've built your entire worldview around fear and resentment. You avoided marriage because of "divorce settlements," you're bitter about "feminism", and you won't even try writing because you've decided the audience is wrong somehow.
This is self-imposed prison thinking. You have the resources to do literally anything, but you've talked yourself out of human connection and creative expression based on internet grievances. Get therapy. Work on why you're so angry at half the population and why you see relationships as financial threats rather than human connections.
Your life sounds like its own punishment right now. Fix the internal stuff first, then figure out what you actually want to do with all that freedom.
I'm angry at the political movement, not the gender. I don't have that much in common with the other half of humanity and our interests are diametrically opposed in many cases, but that's not really their *fault*, and both can be pretty much chalked down to the exigencies of biology. The battle of the sexes is older than humanity and will likely continue after it in whatever inherits the earth after us (unless it's the paperclip AI after all.)
As for financial threats--I mean, I agree lots of people enjoy and find huge amounts of value from relationships. I think I have a bad combo of poor character judgment and high but difficult to replace income--basically, I'm a great target for the divorce scam. Just because I shouldn't get married doesn't mean NOBODY should.
Thanks for weighing in--critical responses are just as useful as supportive ones!
So I was doing some pithy joking below about not marrying Hot But Crazy, but real talk: Your assumption here that Bitches Be Golddiggin' isn't necessarily supported by the data about your single female ***PEERS***, who:
* are more likely to be homeowners than their single male counterparts (https://nypost.com/2025/01/29/real-estate/single-women-own-millions-more-homes-than-single-men-in-the-us/)
* are more likely to have college degrees (https://www.foxnews.com/us/college-degree-gender-gap-widens-with-younger-gen-xers-study-finds)
* are actually slightly more wealthy than their single male childless counterparts (https://www.pewresearch.org/short-reads/2024/11/04/among-unmarried-adults-women-without-children-have-as-much-wealth-as-single-men/)
* are sometimes anticipating large inheritances from wealthy Boomer parents (https://fortune.com/2025/03/28/millennials-richest-generation-on-record-great-wealth-transfer-from-baby-boomers/)
Which is to say, amongst your single female ***PEERS***, there are many who might want to protect their assets from you, to the degree of wanting ironclad prenups or to avoid the hassle of potential divorce by not getting married at all. Marriage is an institution which primarily exists to nurture children, so there's an extremely good argument that it's not worth the risk of getting married for a childless union.
You may have noticed my emphasis on the word "***PEERS***.
That's the critical word here.
Your single female peers are the women +/- seven years of your 45 years, the baby GenX and elder Millennials aged 52 to 38. These are women who began working in an okay-ish economy and have had enough decades to build the kind of careers, personal wealth, and experience which would prevent them from coveting yours, and, as a bonus, who no longer have the issue of future children looming on the horizon.
Single female 31 year olds and especially single female 23 year olds are not your peers. You shouldn't even be considering anyone young enough for you to have conceivably fathered (or grandfathered!!!) as real "adults." Getting into a partnership with one of them - particularly a debt-laden poor one - is *adoption,* not marriage, and when you adopt anything, pet or child, you have the obligation to pay to support it.
Good thing you don't need to adopt!
Perhaps you're horrible in person, but on paper, you are a goddamned unicorn to your single female ***PEERS.*** If you give dating apps or a matchmaker a parameter of "single, child-free career women 38-52," you will be absolutely inundated by wealthy-ish to legitimately wealthy single women who roughly want the same kind of life you do. It's not that hard; the odds are very, very, very much in your favor.
But that?
That feeling right there?
That feeling of resistance upon being asked to seriously picture yourself partnering with a 38-52 year old woman, because you're used to seeing 19-24 year olds in your porn?
That is most of the reason why the "divorce scam" you're so afraid of even exists.
I'm not looking right now. But...
Your assumption I was looking for 19-24 year olds is quite well statistically supported given lots of evidence about the average male preference...but I don't watch porn (believe it or not)....some of us have good imaginations. If one expressed interest in me I would assume it was exactly as you claim!
If I were listening to the pronatalists (which if you read my responses you see I am not) I'd probably be going early 30s...after getting into shape so they assume I won't drop dead in 10 years. And even then it would be an extremely uphill struggle. I am not under any illusions as to my own attractiveness!
FWIW when I was on OKCupid (and, uh, Feeld) I never matched more than 5-6 years below my age (and in fact went older on multiple occasions), and never found someone where the money thing wasn't a concern.
Well, I wasn't necessarily assuming you were *actively* looking for hot 19-24 year olds; they just happen to be the most visible and common demographic in porn and as the prize owed to male protagonists in all of media (as observed here: https://www.cracked.com/article_19785_5-ways-modern-men-are-trained-to-hate-women.html).
(Sidebar: I'm occasionally amused when people make sweeping universal observations like, "there are no women on this dating app!" or "women get hit on in public by men all the time!" The large population of all the *other* women on that dating app and the women who don't get hit on in public aren't even considered "women" because they aren't sufficiently aesthetic to be fundamentally *visible.*)
Anyway.
I'll offer a apology for assuming you were seeking partners 10+ years younger, but I'm still giving a side-eye to the idea that *everyone* is a threat to your wealth. When was the last time you were looking? Five years can make a big difference...
Apology accepted.
(Sidebar: Yeah, I had heard something similar and was taking it into account; I view it as the other side of 'all the monogamous guys are short' (an actual line in a New York magazine article). Basically, people who don't meet the speaker's frequently-unrealistic standards might as well not exist. )
Sadly, I actually have to do work now. Good luck!
>You avoided marriage because of "divorce settlements,"
That is entirely reasonable. The odds are very roughly 50:50.
>You have the resources to do literally anything,
False. In fact, simply retiring at this point with his finances is possible, but somewhat risky (as other commenters have noted), let alone any grander choices.
> The odds are very roughly 50:50
"The odds" are not your personal odds. Your personal odds are substantially better than 50-50 if you do some simple things, like:
1. Don't get married aged <25
2. Don't marry someone you've dated for less than (say) 3 years
3. Don't be poor, don't be uneducated, don't be a drug addict, don't be violent or prone to angry outbursts, don't be a criminal, don't be a jerk, and make sure your spouse isn't those things either
4. Don't marry someone you only kinda-sorta love or someone who only kinda-sorta loves you
and a bunch of other common-sense things.
That still doesn't lower your risk to zero. But... is divorce the end of the world? Seems to me that you just wind up back where you started, being single. Maybe a bit poorer.
Simplest thing:
0. Don't marry Crazy just because it's really, really hot.
...but what if it's really, really, *really* hot?
Also what makes them hot is something in us that's missing, and desperately yearns for something that person is playing up to. It might be an act, it might be they were trained, accidentally or deliberately, by a parent or just the world. But chances are the person's hotness is a dark reflection of our undeveloped stuff.
Develop that stuff, and yeah, you'll find fewer people desperately hot, but the whole world is a lot more peaceful and enjoyable.
Whatever is making it really, really, *really* hot is almost certainly also what's making it really, really, *really* crazy, so think about the other 23 hours and 45 minutes a day you're going to be spending with Crazy.
Many Thanks!
FWIW, for my first wife, we actually met all of those conditions save that I was 24 at the time, but nonetheless she divorced me 5 years later (she ultimately said she wanted kids (she never did get them) and I didn't).
>Maybe a bit poorer.
Or a lot poorer. Prenups are a thing, but I don't know how much they help. In Anonymous Dude's case, he has the savings from half his working life to protect.
Divorce is _NOT_ a small hazard.
In the grand scheme of things it’s not that big of a deal. Better than spending your life shooting your self in the foot because, you know, …every one else sucks eggs.
>In the grand scheme of things it’s not that big of a deal.
Many Thanks! The potential to lose half one's savings might not be that big a deal in the grand scheme of things, but we live our lives on a human scale, not a grand scale. And losing half one's savings _is_ a big deal on the scale of one's life. Very early in one's career one can recover. Halfway or more through one's career - not so much.
As I wrote, prenups are a thing, but I don't know how much they help.
I thought this was a considerate, practical comment that points to the bright blue elephant in the room, fwiw.
Yeah, I wasn't offended. You lay yourself out like that you're going to get some tough love and I think it's not even necessarily wrong.
Less of this please.
No, I think it's a good comment, in the true-and-necessary category.
While I empathise with the OP's general point of view, I do agree that he should probably work on his resentment if he wants to enjoy the rest of life rather than waste it being grumpy.
One small attitude adjustment that helped me become less grumpy was to realise that yes, the world is imperfect, I can't possibly fix everyone else's flaws, but perhaps the best thing I can do is to set a good example. Instead of focusing on what everyone else is doing wrong, let me just be the best darn me that I can be, and maybe I'll be a good influence on someone else.
Anybody (including you, I see) who responds to a comment like this with some version of "Yep, these are all totally normal feelings, have you considered travel?" is doing this man a disservice. The fact that this guy doesn't know what to do with his life isn't the problem, it's downstream of the problem.
OP, you need to get over your hangups and resentments. Seek out some help in doing so, and get to work. Then the rest will take care of itself.
I recommend moving to a European country that's well-insulated from wokeness. Switzerland maybe? Slovenia? American culture is sort of over, at least for men of our generation. It's only going to get worse over the next 20 years. Move someplace you can enjoy participating in society. Spend the next couple years getting fluent in French or German or whatever and then retire and leave.
Start your own Substack that's anti-feminist but not meathead. The fact that you don't really fit into the existing ideological taxonomy means that you can create your own niche. There's a big appetite for outsider voices right now. See if you can build it into something.
I'm really, really hesitant about the whole politics thing now. I don't really see a comfortable 'tribe' to join--I'm too cynical and market-skeptical for Grey, too identity politics-skeptical for Blue, and just don't have the rural experience for Red.
I have thought about languages and exile, but I'm not sure which one to pick--Hungarian's hard as hell to learn. I do think you are right about this country going downhill.
> I don't really see a comfortable 'tribe' to join
Then create your own. That's the point.
Fuck Hungarian. Learn German, move to Switzerland. Most of Europe speaks English anyway.
Hungary, if you can learn not to smile.
And learn Hungarian? No thanks.
You have enough money for early retirement, so you can just quit. It seems you don't know yourself very well if you're asking a question like "what will I do for the rest of my life?". Time to follow up on any passing fancy you ever had, try things you haven't tried before, until you find your favorite thing in the world, then you can just do that.
But if his favourite thing in the world turns out to be expensive then he'll wish he'd hung around for a few more years to get a lil' bit richer. At $3M you can easily retire if you love kayaking and reading, but if it turns out your favourite things are race cars then you might have problems.
If his favorite thing is race cars it’s too late already at 45. Unless he just wants to own one.
It's too late to be a professional race car driver but there's plenty of middle aged dudes having fun racing cars.
