A man in the UK has been sentenced to two years in prison for maintaining an online library of far-right political stickers.
I wonder if clowns like Freddie deBoer who were hysterically crying about people facing negative consequences for supporting palestine are even opposed to this. Or Scott, for that matter, who was quick to cry about some incompetent rich Haitian diversity hire at harvard get fired but has not to my knowledge ever talked about europe imprisoning people for political speech.
Anybody still following up on Ivermectin as a COVID-19 treatment, there's a new study using Bayesian statistics. It's a randomized controlled trial with N = 8811, offering this interpretation of the results: "Ivermectin for COVID-19 is unlikely to provide clinically meaningful improvement in recovery, hospital admissions, or longer-term outcomes. Further trials of ivermectin for SARS-Cov-2 infection in vaccinated community populations appear unwarranted."
So I'm in the middle of listen to Tucker Carlson interview Putin. And... well I can only describe this as 'liking' Putin's narrative of history, compared to the hawkish government here in our country. Wouldn't it have been good for Russia to join the west as a full economic partner? It would have been good for everyone. No? Amazing number of times (Putin says) our presidents made a tentative deal, only to be talked out of it by their advisors. Isn't peace with Russia better than war?
Just writing here to let anyone in the Minneapolis-St Paul area know that I've made a local ACX community discord for the region to plan local events and chat generally. The link is here: https://discord.gg/hySwpphmdN
In one of the previous Open Threads I complained about how both Google Translate and DeepL screw up translations between Slavic languages, basically by erasing every information that does not make sense in English -- such as grammatical genders, flexible word order, or even distinguishing between words that happen to be homonyms in English -- so much that the result looks as if someone translated first to English and then from English to the other language (this is most likely not a technical description of what actually happens, but the results look the same for all practical purposes).
It just occurred to me to try Yandex Translate https://translate.yandex.com/ and yes, it seems better in this aspect... but worse in some other aspects... so I suppose I will actually use both Yandex and Google/DeepL translations side by side.
Posting here, because it took me surprisingly long time to think about a solution that is obvious in hindsight, so perhaps someone else may benefit from this idea, too.
Compared to other religious beliefs, why is it so hard to convince people that pantheism is true? Spinoza makes a great case that God, whatever God is, must include everything, because God is perfect.
It's a simple argument, if you want to believe in God.
I understand that CS Lewis argues that things must complicated and therefore we should reject simple explanations. I believe the opposite.
> Hitler couldn't exist without WW1, which was the product of Bismarck's wars, which were a German answer to Napoleon, who took over because of the revolution, which was inspired by the American revolution, which was ideologically rooted in Dutch and English anti-catholic activity, which was caused by Catholic absolutism emanating from the wealth of Spain, which came from the new world. So really, the discovery of America is to blame.
Neat little comment isn't it? Discuss its validity.
Posted v3 of the Eve Theory of Consciousness: https://www.vectorsofmind.com/p/eve-theory-of-consciousness-v3 Briefly, it argues that recursive self-awareness is what separates humans from the rest of the animal kingdom, is what allowed us to conquer the world, and has evolved fairly recently. There is the first strong evidence about 50kya, but even then the experience of self could have been fractured. "Recursive" culture spread, which changed the selective gradient for those that experienced a seamless construction of self from a young age.
The main thrust of the argument is about evolution, but the evidence I put forward for recursive culture spreading is interesting even if it didn't figure in our evolution. The bullroarer is a religious instrument used in male initiation ceremonies from Australia to the Americas, to Africa, to Europe. An implement of the primordial snake cult?
Scott linked earlier versions of this argument: the Snake Cult of Consciousness, and the Unreasonable Effectiveness of Pronouns. The Snake Cult piece looked at the Sapient Paradox, which asks why sapient behavior is not widespread until about 12,000 years ago. The pronoun piece looked for linguistic evidence of Julian Jayne's theory about self-awareness being recent.
Kids utilizing AI to create nonconsensual deepfake sexual photos of their peers was a perfectly predictable result of generative AI. This current story is probably just the tip of the iceberg. Outraged parents will be suing school districts (at the taxpayers' expense) and demanding that school staff and administrators use (waste) their time and limited resources to address the problem. Parents will be suing other parents (possibly financially ruining other parents who likely never head of generative AI or know how to search for the app on Little Johny's smartphone). The resources of police departments and prosecutors will be wasted investigating and prosecuting juveniles with unpredictable results (since the laws don't really cover this twist in the evolving world of kid-generated kiddie porn). The most sensible solution to me would be that parent and teacher groups launch a class-action suit against generative AI companies to stop allowing nude bodies to be merged with non-adult faces.
Ultimately this will be a much bigger deal than Google Gemini generating racial diverse photos of Wehrmacht soldiers. But kids will be kids (and I mean that in a creatively generative negative way)...
Is Google no longer gray tribe? On my twitter feed right now:
"In Google's 2023 Annual Report, the terms "unbiased", "objective" and "accurate" did not appear even once. Nor did the "Don't Be Evil" motto- it has largely been retired."
This is in addition to the leading designer for Gemini having some *very* Blue Tribe opinions on Twitter.
I'm not anywhere near SF or Silicon Valley and was interested if anyone could tell me how the culture has changed over time.
I'm a biomedical science student who will be interning with some scientists in Toronto this summer. The current thinking is that I'll use publically available data to look for some sort of link between air pollution or maybe weather anomalies and heart disease. With that in mind, what are some databases I should investigate, either in Canada, the US, or elsewhere?
Domestic terrorists and mass shooters in the US often show similar psychological characteristics and behavior in terms of their plans. Seeing as this happens so often in the US, what's the base rate that potential perpetrators get caught? If the base rate is low, why is there a blind spot like this in our surveillance?
If it's framed as a search problem, maybe ten thousand out of three hundred million people in the US are serious suspects:
- Nearly all have anti-social, autistic characteristics or something in the same category of diagnosis (~1 percent of the population) – 3 million people
- Nearly all are male (~50 percent of the population) – 1.5 million people
- Nearly all reveal intention or potentially violent instability ahead of time (as a personal estimate, ~1 percent of the population) – 15 thousand people
Those numbers are obviously rough, but I think the ~10 thousand figure is accurate for the US. What are defense agencies not be able to do that lets this continue to happen?
Looking for a good book about the energy industry. Something that covers most everything (carbon based, nuclear, solar, wind, etc). Some technical depth would be good but obviously it will be limited if it covers everything. Any recommendations?
I see the UAP story pop up from time to time ... and remain a complete disbeliever.
Reports come from only military aircraft, typically from complex computer controlled aircraft equipment. Reports never come from commercial or private aviation despite these flights outnumbering military flights by some staggering amount.
Consider, that we're not using Isaac Newton's Radar anymore. Every image presented is a collection of data. Objects presented on a screen are assembled from data in a table or database, i.e. direction, strength, etc. Most of the equipment isn't on it's first generation, but has undergone several hardware and software upgrades. That means I'm writing code to interface someone else's system, on proprietary interfaces, which means we're writing the handshaking code too. If my software doesn't properly address memory, or someone else's software walks on my storage locations, we get garbage ... garbage which may just look like another 'flying' object. Did you think there are 10,000 pairs of eyeballs combing through this stuff?
Commercial aircraft just never seem to report UAPs, despite having hundreds of bored passengers, carrying cameras, taking pictures of clouds and what-not ... yet never seem to report a UAP. Private aviation (pop in his Cessna) and carrying a phone with a camera never seems to see a UAP either.
Are there any examples of moves from famous human chess games that were previously considered brilliant but are now known to be unsound due to computer analysis?
apropos of nothing, here is a wonderful example of (unless I miss my guess) an unsung hero of words, a laborer in the trenches deep in the dictionary yet spinning their own beauty into the world of words, just in case it happens to catch and inform some hapless writer looking for advice on capitalization like I did. Fortunately I can share a little of my joy. Worth a click:
I know the verses in The Bible that say "thou shalt not kill" are more correctly translated as "thou shalt not murder". Is the same generally true for the commandments against taking life in Buddhist and Hindu scriptures, or are those actually meant to be about killing in general?
There is a lot of boredom these days, which Hoffer mentions as well: "There is perhaps no more reliable indicator of a society’s ripeness for a mass movement than the prevalence of unrelieved boredom." Are we in the midst of a mass movement, or is one on the horizon?
My partner and I are in our 30's, both ACX readers who are married due in part to this blog. We were planning to have children this year but recently found out that we both technically qualify for an Asperger's/high-functioning ASD diagnosis. (We weren't looking to get diagnosed; one of us was seeking ADHD treatment and got diagnosed that way, and the other of us found out about a childhood Asperger's diagnosis that had been kept a secret.) We are both high-functioning and independent, if a little neurotic, but we are worried that combining our genes could result in a greater chance of us having a child who is autistic to the point of being extremely disabled. Can any doctors/scientists here help us think through the following:
1) What is the increase in our risk of having a severely disabled, low-functioning child by chance, given we both have mild ASD?
2) Is it worth doing IVF to select for genes less likely to cause developmental disorders up front?
3) Can anyone recommend genetic counselors in the Bay Area who are smart and will take us seriously? So far, it seems like most medical providers do not know how to deal with the concerns of autistic adults, especially prospective parents. It feels like doctors are either brushing us off or patronizing us.
In Eastern Canada, there are groups of individuals falsely claiming to be part-Indian, known as the Eastern Metis. They've organized "tribes" with "chiefs" and "membership cards;" the whole deal. They've built a whole conspiratorial worldview, in which their ancestors were listed as white in censuses because they were trying to avoid the Residential School Genocide. Many are obvious grifters, looking to benefit from affirmative action. I wonder what it's like for the second generation. I bet many were told by their parents "this is just a fun cultural thing, we're basically just white people, don't take it too seriously." But I bet many others grew up believing it wholeheartedly. I'm indigenous, part of a marginalized minority; settler colonialism is the reason my family has little money, etc. Then one day they're browsing the net and they learn that, not only are they not the identity they thought they were, their whole community is fraudulent. They aren't indigenous, their tribe is not a real tribe, their chief is not a real chief, their membership card is just a worthless piece of plastic. Real Truman Show s***.
Imagine that two people are lost at sea, with only enough food for one. One is the world’s most promising cancer researcher while the other, to be polite, has far less potential to improve humankind. Most people would wisely choose the cancer researcher.
Yet the fashionable Effective Altruism movement focuses on lower-potential people—typically, people who are struggling in poor countries. That implies the belief that all lives have roughly equal value. Of course, our hearts go out to “the least among us,” but they face so many barriers. If we care to maximally benefit humankind, we’re wiser to invest in people with great potential for ripple effect who, importantly, would not otherwise get fully funded.
For example, leading lights in the Effective Altruism movement are far from downtrodden, for example, William McCaskill, Holden Karnofsky, Peter Singer, and Zvi Moshowitz. Effective altruism might fund such people to develop ever more “ripply” altruism.
Some other possibilities for more ripply and thus more effective altruism:
— SuperCourses: online versions of standard school courses taught by dream teams of transformational instructors, augmented by vivid demonstrations and gamification. Of course, instruction would be individualized, not just in pace but in teaching approach. Machine learning would make that individualization ever better, and automatic translation would make SuperCourses available in many languages. The development of SuperCourses would enable every student, rich and poor, kindergarten through college, Alabama to Zululand, to get a world-class education. The private and government sectors haven’t funded this—I have proposed SuperCourses to top U.S. and California education officials and gotten nowhere. One reason is the fear that the teacher’s union would use its might to try to stop it to preserve teacher jobs. But if developed and disseminated worldwide, SuperCourses could be very effective altruism indeed.
