We had plumbers in this week, they said they would be finished Wednesday, now say Friday. The meetup is scheduled for Saturday and we don't think we can get things sufficiently back in order to do it. I have sent the cancellation out on my mailing list, asked Scott to post it, and am putting this up as well.
When I visit https://astralcodexten.com, I'm taken to a sort of generic business-looking website. When I visit https://www.astralcodexten.com, I'm taken to the normal blog. Is this intentional (as some sort of ARG or defense against the public internet?) or accidental? Seems strange to me.
What traits does the aesthetic known as "classiness" entail? What exactly does it look like / feel like to have a classy home, a classy outfit, etc?
Most importantly: Is the defining feature of classiness simply that it signals status? Is classiness actually 100% a matter of social signalling, or does it have aesthetic qualities independent of signalling ability, at all?
Hi everyone! EA for Jews is running another round of the EA and Judaism Intro Fellowship this October (https://eaforjews.org/take-action/fellowship/). I thought it might be of interest to some folks here. Please help spread the word to any Jewish communities you are part of!
Today I was thinking: coupons and loyalty programs basically use pointless makework as a form of costly signalling so that companies can get information to more effectively price-discriminate among their customers. But their existence is probably bad for economic efficiency overall because it distorts price signals and diverts labor into pointless makework. The cost of providing the product doesn't actually depend on who's buying it in the vast majority of these cases of price discrimination. Price discrimination transfers some of the surplus from consumers to producers but never increases the total surplus and it might create some deadweight loss if done imperfectly so banning price discrimination would probably not do any net harm.
Advertising wastes people's time conducting emotional manipulations that have no connection to the quality of the product. The information value of advertising is negligible except for obscure products, but most advertising is brand advertising for things we've seen before. Customers can gossip amongst themselves or use trusted third parties like consumer reports to gain much better information than advertising can provide. Most advertisers are incurring a lot of expense in an arms race that provides no value to anyone except zero-sum market share gains.
I think a high tax on advertising and coupon/loyalty programs would probably be a good thing even though I am generally way out on the right on economic issues.
There was a article years ago - 10 proposals for the US that make sense but will never pass or something like that. One being self valuing your house to determine property taxes but can be bought for that price by anyone at anytime. Have been searching in vain for a while - any leads?
Well, I got my covid shot today. My flu shot was last week, so hopefully I'll be good for this season.
Here is a thing that I can't say to my friends, because it codes as anti-vax: I sometimes wonder if vaccines might be bad for you. Not in a "vaccines cause autism" way, but more like this: the covid vaccine will make me feel exhausted and feverish tomorrow, as my immune system kicks into overdrive and frantically converts all available resources into antibodies. I feel like that probably counts as a high-stress event? How many times can my immune system do that before it starts to get worn out?
I got the covid shot anyway, because covid is alarmingly common in my friend group, and because it's all sorts of bad for you, and I really don't want to get it. But I kind of resent that they made it this strong. I resent that nobody has researched whether a one-quarter-strength covid vaccine would work approximately as well, and be less of a strain on people's bodies.
This paper argues that all the different cognitive biases identified by psychologists can be reduced to a small number of fundamental beliefs coupled with confirmation bias. Of course, as rationalists, we're all aware of our cognitive biases and thus we are able to set our biases aside when looking at the evidence. ;-)
I'm writing a post on The Adventures of Huckleberry Finn, just would like to hear what some of you, any of you, think about the novel. An informal survey but your quick thoughts would be appreciated.
How do you all handle diversity statements? For context, I'm a STEM undergraduate applying to graduate programs, and this part of the process really has me stumped. I struggle somewhat with writing to begin with, so statements of interest are hard enough, but these truly challenge me, not least because I worry that these are designed as filters to keep out people with more milquetoast liberal views on diversity like myself. Not to mention, I have a pretty privileged background, all things considered. Have others been able to even find programs that don't require these? Have commenters here found success with any particular strategy in writing diversity statements?