OK. I suppose first we would have to define what we mean by a race car. Out of curiosity I did some googling and the whole Reddit thread about pretty much what we’re talking about here (middle-age guys getting into car racing for fun.) seems like the cost of getting a machine ranges from about $8000 up to $100,000 depending on what kind of racing you want to get into. Obviously, you could spend a lot more if you wanna go to Maserati‘s Ferrari‘s Lamborghini‘s McLaren’s etc.
it’s all kind of peripheral to the point, though, which is that the poor fellow doesn’t know what to do with himself.
1. Set a time limit for staying at the current well-paying woke job you hate. (4 years, for example.) Maintain the personal commitment to stick to that time limit.
2. Quickly come up with a plan for what you will do when you quit. Example: a one year sabbatical in a cabin in upstate Maine, followed by establishing a career as a freelance consultant/paladin in whatever field you have expertise in from your current job, but working for the other side.
3. Use the time before the limit runs out to solidify this plan or come up with a better one. My guess is that you will want to be self-employed as a service provider with a very small client list. Something like a private eye, but not a private eye.
4. When the time runs out, just do it.
Not a bad idea. I never quite got the outdoors thing, though I know a lot of people enjoy it. I'm not going to go into too much detail on what I do for obvious reasons, but an exit plan involving some of my prior skills is a good idea.
Thank you!
I'd be traveling the world if I had significant surplus funds. Or, were you one of the people who didn't understand the need to travel that some people have? (We may have had this discussion a few open threads back.) Anyway, seeing places and people you've never seen will give you a greater horizon of opportunities than sitting at home being bored and introspective (that's not to say you can't get bored and introspective traveling, but moving on to the next place will generally distract you out of it).
Yeah, not a huge travel person historically. But it's something to think about.
Thank you for chiming in!
Don’t get married. Find someone you hate and buy them a house.
You realize you are basically preaching to the choir here, right?
But thanks for the gag! I can always use it if I need an excuse.
Yeah I had a feeling….
Lol. The Nice Guys?
Just a joke I heard a long time ago that for some reason stuck with me. I have no idea where it originated.
Then the movie stole it. “Marriage is buying a house for someone you hate.” Underrated film actually
I prefer the formation that I know to that one. For some reason, it's just not as funny to me. I told it to my divorce lawyer a long time ago, and I was very impressed at how she managed to keep a straight face. I was also very surprised she had never heard it. It kind of reminds me of something Henny Youngman might've said. "Take my wife… Please!"
Some people attribute it to Johnny Carson, which I find easier to believe. It's kind of Borsch belt.
The quote is always attributed to Rod Stewart. He's recent enough that it could really be him.
A source for the quote is never provided, though.
I have a hard time believing that but who knows
Well, I'll tell you what. You try to find it attributed to Johnny Carson, and I'll try to find it attributed to Rod Stewart.
No, it’s not worth the trouble is it.? I asked ChatGPT and Rod Stewart is the number one candidate. It just doesn’t work for me if he said it. Let it lie. I’m sure someone said something a lot like it a long time ago.
Help Ukraine.
I appreciate your concern, but I don't really think getting involved in a war on the other side of the world is a great idea.
No it isn’t.
Goodidea. Murder, however, is a rather dicey proposition.
Ukraine is currently in the midst of getting flattened by Russia.
(Okay, in all seriousness, it is becoming shrubberies, which, yes, are not mentioned in the NATO Combined Military Services Doctrinal Handbook -- our tanks are not doing well.).
It's not to late to get married and have kids. "The purpose of life is grandkids." (George H.)
What is it with the pronatalists? I know you guys want high-IQ people to breed but my sperm are carrying more mutations than the Ninja Turtle franchise at this point!
Thanks for the thought though!
" (I really need to give up the book-buying; I did *not* need that Arkham House first edition)."
I was shrugging along until I hit this. Had I pearls, I'd be clutching them right now.
Need? NEED???? Sure, you don't *need* it, but you don't *need* a lot of things. Being a cow grazing in a field can be a happy enough life, but do you really want to live on the level of "time to graze, no other needs at present"?
You're going to make me invoke Maslow's goddamn Hierarchy and I hate that thing, but an Arkham House first edition is one of the levels above the base "graze in the field" level.
Are you human or not? Give up buying books, might as well give up breathing while you're at it!
Eh, I was running out of space.
It might make a passable hobby, I suppose, but I'm less good about reading them than buying them.
Thanks for chiming in though!
> Eh, I was running out of space.
The traditional solution at this point is to open a second-hand bookshop.
There's no evidence Maslow's "hierarchy" is an actual hierarchy https://entitledtoanopinion.wordpress.com/2010/07/18/what-evidence-is-maslows-hierarchy-of-needs-based-on/
Give nature that which nature only needs, man’s life’s as cheap as beasts
I have, in my time, bought books rather than food so someone saying "yes I'm comfortably off though not rich rich, what should I do with my money (maybe I'll cut back on book-buying)" is stabbing me through the heart. Through the heart, Horace!
Who is Horace?
Is there anything that grieves you about the state of the world? Not politics, that'll just depress you. But does it bother you that science isn't progressing quickly enough? That architecture in your city isn't beautiful or functional? That people die of preventable disease? That we don't have some awesome technology you dreamed about as a kid? Or something much smaller would work too. Quit the job you hate, and find a way to work on it.
> Not politics, that'll just depress you.
> But does it bother you that science isn't progressing quickly enough? That architecture in your city isn't beautiful or functional? That people die of preventable disease? That we don't have some awesome technology you dreamed about as a kid?
Political decisions are a significant part of the reason all those other things are as they are.
That's true, but they're also all problems that we can do something about, without needing to solve the political problems first.
Politics is a uniquely frustrating thing to spend your time and effort on. Despite your best efforts you'll probably never be anything other than one voice out of millions. Sometimes good things will happen and sometimes bad things will happen, and it will almost certainly turn out that none of the effort or brain power that you applied to the problem made any difference at all.
It's much better for your life satisfaction and mental health to focus on meaningful smaller-scale changes that you can actually make, rather than dreaming of the large scale changes that you'd love to make if only someone would let you. I can't change the course of global architecture, but I can build a pretty house for myself. And maybe that will inspire someone else.
Yes.
“All politics is local.” - Tip O’Neill
Science? All y'all are working on that and much better knowledgeable and connected than me! Also with the way the internet and social media turned out I'm not sure I want more progress, to be honest!
Architecture? Yeah, but that would require a lot more dough than I have. The point is I've probably got enough to *live on*, but not enough to *do anything with*.
Preventable disease? I kind of feel like the EAs are on that one and there's not much I'm going to do. (That and I can't take anyone talking about shrimp welfare seriously enough to give money to.)
Re: preventable disease, instead of thinking about what percent of the world you can help, think about the individuals. You could afford to save the life of at least one kid for a pretty minimal amount of money. Think about a 10 year old kid who is about to go blind (or die!) from vitamin A deficiency. Maybe he is the only son in a family, who will rely on him as he gets older for income.
Yeah awesome, you can save that kid so that he can grow up and father three kids that will go blind. Don't invest in failure, that just leads to more failure.
That's a $20 dollar fix, for the kid about to go blind. Seriously, it's cheap!
Generic advice follows, but maybe you need some...
A steady stream of minor achievements or acheiving mini goals can be pleasurable. Have fun! In some sense, a sense of progress = fun.
If your work is actively making you unhappy, and you have achieved financial independence, consider optimizing your life around more meaningful work.
If you actively *hate* your work, you are probably going to be somewhat unhappy. To mangle a quote, happiness is when your thoughts, feelings, and actions are aligned.
Join a writers club near you! Take workshops along the way to understand structuring, techniques, and get practice. If you see a novel as the way to influence the world, then maybe recognize that and find better ways to do so. Why do you want to write? Fiction? Character exploration? If you *really* want so see a novel happen, you can lower your expectations and write a bad novel. And build from there.
I don't think I'd influence the world; I'm not that grandiose. (It's quite possible I'm not egotistical enough to be a great artist.) I used to look at all the sci-fi and fantasy paperbacks as a kid in Barnes and Noble and dream of seeing my name up there with a fun colorful cover, but now the market has shifted.
Have you ever played table top role playing games? There are professional game masters now. This is another avenue for story telling. Also meeting people.
I am one of your feminist female bugbears and I would love to read more new arkham house weird scifi with strange cover art written by childless men, no sarcasm. bring it on, i dare you
You might not actually be a bugbear depending on your precise views, but it's probably not worth getting into a political argument here ;)
Thanks for the encouragement though. Arkham House no longer exists...but if you mean fiction by weirdos...I don't know. You like the ones by childless men for some reason? I could see that, more adventure or something, or a consciously childfree woman might like to see the other side.
I think if I ever finished I'd at least pay for an artist! Sadly now that everyone's reading electronically there isn't much call for the great Michael Whelan or Rowena pieces anymore. It's another part of the world that's gone cold.
> but now the market has shifted.
If you really think about it, this does not have much to do with your urge to write a book. It says a lot about the outcome you would want if you were to write a book, but don’t you have some personal inner reason to want to write a book which has nothing to do with the market? If not, I can only think of two reasons. You need to make money writing a book or you’re letting your grandiosity get in the way of doing something that is personally significant to you.
Yeah, there was definitely a certain amount of desire for glory in there. Which I don't feel that guilty about--we're all at least a little ambitious.
Certainly no need to feel guilty about it. Ambition is not a bad thing, as long as it doesn’t get in your way.
I think you're making excuses for abandoning the author dream. Yes, there's far more adversity to overcome, and if you had to go through the traditional gate-keepers, it's probably harder for you to get published than for anyone else since the invention of the printing press, but self-publishing is a perfectly workable alternative. If you're not relying on it for income, and just want to have your work read by others, the usual arguments against it don't seem to apply.
EDIT: I originally said "anyone else in history," but I don't think that's quite true.
Does chatgpt actually induce psychosis in people?
It seems it can push vulnerable people over the edge. This isn’t surprising. Imagine an evil psychologist who wanted to make his or her clients psychotic - it seems likely this would meet with *some* success. AI is great at agreeing with people, which can confirm their delusions.
I’m interested in how many of the cases mentioned in several recent articles were specifically ChatGPT during that one week of extreme sycophancy in late April, as opposed to other LLM assistants. That is, I’m interested in whether that particular model was extremely effective at bringing out psychotic ideas, while others have been better safety trained to avoid that.
I suspect that "induce" is probably wrong, but "amplify" would be correct.
yeah, give a potential psychosis a machine that is both socially considered wise and has a knack for always finding a way to agree with you and youll finally know what mass scaled SaaS is - Schizophrenia as a Service
This is the seed for one hell of a variation on the old zombie apocalypse story.
Maybe of interest to the Georgists here - I wrote about the heterodox interpreters of Henry George's ideas (Theodor Herzl, Peter Thiel among others) for Asterisk: https://open.substack.com/pub/asteriskmag/p/the-georgist-roots-of-american-libertarianism.