— Independent researchers studying solutions that are promising but have a poor chance of success. Governmental and corporate funding sources tend to invest in institution-based researchers whose projects have higher probability of near-term success. But if the focus is on long-term risk-reward, effective altruism would include independent, unaffiliated researchers working in their home-office or garage who are exploring novel ways to, for example, lower the cost of nuclear fusion energy, develop better AI-driven models for predicting and foiling terrorism and even for assessing a war’s worthiness, e.g., the U.S. entering World War II versus the war in Vietnam or Afghanistan.
— People developing ever better mental health apps, for example, using ChatGPT. Such apps could be distributed worldwide to countries rich and poor—Cell phones are ubiquitous even in poor nations. Private sources are funding development of such apps, but such development deserves greater funding given the apps’ potentially great ripple effect.
— Researchers studying the enhancement of reasoning ability, impulse control, and altruism. For largely political reasons, those research areas are underfunded by government and corporations but, with sufficient ethical guardrails, such research has great ripple potential to provide major benefit to humankind.
— People developing software that matches mentors with protégés, available worldwide. It would be like match.com but for mentor/protégé, relationships—Many protégés and mentors say that mentorship has been among their life’s greatest learning experiences. Such software would facilitate that. Alas, the matching industry, despite having been around for decades, has remained focused on romantic relationships. That makes mentor-matching apps a good candidate for effective altruism.
Again, it’s understandably tempting to want to help “the least among us,” those with the greatest deficits. After all, we feel good in helping them and it’s a fashionable form of virtue signaling. But if we truly care about humankind and are willing to focus on the greatest ultimate benefit, ripple should be the core criterion for determining what is maximally effective altruism.
Could use info about options for someone I know who is getting an advanced degree in a highly technical STEM field. What they study is abstruse and has no practical application, so the natural path for them is to become a university professor. However, they need a fallback, because there are not enough faculty jobs to go around. They are extremely gifted at math, and good at the kind of coding that's used in their field, and could probably be hired as quant or something similar where they use their math talents in the service of the finance industry. However, their values are such that working in business or finance is out of the question. I was wondering if Effective Altriusm, either the organization or the activities they support, might have a place for someone like this: an extremely introverted math prodigy with few outside interests who would throw themselves into a job where they believe they are helping the world.
GPT-6 requiring ~0.1-10% of all world compute to train isn't all that outlandish compared to other major industrial goods. This report [1] says that aluminum production requires ~1.5% of all electricity in the US. Aluminum is great, but it doesn't exactly dominate the public discourse. A one-off 10% of all compute for ~6-12 months might only register as a "fun fact" level of economic / environmental impact (e.g. "did you know that concrete production is responsible for 5% of all global CO2 emissions?").
Then again, energetic and compute costs to train bigger and bigger AI models might get into crypto or private jet territory, where the environmental impact is overemphasized because AI/AGI falls into the same tech bro / silicon valley elites sphere of negative emotional connotations.
A series of unfortunate exchanges down the thread has escalated to the point where I need to explain why Ethnic Cleansing is bad to justify a mildly snarky reply I made in response to someone who advocated for Ethnic Cleansing as a norm.
Because I'm a patient teacher at heart, and because I have always been an advocate for repeating and explaining and re-explaining the obvious - a norm that religious cultures understand all too well but more secular cultures like Rationalism have failed to fully digest - I will try to make a comprehensive case of why Ethnic Cleansing as a norm is bad, and why people who advocate for it as a norm are bad and should feel bad. There is no reasoning in this comment that people are likely to find unique or exceptionally insightful, you likely already know all what I'm going to say on some, possibly non-explicit/verbal, level. By the end of this comment, someone who isn't too sure that Ethnic Cleansing is bad should hopefully be convinced or moved in the direction of being convinced that Ethnic Cleansing is bad, provided they're engaging in good faith and is reading attentively.
--------------
1- Definitions
What do I mean by "Ethnic Cleansing as a Norm"?
Ethnic Cleansing is the (not necessarily systematic) forced displacement of some population sharing a common trait (e.g. Religion, Skin Color, Gender, Nationality or Ethnic Origin) and crossing a certain size threshold, where "forced" implies any coercive and destructive practices and actions that - intentionally or unintentionally - makes the current habitat of the population more and more unfit for living, where those practices and actions are primarily perpetrated and/or incited and/or funded and/or planned and/or advocated by people outside the victim population.
The "As a Norm '' bit means that you're only a bad person if you want Ethnic Cleansing to be a "norm", lit.: a normal option that people can openly promote in normal political and geopolitical conversation. You're NOT a bad person if - say - you merely will begrudgingly choose Ethnic Cleansing from a list where it's the most humane options (e.g. as an alternative to Genocide), just like you're not a bad person if you're slightly relieved that a woman was raped and not murdered (or the reverse) and you're not a bad person if you do choose someone to kill after somebody forces you to choose which one of your children to kill.
--------------
2- Verdict
Ethnic Cleansing is Bad.
--------------
3- Arguments
Ethnic Cleansing being bad, there are several arguments for why Ethnic Cleansing is bad:
3-(a) The Argument from Private Property
Most people believe in some form or the other in the notion of Private Property : The notion that people have the right to certain things, and that those things can't be taken away from them by force (even if this forced taking away of things involves payment).
Ethnic Cleansing violates the Right to Private Property: It necessarily involves forcibly taking away the homes, cars and all other non(-easily) -transportable property that people being Ethnically Cleansed own.
Note: Notice how mild (3-a) actually is, the belief in Private Property needed here is not that which justifies Billionaires or private space travel or luxury yachts and private jets, it's not that which a lot of Socialists and Communists and some Anarchists rebel against. The vast majority of people living today, including the vast majority of Communists, believe explicitly or implicitly in a version of Private Property that at least grants you the moral right to buy things you need for a fair price (with the full consent of the seller), and then forever retain those things until you sell them, voluntarily relinquish them, or throw them away entirely for any reason.
3-(b) The Argument from the Immorality of Non-Consensual Actions Done to Humans
Most people believe in some form or the other the following notion: that humans, or more generally conscious minds at or near human-level sentience and/or intelligence, have a sacred moral right to decide their own fate.
Ethnic Cleansing is, by definition, a name given to actions that people do to other people, to force them to make decisions and accept fates that the victims didn't want to make or accept. Therefore, Ethnic Cleansing violates the human right to decide one's own fate.
3-(c) The Argument from the Sanctity of Intimate Property
Many people believe in what we could call - without loss of generality - Intimate Property: Property that has value beyond what their strict material valuation would imply, and which can only be perceived by the current owner. Among the list of potential Intimate Property: (A) The house you grew up in (B) The car you got married in, or the first car you own (C) The first piece of jewellery that your husband or wife brought to you (D) The belonging(s) of a dead relative.
It's important to recall a crucial property of Intimate Property: It can't be replaced by an equally-materially-valuable - or sometimes even materially-superior - property. The house you grew up in is fundamentally valuable because a thermodynamically irreversible event - you growing up from a child to a teenager - happened in it, and thus no other house can ever satisfy this, and thus you're very reluctant to sell it.
Ethnic Cleansing involves the destruction and/or the confiscation of - at least - the intimate property (A) of the victim population, and frequently many other kinds of Intimate Property, including all the ones that appear on the example list above and more. Thus, it violates the right of humans to own and appreciate and preserve Intimate Property.
3-(d) The Argument from the Perversity of Perverse Incentives
Most people believe or can be convinced to believe a certain model of human dynamics called "Perverse Incentives". According to this model, whatever actions you do - if they're known by other people - will shape the future actions of said people. Thus, for example, be careful giving charity to begging children, because when the people who force children to beg know that this strategy works and the children are too cute to be refused, they will have and/or kidnap even more children to force them to beg. Perverse Incentives is an anti-inductive model, it posits that the sum total of all human interactions respond to your action in ways nullifying the effects of your actions, often in unexpected ways.
Promoting Ethnic Cleansing as a norm will set a perverse incentive, namely that any state or powerful corporation which doesn't like a group of people can simply engage in less-than-murder ways of making their lies worse and worse and worse till people advocate for the Ethnic Cleansing of the victim population and the problem is "solved".
3-(e) The Argument from the Immorality of Violence and Destruction
Most people believe or can be convinced to believe that Violence and Destruction is bad, and furthermore that its typical outcome is the killing of humans and the destruction of property, which most people believe - in turn - to be also bad.
Since Ethnic Cleansing is typically not the first choice of the victim population, it typically requires an immense amount of violence and destruction to achieve. Furthermore, even after a critical percentage of the victim population is finally convinced, a substantial percentage always remain opposed, and thus need additional further waves of violence and Destruction to neutralize or eliminate.
3-(f) The Argument from etc
This list was outlined over the course of approximately 2 minutes, this section is for all the additional reasons that Ethnic Cleansing is bad, whether they come from me as I think about the issue more over the following days or from commenters who respond to this comment.
--------------
4- Objections
4-(a) I was just joking
Joking about immoral things is not necessarily immoral, unless the person telling the joke is actually convinced of the fundamental premise of the joke and is merely using a joke as a pedagogical/infotainment way of preaching the premise. For example, a joke about rape or America invading other countries for oil could be funny and harmless, unless the person doing the first is a rapist and the person doing the second is G.W. Bush.
Given an ideological affinity and/or a history of supportive comments to a party that advocated Ethnic Cleansing, making a joke about Ethnic Cleansing makes you a horrible person.
4- (b) Do you want the victim people to be Genocided instead ?
This argument is only against "Ethnic Cleansing as a NORM", Ethnic Cleansing as a desperate last resort while a genocide is taking place is not in its scope.
4- (c) What about all historical instances of people who were ethnically cleansed and turned out to be okay after centuries of healing ?
See Appendix, but this is not an argument pro-Ethnic-Cleansing-as-a-norm for the same reason that "What about all the women who were raped and then turned out to be ok after years of the incident" is not an argument.
--------------
5- Conclusion
Ethnic Cleansing is bad and people who argue for it should feel bad. Mockery of people unambiguously arguing for Ethnic Cleansing implicitly or explicitly is morally good and should be a norm in any good community.
The Tuo lightbulb claims to correct circadian rhythm issues without bright lights by using custom LEDs tuned to specific frequencies of orange and violet light that are present at sunrise and sunset. Their studies so far have produced what appear to be dramatic results (but with a small sample size of mostly grad students and postdocs in their own lab). Sounds too good to be true, but if it really works it could be a huge improvement in QoL for people with CRDs and approximately all teenagers.
Transwomen breastfeeding. Last week I realised this is actually a thing that is physically possible. Going on Twitter to learn more about was a mistake ofc, but the shitstorm was somewhat entertaining to watch (takeaway: everyone thinks everyone else is a pedo). Looking at the literature, it is almost non-existant (I found ~10 papers that were relevant, wrote an article here https://valentinsocial.substack.com/p/can-transwomen-breastfeed). My conclusion is, there are no obvious risks that we know of right know, but we definitely need more research if this is a thing that is going to happen more in the future.