I am no fan of Donald Trump. I am certainly no fan of Ron DeSantis. When I read this, I had the feeling I needed to choose between an opportunist or an ideologue. When Trump casually says that he can cut a deal about one of the most profound schisms between right and left. It’s very tempting to believe him. I think overall I prefer opportunism to ideology.
I mean, I know they are all playing games with each other but still…
I remember reading about an author's fictional characters being expressions of 'personified mental constructs'[my words, probably]. Not sure how to do a search for this - it's like misspelling a word so badly spell check shrugs.
Does anyone have any solid starting point I can work off of? You would have my gratitude and a gift of 1000 points!
Is anyone doing analysis of the efficiency of the incentives being given for the CO2 emissions avoided or atmospheric CO2 removed. I've seen some high numbers but from sources that are opposed to any incentive, so I distrust them, but nothing from the loyal opposition. In principle I know that the form of the subsidy -- for investment in instead of output from the target technologies -- is wrong, but principles do not give magnitudes.
Some organizations in Europe signal their wokeness by having a token Black person at an important position. Other organizations signal their wokeness by having a token Gypsy person at an important position. I haven't seen an organization that would have *both*.
The general pattern seems to be that commercial companies prefer Black people, while non-profits prefer Gypsy people.
LLMs are basically trying to do to human writing what the book review submitters were trying to do to Mr. Alexander's writing. For LLMs, the metric is how much their writing can seem like that of a generally skilled human. For book reviews, the metric is whether an audience selected for liking the Astral Codex Ten house style likes your kind of writing.
It's been established that Mr. Alexander is the best at writing like himself (1st place, plus a good-but-not-great finish for the second entry). Currently, humans are also the best at writing like humans more generally, although it seems like the difference is getting small for some more prosaic tests.
This suggests to me that it's going to be very, very hard to get LLMs to write better than humans. Obviously it's not open-and-shut, but I would presume some of the competitors spent much longer on their entries than Scott spent on the actual Njal's saga review (minus the gimmicky parts that I utterly ignored), or otherwise had more resources at their disposal.
In particular, I think this is a sign that we shouldn't be particularly worried about LLMs ripping off specific authors and being successful at doing so. For one, this demonstrates that it's a pretty tough job to outdo the literal benchmark on a task. But also, a lot of people are complaining about various aspects of Scott's style showing up too much in the book reviews; I'd expect complaints about there being too much Jane Austen in some hypothetical JaneAustenBot's works. At least for me, books by a specific author have clearly diminishing marginal returns.
I wouldn't take a random contest on a blog too seriously, but it's an interesting antidote to some of the more extreme forms of concern about LLMs rapidly becoming superhuman.
I'm trying to understand partner dances. I was forced to learn some foxtrot back in middle school and I've picked up some country dance as a happy amateur now when I've matured. I'm trying to branch out into some Latin dances now as well. But I don't feel like I "get it".
1. Most (well-known) dances seem to have a "social" scene and a "show"/"competition" scene. To me as a casual observer, the "show" dance seems very far divorced from the "social" dance. I feel like if there's this sport called "golf" and when I try to learn it my teachers tell me to practice hitting this white ball with this club to make it go into a hole, but when I watch a golf competition on TV they're playing soccer. As a beginner, I try to learn the steps, and I try to learn the "feel" and posture of the dance. But are those things even important? Can I show up at a shag dance and dance the whole night without doing any shag basics steps?
2. Related: Often when I watch a video from a dance show, I don't see the dancers do any of the basic moves that I learn as a beginner. Often I wouldn't know what dance they were dancing unless the video title told me. How do you identify dances? Can you do it from the music alone ("this is a bachata song so they're dancing bachata")? Is there's some good flashcard-style way to learn to identify different dances? Like, I get a ten second clip of people dancing and I've to identify it. I guess I should hack it together myself using YouTube... Do dances have signature moves that makes them recognizable (but "advanced" dancers don't do the basic moves per 1., and also couldn't I do the signature move of one dance while dancing another dance just because I'm cool?)? How do I spot what dance someone is dancing?