I read that article when Tyler Cowen recommended it. I enjoyed it more than most long form articles I’ve read in a long, long time. Thank you for writing it.
What do you think the prospects of accurately and fairly estimating land value in urban areas actually are? This is the part I despair of, particularly distinguishing autogenous value increase and value increased by ancillary services created by other actors.
I also worry that the incentive to vertically integrate would be enormous, to internalize more of the autogenous value increase.
That's very kind! Lars Doucet has written about the former question for ACX here: https://www.astralcodexten.com/p/does-georgism-work-part-3-can-unimproved and re: the latter I think that's probably right, as David Friedman describes here: https://daviddfriedman.blogspot.com/2023/01/a-problem-with-georgism.html. There's a very good list of the other major critiques in this LessWrong post if you're interested: https://www.lesswrong.com/posts/CCuJotfcaoXf8FYcy/some-arguments-against-a-land-value-tax.
I enjoyed Lars’s article, but I’m not sure he dealt with the problem of autogenous increases in land value.
If I build a 50 story office building, and your development corp builds 500 units of mixed use retail/housing around it, I just don’t see how you separate my office building’s gravitational pull from the amenity value of of all the mixed-use nearby. If we don’t want the landlord a black away from the retail to get a windfall, I don’t see even theoretically how we fairly assign the increased land value to all three actors.
Lars’s article seemed to basically assume land value is a continuous function of such a large number of actors that you can basically just accept that my entity’s effect on the value of my own land is minimal, or if it’s large, I’m probably using the land so intensively that the autogenous increases is a very small drag on my improvement incentives.
Anyway, thanks for other links, looking forward to reading them.
What to you makes the autogenuous value increase a problem. Is it the incentive, you are getting taxed on value you create which is the problem lvt is trying to avoid? A secondary effect is still much better than a primary effect (income, property tax etc).
I live in a city where transactions on empty land is common so measuring land value autogenuous or otherwise doesnt seem difficult. Seems like with better land use, which lvt is meant to incentivize, empty land transactions will be more common as land regularly gets repurposed, so itll get easier to measure overtime.
Does anyone know how to apply for the Astral Codex grants? The last post I can find about it is from 2024. Do you need to be a subscriber?
It's not continuous, Scott announces it when it happens. You don't need to be subscribed.
Sincere thanks.
What is happening with SpaceX Starship?
I understand that failures are a normal part of a rocket testing process, and SpaceX "iterative development" philosophy in particular is bound to have a lot of them. I am not implying neither that the Starship is "doomed", nor that it "should" progress much faster. But it looks to me like there are two very different trend lines here.
From the beginning of Starship high-altitude testing and then for integrated flight tests 1 through 6 we see steady improvements, where basically each flight goes a bit better than the previous one. Then for flight tests from 7 onwards we switch to a different mode, where it looks like each flight only discovers additional problems (flights 7-9 all had various problems with the ship, and the ship for flight 10 blew up during the static fire).
I notice that I am confused here. It looks like there is some latent variable which changed between flights 6 and 7, switching them from one mode to another. Does anyone have reasonable ideas for what this could be?
Some hypotheses (not necessarily plausible, just things that *would* explain the change if they somehow turned out to be true):
* Starship block 2 problems. Basically, there were some serious design problems with block 2, leading to it having not one but at least 3 different failure causes. If this is the case, why was their design *so* bad, much worse than block 1?
* Musk problems. Maybe for some reason Musk's supervision and decision making are crucial to SpaceX success. Then between November 2024 and January 2025 he became much more busy with the government, didn't have enough time for SpaceX, and the problems followed.
* China intervention. Maybe the CCP decided that SpaceX was progressing a little too smoothly a little too fast and gave an order to sabotage them somehow? (Again, I don't have any model for this, I really have no idea whether this is even possible, let alone plausible)
* The gods are angry with them. Maybe the coin really happened to first land on heads 6 times, and then land on tales 4 times after that.
Does anyone have a good model for this?
As others have pointed out, it's the difference between Starship Block 1 and Starship Block 2.
As for why they don't just go back to the version that sort of worked, Block 1 worked tolerably well but was heavy enough that it basically couldn't carry any useful payload to orbit; it was barely capable of lofting its empty self into the lowest practical parking orbit. My model gives it 10,176 kg of payload to a 185 km orbit, a bit over half of what SpaceX's workhorse Falcon 9 can do and not even close to being worth the bother of a Starship.
Block 2 might be able to deliver a hundred metric tons to Low Earth Orbit, and that's definitely worth doing. Block 1 is only valuable as a proof of concept for the basic system; SpaceX absolutely has to get Block 2 (or Block 3) working or the whole thing is a monumental waste.
Block 2, the vehicle that matters, is currently 0 for 3. But the Falcon 1, SpaceX's first rocket, crashed and burned on its first three launches and made it to orbit on the fourth try. That's SpaceX's style. Fingers crossed that it will keep working.
This pattern seems par for the course in SpaceX development and even in Starship development IMO. If you go back to the high altitude hop testing you mentioned, SN8-SN11 all exploded before SN15 kind of worked. Same thing with F1 dev and F9 landing - a string of failures where not much progress seems to be happening, people start getting worried, and then all of a sudden they whack enough issues and the whole thing starts working pretty well. If ships are still blowing up in 6mo to a year I'll start getting concerned.
Starship Block 1 was robust, but not as performant as SpaceX would like it to be.
They are pushing the envelope with Block 2, trying to get 25 per cent more payload capacity from it. That is a huge capability leap, and the envelope is not very generous.
You missed:
* Rockets are stupid hard. Starship is the biggest and eventually most powerful rocket ever built, with by far the biggest payload. It is also intended to be fully reusable and hence extremely economical. These are competing design goals that require stretching envelopes far, far beyond what was ever previously thought possible. Failures during development are virtually guaranteed, and its better to fail fast and learn fast.
Sol has it broadly right. The booster part has gone pretty well and the issues that they had were ironed out in the first few flights. That makes logical sense - SpaceX is a rocket company, and though the booster is bigger and more complex than any they had flown before, they likely had a reasonable idea of how to get it to work.
The Starship upper stage is a different kind of problem. There have actually been numerous problems with it since the beginning, but they are often not easy to spot and SpaceX has tended to downplay them from the beginning. Attitude control in orbit seems to be a persistent problem. Reentry itself is a massive challenge, and - even though they did manage a few controlled landings - they are clearly still far from solving it fully. And notably, they are still doing only the simplest suborbital test.
So I think a reasonable take is that the first few flights showed good progress with the booster, which allowed casual observers to overlook the issues with the upper stage. Since the booster is now looking good, people are paying more attention to the harder part. And simultaneously, SpaceX is trying to redesign it to fix the existing issues, but are then running into other (very visible) problems.
"Block 1" was the first iteration of starship that was used for the first 6 flight tests. "Block 2" which is slightly larger with significantly more ambitious internals, has been used from flight 7 onward.
I think it's too much of a coincidence for the first 6 flights to go incrementally better each time, then the next 3 launches to all have unique mission-ending problems with the updated starship design. It looks like they got the booster right judging by its incredible success at both catching, and reusing it, but something is definitely wrong in Block 2.
I wouldn't update too much on it. They are extremely ambitious with every new development, and although SpaceX innovates quickly, at launch you're looking at rockets that were designed at least a year earlier. S35, the Starship that launched most recently in May began construction back in October 2024, about 3 months before the first Block 2 Starship, S33 Failed. Thus, very few lessons from S33 could have been used for S35, if any at all. And certainly no lessons were learned from after RUD, which for flight 7 and 8 was before the problems seen in flight 9.
It's of course a bad sign that they have had multiple failures in a row, as any test failure is. However, if you look at SpaceX as designing a Starship manufacturing system as much as Starship itself, then the loss of any individual rocket isn't actually that big of a deal. Combine that with infinite money for all practical purposes, and while each failure sets the mission back, it doesn't move it meaningfully towards cancelation. As is, they have the largest rocket booster on the planet that they've successfully reused (this is truly an insane accomplishment), the highest output rocket engine production facility on the planet (with so far high demonstrated reliability, at least on the booster), and a factory capable of producing an accelerating number of Starships and boosters. They are also building literally a second version of the entire setup they have in Texas, in Florida, which will add some serious redundancy, an internal spirit of competition, and allow for significantly faster launch cadence.
What I read (and maybe someone more into the subject can comment whether this makes sense) is that the first couple of flights revealed that the engines could not deliver the expected thrust, and hence, deliver the expected payload to orbit. So the rocket's design was revised with an emphasis on saving weight. Designing a rocket always means operating on the very edge of what's possible in terms of materials science, and trying to shave off a few percent of weight may have pushed "Starship" from "yeah, with a little luck" into "nope" territory.
While you're right the new design, Block 2, was largely designed with weight savings in mind (anything saved on the reusable 2nd stage is a fraction of the gained payload due to the rocket equation), lack of engine thrust is not the primary driver of that.
The engines have actually over performed on thrust and SPI. They've adjusted the design to make it larger, with a higher target mass to orbit to accommodate for that increased thrust.
The only comment I have is that,
reviewing the history of Starship tests, I don’t see the pattern you describe.
My only comment is 'neither do I'
When I became a parent, two things happened (well more than two really but there's two I want to talk about now):
- I appreciated my own parents more
- I became familiar with dealing with the disgusting and unpleasant parts of daily life in order to support another human that I loved who couldn't take care of themselves.
Now my mother is physically disabled and my father has dementia, and I find I rely on these two benefits of having children in the care that I now help to provide for them.
Which makes me wonder: is there any research on the different attitudes to caring for elderly relatives among child-free people compared to people who are parents themselves?
I knew from a fairly young age I wouldn't have my own children and married a woman who biologically can't do so, which has worked out for both of us. A large part of my decision was the reality of my own childhood.
I don't think my parents should have had kids or even gotten together as a couple at all. My father could have made it work with a different woman I think, my mother was largely the problem. She was far too consumed by her own problems with mental health and addiction to do much parenting at all. She ditched my father for vague and changing reasons when I was 2 so it was just my sister and me at her mercy. Fortunately she was easy to avoid, assuming she was even around to need to avoid.
My wife and I will never get the chance to answer this question though. All of our parents died relatively young. My dad was the oldest to die at 65, her father was the youngest at 48. While I spent most of my childhood (And adult life) avoiding my mother as much as I could, my wife genuinely loved her mother and was devastated when she died at 52. Her reaction was both very touching and entirely alien to me. My own mother’s death was a moment of quiet relief for everyone that knew her. Simply being alive was often too much for her, and her suffering was finally over. I like to imagine she had some sort of positive influence on at least some of the people that knew her, but I struggle to think of anything.