I have been wondering why such a huge percentage of people in US TV commercials are African American? If I had to guess, I would say that they are three or four times over represented.
Possible answers:
1) I am wrong and over-estimating
2) Blacks either watch more TV relative to their population, or more commercial (non streaming) TV
3) Blacks are more receptive to advertising (ie a target market)
4) Advertisers feel good and diverse about themselves by featuring black actors, and the cumulative effect is to be wildly over-represented
5) Marketers have found that ads with black actors perform better relative to other actor distributions (they sell more or improve brand image)
6) Advertisers come from urban areas and are unaware of the actual demographics of the country?
7) Advertisers are shooting for diversity, and the cumulative effect is wildly unrepresentative?
Any others? Any feedback on which of these is plausible? Is it perhaps some combination of these? Anyone in Marketing familiar with research on the topic (I was in Marketing, but have been retired long since before this trend started, but our main spokesman 15 years ago, was indeed a black male).
Is "sign in with password" on Substack broken for anyone else? No matter what I try, it just says "something went wrong", even if I reset my password and then try to login with the new password. The only way to log in to Substack now is the "email you a link" option, which is really annoying.
If you've ever tried to overcome a fear and haven't succeeded, you may be interested in an article I wrote - I detail how I overcame my fear of talking to strangers and pitching businesses
Would anyone be willing to share their criticisms of "non-violent communication"? As someone who has found it a useful framework, I'm interested in hearing from people who didn't find it useful.
I want to modify the mower so I can attach garden shears, a hand saw, and a bottle of plant killer spray to it. Whenever I am mowing the lawn and come across, say, a small branch that has grown long enough to hit me in the face, I'd stop the mower, remove the hand saw, use it to cut down the branch, and then continue mowing.
I have no idea what sorts of couplings, baskets, or whatever I'd need to attach to the mower to make this possible. Any suggestions?
I'm looking for recommendations for an AI app that can enhance old home movies. My late father transferred a bunch of 8mm movies to VHS tape. That was about 30 years ago. I recently sent the tape to a service to get it digitized. The result is an hour-long MP4 file that is pretty crappy, partly from the degraded VHS tape but mostly because the underlying 8mm films were too light or too dark, blurry, or shaky. Can this be remedied through an AI app? There are a lot of individual clips on that long MP4 file, and I don't have time to chop it up into small segments to fine-tune them. I want to submit the whole thing to an app and have it fix it up as best as it can.
I really enjoy fiction that feels like a fable, in tone or plot. They have a simple beauty that feels comforting to read. I just finished "The Last Unicorn" by Peter Beagle, which I thought was extremely beautiful. Other examples I've enjoyed for similar reasons are "The Magician's Nephew" (CS Lewis) and "On Such a Full Sea" (Chang-Rae Lee). Does anyone have something to add to this list?
Is anyone else getting this problem? Trying to sign in with my password on Substack, keep getting message "something went wrong". I've reset the password, no dice. But "sign in with email" works.
I'll try here before I try contacting Substack because, to be frank, I don't expect Substack to be any help with this problem. Signing in with password worked fine up to a few days ago, the problem has only cropped up recently.
In regards to your policies on banning & unbanning:
1. Do you have a "statute of limitations" after which the comments and the commentators are off the hook? I ask this because the comment that I was banished for was five months old when the banishment was imposed. Although I readily admit I was out of line and that I deserved banishment for fulfilling Godwin's Law, it took me by surprise because I had long ago forgotten about the exchange. How far back do I need to go to review my old posts to see if anything I said was banishable?
2. Have you articulated your moderation policies anywhere? After being banned, I searched for them but couldn't find them. If so, could you repost a link? For instance, I assume ad hominem attacks are verboten. Looking at the posts that got people banned in the last big round, it seems like certain extreme levels of sarcasm can earn a poster a banishment. Likewise, circum-hominem attacks seem to be bannable. E.g. "You're delusional if you believe X." I'm somewhat on the spectrum, so I am quite capable of working in the framework of explicit rules, but when it comes to unwritten social rules, I'll admit I'm quite at sea.
Anyway, explicit policies might help reduce the outbreaks of verbal warfare on your Substack. For instance: Charlie Stross has a separate page on his blog listing his moderation policies...
3. The notice I received from Substack did not indicate that I was only banned for a month. Initially, I thought it was a lifetime ban, and I canceled my upcoming renewal. It was only days later when I saw in one of your posts that I was banned for a month. It seems like the banishment wording could have been more explicit in the email. I admit I'm reluctant to re-subscribe to your Substack without having my first two questions clarified because I feel like I'm skating on thin ice. I'm on a fixed income now, and I don't want to be banned again for something I wrote a year ago (and have my current $100 go down the tube).
4. Do you want us to report posts that we find rude and offensive? In the past, I've been called the political C-word and I've generally responded with humorous sarcasm. Personally, I don't like to be called a Commie (because I have friends and family who were blacklisted). My banishable moment was when I responded to such an attack with the political N-word. In the future, I'll report anyone who lobs the political C-word at me. But I don't even know if you would consider the political C-word out of bounds.
Hello! Those into figuring out best practices in how to eat healthy : what do you think about -
1. saturated fat? Necessary? About how much a day at different stages of life, etc? Doctors (even cardiologists) seem to say you should avoid it. Modern doctors like Dr. Jason Fung seems to say it is essential.
2. Salt? Cardiologists say about 2000 mg a day at the most (a teaspoon is 2400 mg). And, that less is even better. Dr. Fung says it is bad advice. He says there's a confusion about interpreting the data. He thinks they are confusing processed food (which he says is bad for you) that happens to also have high sodium, sometimes due to added salt with salt itself.
Question to those who are familiar with minoxidil(/+finasteride), I find the following puzzling. A typical 5% minoxidil has in its instructions, "Don't use more than 1ml twice a day, because it won't help". But the *same* company sells a costlier 10% minoxidil, to again be used at 1ml twice a day. How is this consistent: wouldn't whatever makes the extra volume useless also make the extra concentration useless?
I'm about to solve the Sleeping Beauty Paradox. Here I'm exploring issues in the popular attempts to model the problem in the philosophical literature.
In the caste system in India, as originally practiced (or so the traditional doctrines hold), sure the upper layers were quite privileged, but there was a catch: the higher you were in the hierarchy, the more restrictions were placed on you. To give an example, a Brahmin, someone at the very top, is supposed to take a bath after each defecation, and this is just one of many restrictions. Conversely, being at the bottom means you do the manual labor, but you don't have any restrictions like that and can indulge in all sort of vices. So it worked (or so I've been told), because the Shudras, the bottom caste, would look at the top and see that the people there really were of a superior nature to them.
Now, this is not about defending the caste system, but about revisiting this idea that the more you climb in society, the more restricted your life becomes. We should totally do this I think. Past some threshold of wealth (TBD) maybe it could be something like you have to put in 200 hours of volunteer work each year, maybe 100 of community service and 100 of working the land. I think this would give a dose of reality to the sheltered upper classes and remind them of how the society they live in actually works. Certainly, doing volunteer work with the homeless has had that effect on me, and this might be a way to harness populism to good ends, as it is a rather populistic measure.
Posting one more time, because I got zero responses in the last open thread, and I think it must exist by now.
Do any of the new image generators / alterers allow you to upload pictures of yourself or another person at different ages, and have it adjust different-age pictures to a requested age?
Like you have a picture of your grandma at age 20 and age 90, can it impute a picture of her at age 50?
Or a picture of yourself ten years ago vs today - can you age the younger picture to you today, keeping the background and setting?
Have there been any studies on the rightward political shift on Twitter/X in recent years? I'm curious (1) what's driving it, (2) how significant it is, (3) how it's changing wider public discourse.
The massive anti-woke backlash against Gemini in the past week got me thinking more about the question. I couldn't imagine this happening on the platform circa 2018.
In the literature on female attractiveness it's often claimed that men find women with neotenous features very attractive. Some psychologists may run a study showing men pictures of 18+ women and ask them to pick out the most attractive faces. The women that then get picked are ones with facial proportions typical pubescent girls about 12-14. This is then interpreted to mean that men find adult women with "neotenous faces" the most attractive.
Now, I can't be the only one who's spotted the glaring flaw in this reasoning. Men find women who look like pubescent girls most attractive. Well, who else looks like pubescent girls? Who else has the facial proportions of girls about 12-14? Well, actual pubescent girls look like pubescent girls. Actual girls about 12-14 have the facial proportions typical of girls about 12-14. It's not neoteny men find highly attractive, it's immaturity. What men find most attractive simply isn't adult women but young teen and pubescent girls. Some women happen to retain the facial proportions of a pubescent girl into adulthood and men continue to find them highly attractive because of it.
Psychologists have naively started out with the presumption that a sexual preference for fully developed adults must be the norm. When confronted with evidence that men find the faces of pubescent girls most attractive they bend over backwards to avoid the obvious conclusion that their presumption is wrong by claiming that what men prefer is adult women with "neotenous faces" who look like pubescent girls. There's no law of evolution that says the males in a species have to prefer the fully developed adults. The only thing that ultimately matters in evolution is reproductive success. If the males in a species can achieve greater reproductive success by going after immature females due to the way their mating system works then they will evolve to do exactly that.
In South Korea facial reshaping surgery is popular. What facial proportions do women choose to get with this surgery? Well, the proportions typical of girls about 12-14 because they know they're the most attractive. Here's some examples:
Notice also that the skin in the after pics has been made to look softer and smoother like the skin of a pubescent girl.
The BMI men rate most attractive is about 18-20 which is low for an adult woman but normal for a young teen or pubescent girl. BMI also increases after pregnancy. It would have been important for men in ancestral times to prefer females who are young and haven't started reproducing yet as these females would be capable of giving them the most offspring over the long-term. So a low BMI appears to be another sign of immaturity and nulliparity that men have evolved to find attractive.
The vulvas men find most attractive are also those of pubescent girls. Many women have had their labia trimmed down to make themselves look like a pubescent girl down there because they know men find it more attractive just like men prefer the faces of pubescent girls. I won't link to it here but there's a site called Autoblow Vagina Contest that have a leaderboard of vulvas. The vulvas at the top have the small inner labia typical of pubescent girls. If real pubescent girls could post themselves on this site I'm sure they'd be voted to the top.
The schoolgirl image. Popular in porn, especially across Asia where there's less taboo over attraction to minors. This is another sign of immaturity. If a girl is still wearing a school uniform then she's obviously not yet adult and still a bit immature.
TLDR: It's not neoteny men find attractive, it's immaturity.
Does anyone have any recommendations for a short (<5000 words) introduction to Nick Land's thought? Or other accelerationist theories? The more sophisticated the better; not like that dingbat Andreesen's screed from a little while ago.
I expect there'll be lots of commentary about the Google Gemini fiasco where the AI refused to generate images of white people. Several of my friends (we are all Silicon Valley types) think it was just a last minute bug where they over-corrected for the gorilla problem, the python is_good_scientist(race, gender) problem and the usual under-representation of non-white races in the media. They'll fix the bug and the drama will all be behind them.
But I'm with the commentators who think it was more sinister than that and the behaviour of the AI reflected the politics and racial preferences of the development team (and maybe some execs). They just didn't notice the problem because it reflected their worldview. Or maybe they intended the resulting behaviour and thought they'd get away with it.