I really don’t know what I would have done if any of them lived long enough to need care. I doubt I would have taken an active part in caring for my mother if she were elderly to be honest. Most well-adjusted people seemed to have been effortlessly bonded with the parents, especially their mothers. This wasn’t my experience. A large part of my decision to not have children stems from having no idea what a good, or even acceptable, childhood and child care even looks like. I’ve heard my friends talk about the joy of reliving childhood experiences through their own children. I have no conception of what this even means; I hated being a child, and I hated myself when I was a child. Things got better as I got a little older and realized what I could get away with as a teenager as long as I had good grades and adult authority figures slotted me into the “good kid” bucket. In retrospect I learned some very antisocial ways of getting through life then and very little to prepare me for being a adult net-contributor to society.
As I think about it more, I don’t think I would have. I’d probably give my sister money to help if she was willing to.
Thank you for sharing something so deeply personal.
I don’t know about the relevant research, but my Dad has recently had a brutal descent into Alzheimer’s, and frankly, the fact that 2/3 of his kids don’t have kids seems like one of the handful of blessings that has helped keep our heads above water. My sister, with two young children, is drowning in caretaking; I can’t imagine what we’d be doing if all three of us were in that situation. I think ultimately the care required in these situations will expand to fill (and then almost capsize) whatever is available, but the wider familial benefits of a handful of childless aunts and uncles who can swing in for backup suddenly seem very visible to me.
Has your family considered putting him in a nursing home at all? I am very curious to see how the opinions of people about nursing homes does over time. Most people I talk to still have a visceral "no" reaction, but Alzheimer's is often the one exception.
He is now in memory care, but the process of getting him there was wild + excruciating and consumed our lives for more than a year (there were complicating factors.) And even now, with his safety and basic daily needs taken care of—he is in an excellent facility; collectively my family is better off in terms of resources than 99% of people in our situation—I think we all *still* feel there is so much more we could be doing. Which really does speak to that sense that, probably much as with kids, half the battle comes from accepting your limitations and knowing that you’re going to do it imperfectly. But man is it tough.
- half the battle comes from accepting your limitations and knowing that you’re going to do it imperfectly
Yes this, absolutely. Just like it's generally a sign of being a good parent that you occasionally wonder if you're a bad parent.
Very familiar with the difficulty of that placement - it similarly upended our lives for probably the same amount of time. There's a lot of prep work if you're doing it well. There's a picture somewhere my dad took of all the paperwork he had to do, and if the all the pages were bound in a book, it'd probably be heavy enough to kill someone with.
My mother has been in her facility for 21 years and I still get that feeling all the time, even though she's at the point where there's just objectively so little I can do. My father/her husband would visit her 4-5 times a week, and her parents would cover the other days. They either didn't work or had non-traditional jobs with a lot of flexibility. Now that they're all dead and I'm working a traditional office job, not being able to keep up that routine really hurts. I guess it is good I am child free - it's already so easy for life to get in the way and prevent me from visiting.
Yes that's a really good point. My kids are older so less of an ongoing demand (and even able to help out in some regards). And I also recognise the idea of care needs expanding to fill availability.
I'd be very interested in any research like that. Anecdotally, I suspect there's a large difference.
I'm child free and have a disabled parent who lives in a care facility. I didn't put her there, but after living with her in both circumstances, I found it a lot easier to participate in the process when she was in a care facility.
People who've had children are absolutely more confident in handling the disabled, in my experience, but I do also sometimes think that's a bad thing.
Disabled people are not children or perfect innocent angels but seem to be commonly bucketed into one of those mental categories by the able bodied who are not familiar with disability in their personal life. I think some caregivers never update how they care for someone and default to whatever they learned as a parent taking care of their baby, and get frustrated with disabled adults being "difficult".
I think caregiving is going to be very important in the years to come and do worry that some child free people who have not challenged themselves in other areas of life will find themselves failing to help people they love because they never got over their "yuck" reflex or learned how to do boring, tedious, necessary things for the benefit of someone less capable.
Yes, that's what I'm wondering about, but I don't want to just assume it's the case. One thing that gives me pause - although admittedly going purely on stereotype and cliche - is that some east Asian countries with very low tfr are seen as having a more positive attitude towards elder care.
Interesting. I’m not sure I share either of those experiences.
> - I appreciated my own parents more
I wouldn’t say that becoming a parent has made me appreciate my parents *less* – both before and after I became a parent, I appreciated them a lot. but I do think becoming a parent has made me focus more on parenting choices, and realize areas where they made parenting choices that I think were incorrect. (At least, according to my value system.) As a result, while I still appreciate the effort they took, I now have a slightly more mixed view of their parenting in general.
> - I became familiar with dealing with the disgusting and unpleasant parts of daily life in order to support another human that I loved who couldn't take care of themselves.
as far as this goes, although I haven’t faced the issue of elder care yet, I anticipate that the largest difficulty there will be emotional rather than physical. For example, if I need to change a diaper on one of my parents, I don’t anticipate that the physically gross nature of the task will be much of a problem compared to the emotional aspect of them, being embarrassed by the assistance/feeling shame.
And that’s obviously not an issue I face when changing my baby’s diaper, so I’m not sure it has prepared me for elder care in that meaningful of a way.
This is closer to my wife's experience, where becoming a parent ... clarified her judgment with regard to some of her mother's actions.
For me I think it's hard to truly understand the emotional weight that parents bear until you've been through it. I don't think I was a difficult child but I didn't understand at the time some of the things I put them through, that seemed trivial to me.
There has been a really interesting discussion about school in the book reviews, so now I am curious. In m’y European country, not only are options severely constrained, people typically view teaching as something that doesn’t mix well with a parenting relationship, so you delegate it all.
So, if you have school-aged children, do you teach them reading / math / Anglo-Saxon poetry / oboe playing yourself (in addition, or instead of teachers, whether public or private) ?
How did it go ? Do you use AI ? How much time does it take ?
Thank you !
I tend to think that humans of any age don't want people they have close relationships with to cosplay the teacher role. That doesn't mean humans can't take direct instruction from such people, just that we don't want to be bossed around by them. We always planned to homeschool so we always went out of our way to not make learning a power struggle and instead approach academics like vegetables. We model lifelong learning and seek out resources that are worth our kids' time and have a daily routine that includes a sustainable amount of academics rather than cramming them into 180 days and don't make unnecessary demands. It's a blast and I feel so lucky to get to do this.
> So, if you have school-aged children, do you teach them
> reading / math / Anglo-Saxon poetry / oboe playing yourself
> (in addition, or instead of teachers, whether public or private) ?
I can't tell at whom this question is directed.
I live in the US rather than Europe.
My wife and I homeschooled our (not particularly academically gifted) child K-12. We outsourced some of the stuff we weren't good at (e.g. sports). We did teach him math, reading, literature (he read "Beowulf" and "Gawain and the Green Knight" in translation; some of the Canterbury Tales in the original Middle English), history and science ourselves.
Thanks, this is really cool ! I am just interested in a different perspective because, in my social circle IRL, people teach very few academic subjects to their children themselves*.
I read that a lot of the commenters here have a different perspective, so I really appreciate learning about it ! It’s fascinating to see that it seems totally doable to a lot of you, instead of completely alien (wonders of the Internet, I suppose !)
*They, and those children, then go on to have mostly successful lives as high-level civil servants/ engineers / bankers / etc, but apart from the occasional 28-years-old crisis reconversion of finacial auditord as painter / baker / bricklayer, this is not a very creative or independent society, as you can imagine.
It is common for homeschoolers to stop at high school (so the professionals handle grades 9 - 12).
For a bunch, it is also sorta considered on a year-to-year basis. "Yes, I think I can teach a single 1st grader what he/she needs to learn," a parent might say while thinking about 1st grade. No need to decide right now if they can teach 8th grade as well. Or even 2nd grade.
I do believe that America is an outlier about this, though.
I have children that are not yet ‘school age.’ I’ve taught them to read (using phonics) and they’re working on arithmetic. I intend to teach them all I can myself, and get tutors for things like music where I’m not very talented. I don’t intend to send them to school.
Teaching them to read took about 20-30m per day for about a year. It’s occasionally frustrating but the rewards greatly exceed it. I don’t have any idea what I would use an LLM for in teaching my kids (besides the usual stuff I use them for, like finding citations/sources and conducting starting research on subjects).
That sounds great ! (I asked the question about AI because everybody here seems to think it will revolutionise education, but I would have no idea how to use it either.)
If it’s not indiscret, does that mean that you stopped working to take care of your children ? Or do you manage to do both ?
I do both. Work from home is very liberating in that way. I can shuffle my own schedule around.
> people typically view teaching as something that doesn’t mix well with a parenting relationship, so you delegate it all.
I would be very interested in hearing more about this. From my (admittedly American) perspective, so much of inherent parenting seems to basically be teaching e.g. potty training, teaching manners, teaching how to eat food and not throw it on the floor, etc. How does that fit in with a social norm that teaching doesn’t mix well with a parenting relationship? Or am I misunderstanding you?
Whenever I talk about teaching my children to read / music / a language I speak / whatever, parents here tell me that 1) school will do that them better than me 2) they should enjoy their childhood and not learn to read, etc too soon 3) it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
This mostly means that you have to fight with the children in order to make them listen and focus, and that a third person (like a teacher), whose relationship with your children is less emotional, can teach them more effectively.
After Covid and school closures, mort parents I know had a horrible time trying to force their kids to watch Zoom class while working from home and never wanted to do it again. So all you people teaching your kids must have tools or skills that they don’t !
I am also really curious because it seems so hard to find the time to have regular learning sessions with the children during the week while working, so I wondered how you fit that in your schedules. But that must be very individual and family-dependent.
> school will do that them better than me
I taught my children how to use a computer, and I certainly did a *much* better job than the school did so far. My older child is in 4th grade; at school they teach how to use MS Word and PowerPoint. At home, we already did vector graphics in Inkscape, and some simple programming in Scratch.
> they should enjoy their childhood and not learn to read, etc too soon
This is one of those stupid opinions that come from people who can't tell the difference between "forced to do something" and "genuinely curious about something". (Perhaps they were never genuinely curious about something themselves? Or maybe school destroyed all their natural curiosity?) How is childhood less enjoyable if - in *addition* to everything else - you can also read a book? Or a sign on a street, or anything.
Is it perhaps the very process of learning that is supposed to be unenjoyable? Well that's easy: only teach your children when they are curious, and stop when they get annoyed.
> it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
As long as you don't give them grades and homework, the roles are definitely not the same.
> you have to fight with the children in order to make them listen and focus
Perhaps that depends on the child. The school review article also assumes that most kids are unwilling to listen and learn. I have no idea how much that is true... my kids have inherited my genes, they *are* curious about stuff. Not all stuff, and not all the time, but that's okay.
> So all you people teaching your kids must have tools or skills that they don’t !
I think the required abilities are intelligence (that's mostly given in this thread), knowledge of the subject (shouldn't be a problem while your kids are in the elementary school; plus you can read their textbook), and having a good relation with your kids. Then if you tell them something, they will listen. Also, don't push it; if they are not in the mood today, try tomorrow. If they are in the mood only for 10 minutes, stop after 10 minutes.
> it seems so hard to find the time to have regular learning sessions with the children during the week while working
This is true. Most homeschooling families I know, there is a stay-at-home mom.
But you don't have to compare yourself to school. For example, if you are teaching your child to read, that is only one subject; the school teaches ten subjects at the same time. Also, school needs to proceed at a certain speed; you don't have to. Finally, school wastes a lot of time on classroom discipline and similar stuff; if you only teach your children when they are in the mood, this is not an issue.
So, homeschooling (all subjects, at the same speed as school) is difficult. But teaching your children a thing or two, at their preferred speed, is not.
> it is dangerous to take on the role of teacher when you are a parent, because those are not the same.
What I find fascinating about this from my own limited, cultural perspective, is that I view teaching as inextricably part of parenting, especially at young ages.
For example, I very clearly view myself as teaching my daughter to speak, to eat using her hands, to eat using a knife and fork, not to put her hands or fingers in electrical sockets, to use the bathroom rather than be in diapers, to dress herself, etc., etc. So from my cultural perspective, it sounds like you’re saying parent should teach their kids a bunch of things and then suddenly stop teaching their kids things (like reading).
Are you saying that in your culture those earlier parts of parenting aren’t conceptualized as teaching?
> This mostly means that you have to fight with the children
> in order to make them listen and focus, and that a third
> person (like a teacher), whose relationship with your
> children is less emotional, can teach them more effectively.
I have encountered this claim. I think it depends on the child, the parent and the topic. I rarely had to fight with my child to get him to listen and focus (*).
I did find that having a "academic education" spot made it easier to settle into the "now we are formally schooling" mindset. But it wasn't necessary.
(*) One thing that probably helped a lot with the focus was that he took Tae Kwon Do lessons starting when he was six. A *lot* of the repetition was "focus and concentrate." This carried over to other things (e.g. baseball) and also to schooling where we could tell him this and (a) he understood what he was supposed to do, and (b) had a lot of experience DOING it for someone else, and (c) saw how it helped in the Tae Kwon Do (or baseball) environment so maybe it actually mattered.
Those are interesting arguments. I was raised in a European country and my parents did not teach me math or reading, but I don't know if their reasoning follows along the same lines. The argument that viscerally resonates is the one about roles: teachers teach, parents parent, etc. so mixing responsibilities is not good.
Maybe it's my time living the USA now, but my gut reaction to these points is bewilderment. 1) maybe, maybe not, who knows what kind of teacher your kid will get; 2) reading and numbers can bring great joy, and additionally they can help the child be more part of society which can also bring joy; 3) like the other comment already said, parents already teach a great deal, so not sure how reading is different than teaching to fish or clean the house or take care of the dog.
How was your own childhood like in these terms? What is your stance on these questions?
Oh, I hated school, never learnt anything there, and was systematically bullied for participating in class, so I am in favour of burning all schools, if possible yesterday. But this is is considered very original here, and every time I say so people tell me that fighting bullies must have taught me to stand up for myself so I should not deprive my children of this socialising experience.
My father taught me to read on cereal boxes at breakfast, I then learnt most of what I know from books at home / family debates at meals during the weekends. So I don’t really have a model of a structured way to teach children in the family. I would love to, however, because I think it might be great fun ! (The eldest is 5, so we have a great time of answering weird questions about the world at the moment).
Around me, parents (mostly mothers) spend what seems to be a very unpleasant time repeating school lessons and exercises every evening, but any actual teaching (for example, music or language lessons) are delegated to private teachers, even for the basics.
When children are older, it is very common to delegate the « repeating the lessons and doing the exercises » part to tutors (usually students at university), precisely because parents find there’s too much conflict when they do that with teenagers.
Been thinking about the entire Alpha School vs normal school thing. When my parents went to school in rural Denmark in the 1970s they had 900 hours of teaching in 7th grade (so round 4.5 hours per day + homework ), and used a lot of the same motivation methods as the Alpha school dies . Currently it seems the US public schools have 1000 to 1100 hours per year. In 2013 we tried in Denmark to demand 1200 hours (so 6 hours per day). Despite this academic standards kept slipping, and the reverse Flynn effect are still taking away 3 IQ points per decade.
So it seems that actually the amount of hours above some limit is not very important. Perhaps with some simple incitament system even normal schools could go down to 800 hours per year + some homework and keep their current standard. That would not be very far from the ~600 hours studying that Alpha seems to be using.
Maybe culture in general has become stupider and schools have to try to compensate for this. Culture both in terms of medium and content. Eg people spontaneously read less because of TV or the internet. Music becomes less complex. And culture values ”coolness” over accomplishment (which is ”square”).
People spontaneously read more because of computers and smartphones.
Not necessarily, or at least, not "reading" as I imagine Kristian was describing it.
There's this, by a college professor:
https://hilariusbookbinder.substack.com/p/the-average-college-student-today
And that certainly aligns with what I'm seeing amongst my younger Millennial and Gen Z coworkers. They can't seem to absorb basic written information from emails or shift reports, nor can they adequately convey extremely basic information in writing. We are a high-end, service-centered luxury product, but they literally aren't able to put together so much as, "On Monday July 7 approx 9 AM, client complained that widget wasn't performing third state of widgeting. Widget analyst checked and says widget needs to be replaced, not repaired, will perform between noon and 1p. Offered client 50% off this widgeting as apology, client seemed satisfied."
I'm not exaggerating.
You know, there seems to be a really weird synchronization between the timing of this stuff and the emergence of AI. Everyone's worried that AI will cause mass unemployment but what if the opposite is true: what if AI is appearing just in time to save us from a catastrophic failure to adequately train the next generation of adults? If so, what could possibly explain that coincidence? It's almost as if people are taking their cognitive foot off of the gas because the AI has stepped in, but the order is backwards. The AI hasn't actually stepped in yet and this de-literacy trend has been going on for 10 years now. Is it some weird historical synchronicity?
> this de-literacy trend has been going on for 10 years now
I wonder whether people are actually getting less competent, or they are mostly the same but now more incompetent people are getting the credentials.
Like, in the past: 10% of people at university, all of them can read; today: 50% of people at university, one in five can read. The same number of people can read, but the quality of the university declines.
They’re not getting dumber because IQ is mostly genetic and genes aren’t changing that fast. However our insane obsession with equity means that we’ve stopped identifying the best and brightest. So our leadership and managerial classes are increasingly composed of poorly trained midwits who don’t really know what they’re doing. I suspect that means that the next 2008-level generational crisis we face will cause unrecoverable damage and it will shift us to a new and worse social equilibrium. In 50 years I think the US will be a richer version of Brazil.
To the extent that there is a failure to train the next generation of adults, it is because increasingly sophisticated algorithmic media (social and otherwise) is feeding them a steady stream of engrossing but fragmentary content that degrades focus, attention span, and the like. Not out of malice, but because their owners are paid by the click and if you're paying attention you'll realize that you don't need to click on something else just yet.
So you're basically hoping that the first generation of "good" AI will compensate for the damage caused by fifteen years of crappy proto-AI.
>If so, what could possibly explain that coincidence?
Perhaps the internet? The vast volume of online text is an essential input to the training of LLMs. And the same infrastructure lead to pervasive social media - and, arguably, shortened human attention spans.
Is the issue attention span? Or is it actual inability to /write/ as opposed to a different intellectual deficiency? (pre-frontal cortex issues can come from a lot of places, but we could blame schools for failure to teach leadership if we want.)
If the person is unable, without significant prompting, to relay that series of events (that you describe), then the issue isn't writing or reading.
Now, if you want to get really fun, recognize that the IQ Test Scores aren't going down.
>Now, if you want to get really fun, recognize that the IQ Test Scores aren't going down.
Why would IQ scores go down? IQ is primarily genetic and our genes aren't changing. We wouldn't expect a reduction in education to be reflected in lower IQ scores.
At the very least, whatever caused the Flynn effect could presumably be reversed. IQ is clearly not *only* genetic.
I find I can't absorb information from watching videos, I have to read the text to understand what's going on (apart from watching instructional videos about "okay so when you're taking apart the thingummy, *here* is where the release button is hidden that they don't tell you about in the manual").
I can imagine younger generations being the opposite, and getting all their information visually but not from texts. Which is great, they can learn the material. But then when they have to write down a message or report, as Christina says, they're sunk. Watching a video and listening to people talking is not at all the same as "this is how you write down the details of what happened, including all relevant information and leaving out what isn't necessary". So if it was talking face-to-face they probably could communicate "Mr Jones in Room 366 complained about the taps in the shower sticking" but having to write it down they can't work out that no, they don't need to include that they were sweeping the floor at 8:00 a.m. when the call came and yes, they do need to include the room number.
This is not a writing issue, this is a "leadership" issue, in that they are not able to put themselves into someone else's shoes. This is the sort of "basic communication" that one should learn before getting a job. I'd be willing to think of anyone who can't filter information, even when prompted to filter information, as fundamentally retarded. This seems like the sort of question that might work on an IQ test.
I know a guy whose native language is pictoral. He had severe issues with arithmetic because he'd keep on losing track of the pebbles (also had issues with symbols). He does NOT have issues with communication. It's not a visual versus textual difference.
(If you leave subtitles on, can you grasp the video better? Or do you have to read the text apart from the video?)
Maybe they read more words but fewer challenging texts. I have understood college professors have experienced a decline in the reading comprehension skills of undergraduates over the last few decades.
But the same was being reported in the 1980's.
My expectation is that they're learning just as much, but different things. Certainly reading extended text with a plot is different from browsing blogs.
I honestly doubt that they're learning more than rats who run mazes for cheese. Less, in that they have less patience and attention spans.
What you're looking at is deliberate addiction to quick-stimulus, leading to dramatically decreased attention spans.
What is the reputation of the FAR AI lab in the alignment circles?
Positively regarded by AI safety folks I know and work with. Well regarded as a space. Great staff and well run events, locally and internationally. I haven't heard anything negative about them at all.
I might be *slightly* biased towards them as I've worked with a couple of their staff, but I don't think overly so.
I've been wresting for decades whether libertarian ideas are right and wrong, eventually I ran across an interesting one. Do not see the market as an institution, like a bunch of businesses, but as a meta-institution evolving other institutions, it can evolve something like a government if that is best. Okay, but in that case shouldn't there be a partnership with the real government, and the real government helping that kind of market-government out?
Imagine the kind of town where half the population is working in the same privately owned mine. I think it is necessary that the town government and the mine will have a complicated partnership. The government needs to regulate against some abuses, but cannot possibly treat the mine as some sort of an unfriendly entity, they have to help them a lot, in order to secure those jobs. What I cannot imagine is being entirely neutral and indifferent to the mine, libertarian style.
All ideologies ignore a part of reality, you need to know which one.