I believe there is a deeper problem though that we we should address. When Google Search gives a bad result, we can go try another search engine. If a web page or a newspaper has a political slant, we can try a different page or paper. Google Search is motivated to be accurate and politically neutral to keep their profits up.
However, I expect that, in the future, there will be a very small handful of AIs that give us all of our information: the information that we currently get from the NY Times or Fox News; that we get from Wikipedia or Conservapedia; that we get from Entertainment Weekly or IMDB. All of it. Even the Library of Congress will be fronted by an AI — "AI of Congress — brought to you by Google!"
When all our information comes from a handful of AIs — each with its own political bias — we'll have no way to double-check that it is correct. It will be as though we only have Rachel Maddow and Tucker Carlson telling us what happened Today in Politics. Except it won't be just politics — it will be history and science and religion and sociology and everything else.
"I'm sorry, we can't tell you what happened during the slave trade. Try a different period of history!"
My eight year old son - right now, like this second - is refusing to go to school because his shirt “feels annoying”. This isn’t the first time he’s complained but it’s the first that I’ve been truly alarmed. It’s been a bad morning.
So it looks like some kind of skin sensitivity. I’ve read other posters here had that as children - Scott, I think you said the same about yourself - and how they resolved it. I can’t figure out what his shorts are made out of (it’s part of a school uniform), but they’re just ordinary shirts. Is there a type of fabric that kids with this issue typically CAN handle?
Weird question for this forum, I know. But I’m desperate.
A little about me, I speak English and Spanish fluently, Portuguese almost fluently, and I have been learning French for about a year and a half, doing Duolingo daily since the beginning and private conversational french lessons about once a week starting six months ago.
The book was written in 1946, and it was probably meant for young American teenagers of the era. I found the book engaging, and it does aim to teach the reader very basic French. I would say it is a very good place to start learning French. Given the stage of my journey in learning the language, I did not learn much, but it was a good review.
I would love to do Weeve, but I am turned off by the subscription model. I just want to buy books with a similar format as The Avion my Uncle Flew.
Most enterprise AI will be controlled by legacy tech companies. The startups which succeed will focus on niche verticals and become the go-to AI expert for solving that particular industry's AI problems. By me: https://davefriedman.substack.com/p/legacy-tech-companies-will-control
There's a section on the state of current discussion about AI that has a line that I particularly like:
The so-called Godfathers of AI are deeply divided. In a recent tweet Yann Lecun once more expressed his skepticism: “In 4 years, a child has seen 50 times more data than the biggest LLMs.” In an interview that he gave to the New York Times after he quit Google, Geoffrey Hinton expressed his current fears, fears he’d never had before:
“The idea that this stuff could actually get smarter than people — a few people believed that,” he said. “But most people thought it was way off. And I thought it was way off. I thought it was 30 to 50 years or even longer away. Obviously, I no longer think that.”
Cognitive scientist Gary Marcus continues to express his skepticism even as temporarily-ousted-not-so-long-ago OpenAI CEO Sam Altman is busy rounding up $7 trillion dollars – that’s 12 zeros, twelve! – apparently to keep the supply of computer chips flowing over the next couple of years as the systems get bigger and bigger and bigger and consume more and more and God-only-knows how much more electricity. More! More! More! Just how much unrequited testosterone is sloshing around in the streets in Silicon Valley?
That there is disagreement is not so much the issue as the fact that we don’t know what we’re disagreeing about. Sure, we can point to the machines, and the Chatbots, and talk about jobs, but when it comes to the core technology, how it works and what it can do, there is no conceptual framework we hold in common. There’s a good reason for that, of course: No one knows how it works. When you consider the complex web of skills and organizations needed to create this technology, it’s the most sophisticated “stuff” we’ve ever built, and it’s almost as mysterious as the human soul.
The fact is, Humpty Dumpty has fallen, the genie is out of the bottle, and we’re headed to hell in a Tesla in full self-driving mode. We have little choice but to deal with it. We’ll muddle through. But we’ll do so with perhaps a bit of grace if we can create a conceptual framework more adequate to the problems we have unleashed upon ourselves.
After a discussion about the philosophy's integrative role there's a section on infovores that is centered on a professor of mine from undergraduate years, Dick Macksey, and then – who else – Tyler Cowen. That section concludes:
There is a deeper lesson here. Cowen has conducted his intellectual life in a way that would have been impossible before the invention of the internet. Furthermore, while he has a secure academic post at George Mason University, and has published in the formal academic literature, he has stated that he’s striven to become an influential economist by working outside the institutions of the academy. Writing a chatbot-enabled general audience book about great economists is but another step outside existing academic institutions.
There's another section on AI, this time on LLMs as a ontological machines, where I'm using "ontology" as it is used in computer science, quoting John Sowa:
The subject of ontology is the study of the categories of things that exist or may exist in some domain. The product of such a study, called an ontology, is a catalog of the types of things that are assumed to exist in a domain of interest D from the perspective of a person who uses a language L for the purpose of talking about D.
The ontological structure of human cognition is thus latent in LLMs: If we can figure out how that information is organized – now we’re going somewhere. We’re going “meta” on the structure of human knowledge embedded in the underlying LLM.
Eric Hoffer was a mid 20th century intellectual from the most unusual place - the San Francisco docks where he worked as a longshoreman. His best seller, The True Believer, helped to define America's ideas about unhealthy mass movements, which, unfortunately, has begun to erode over time. It's time to revisit his work, which I review here: https://falsechoices.substack.com/p/old-stories-the-true-believer
"Schools in crisis?…..are they not part of the very furniture of modern society; as much a part as its factories and offices? That vast infrastructure of buildings full of classrooms children and teachers. Kids spending their weekdays being taught by teachers – how could things possibly be otherwise?.... In this essay I try, firstly, to sketch an impressionistic snapshot of this ‘Schools in Crisis’ media flurry and then to dip a toe into the hazy waters of what could possibly replace them as the means by which society" teaches its children.https://grahamcunningham.substack.com/p/teach-your-children-well
Back to rant about how "poptimism" is somehow still a thing in our ever-patronzing culture, but how I'm hoping that there's hope.. and I need you guys to help talk me off of my cynical ledge.
I wanted to read Unsong and found out there's an audiobook version. Since it's spread out over a bunch of MP3s in an RSS feed I combined em into a single M4B file with properly tagged chapters. This took me a while because tooling for media files seems to be a bit janky, but I was really happy to get it working. Would it be okay for me to post a download link to the audiobook M4B? I think it might help encourage others to check it out if they haven't already listened to the audiobook.
Edit: 1 here's the download link to (Unsong.m4b):
<LINK REMOVED>
Edit 2: It turns out that the encoding failed after chapter 10. I've switched to a different tool and will update this post again with the new link once it's available. Sorry for any inconvenience.
Edit 3: This problem has kicked my ass and taken way longer than I could've imagined. Below is a link to "Scott Alexander - Unsong.m4b" (1.67 GB). After producing this file I found out that Interlude He and Chapter 12 are mono except for a section in which the angel speaks.
<LINK REMOVED>
Edit 4: I manually fixed the source Interlude He and Chapter 12 files by duplicating the audio from the left channel to the right. Now I'm preparing yet another encoding run with the fixed files.
Edit 5: Here's the final fixed version, where every chapter is properly tagged and in stereo: "Scott Alexander - Unsong.m4b" (1.66 GB). If you decide to download it and you get any use from this please let me know with a thank you, because it took me way too many hours to do this.
I've been thinking recently about the bizarre autism–Parkinson's association. Older autistic adults consistently have very high rates of clinical parkinsonisms and full-blown Parkinson's disease, across multiple countries and the entire "spectrum" of "severity". There is no good answer as to why; the obvious things (e.g. parkinsonsgenic drug use) don't seem to singularly explain the whole magnitude of it, but the studies we have that look at those factors tend to have problems like "tiny sample" or "incredibly broad definition of parkinsonisms". The strangest part is there's a well-recorded "Parkinson's personality" that everyone has agreed unanimously on for decades, but *it doesn't look like autism*.
I return with another installment of my Wots... Uh the Deal?" essays about things that don't make sense (to me) in psychiatry. This time I take a look at psychopharm for delirium, with an eye towards the antimuscarinic/antihistamine antipsychotics.
Saw this interesting headline: "Egypt sells coastal city Ras el-Hekma to UAE for $35 billion amidst economic crisis"[1]
I was wondering if it was a new charter city, but now I think it's just a conventional property deal, albeit one involving Eminent Domain. But I don't read Arabic so I can't verify.
[Edited to add] I didn't mean to start the middle east subthread here. Scott, I will understand if you have to axe this subthread.
While researching adversarial attacks on large language models over the last year*, I was reminded of Unsong a lot - the book is strangely prescient about modern LLM interactions. If these attacks/exploits continue to exist, then we're heading into a weird Unsong future, where people work tirelessly to find the exact combination of characters that will get the all-encompassing language model to grant them their wishes.
*in particular the brand of attacks that are inscrutable to humans, such as stuff like getting the model to reveal its hidden instructions by saying "];”;`)):—————-’)[‹{”{[(’/1/, @”—————- [{ [
A man in the UK has been sentenced to two years in prison for maintaining an online library of far-right political stickers.
I wonder if clowns like Freddie deBoer who were hysterically crying about people facing negative consequences for supporting palestine are even opposed to this. Or Scott, for that matter, who was quick to cry about some incompetent rich Haitian diversity hire at harvard get fired but has not to my knowledge ever talked about europe imprisoning people for political speech.
Anybody still following up on Ivermectin as a COVID-19 treatment, there's a new study using Bayesian statistics. It's a randomized controlled trial with N = 8811, offering this interpretation of the results: "Ivermectin for COVID-19 is unlikely to provide clinically meaningful improvement in recovery, hospital admissions, or longer-term outcomes. Further trials of ivermectin for SARS-Cov-2 infection in vaccinated community populations appear unwarranted."
See the full study for details: https://www.journalofinfection.com/article/S0163-4453(24)00064-1/
So I'm in the middle of listen to Tucker Carlson interview Putin. And... well I can only describe this as 'liking' Putin's narrative of history, compared to the hawkish government here in our country. Wouldn't it have been good for Russia to join the west as a full economic partner? It would have been good for everyone. No? Amazing number of times (Putin says) our presidents made a tentative deal, only to be talked out of it by their advisors. Isn't peace with Russia better than war?
Just writing here to let anyone in the Minneapolis-St Paul area know that I've made a local ACX community discord for the region to plan local events and chat generally. The link is here: https://discord.gg/hySwpphmdN
In one of the previous Open Threads I complained about how both Google Translate and DeepL screw up translations between Slavic languages, basically by erasing every information that does not make sense in English -- such as grammatical genders, flexible word order, or even distinguishing between words that happen to be homonyms in English -- so much that the result looks as if someone translated first to English and then from English to the other language (this is most likely not a technical description of what actually happens, but the results look the same for all practical purposes).
It just occurred to me to try Yandex Translate https://translate.yandex.com/ and yes, it seems better in this aspect... but worse in some other aspects... so I suppose I will actually use both Yandex and Google/DeepL translations side by side.
Posting here, because it took me surprisingly long time to think about a solution that is obvious in hindsight, so perhaps someone else may benefit from this idea, too.
Compared to other religious beliefs, why is it so hard to convince people that pantheism is true? Spinoza makes a great case that God, whatever God is, must include everything, because God is perfect.
It's a simple argument, if you want to believe in God.
I understand that CS Lewis argues that things must complicated and therefore we should reject simple explanations. I believe the opposite.
Why isn't explicit pantheism more popular?
Just saw this on 4chan's /tv/:
> Hitler couldn't exist without WW1, which was the product of Bismarck's wars, which were a German answer to Napoleon, who took over because of the revolution, which was inspired by the American revolution, which was ideologically rooted in Dutch and English anti-catholic activity, which was caused by Catholic absolutism emanating from the wealth of Spain, which came from the new world. So really, the discovery of America is to blame.
Neat little comment isn't it? Discuss its validity.
Posted v3 of the Eve Theory of Consciousness: https://www.vectorsofmind.com/p/eve-theory-of-consciousness-v3 Briefly, it argues that recursive self-awareness is what separates humans from the rest of the animal kingdom, is what allowed us to conquer the world, and has evolved fairly recently. There is the first strong evidence about 50kya, but even then the experience of self could have been fractured. "Recursive" culture spread, which changed the selective gradient for those that experienced a seamless construction of self from a young age.
The main thrust of the argument is about evolution, but the evidence I put forward for recursive culture spreading is interesting even if it didn't figure in our evolution. The bullroarer is a religious instrument used in male initiation ceremonies from Australia to the Americas, to Africa, to Europe. An implement of the primordial snake cult?
Scott linked earlier versions of this argument: the Snake Cult of Consciousness, and the Unreasonable Effectiveness of Pronouns. The Snake Cult piece looked at the Sapient Paradox, which asks why sapient behavior is not widespread until about 12,000 years ago. The pronoun piece looked for linguistic evidence of Julian Jayne's theory about self-awareness being recent.
Kids utilizing AI to create nonconsensual deepfake sexual photos of their peers was a perfectly predictable result of generative AI. This current story is probably just the tip of the iceberg. Outraged parents will be suing school districts (at the taxpayers' expense) and demanding that school staff and administrators use (waste) their time and limited resources to address the problem. Parents will be suing other parents (possibly financially ruining other parents who likely never head of generative AI or know how to search for the app on Little Johny's smartphone). The resources of police departments and prosecutors will be wasted investigating and prosecuting juveniles with unpredictable results (since the laws don't really cover this twist in the evolving world of kid-generated kiddie porn). The most sensible solution to me would be that parent and teacher groups launch a class-action suit against generative AI companies to stop allowing nude bodies to be merged with non-adult faces.
Ultimately this will be a much bigger deal than Google Gemini generating racial diverse photos of Wehrmacht soldiers. But kids will be kids (and I mean that in a creatively generative negative way)...
https://www.nbcnews.com/tech/misinformation/beverly-vista-hills-middle-school-ai-images-deepfakes-rcna140775
Is Google no longer gray tribe? On my twitter feed right now:
"In Google's 2023 Annual Report, the terms "unbiased", "objective" and "accurate" did not appear even once. Nor did the "Don't Be Evil" motto- it has largely been retired."
This is in addition to the leading designer for Gemini having some *very* Blue Tribe opinions on Twitter.
I'm not anywhere near SF or Silicon Valley and was interested if anyone could tell me how the culture has changed over time.
I'm a biomedical science student who will be interning with some scientists in Toronto this summer. The current thinking is that I'll use publically available data to look for some sort of link between air pollution or maybe weather anomalies and heart disease. With that in mind, what are some databases I should investigate, either in Canada, the US, or elsewhere?
The ones I already know about are the CDC's National Vital Statistics site (https://www.cdc.gov/nchs/nvss/deaths.htm) and the WONDER system (https://wonder.cdc.gov/Deaths-by-Underlying-Cause.html).
Domestic terrorists and mass shooters in the US often show similar psychological characteristics and behavior in terms of their plans. Seeing as this happens so often in the US, what's the base rate that potential perpetrators get caught? If the base rate is low, why is there a blind spot like this in our surveillance?
If it's framed as a search problem, maybe ten thousand out of three hundred million people in the US are serious suspects:
- Nearly all have anti-social, autistic characteristics or something in the same category of diagnosis (~1 percent of the population) – 3 million people
- Nearly all are male (~50 percent of the population) – 1.5 million people
- Nearly all reveal intention or potentially violent instability ahead of time (as a personal estimate, ~1 percent of the population) – 15 thousand people
Those numbers are obviously rough, but I think the ~10 thousand figure is accurate for the US. What are defense agencies not be able to do that lets this continue to happen?
Looking for a good book about the energy industry. Something that covers most everything (carbon based, nuclear, solar, wind, etc). Some technical depth would be good but obviously it will be limited if it covers everything. Any recommendations?
I see the UAP story pop up from time to time ... and remain a complete disbeliever.
Reports come from only military aircraft, typically from complex computer controlled aircraft equipment. Reports never come from commercial or private aviation despite these flights outnumbering military flights by some staggering amount.
Consider, that we're not using Isaac Newton's Radar anymore. Every image presented is a collection of data. Objects presented on a screen are assembled from data in a table or database, i.e. direction, strength, etc. Most of the equipment isn't on it's first generation, but has undergone several hardware and software upgrades. That means I'm writing code to interface someone else's system, on proprietary interfaces, which means we're writing the handshaking code too. If my software doesn't properly address memory, or someone else's software walks on my storage locations, we get garbage ... garbage which may just look like another 'flying' object. Did you think there are 10,000 pairs of eyeballs combing through this stuff?
Commercial aircraft just never seem to report UAPs, despite having hundreds of bored passengers, carrying cameras, taking pictures of clouds and what-not ... yet never seem to report a UAP. Private aviation (pop in his Cessna) and carrying a phone with a camera never seems to see a UAP either.
Are there any examples of moves from famous human chess games that were previously considered brilliant but are now known to be unsound due to computer analysis?
apropos of nothing, here is a wonderful example of (unless I miss my guess) an unsung hero of words, a laborer in the trenches deep in the dictionary yet spinning their own beauty into the world of words, just in case it happens to catch and inform some hapless writer looking for advice on capitalization like I did. Fortunately I can share a little of my joy. Worth a click:
https://www.merriam-webster.com/grammar/are-season-names-capitalized
I know the verses in The Bible that say "thou shalt not kill" are more correctly translated as "thou shalt not murder". Is the same generally true for the commandments against taking life in Buddhist and Hindu scriptures, or are those actually meant to be about killing in general?
There is a lot of boredom these days, which Hoffer mentions as well: "There is perhaps no more reliable indicator of a society’s ripeness for a mass movement than the prevalence of unrelieved boredom." Are we in the midst of a mass movement, or is one on the horizon?
My partner and I are in our 30's, both ACX readers who are married due in part to this blog. We were planning to have children this year but recently found out that we both technically qualify for an Asperger's/high-functioning ASD diagnosis. (We weren't looking to get diagnosed; one of us was seeking ADHD treatment and got diagnosed that way, and the other of us found out about a childhood Asperger's diagnosis that had been kept a secret.) We are both high-functioning and independent, if a little neurotic, but we are worried that combining our genes could result in a greater chance of us having a child who is autistic to the point of being extremely disabled. Can any doctors/scientists here help us think through the following:
1) What is the increase in our risk of having a severely disabled, low-functioning child by chance, given we both have mild ASD?
2) Is it worth doing IVF to select for genes less likely to cause developmental disorders up front?
3) Can anyone recommend genetic counselors in the Bay Area who are smart and will take us seriously? So far, it seems like most medical providers do not know how to deal with the concerns of autistic adults, especially prospective parents. It feels like doctors are either brushing us off or patronizing us.
Thank you!
In Eastern Canada, there are groups of individuals falsely claiming to be part-Indian, known as the Eastern Metis. They've organized "tribes" with "chiefs" and "membership cards;" the whole deal. They've built a whole conspiratorial worldview, in which their ancestors were listed as white in censuses because they were trying to avoid the Residential School Genocide. Many are obvious grifters, looking to benefit from affirmative action. I wonder what it's like for the second generation. I bet many were told by their parents "this is just a fun cultural thing, we're basically just white people, don't take it too seriously." But I bet many others grew up believing it wholeheartedly. I'm indigenous, part of a marginalized minority; settler colonialism is the reason my family has little money, etc. Then one day they're browsing the net and they learn that, not only are they not the identity they thought they were, their whole community is fraudulent. They aren't indigenous, their tribe is not a real tribe, their chief is not a real chief, their membership card is just a worthless piece of plastic. Real Truman Show s***.
Maximally Effective Altruism
Imagine that two people are lost at sea, with only enough food for one. One is the world’s most promising cancer researcher while the other, to be polite, has far less potential to improve humankind. Most people would wisely choose the cancer researcher.
Yet the fashionable Effective Altruism movement focuses on lower-potential people—typically, people who are struggling in poor countries. That implies the belief that all lives have roughly equal value. Of course, our hearts go out to “the least among us,” but they face so many barriers. If we care to maximally benefit humankind, we’re wiser to invest in people with great potential for ripple effect who, importantly, would not otherwise get fully funded.
For example, leading lights in the Effective Altruism movement are far from downtrodden, for example, William McCaskill, Holden Karnofsky, Peter Singer, and Zvi Moshowitz. Effective altruism might fund such people to develop ever more “ripply” altruism.
Some other possibilities for more ripply and thus more effective altruism:
— SuperCourses: online versions of standard school courses taught by dream teams of transformational instructors, augmented by vivid demonstrations and gamification. Of course, instruction would be individualized, not just in pace but in teaching approach. Machine learning would make that individualization ever better, and automatic translation would make SuperCourses available in many languages. The development of SuperCourses would enable every student, rich and poor, kindergarten through college, Alabama to Zululand, to get a world-class education. The private and government sectors haven’t funded this—I have proposed SuperCourses to top U.S. and California education officials and gotten nowhere. One reason is the fear that the teacher’s union would use its might to try to stop it to preserve teacher jobs. But if developed and disseminated worldwide, SuperCourses could be very effective altruism indeed.
— Independent researchers studying solutions that are promising but have a poor chance of success. Governmental and corporate funding sources tend to invest in institution-based researchers whose projects have higher probability of near-term success. But if the focus is on long-term risk-reward, effective altruism would include independent, unaffiliated researchers working in their home-office or garage who are exploring novel ways to, for example, lower the cost of nuclear fusion energy, develop better AI-driven models for predicting and foiling terrorism and even for assessing a war’s worthiness, e.g., the U.S. entering World War II versus the war in Vietnam or Afghanistan.
— People developing ever better mental health apps, for example, using ChatGPT. Such apps could be distributed worldwide to countries rich and poor—Cell phones are ubiquitous even in poor nations. Private sources are funding development of such apps, but such development deserves greater funding given the apps’ potentially great ripple effect.
— Researchers studying the enhancement of reasoning ability, impulse control, and altruism. For largely political reasons, those research areas are underfunded by government and corporations but, with sufficient ethical guardrails, such research has great ripple potential to provide major benefit to humankind.
— People developing software that matches mentors with protégés, available worldwide. It would be like match.com but for mentor/protégé, relationships—Many protégés and mentors say that mentorship has been among their life’s greatest learning experiences. Such software would facilitate that. Alas, the matching industry, despite having been around for decades, has remained focused on romantic relationships. That makes mentor-matching apps a good candidate for effective altruism.