Libertarians typically deny the parts where cooperation on a large scale is needed (for example, many are covid denialists, because pandemic is the kind of a thing where the market does not have much of a solution). Also, their claims how governments cannot do anything right are sometimes exaggerated (for example, many countries have a decent mass transit system), and at worst become a self-fulfilling prophecy (when people stop even trying to do things right, because they believe that "science proved it impossible", that's when even more things fall apart).
It is best to think about them as an antidote to the opposite fallacy -- that government always does everything right, and that everything needs to be micromanaged by politicians.
The market is a mechanism for organizing human cooperation more efficiently than anything else we've ever come up with. But it only works when the people doing the cooperating are all on board with the desired outcome; you can't round up the dissenters and put them to work on the project you think is the Best Thing Ever. Some people hate it for that. The rest of us, make it work for us. Together. Cooperatively.
Things that markets are great at:
* Pay attention to million different things at the same time. A central planning committee, or even a group of citizens deciding democratically, would only pay attention to a few topics that happen to seem most important to them at the moment. On market, someone *also* notices an opportunity to make a profit by improving something that is relatively trivial; this is how we get the relatively trivial things to work well, too. (This is why market economies typically do not have shortages of toilet paper, that planned economies are famous for.)
* Respond flexibly to changes in supply and demand. Things get more expensive before you run out of them; and as a consequence you often *don't* run out of them, because people respond to the increased price by switching to a substitute or finding a way to spend less.
* Provide relative freedom and autonomy. When there are many possible things to do, you don't have to do any specific one. If you hate X for some reason, you can choose to do Y instead. (If too many people do that, X will probably start to pay better, so some people who don't mind will switch from Y to X.)
* ...other things I forgot.
Things that markets are not so great at:
* Doing things that are not profitable, such as feeding the homeless.
* Avoiding things that are profitable but otherwise wrong, such as slavery. Reducing negative externalities, such as global warming, robotic spam calls, etc.
* Generally, following the law. The market always looks at a punishment for breaking the law as a price sticker and asks: "given the punishment, and the probability of actually being caught and punished, is it worth doing?" and uncomfortably often, the answer turns out to be "yes" (or "yes, if we spend tons of money on lawyers").
* Lots of resources are burned on zero-sum games, such as advertising trying to switch customers from product X to a mostly identical product Y.
* ...other things I forgot.
I see you're making multiple assumptions you haven't justified.
One is that the market is run by some sort of management capable of entering into partnerships. There is no such management, so there can be no partnership with the real government or anything else.
A second is that there's a government (I'm assuming you mean an institution with a monopoly on the use of force). There's an entire school of economic thought, known as anarcho-capitalism (or ancap for short, as I like to use), that discusses capitalism in the absence of a government. A market does not have to evolve one, although it's possible that something about human biology makes governments very likely. And I see use in discussing the market situations where governments do arise.
A third is that your mining town example exemplifies a market. It sort of does, but it's not free, and so many of the usual free market principles won't apply. The most obvious non-free market is the labor market: you're implicitly assuming the mine is the only customer for labor. It probably isn't, in a stable town - the people who work inside the mine have a demand for food, shelter, clothing, entertainment, and various other creature comforts, and division of labor principle strongly encourages them to purchase those things from yet other people. So, anyone in town deciding who to sell labor to, will have multiple employers to choose from. This is good news for people selling labor.
It is possible - and perhaps even likely - that townfolk will have an incentive to keep the mine open and employed. But a free market has a way of signaling it: a Help Wanted sign, with a salary offer attached. If the salary is high, more people will apply; they will work in the mine instead of farming, ranching, fishing, logging, weaving, and so on, so labor in those jobs will become more scarce, until those salaries rise as well (due to miners suddenly wanting food and houses and cloth badly enough to offer more money for them). Equilibrium occurs about when the salary for each job corresponds to the demand, modified by the training and talent required, physical strain, and risk of hazards on each job.
If the salary for mining goes up as high as it can go (the ore produced by a miner cannot be sold for enough to pay for that miner), then the mine is at capacity; if those miners aren't enough to sustain all the other labors, then laborers leave. Either way, equilibrium is reached.
Why should govt hating business even be a default?
Channeling Churchill - "capitalism is the worst form of political economy except for all those other forms that have been tried from time to time."
I am neither a tankie nor a Randian. Capitalism is the best way to create wealth, but not an infallible god. It's humanity, occasionally at its best and occasionally its worst. Most often muddling through somewhere in between.
I think you would find that the libertarian does not live in the town with the mine. He lives in a think tank. But perhaps that may be because I’m currently driving through a stinky area of oil and gas production - an area that fuels our economic growth, enables Amazon to zoom our presents around, enables rappers to rap, etc.
If you think of 2 original sins, pride and envy, the libertarian starts with the former: don’t tell me what to do. Sometimes that attitude may be informed by experience and justified.
But when so, that is because the other strand of liberalism - egalitarian liberalism, driven by envy - has overreached via government, become too powerful.
It is best to see libertarianism as a counterbalance. Otherwise it is an extreme, radicalized response to an extreme situation. We need government to keep the mine from polluting our drinking water, but we don’t need it to run the mine or redistribute the mine’s profits.
The short version is competition, incentives and safeguards. It's not about companies being better than states; All large organizations eventually suck if they have no competition, no good incentives, and no safeguards.
A market of many, minimally regulated companies offering similar products and employing similar workers will naturally gravitate towards efficiency to extract more money from each product and/or get a greater share of the market by undercutting the competition, and naturally gravitate towards better pay and better conditions for the workers because otherwise those will go elsewhere. Not perfectly so, and if you let them get away with it they will probably try to employ cartel tactics to get more, but as a general direction they will do these things. Markets also have the advantage that accommodating different people is a legitimate way of carving out a safe niche for yourself (though again there are some shitty tactics here).
Democracies also have, in principle, these components. But they tend to select most strongly for the things that sound good, as opposed to the things that actually work well, and they have a tendency to enforce the same thing for everyone. Nevertheless, you need some entity to oversee that the companies do not employ cartel and other shitty tactics, and democracies seem to be the least bad of the bunch.
The very worst is bureaucracies. Bureaucracies exist in most states, they are not voted in, they do not have to survive competition, their only incentives are rent seeking, entrenching themselves and insulating themselves. In many states, bureaucracies have successfully managed to make the democratic state dependent on itself as necessary middle managers. Likewise, large companies have realized that they can enter a symbiotic relationship with the bureaucracy through excessive regulation, wherein they insulate themselves from new competition by making entry into the market so prohibitively expensive and difficult that very few dare trying. This is were the idea of deep states, BigCorps, etc. come from.
As far as I can see, pure ancap libertarians are very rare. But it's imo reasonably to look at the current state of the, well, state and conclude that we need to regulate less, reduce the state coefficient and try to spread out power from a few big corps to many smaller again as much as possible. The main point of contention is usually whether the latter happens mostly automatically in a minimally regulated market, or whether you really need to outright smash the large corps.
For your example, there isn't really a market, so I expect no good outcomes. Maybe if you aggressively enable other smaller competitors.
If you have to introduce a very simplified and kind of unintuitive thought experiment to understand something (when you already have lots of direct evidence you could use), you probably have some preconceived notions about how the subject should work that the existing real world data doesn’t justify for you.
It’s difficult for me to understand why we should follow your intuitions about what kind of government policy we can imagine, right? Some people cannot imagine a government which doesn’t have a policy prohibiting hate speech, and yet I find it very easy to imagine a place where nobody gets arrested for saying things, just for doing things.
Especially in the case of a massive employer, the laissez-faire attitude is to offer no enticements whatever, partially because the more you do so the more the mine has captured the town and if the mine ever closes the town’s industry will die. Rather you should make sure the market is open to new entrants which can compete against the mine. The best way to do this is lowering trade barriers. By default, the mine has no power to do anything ‘evil’ to the town except leave, which it can do regardless of your attempt to bribe it to stay, and history says that’s exactly what will happen.
In your example, I do not see any markets.
An entity which employs 50% of the working population directly has just about cornered the employment market: those who do not work for them likely work in service industries which depend on the miners -- stores selling stuff to them, schools full of miner kids, doctors treating their families, policemen stopping their brawls and so forth.
Likely, it would be common knowledge that without the mine, the natural size of the town would be three farms and a graveyard.
In this environment, the natural state of affairs is that the owners of the mine end up running the town, e.g. through bankrolling officials. Likely the mine owners will get away with pretty much anything short of hunting townfolks for sport. They will not get pulled over for speeding, there will be no town council hearings about elevated cadmium levels in the creek and so on. Any oversight will have to happen at a higher government level where 99% of the constituents are completely indifferent to the fate of the mine.
If the same happens on a national level, whatever polity is dependent on that single source of employment is in deep shit. Private ownership of the means of production is only good in so far as it leads to a working market, a private monopoly is not better than a state-run monopoly, so they could just opt to nationalize their mining industry. A better approach would probably be to split the mining company into different mining companies though, but even this will leave them beholden to overall industry interests. If you want to regulate mine pollution, don't get yourself into a situation where half of your population work in mines.
>"…a private monopoly is not better than a state-run monopoly…"
Private monopolies (when not due to explicit state-granted privileges) still face (an attenuated degree of) market discipline; state-run monopolies can be nigh-indefinitely sustained despite persistent value-destroying behavior, due to taxing power.
Commanding heights of economy will be inevitably coordinated with the state.
Suppose you think labor unions are bad because they slow progress. For example, they ban automation in order to protect jobs, or they ban more efficient ways of doing things.
How does this work on a theoretical level? It seems like you should have three parties who are all fighting to collect the surplus from a transaction: company management/shareholders, workers, and consumers. Even if we grant that "public interest" = the interest of the consumers, that seems to leave management/shareholders and workers on an equal footing as potential rent-seekers. What's the theoretical argument for why management/shareholder interests are aligned with consumers' (in increasing efficiency / bringing down prices) but workers are against?
That’s a very big assumption up front; the public are mostly workers and workers are mostly citizens. That’s the Achilles heel of neoliberalism; at some point the citizens get mad.
Management and shareholders have highly different interests from each other, as well.
Lots of other good responses already, but I want to point out that I'd model the union as a fourth party in competition with the workers, shareholders and consumers.
Owners maximiz present value of all future profits. Workers are more short term cash flow interested. Hence the common dynamic of opposing automation.
Secondly owners are more agnostic of the form of that future profit. More low margin goods fewer high margin goods are both candidates. Workers need the costs(funded from profits) to come in the form of employee costs.
Short term vs long term
Maximizing one form of cost, rather than all sources of enterpise value
Management is in constant negotiation with the consumers, via prices and advertising. Management is in constant negotiation with labor, possibly through a labor union but definitely by offering wages and benefits. But there's no channel for consumers and labor to negotiate directly with one another. So if the process is viewed as "management and labor are both seeking rent from the consumers", then management is the party that will actually be collecting that rent and they're going to start off in a privileged position when it comes to deciding how the rent will be divvied up.