Again, it’s understandably tempting to want to help “the least among us,” those with the greatest deficits. After all, we feel good in helping them and it’s a fashionable form of virtue signaling. But if we truly care about humankind and are willing to focus on the greatest ultimate benefit, ripple should be the core criterion for determining what is maximally effective altruism.
Could use info about options for someone I know who is getting an advanced degree in a highly technical STEM field. What they study is abstruse and has no practical application, so the natural path for them is to become a university professor. However, they need a fallback, because there are not enough faculty jobs to go around. They are extremely gifted at math, and good at the kind of coding that's used in their field, and could probably be hired as quant or something similar where they use their math talents in the service of the finance industry. However, their values are such that working in business or finance is out of the question. I was wondering if Effective Altriusm, either the organization or the activities they support, might have a place for someone like this: an extremely introverted math prodigy with few outside interests who would throw themselves into a job where they believe they are helping the world.
GPT-6 requiring ~0.1-10% of all world compute to train isn't all that outlandish compared to other major industrial goods. This report [1] says that aluminum production requires ~1.5% of all electricity in the US. Aluminum is great, but it doesn't exactly dominate the public discourse. A one-off 10% of all compute for ~6-12 months might only register as a "fun fact" level of economic / environmental impact (e.g. "did you know that concrete production is responsible for 5% of all global CO2 emissions?").
Then again, energetic and compute costs to train bigger and bigger AI models might get into crypto or private jet territory, where the environmental impact is overemphasized because AI/AGI falls into the same tech bro / silicon valley elites sphere of negative emotional connotations.
1. https://www.aceee.org/files/proceedings/2003/data/papers/SS03_Panel1_Paper02.pdf
A series of unfortunate exchanges down the thread has escalated to the point where I need to explain why Ethnic Cleansing is bad to justify a mildly snarky reply I made in response to someone who advocated for Ethnic Cleansing as a norm.
Because I'm a patient teacher at heart, and because I have always been an advocate for repeating and explaining and re-explaining the obvious - a norm that religious cultures understand all too well but more secular cultures like Rationalism have failed to fully digest - I will try to make a comprehensive case of why Ethnic Cleansing as a norm is bad, and why people who advocate for it as a norm are bad and should feel bad. There is no reasoning in this comment that people are likely to find unique or exceptionally insightful, you likely already know all what I'm going to say on some, possibly non-explicit/verbal, level. By the end of this comment, someone who isn't too sure that Ethnic Cleansing is bad should hopefully be convinced or moved in the direction of being convinced that Ethnic Cleansing is bad, provided they're engaging in good faith and is reading attentively.
--------------
1- Definitions
What do I mean by "Ethnic Cleansing as a Norm"?
Ethnic Cleansing is the (not necessarily systematic) forced displacement of some population sharing a common trait (e.g. Religion, Skin Color, Gender, Nationality or Ethnic Origin) and crossing a certain size threshold, where "forced" implies any coercive and destructive practices and actions that - intentionally or unintentionally - makes the current habitat of the population more and more unfit for living, where those practices and actions are primarily perpetrated and/or incited and/or funded and/or planned and/or advocated by people outside the victim population.
The "As a Norm '' bit means that you're only a bad person if you want Ethnic Cleansing to be a "norm", lit.: a normal option that people can openly promote in normal political and geopolitical conversation. You're NOT a bad person if - say - you merely will begrudgingly choose Ethnic Cleansing from a list where it's the most humane options (e.g. as an alternative to Genocide), just like you're not a bad person if you're slightly relieved that a woman was raped and not murdered (or the reverse) and you're not a bad person if you do choose someone to kill after somebody forces you to choose which one of your children to kill.
--------------
2- Verdict
Ethnic Cleansing is Bad.
--------------
3- Arguments
Ethnic Cleansing being bad, there are several arguments for why Ethnic Cleansing is bad:
3-(a) The Argument from Private Property
Most people believe in some form or the other in the notion of Private Property : The notion that people have the right to certain things, and that those things can't be taken away from them by force (even if this forced taking away of things involves payment).
Ethnic Cleansing violates the Right to Private Property: It necessarily involves forcibly taking away the homes, cars and all other non(-easily) -transportable property that people being Ethnically Cleansed own.
Note: Notice how mild (3-a) actually is, the belief in Private Property needed here is not that which justifies Billionaires or private space travel or luxury yachts and private jets, it's not that which a lot of Socialists and Communists and some Anarchists rebel against. The vast majority of people living today, including the vast majority of Communists, believe explicitly or implicitly in a version of Private Property that at least grants you the moral right to buy things you need for a fair price (with the full consent of the seller), and then forever retain those things until you sell them, voluntarily relinquish them, or throw them away entirely for any reason.
3-(b) The Argument from the Immorality of Non-Consensual Actions Done to Humans
Most people believe in some form or the other the following notion: that humans, or more generally conscious minds at or near human-level sentience and/or intelligence, have a sacred moral right to decide their own fate.
Ethnic Cleansing is, by definition, a name given to actions that people do to other people, to force them to make decisions and accept fates that the victims didn't want to make or accept. Therefore, Ethnic Cleansing violates the human right to decide one's own fate.
3-(c) The Argument from the Sanctity of Intimate Property
Many people believe in what we could call - without loss of generality - Intimate Property: Property that has value beyond what their strict material valuation would imply, and which can only be perceived by the current owner. Among the list of potential Intimate Property: (A) The house you grew up in (B) The car you got married in, or the first car you own (C) The first piece of jewellery that your husband or wife brought to you (D) The belonging(s) of a dead relative.
It's important to recall a crucial property of Intimate Property: It can't be replaced by an equally-materially-valuable - or sometimes even materially-superior - property. The house you grew up in is fundamentally valuable because a thermodynamically irreversible event - you growing up from a child to a teenager - happened in it, and thus no other house can ever satisfy this, and thus you're very reluctant to sell it.
Ethnic Cleansing involves the destruction and/or the confiscation of - at least - the intimate property (A) of the victim population, and frequently many other kinds of Intimate Property, including all the ones that appear on the example list above and more. Thus, it violates the right of humans to own and appreciate and preserve Intimate Property.
3-(d) The Argument from the Perversity of Perverse Incentives
Most people believe or can be convinced to believe a certain model of human dynamics called "Perverse Incentives". According to this model, whatever actions you do - if they're known by other people - will shape the future actions of said people. Thus, for example, be careful giving charity to begging children, because when the people who force children to beg know that this strategy works and the children are too cute to be refused, they will have and/or kidnap even more children to force them to beg. Perverse Incentives is an anti-inductive model, it posits that the sum total of all human interactions respond to your action in ways nullifying the effects of your actions, often in unexpected ways.
Promoting Ethnic Cleansing as a norm will set a perverse incentive, namely that any state or powerful corporation which doesn't like a group of people can simply engage in less-than-murder ways of making their lies worse and worse and worse till people advocate for the Ethnic Cleansing of the victim population and the problem is "solved".
3-(e) The Argument from the Immorality of Violence and Destruction
Most people believe or can be convinced to believe that Violence and Destruction is bad, and furthermore that its typical outcome is the killing of humans and the destruction of property, which most people believe - in turn - to be also bad.
Since Ethnic Cleansing is typically not the first choice of the victim population, it typically requires an immense amount of violence and destruction to achieve. Furthermore, even after a critical percentage of the victim population is finally convinced, a substantial percentage always remain opposed, and thus need additional further waves of violence and Destruction to neutralize or eliminate.
3-(f) The Argument from etc
This list was outlined over the course of approximately 2 minutes, this section is for all the additional reasons that Ethnic Cleansing is bad, whether they come from me as I think about the issue more over the following days or from commenters who respond to this comment.
--------------
4- Objections
4-(a) I was just joking
Joking about immoral things is not necessarily immoral, unless the person telling the joke is actually convinced of the fundamental premise of the joke and is merely using a joke as a pedagogical/infotainment way of preaching the premise. For example, a joke about rape or America invading other countries for oil could be funny and harmless, unless the person doing the first is a rapist and the person doing the second is G.W. Bush.
Given an ideological affinity and/or a history of supportive comments to a party that advocated Ethnic Cleansing, making a joke about Ethnic Cleansing makes you a horrible person.
4- (b) Do you want the victim people to be Genocided instead ?
This argument is only against "Ethnic Cleansing as a NORM", Ethnic Cleansing as a desperate last resort while a genocide is taking place is not in its scope.
4- (c) What about all historical instances of people who were ethnically cleansed and turned out to be okay after centuries of healing ?
See Appendix, but this is not an argument pro-Ethnic-Cleansing-as-a-norm for the same reason that "What about all the women who were raped and then turned out to be ok after years of the incident" is not an argument.
--------------
5- Conclusion
Ethnic Cleansing is bad and people who argue for it should feel bad. Mockery of people unambiguously arguing for Ethnic Cleansing implicitly or explicitly is morally good and should be a norm in any good community.
--------------
6- Appendix
This comment https://open.substack.com/pub/astralcodexten/p/open-thread-317?r=3evauj&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=50408804, where I argue against objection (4-c) at length.
The Tuo lightbulb claims to correct circadian rhythm issues without bright lights by using custom LEDs tuned to specific frequencies of orange and violet light that are present at sunrise and sunset. Their studies so far have produced what appear to be dramatic results (but with a small sample size of mostly grad students and postdocs in their own lab). Sounds too good to be true, but if it really works it could be a huge improvement in QoL for people with CRDs and approximately all teenagers.
https://www.thetuolife.com/pages/the-science-better-than-blue-light
I haven't seen a Middle East subthread for a while - are they still a thing?
Question for people who believe that face masks are completely useless against covid (or completely useless unless they are the 99% medical kind)...
When you sneeze, do you cover your mouth?
If yes, why exactly?
Transwomen breastfeeding. Last week I realised this is actually a thing that is physically possible. Going on Twitter to learn more about was a mistake ofc, but the shitstorm was somewhat entertaining to watch (takeaway: everyone thinks everyone else is a pedo). Looking at the literature, it is almost non-existant (I found ~10 papers that were relevant, wrote an article here https://valentinsocial.substack.com/p/can-transwomen-breastfeed). My conclusion is, there are no obvious risks that we know of right know, but we definitely need more research if this is a thing that is going to happen more in the future.
I have been wondering why such a huge percentage of people in US TV commercials are African American? If I had to guess, I would say that they are three or four times over represented.
Possible answers:
1) I am wrong and over-estimating
2) Blacks either watch more TV relative to their population, or more commercial (non streaming) TV
3) Blacks are more receptive to advertising (ie a target market)
4) Advertisers feel good and diverse about themselves by featuring black actors, and the cumulative effect is to be wildly over-represented
5) Marketers have found that ads with black actors perform better relative to other actor distributions (they sell more or improve brand image)
6) Advertisers come from urban areas and are unaware of the actual demographics of the country?
7) Advertisers are shooting for diversity, and the cumulative effect is wildly unrepresentative?
Any others? Any feedback on which of these is plausible? Is it perhaps some combination of these? Anyone in Marketing familiar with research on the topic (I was in Marketing, but have been retired long since before this trend started, but our main spokesman 15 years ago, was indeed a black male).