A good union can figure out what management's game is, what the underlying numbers are or ought to be, and negotiate a better deal for its members. But, as you note, they can also block improvments in the business - and if you are looking at this from the POV of a consumer wondering "why can't I have the results of the more efficient process", then it's a problem that you can't negotiate with the union. Management got where they are the old way, so they're probably OK with the old way continuing so long as they aren't being outcompeted by someone with a more efficient model, and it's the same union blocking the same improvements across the whole industry if it comes to that.
Unions can block *changes* to the business, some of which *might* be improvements. Many changes made to businesses are deleterious for the consumer, labor, or, in the medium and long term, management.
If people *knew* what changes were going to improve businesses, the world would be a different place.
Two things are off on this framing. The nature of the struggle, and the parties of the struggle.
First, the nature of the struggle is not a fight, it is a competition. Specifically it is a competition to cooperate, where the terms of the cooperation are determined by the competition.
Second, the competition is not between parties, it is within them.
Consumers compete with other consumers to buy the best value products from producers (a form of cooperation) according to the rules of the marketplace
Producers compete with other producers to sell products (at a profit)to consumers (cooperation again) according to the rules of the marketplace.
Prospective employees compete with other prospective employees for jobs with employer/producers (cooperation again) according to the rules of markets.
A cartel would be where producers collude to no longer compete, thus violating the rules of the market, and allowing producers to either take advantage of consumers or employees (if the collusion is on wages or working conditions)
I am not sure what the term is for consumers to collude, as they are too numerous and disconnected to do so (a consumer cartel succumbs to Mancur Olsons logic of the difficulty of collective action.) I can imagine it can occur in auctions though.
A labor cartel is known as a union, or closed shop union. This is where existing employees collude to exclude or hinder prospective employees from competing with them.
A rent seeker is someone who extracts more from the transactions than would be expected in an open competitive market. In this case it would be cartels of producers, or closed shop unions of employees. Again, I don’t know what the term is for colluding consumers, if there even is one.
So to answer your question, neither form of cartel (producer or employee) is aligned with consumers needs. Both interfere with the dynamic nature of markets to incentivize lower prices, better quality, and more efficient production.
A company’s incentives don’t really leave much room for “rent seeking” in this negotiation. They want to pay the workers as little as possible and extract the maximum profit for themselves. The amount of profit they make is determined by 1) the cost of inputs, including labor, and 2) their profit margin. There is never a negotiation between a union and management in which companies finagle higher prices and higher profits as a result of the bargain struck. Unions negotiating higher wages in various forms for themselves is the reverse: basically always higher prices, at the expense of the company’s profit margin and the public’s wallet.
The right comparison is of a cartel, in which all sellers of a certain widget collude to keep prices artificially high and output restrained.
I think Adam Smith noted the terrain here in The Wealth of Nations. At least, I recall some discussion of why the employer typically has the upper hand over the employee; I don't recall any discussion of collusion with the consumer. It's been years since I read it, and I get the sense that in Smith's time, the collusion might well have been consumer+laborer vs employer, due to UK only just beginning the Industrial Revolution, and seeing the concept of division of labor as a powerful artifact suddenly falling in the lap of the employer.
We can assume for starters that in a market for widgets, the consumer's enjoyment of that widget, and the laborer's expenses in producing the work to make that widget, are fixed. The former is much higher (or else no one would make that widget), and consumer, employer, and laborer compete for their share of that difference.
AIUI, employers had a natural advantage in information. They knew better than any laborer how to squeeze down the price of that labor, and knew more sources of labor, than any laborer; so if an individual laborer couldn't deliver on that price, the employer knew where to get another laborer who could. This is turn created an incentive for laborers to conceal their tricks for productivity from each other (just as suppliers of any commodity hide their trade secrets), but there was only so much to know about manual labor.
So the next most ready solution was to coordinate a representative for laborers, that knew how much to negotiate for labor, how to standardize productivity, etc. In other words, unions. The catch is that for reasons I'm not completely familiar with, unions are basically monopolies - any employer in that business has only one union they can go to to purchase labor, and in some cases, they're not allowed to seek free agents. (It's even morally condemned in several of those cases.)
Meanwhile, the analog to unions among employers - cartels - are likewise morally condemned (arguably because it would just reinstate the natural information advantage). The analog among consumers is the state, manifested as product safety regulations, consumer protection agencies, etc. - and is not condemned. Probably because such institutions view customers as otherwise disunited individuals, just as laborers.
At this point, it's important to consider consumers who aren't disunited - some consumers are themselves huge businesses, and have little trouble negotiating for their share of the surplus. Think supermarkets buying produce from local farms; manufacturers buying inputs from raw material suppliers; retailers buying from manufacturers; governments buying from defense contractors. Some labor is rarified enough to be in a similar position (think doctors, lawyers, software engineers, middle managers, or any specialist creating anything that doesn't have to be a public good in order to be valuable).
Notice how rent-seeking doesn't really enter the picture? To me, who teams up with who is largely a function of who has more of the relevant information. Whoever has the most, has the natural advantage, and everyone teams up against the king of the hill. In the Industrial Revolution, the employer was king, so consumers and labor teamed up on employers. If the state, in its capriciousness, favors one of those two, the team-up changes.
After the World Wide Web Revolution, things got more chaotic. In the 1980s, Reaganomics tapped suppliers as the winners, so consumers and labor were the underdogs, until New Deal economics reasserted itself and labor got the upper hand - for a while, and in some sectors. Nowadays, tech is the king of the hill, so the coalition is small business consumers with non-tech labor and a few management concerns. Plus, there's a whole market industry that sort of loosens the assumption about enjoyment per widget being fixed, and now an AI revolution starting to weaken the other end.
Both the workers and the owners have aligned interests against customers to create a monopoly to increase total profits. So they will agree on something like you can only hire union labor that reduces the ability of new companies to enter the market. The lack of efficiency isn’t a problem for the owners because they still make more profits due to the monopoly power, even if they have to share some of that with the workers. To have an effective union, you really need your whole industry to participate, and a major task of unions is to try to prevent non-union competition from emerging. This is also why companies in competitive markets are so opposed to unions—if they have to deal with unions but their competitors don’t, they’re in big trouble, because they won’t be able to build the monopoly power that makes it possible to succeed even with inefficiencies.
> What's the theoretical argument for why management/shareholder interests are aligned with consumers' (in increasing efficiency / bringing down prices) but workers are against?
I don't think workers are particularly against. The typical worker is willing to take a coinflip as to whether they are paid 50% more in a more efficient industry, or send a few months unemployed seeking another job.
But union leadership has different motivations. They want to maintain a large dues-paying union membership.
Two dynamics. Firstly, shareholders desire profits which increase the value of their ownership and workers desire salaries which they take outside the company. Secondly, shareholders' interests are aligned with the company as a whole while workers' interests are specific to whatever job they do.
This means that disruption that increases the value of the company is in shareholders interests, who benefit from the rise in company value, but can be against the workers' interests even if it might increase salaries and positions overall since it might disrupt the specific current job workers do. Thus shareholders have an interest in automating and reorganizing to achieve maximum efficiency because it maximizes their pay out. Unions have the opposite incentive, to ideally increase payroll as much as possible and importantly to keep existing jobs as they are so there's minimal disruption to members.
You're right that owners can also work against the interest of consumers. The most famous example of this are guilds which were not groups of workers negotiating with capital but business owners who often coordinated to keep employee's wages down and prices high. But it's also how monopoly works. However, the corporate coordination is banned in the US. They are not banned in Europe which is part of how their system works and part of why there's less labor-capital conflict. But it's also part of why they have a larger grey economy and less automation. So in the US it's mostly labor unions that do it.
Basically, the New Deal decided it was going to ban coordination and collaboration between businesses and then give unions a special set of privileges and allow them to coordinate. This was supposed to permanently tilt things in favor of workers. And it did in those sectors but it also stagnated the unionized sectors over time.
This is why communism was supposed to work, by the way. Worker's cooperatives were supposed to be more efficient and have no conflict, despite having no unions, because the workers and owners were one and the same. However, what actually tended to happen even in the most successful cooperatives is that without someone to push for firm level efficiency things stagnated. The way communism solved this was the state pushing for economy wide efficiency. Which resulted in the command economy, the calculation problem, and is now widely considered a bad idea.
So to answer your question directly: Free-markets are pareto-optimizers for the dyad of producer/supplier of the original factory components but only indirectly affect worker interests. Separately in the dyad of company/employees a free-market maximizes their benefits. Unions act as monopolies and benefit those that are covered, but have adverse impact to the other parties.
Free market model:
The original product sale is a two party transaction (company/individual) in a competitive market with multiple possible suppliers/consumers. Markets clear when marginal price = marginal demand, and utility amongst those two classes of parties is maximized.
Within the company, there are a large number of factors of production: labor of different types, different parts, services, rent, etc. Each of those presents a market. Those unique markets each maximize utility between the associated classes of parties. In this way, labor is optimized.
If the original product sale is a monopoly, the product-seller can gain excess returns, but does so at the expense of its purchasers (and to an extent in $ that exceeds the loss of the $, meaning it's net harm). A minor side effect: volume of products necessarily decreases, so the suppliers-of-factors-of-production are also shorted volume (less widgets, labor, etc.).
If the labor-sale interior to the company is covered by some sort of collective bargaining relationship (labor monopoly), the product-seller (people selling their time to the company), can gain excess returns, but will do so at the expense of its purchasers (the company, in the same net-harm context). A minor side effect: volume of labor necessarily decreases, so fewer employees should be employed. A new minor side effect: volume of products also decreases, so product-buyers and other-factor-of-production suppliers are also shorted volume.
Very few things are only good or only bad. Labor unions are a mix, just like almost everything else. They aren't even ALWAYS good for their members. But they are a way of evening the balance of power between the workers and the employers that "sort of works". If you want to replace them without seriously injuring lots of people, you need to find a better tool for that purpose. (But be aware that the government is generally on the side of the employers.)
Alcoa pitted both the managers and the union against each other, in a crusade to decrease workplace injuries. Worked like a charm. This suggests that both managers and the union have obvious deficiencies in "preventing injuries."
Improvements in technology that might make the firm more profitable and competitive, might reduce the demand for labor within the firm. While the firm is motivated for profit, it's also motivated to cut unnecessary costs, which may include a large number of longshoremen preventing automation in ports so as to protect their own jobs.
I think the argument is just that monopolies in any necessary input of production, whether capital or production stifle innovation. Whether that's Standard Oil buying up all the competition, or a union that covers all the available ports, you're going to end up with significant deadweight loss. Compare the longshoreman's union to the auto unions. Tesla can outcompete ford, partially because it's significantly more flexible in how it can allocated labor.
I believe it's a principal agent problem, although I'm open to correction. From Adam Smith (1)
"
The interest of the second order, that of those who live by wages, is as
strictly connected with the interest of the society as that of the first.