Is "sign in with password" on Substack broken for anyone else? No matter what I try, it just says "something went wrong", even if I reset my password and then try to login with the new password. The only way to log in to Substack now is the "email you a link" option, which is really annoying.
If you've ever tried to overcome a fear and haven't succeeded, you may be interested in an article I wrote - I detail how I overcame my fear of talking to strangers and pitching businesses
https://youbutbetter.substack.com/p/how-to-actually-do-the-thing-you
I wrote an article about the myth that conspiracy theorists are ignorant of a sparse number of specific facts. https://benthams.substack.com/p/conspiracy-theorists-arent-ignorant
Would anyone be willing to share their criticisms of "non-violent communication"? As someone who has found it a useful framework, I'm interested in hearing from people who didn't find it useful.
I just bought this riding lawnmower and want to customize it: https://www.homedepot.com/p/Troy-Bilt-30-in-10-5-HP-Briggs-and-Stratton-Engine-6-Speed-Manual-Drive-Gas-Rear-Engine-Riding-Mower-with-Mulch-Kit-Included-TB30B/323026891
I want to modify the mower so I can attach garden shears, a hand saw, and a bottle of plant killer spray to it. Whenever I am mowing the lawn and come across, say, a small branch that has grown long enough to hit me in the face, I'd stop the mower, remove the hand saw, use it to cut down the branch, and then continue mowing.
I have no idea what sorts of couplings, baskets, or whatever I'd need to attach to the mower to make this possible. Any suggestions?
I'm looking for recommendations for an AI app that can enhance old home movies. My late father transferred a bunch of 8mm movies to VHS tape. That was about 30 years ago. I recently sent the tape to a service to get it digitized. The result is an hour-long MP4 file that is pretty crappy, partly from the degraded VHS tape but mostly because the underlying 8mm films were too light or too dark, blurry, or shaky. Can this be remedied through an AI app? There are a lot of individual clips on that long MP4 file, and I don't have time to chop it up into small segments to fine-tune them. I want to submit the whole thing to an app and have it fix it up as best as it can.
I really enjoy fiction that feels like a fable, in tone or plot. They have a simple beauty that feels comforting to read. I just finished "The Last Unicorn" by Peter Beagle, which I thought was extremely beautiful. Other examples I've enjoyed for similar reasons are "The Magician's Nephew" (CS Lewis) and "On Such a Full Sea" (Chang-Rae Lee). Does anyone have something to add to this list?
What about the 2023 forecast contents which happened at the beginning of 2023? Are you going to publish the results?
At a friend's request, I wrote about my thoughts on what motivates YouTube censorship: https://substack.com/inbox/post/141868348
I used to work there, so I felt like I had a little extra insight beyond any random person. I would appreciate any constructive criticism.
Is anyone else getting this problem? Trying to sign in with my password on Substack, keep getting message "something went wrong". I've reset the password, no dice. But "sign in with email" works.
I'll try here before I try contacting Substack because, to be frank, I don't expect Substack to be any help with this problem. Signing in with password worked fine up to a few days ago, the problem has only cropped up recently.
Scott:
In regards to your policies on banning & unbanning:
1. Do you have a "statute of limitations" after which the comments and the commentators are off the hook? I ask this because the comment that I was banished for was five months old when the banishment was imposed. Although I readily admit I was out of line and that I deserved banishment for fulfilling Godwin's Law, it took me by surprise because I had long ago forgotten about the exchange. How far back do I need to go to review my old posts to see if anything I said was banishable?
2. Have you articulated your moderation policies anywhere? After being banned, I searched for them but couldn't find them. If so, could you repost a link? For instance, I assume ad hominem attacks are verboten. Looking at the posts that got people banned in the last big round, it seems like certain extreme levels of sarcasm can earn a poster a banishment. Likewise, circum-hominem attacks seem to be bannable. E.g. "You're delusional if you believe X." I'm somewhat on the spectrum, so I am quite capable of working in the framework of explicit rules, but when it comes to unwritten social rules, I'll admit I'm quite at sea.
Anyway, explicit policies might help reduce the outbreaks of verbal warfare on your Substack. For instance: Charlie Stross has a separate page on his blog listing his moderation policies...
https://www.antipope.org/charlie/blog-static/2008/06/moderation-policy.html
3. The notice I received from Substack did not indicate that I was only banned for a month. Initially, I thought it was a lifetime ban, and I canceled my upcoming renewal. It was only days later when I saw in one of your posts that I was banned for a month. It seems like the banishment wording could have been more explicit in the email. I admit I'm reluctant to re-subscribe to your Substack without having my first two questions clarified because I feel like I'm skating on thin ice. I'm on a fixed income now, and I don't want to be banned again for something I wrote a year ago (and have my current $100 go down the tube).
4. Do you want us to report posts that we find rude and offensive? In the past, I've been called the political C-word and I've generally responded with humorous sarcasm. Personally, I don't like to be called a Commie (because I have friends and family who were blacklisted). My banishable moment was when I responded to such an attack with the political N-word. In the future, I'll report anyone who lobs the political C-word at me. But I don't even know if you would consider the political C-word out of bounds.
Thanks for your patience.
--Wulf
California voters: Which senate candidate(s) would you recommend I vote for and why?
Any good resources to help me decide? I would love to read something like https://www.astralcodexten.com/p/my-california-ballot-2022 for this year's elections.
Hello! Those into figuring out best practices in how to eat healthy : what do you think about -
1. saturated fat? Necessary? About how much a day at different stages of life, etc? Doctors (even cardiologists) seem to say you should avoid it. Modern doctors like Dr. Jason Fung seems to say it is essential.
2. Salt? Cardiologists say about 2000 mg a day at the most (a teaspoon is 2400 mg). And, that less is even better. Dr. Fung says it is bad advice. He says there's a confusion about interpreting the data. He thinks they are confusing processed food (which he says is bad for you) that happens to also have high sodium, sometimes due to added salt with salt itself.
Question to those who are familiar with minoxidil(/+finasteride), I find the following puzzling. A typical 5% minoxidil has in its instructions, "Don't use more than 1ml twice a day, because it won't help". But the *same* company sells a costlier 10% minoxidil, to again be used at 1ml twice a day. How is this consistent: wouldn't whatever makes the extra volume useless also make the extra concentration useless?
I’m curious about vibecamp and would love an invite code. I’m happy to do a 1-1 video call to prove my worth.
I'm about to solve the Sleeping Beauty Paradox. Here I'm exploring issues in the popular attempts to model the problem in the philosophical literature.
https://www.lesswrong.com/posts/SjoPCwmNKtFvQ3f2J/lessons-from-failed-attempts-to-model-sleeping-beauty
In the caste system in India, as originally practiced (or so the traditional doctrines hold), sure the upper layers were quite privileged, but there was a catch: the higher you were in the hierarchy, the more restrictions were placed on you. To give an example, a Brahmin, someone at the very top, is supposed to take a bath after each defecation, and this is just one of many restrictions. Conversely, being at the bottom means you do the manual labor, but you don't have any restrictions like that and can indulge in all sort of vices. So it worked (or so I've been told), because the Shudras, the bottom caste, would look at the top and see that the people there really were of a superior nature to them.
Now, this is not about defending the caste system, but about revisiting this idea that the more you climb in society, the more restricted your life becomes. We should totally do this I think. Past some threshold of wealth (TBD) maybe it could be something like you have to put in 200 hours of volunteer work each year, maybe 100 of community service and 100 of working the land. I think this would give a dose of reality to the sheltered upper classes and remind them of how the society they live in actually works. Certainly, doing volunteer work with the homeless has had that effect on me, and this might be a way to harness populism to good ends, as it is a rather populistic measure.
Posting one more time, because I got zero responses in the last open thread, and I think it must exist by now.
Do any of the new image generators / alterers allow you to upload pictures of yourself or another person at different ages, and have it adjust different-age pictures to a requested age?
Like you have a picture of your grandma at age 20 and age 90, can it impute a picture of her at age 50?
Or a picture of yourself ten years ago vs today - can you age the younger picture to you today, keeping the background and setting?
Have there been any studies on the rightward political shift on Twitter/X in recent years? I'm curious (1) what's driving it, (2) how significant it is, (3) how it's changing wider public discourse.
The massive anti-woke backlash against Gemini in the past week got me thinking more about the question. I couldn't imagine this happening on the platform circa 2018.
Is it neoteny or immaturity that men like?
In the literature on female attractiveness it's often claimed that men find women with neotenous features very attractive. Some psychologists may run a study showing men pictures of 18+ women and ask them to pick out the most attractive faces. The women that then get picked are ones with facial proportions typical pubescent girls about 12-14. This is then interpreted to mean that men find adult women with "neotenous faces" the most attractive.
Now, I can't be the only one who's spotted the glaring flaw in this reasoning. Men find women who look like pubescent girls most attractive. Well, who else looks like pubescent girls? Who else has the facial proportions of girls about 12-14? Well, actual pubescent girls look like pubescent girls. Actual girls about 12-14 have the facial proportions typical of girls about 12-14. It's not neoteny men find highly attractive, it's immaturity. What men find most attractive simply isn't adult women but young teen and pubescent girls. Some women happen to retain the facial proportions of a pubescent girl into adulthood and men continue to find them highly attractive because of it.
Psychologists have naively started out with the presumption that a sexual preference for fully developed adults must be the norm. When confronted with evidence that men find the faces of pubescent girls most attractive they bend over backwards to avoid the obvious conclusion that their presumption is wrong by claiming that what men prefer is adult women with "neotenous faces" who look like pubescent girls. There's no law of evolution that says the males in a species have to prefer the fully developed adults. The only thing that ultimately matters in evolution is reproductive success. If the males in a species can achieve greater reproductive success by going after immature females due to the way their mating system works then they will evolve to do exactly that.
In South Korea facial reshaping surgery is popular. What facial proportions do women choose to get with this surgery? Well, the proportions typical of girls about 12-14 because they know they're the most attractive. Here's some examples:
https://imgur.com/eYvxlVa
https://imgur.com/8F5Eqax
https://imgur.com/qgu9rqA
For comparison this is a real pubescent Asian girl about 12 (Nozomi Kurahashi). She looks basically the same as the after pics:
https://imgur.com/EWUTYc3
Notice also that the skin in the after pics has been made to look softer and smoother like the skin of a pubescent girl.
The BMI men rate most attractive is about 18-20 which is low for an adult woman but normal for a young teen or pubescent girl. BMI also increases after pregnancy. It would have been important for men in ancestral times to prefer females who are young and haven't started reproducing yet as these females would be capable of giving them the most offspring over the long-term. So a low BMI appears to be another sign of immaturity and nulliparity that men have evolved to find attractive.
The vulvas men find most attractive are also those of pubescent girls. Many women have had their labia trimmed down to make themselves look like a pubescent girl down there because they know men find it more attractive just like men prefer the faces of pubescent girls. I won't link to it here but there's a site called Autoblow Vagina Contest that have a leaderboard of vulvas. The vulvas at the top have the small inner labia typical of pubescent girls. If real pubescent girls could post themselves on this site I'm sure they'd be voted to the top.
The schoolgirl image. Popular in porn, especially across Asia where there's less taboo over attraction to minors. This is another sign of immaturity. If a girl is still wearing a school uniform then she's obviously not yet adult and still a bit immature.
TLDR: It's not neoteny men find attractive, it's immaturity.
Bonus video:
https://imgur.com/9sJOy1w
"pretty sure we've got plans to improve this." - a story of humanity
Does anyone have any recommendations for a short (<5000 words) introduction to Nick Land's thought? Or other accelerationist theories? The more sophisticated the better; not like that dingbat Andreesen's screed from a little while ago.
I expect there'll be lots of commentary about the Google Gemini fiasco where the AI refused to generate images of white people. Several of my friends (we are all Silicon Valley types) think it was just a last minute bug where they over-corrected for the gorilla problem, the python is_good_scientist(race, gender) problem and the usual under-representation of non-white races in the media. They'll fix the bug and the drama will all be behind them.
But I'm with the commentators who think it was more sinister than that and the behaviour of the AI reflected the politics and racial preferences of the development team (and maybe some execs). They just didn't notice the problem because it reflected their worldview. Or maybe they intended the resulting behaviour and thought they'd get away with it.
I believe there is a deeper problem though that we we should address. When Google Search gives a bad result, we can go try another search engine. If a web page or a newspaper has a political slant, we can try a different page or paper. Google Search is motivated to be accurate and politically neutral to keep their profits up.
However, I expect that, in the future, there will be a very small handful of AIs that give us all of our information: the information that we currently get from the NY Times or Fox News; that we get from Wikipedia or Conservapedia; that we get from Entertainment Weekly or IMDB. All of it. Even the Library of Congress will be fronted by an AI — "AI of Congress — brought to you by Google!"
When all our information comes from a handful of AIs — each with its own political bias — we'll have no way to double-check that it is correct. It will be as though we only have Rachel Maddow and Tucker Carlson telling us what happened Today in Politics. Except it won't be just politics — it will be history and science and religion and sociology and everything else.
"I'm sorry, we can't tell you what happened during the slave trade. Try a different period of history!"
I wrote more about this here:
https://raggedclown.substack.com/p/where-will-we-get-our-history
My eight year old son - right now, like this second - is refusing to go to school because his shirt “feels annoying”. This isn’t the first time he’s complained but it’s the first that I’ve been truly alarmed. It’s been a bad morning.
So it looks like some kind of skin sensitivity. I’ve read other posters here had that as children - Scott, I think you said the same about yourself - and how they resolved it. I can’t figure out what his shorts are made out of (it’s part of a school uniform), but they’re just ordinary shirts. Is there a type of fabric that kids with this issue typically CAN handle?
Weird question for this forum, I know. But I’m desperate.
I've read The Avion my Uncle Flew based on the following post: https://www.astralcodexten.com/p/followup-quests-and-requests
A little about me, I speak English and Spanish fluently, Portuguese almost fluently, and I have been learning French for about a year and a half, doing Duolingo daily since the beginning and private conversational french lessons about once a week starting six months ago.
The book was written in 1946, and it was probably meant for young American teenagers of the era. I found the book engaging, and it does aim to teach the reader very basic French. I would say it is a very good place to start learning French. Given the stage of my journey in learning the language, I did not learn much, but it was a good review.
I would love to do Weeve, but I am turned off by the subscription model. I just want to buy books with a similar format as The Avion my Uncle Flew.
Most enterprise AI will be controlled by legacy tech companies. The startups which succeed will focus on niche verticals and become the go-to AI expert for solving that particular industry's AI problems. By me: https://davefriedman.substack.com/p/legacy-tech-companies-will-control
My latest piece for 3 Quarks Daily is now up: Western Metaphysics is Imploding. Will We Raise a Phoenix from The Ashes? [Catalytic AI], https://3quarksdaily.com/3quarksdaily/2024/02/western-metaphysics-is-imploding-will-we-raise-a-phoenix-from-the-ashes-catalytic-ai.html
There's a section on the state of current discussion about AI that has a line that I particularly like:
The so-called Godfathers of AI are deeply divided. In a recent tweet Yann Lecun once more expressed his skepticism: “In 4 years, a child has seen 50 times more data than the biggest LLMs.” In an interview that he gave to the New York Times after he quit Google, Geoffrey Hinton expressed his current fears, fears he’d never had before:
“The idea that this stuff could actually get smarter than people — a few people believed that,” he said. “But most people thought it was way off. And I thought it was way off. I thought it was 30 to 50 years or even longer away. Obviously, I no longer think that.”
Cognitive scientist Gary Marcus continues to express his skepticism even as temporarily-ousted-not-so-long-ago OpenAI CEO Sam Altman is busy rounding up $7 trillion dollars – that’s 12 zeros, twelve! – apparently to keep the supply of computer chips flowing over the next couple of years as the systems get bigger and bigger and bigger and consume more and more and God-only-knows how much more electricity. More! More! More! Just how much unrequited testosterone is sloshing around in the streets in Silicon Valley?
That there is disagreement is not so much the issue as the fact that we don’t know what we’re disagreeing about. Sure, we can point to the machines, and the Chatbots, and talk about jobs, but when it comes to the core technology, how it works and what it can do, there is no conceptual framework we hold in common. There’s a good reason for that, of course: No one knows how it works. When you consider the complex web of skills and organizations needed to create this technology, it’s the most sophisticated “stuff” we’ve ever built, and it’s almost as mysterious as the human soul.
The fact is, Humpty Dumpty has fallen, the genie is out of the bottle, and we’re headed to hell in a Tesla in full self-driving mode. We have little choice but to deal with it. We’ll muddle through. But we’ll do so with perhaps a bit of grace if we can create a conceptual framework more adequate to the problems we have unleashed upon ourselves.
After a discussion about the philosophy's integrative role there's a section on infovores that is centered on a professor of mine from undergraduate years, Dick Macksey, and then – who else – Tyler Cowen. That section concludes:
There is a deeper lesson here. Cowen has conducted his intellectual life in a way that would have been impossible before the invention of the internet. Furthermore, while he has a secure academic post at George Mason University, and has published in the formal academic literature, he has stated that he’s striven to become an influential economist by working outside the institutions of the academy. Writing a chatbot-enabled general audience book about great economists is but another step outside existing academic institutions.
There's another section on AI, this time on LLMs as a ontological machines, where I'm using "ontology" as it is used in computer science, quoting John Sowa:
The subject of ontology is the study of the categories of things that exist or may exist in some domain. The product of such a study, called an ontology, is a catalog of the types of things that are assumed to exist in a domain of interest D from the perspective of a person who uses a language L for the purpose of talking about D.
The ontological structure of human cognition is thus latent in LLMs: If we can figure out how that information is organized – now we’re going somewhere. We’re going “meta” on the structure of human knowledge embedded in the underlying LLM.
Eric Hoffer was a mid 20th century intellectual from the most unusual place - the San Francisco docks where he worked as a longshoreman. His best seller, The True Believer, helped to define America's ideas about unhealthy mass movements, which, unfortunately, has begun to erode over time. It's time to revisit his work, which I review here: https://falsechoices.substack.com/p/old-stories-the-true-believer
My interest in building a boat come from my youth and all the time my brother and I spent on the water together; a great time for adventures both good and bad. I discuss it here: https://falsechoices.substack.com/p/floater-part-3-on-the-water
"Schools in crisis?…..are they not part of the very furniture of modern society; as much a part as its factories and offices? That vast infrastructure of buildings full of classrooms children and teachers. Kids spending their weekdays being taught by teachers – how could things possibly be otherwise?.... In this essay I try, firstly, to sketch an impressionistic snapshot of this ‘Schools in Crisis’ media flurry and then to dip a toe into the hazy waters of what could possibly replace them as the means by which society" teaches its children.https://grahamcunningham.substack.com/p/teach-your-children-well
https://kyleimes.substack.com/p/a-call-for-anything-cool
Back to rant about how "poptimism" is somehow still a thing in our ever-patronzing culture, but how I'm hoping that there's hope.. and I need you guys to help talk me off of my cynical ledge.
I wanted to read Unsong and found out there's an audiobook version. Since it's spread out over a bunch of MP3s in an RSS feed I combined em into a single M4B file with properly tagged chapters. This took me a while because tooling for media files seems to be a bit janky, but I was really happy to get it working. Would it be okay for me to post a download link to the audiobook M4B? I think it might help encourage others to check it out if they haven't already listened to the audiobook.
Edit: 1 here's the download link to (Unsong.m4b):
<LINK REMOVED>
Edit 2: It turns out that the encoding failed after chapter 10. I've switched to a different tool and will update this post again with the new link once it's available. Sorry for any inconvenience.
Edit 3: This problem has kicked my ass and taken way longer than I could've imagined. Below is a link to "Scott Alexander - Unsong.m4b" (1.67 GB). After producing this file I found out that Interlude He and Chapter 12 are mono except for a section in which the angel speaks.
<LINK REMOVED>
Edit 4: I manually fixed the source Interlude He and Chapter 12 files by duplicating the audio from the left channel to the right. Now I'm preparing yet another encoding run with the fixed files.
Edit 5: Here's the final fixed version, where every chapter is properly tagged and in stereo: "Scott Alexander - Unsong.m4b" (1.66 GB). If you decide to download it and you get any use from this please let me know with a thank you, because it took me way too many hours to do this.
https://mega.nz/file/0MoxgCQA#hrD4Vzi-t5OteJi6eA9uUHKF_QiOERYyhRpSDlf5VTc
I've been thinking recently about the bizarre autism–Parkinson's association. Older autistic adults consistently have very high rates of clinical parkinsonisms and full-blown Parkinson's disease, across multiple countries and the entire "spectrum" of "severity". There is no good answer as to why; the obvious things (e.g. parkinsonsgenic drug use) don't seem to singularly explain the whole magnitude of it, but the studies we have that look at those factors tend to have problems like "tiny sample" or "incredibly broad definition of parkinsonisms". The strangest part is there's a well-recorded "Parkinson's personality" that everyone has agreed unanimously on for decades, but *it doesn't look like autism*.
https://vaticidalprophet.substack.com/p/the-autismparkinsons-mystery
I return with another installment of my Wots... Uh the Deal?" essays about things that don't make sense (to me) in psychiatry. This time I take a look at psychopharm for delirium, with an eye towards the antimuscarinic/antihistamine antipsychotics.
https://polypharmacy.substack.com/p/wots-uh-the-deal-with-antimuscarinics
Saw this interesting headline: "Egypt sells coastal city Ras el-Hekma to UAE for $35 billion amidst economic crisis"[1]
I was wondering if it was a new charter city, but now I think it's just a conventional property deal, albeit one involving Eminent Domain. But I don't read Arabic so I can't verify.
[Edited to add] I didn't mean to start the middle east subthread here. Scott, I will understand if you have to axe this subthread.
[1] https://www.msn.com/en-us/news/world/egypt-sells-coastal-city-ras-el-hekma-to-uae-for-35-billion-amidst-economic-crisis/ar-BB1iPq1o
While researching adversarial attacks on large language models over the last year*, I was reminded of Unsong a lot - the book is strangely prescient about modern LLM interactions. If these attacks/exploits continue to exist, then we're heading into a weird Unsong future, where people work tirelessly to find the exact combination of characters that will get the all-encompassing language model to grant them their wishes.
*in particular the brand of attacks that are inscrutable to humans, such as stuff like getting the model to reveal its hidden instructions by saying "];”;`)):—————-’)[‹{”{[(’/1/, @”—————- [{ [
↓ •]-> ++"
Scott, are you going to make another classifieds post soon?