The wages of the labourer, it has already been shewn, are never so high as
when the demand for labour is continually rising, or when the quantity
employed is every year increasing considerably. When this real wealth of
the society becomes stationary, his wages are soon reduced to what is
barely enough to enable him to bring up a family, or to continue the race
of labourers. When the society declines, they fall even below this. The
order of proprietors may perhaps gain more by the prosperity of the
society than that of labourers; but there is no order that suffers so
cruelly from its decline. But though the interest of the labourer is
strictly connected with that of the society, he is incapable either of
comprehending that interest, or of understanding its connexion with his
own. His condition leaves him no time to receive the necessary
information, and his education and habits are commonly such as to render
him unfit to judge, even though he was fully informed. In the public
deliberations, therefore, his voice is little heard, and less regarded;
except upon particular occasions, when his clamour is animated, set on,
and supported by his employers, not for his, but their own particular
purposes.
"
The fact that you can collectively appoint someone to represent worker interests is a potential solution but one with obvious principal agent problems. Not just corruption amongst labor union officials but...most workers want stability, a lot. Automation resulting in market disruption that increases wages is a difficult sell to people who just hear that they're losing their job and union officials have no incentive to explain that.
(1) https://www.gutenberg.org/cache/epub/3300/pg3300.txt (Conclusion of Book 1)
Imagine a technology or process of sufficient difficulty to master that a worker must invest years of training in order to perform it at a satisfactory level. In such cases the protections and guarantees of a union might be needed to incentivize workers to enter that track and assure them they will not be terminated before they have learned the process and been able to earn the higher wages it commands for some number of years. If I am looking at investing years of my life into an industry, some degree of insurance from the vagaries of the market or management whim might be the factor that tips my decision one way or another. Without this incentive, the shareholders and consumers cannot reach the labor tipping point to manufacture the widget at all (an ‘everyone loses’ scenario). The average ACX reader who is an auto-didact may underrate the difficulty of learning a new trade for the average worker and thus how valuable union protections can be.
What prevents contracts with the same effect from being agreed upon in the absence of unions?
“The average ACX reader who is an auto-didact may underrate the difficulty of learning a new trade for the average worker and thus how valuable union protections can be.”
Good point. I feel like this often in the Slow Boring comments where unusually high agency high efficacy commenters seem to assume everyone has the same potential if they would only make better choices.
I have gatecrashed weddings in India as a child over 30 years ago. It seems possible even now. Just show up, greet the couple, and eat.
My husband has told me he would be walking home from school or sports past a hall in Bombay and just walk in for ice cream (they always had a fancy wedding going on). He says it was pretty much expected some random kids would come to eat.
The hosts gladly accept your wishes and give you a dessert, even if they know you're gatecrashing (which is hard to know if it's crowded). Sometimes a whole meal although kids typically gatecrash for dessert.
Such a different and informal wedding culture :). Partly because each guest costs a lot less than in America.
I thought about this because people we know getting married in America take the wedding event so seriously. They plan for it for a year. They limit how many guests parents can invite, which I find utterly shocking. They say no to kids under 12. Ultimately what limits the number of guests, I figured, in America, is the fire code.
It is changing in India too and the wedding, especially among the wealthy, is now very similar to weddings in America. But the speeches haven't made it into any Indian weddings in India, as far as I can tell. Any delays are often due to the proper alignment of stars in the sky (literally!).
> Ultimately what limits the number of guests, I figured, in America, is the fire code.
Ultimately, Indian weddings are big because there's a lot of prestige and status associated with having a big wedding.
In the West, that prestige doesn't really exist; we might have other equally irrational and expensive ways of signalling status, but not that particular one. So people invite the people they actually care about. My own wedding was fantastic, and adding an extra thousand people that I barely know would only have made it much worse.
The elders decide. Whatever they want - show of wealth with mega events in different venues (!) or a small wedding at a temple with a lunch after the ceremony. The bride and groom don't have much say, traditionally.
The invitation is traditionally written FROM the grandparents. They're the most important part of the host family.
And the traditional wedding invitation says "Come with your relatives and friends, to bless the new couple."
And while the elders decide, it's considered proper form for the grandparents to take into account what the preferences of the bride and groom are.
The bride's parents usually pay and the groom's side often make unreasonable demands. This is all typical.
"Ultimately what limits the number of guests, I figured, in America, is the fire code."
My understanding is that the WEDDING itself isn't where US folks create limits but the reception. The problem is that the reception costs can easily exceed $100/person (catering is expensive, I guess). So a 500 person wedding reception would come to $50K. This is actually a financial hardship for a lot of people :-)
One *can* have a fairly inexpensive reception: Rent space at a park, don't spend $100/person for the food, etc. But lots of American folks (brides, mostly, I think) want the expensive per-person reception so head count limits are required.
But the issue isn't the wedding :-)
Pointless anecdote time:
My friend's wedding was held at a relative's house. They hired a company to come set up a tent and chairs and speakers in the backyard. Dinner consisted of Papa Johns pizza and supermarket food (the cakes were a bit more upscale).
I don't know what the cost was, but I'm sure it was less than $100/person.
This was the groom's second wedding. I've heard that people tend to go a lot cheaper for second weddings. His first wedding was a much fancier one that even had a rented horse-drawn carriage (I believe his ex was the one who pushed for all that).
Why are people so into spending on their wedding? Is it a display of wealth in many cases?
This part is the same in India in recent generations. It's still not as extreme as in America in the average case.
I say (not that anyone should care!) have religious or other ceremony sanctifying the wedding. Why a lot of guests, why anything more than the ceremony (what you call "wedding"), why a party?
But I only heard about birthday parties and the concept of birthday presents from English story books when I was growing up in India. For birthdays, mom made my favorite food, we went to the temple and some years I distributed sweets in class.
As India gets wealthier and more Westernized this birthday celebration is changing. Even the poorest people celebrate their birthdays with a cake ( nice part of it - seeing themselves as individuals).
> Why are people so into spending on their wedding? Is it a display of wealth in many cases?
I think the root of it is a desire to celebrate one of the major milestones in one's life. Getting married is a big deal. It makes sense to me that people want to mark the occasion.
As to why it is so expensive, that's mostly poor or middle-class people wanting to celebrate in the grand style that they see the rich and famous doing. This leads them to blow the budget trying to afford (just once) the things that come much more easily to the actually wealthy.
I knew a guy (in America) who used to gatecrash weddings for free food (he didn't have a big budget for college). He had to get gussied up, but it was easy to get in.
One more point. In India it's considered really rude and even bad luck to turn away a guest who wants to attend your wedding. So Indian Americans struggle with this. Their kids run it all (not the grandparents or parents, like in India) and they (the parents) get a small allowance for number of guests.
(There's no edit button on my comment using the substack app on my mobile device).
In the shtetls, there used to be iterant wedding crashers. It was very bad luck to turn them away.
>Ultimately what limits the number of guests, I figured, in America, is the fire code.
As an European, I am really blown away how extremely regulated America is. I mean surely restaurants have a fire code here too, but I cannot imagine the average half-drunk quarter-educated ex-waiter restaurant owner has any idea how many people that means or really cares about it at all. What is the chance that someone from the state will pop in and count the wedding guests?
If you're renting the venue, then probably the owner or the owner's insurance company will be checking to see if the fire code is being blatantly violated because if anything goes wrong, "...and they were in violation of the fire code" makes them deep-pocketed lawsuit bait.
I haven't specifically seen this with weddings, but I haven't been involved in much wedding planning. It definitely happens with fan-run cons renting hotel space, and I assume the process is pretty much the same anywhere in the "rent event space to enthusiastic amateurs" business.
I think almost any indoor venue in the US is going to have some kind of occupancy limit imposed by local fire code. But also, churches and halls and reception venues have a limited amount of space for people and tables, caterers have some limit on the amount of food they can prepare/serve in a reasonable time, and also whomever is paying for the wedding (usually either the couple or their parents, depending on age and situation) will have to pay some amount of money per person who shows up to the reception.
This sounds correct to me.
As another European I don’t find this at all shocking. The risk is not that someone would come count the guests, the risk is that something would happen and then it would turn out the fire code was violated and the business would be sued or closed or something.
I also don’t think what limits the number of guests is the fire code. I think people just want to control who is at their wedding.
There can be a limit, in the sense that the venue you're renting can at some point turn away further people. If you set up a reception for 100 people and 200 show up, at the very least you're likely to run out of food and drink and probably the last 60+ guests won't be able to find a place to sit down.
But there is a fire code. I guess it could be used as an excuse too.
I guess the fire code is one thing that increases the price of renting a locale.
I don’t know much about India, but from some articles I have read I got the impressions weddings are a bigger deal in India.
But probably Indian culture is more communal than in Europe or America? I guess in the past a village wedding in Europe was something everyone could partake in but nowadays people live increasingly in social bubbles that are detached from local geography.
People stress about wedding expenses from the minute a daughter is born in India. So, yes. But the other things I mentioned are simultaneously true as well.
What kills me is the reason for the large gaps under the doors in public toilets is also for fire code/emergency reasons. "A person" has to be able to crawl under the door. I'm not that fat and could probably still not fit under it (yes I've been to the US). With the US obesity rate...
Surely, if the gap is there for safety reasons, it is to allow access for a first responder in case the person inside needs assistance.
The large gap under the door is to make it easier to mop the floor. Plenty of multi-stall restrooms in the US have full-length doors. They are just harder to clean and more time consuming to install (full height doors have tighter tolerances), so most restrooms do not use them.
Boggles the mind. Why, I wonder, is that door expected to get stuck, but not other doors? Why not mandate that the latch is designed properly?
Wrote a longform essay inspired by Venkatesh Rao’s concept of life intensification, combined with thoughts from Henrik Karlsson, Virginia Woolf, and a few of my own spirals at 18. It’s about how we don’t “find ourselves” so much as we generate ourselves — through feedback loops, relationships, and the people who draw different tones out of us. I touch on Rao’s ghost-to-character metaphor, the idea of “container people” who midwife our full expression, and how selfhood might be less about discovering a hidden essence and more about becoming legible in motion.
Would love thoughts, critiques, or other essays that explore similar territory : https://velvetnoise.substack.com/p/how-to-become-real
So did I: https://sisyphusofmyth.substack.com/p/in-the-garden-of-eden-baby?r=5m1xrv. I'd love your views.
In psychology, it turns out there is a distinction between high self-monitor and low self-monitor people. High self-monitors change or mold their personalities according to who they're with, but low self-monitors show the same personality to everyone. It's apparently not a human universal, that experience of different people drawing different aspects of you.
And of course, neither is wrong, but I think both come with pros and cons.
My intuition is that this is probably more of a continuum than a 1/0. I and one of my kids are over on the low self-monitor side (not so affected by social norms and often not even noticing them), my wife and my other two kids are more on the high self-monitor side, but with some variation between them.