3. Are there a music theory but for dance? Are there a natural limit to the types and styles of partner dances? Could I make up my own partner dance right now from scratch by just deciding on a couple of steps to a 4x4 beat? Or would what I make just turn out to be rumba with a backstep or something? Most (all?) partner dances seem to belong to a "tradition": do dances naturally split and involve like this? Do they fuse? Are new dances invented from scratch sometimes?
4. Related: Can dances be wildly combined? Could someone just smash together rumba and foxtrot and have a new fusion dance? Would it be recognizable as a rumba+foxtrot fusion? Could such a fusion be done in many different ways? Must the dance I'm dancing match my partner? What happens if I dance shag but my partners dance rumba? What happens if I dance lindy hop to bachata music?
(I know I could get an answer to many of these questions myself by just trying it myself, but I'm interested in the deeper answer that I don't really have the terminology to ask about.)
I know Scott has posted stuff about Buddhism in the past, so I assume there are some "secular Buddhists" here I could ask this question to. I know the Buddhist concept of rebirth is different from the traditional conception of reincarnation, that it's only the karma that gets reborn etc., however, suppose that I am a skeptical materialist that denies even that. If we believe that death doesn't bring rebirth into Samsara, but is a full extinguisment of everything one was, does that make it equivalent to Nirvana? Essentially, if we accept the first noble truth that life is Dukkha, and desire to extinguish that Dukkha, if we don't believe death results in rebirth into this world, then wouldn't suicide be a much quicker and more expedient means than the long and difficult road to achieving Nirvana in this lifetime? Or is there something to Nirvana that is more than just extinguishment, making it still preferable to death even in a world with no rebirth? Please note that I'm not actually contemplating suicide based on this line of reasoning, this is more of an academic exercise and something I'm genuinely curious about, whether Buddhism reduces to implying suicide if the belief in life as suffering is maintained but the belief in rebirth isn't. Essentially, death to an atheist materialist looks a lot like what Buddhists describe as the ultimate goal of paranirvana, i.e. permanently dying and not coming back, so if you think regular death brings that by default anyway, why not just die?
What is the general impression of AI generated art? It seems Aella takes flak for using it in her posts, but I generally like them, and I wouldn't even say her pictures are that effective. Are AI-generated images seen as not classy or something? Is it a generational thing?
How's everyone's experience with meditation? Do people here meditate regularly? Do you get a lot out of it? I did some every day during Covid, and eventually had a kinda-transcendental experience after drinking a beer on a beach where I was extremely dialed in to my surroundings- not, off in my own head thinking about something else. I was just very 'there' and not elsewhere for a little while. Despite that, I got away from regular meditation until recently, when I started up again. I'm now averaging probably 4 days a week, every morning after I get out of the shower and before I have coffee. It's a slow but steady learning curve to get better, though I haven't experienced any benefits outside of the actual meditation yet.
Are other people meditating regularly? Seeing benefits?
Piano players: I want to learn a few pieces nearly perfectly. By this I mean that I'd like to be able to play them in front of a crowd with no stumbles/startovers and *maybe* miss a note here and there. It seems that as I practice, my accuracy asymptotically approaches 90% or 95% and takes a long time to even get close to that.
I am convinced that really good pianists still have to practice a lot but they don't have to deal with asymptotic progress and they can progress fairly linearly up to 100%. At piano recitals, I marvel at the way performers will play multiple pieces without a single mistake.
How do I get there? Is there a way to shoot for 110% so that I hit 100% with less effort? For instance, if I want to lift 100 lbs 6 times, I'd work on being able to lift 120 lbs 8 times. At that point, lifting 100x6 would be almost effortless. What's the analogous approach for piano? Learning to play it in the dark?
Ukrainian forces are slowly breaking through Russia's three concentric lines of defenses in southwestern Ukraine. They've now reached the second line.
Why doesn't Russia respond by building a fourth line farther back? Ukraine's men are moving so slowly that Russia probably has enough time to build it before the Ukrainians reach it.
And if the slow pace continues, what's to stop Russia from building fifth, sixth and seventh defensive lines in time?
What would the effect on the economy be if factories were 100% automated and they could shut down without ill consequence during recessions? No furloughed human workers, no increase in unemployment and welfare claims, no families damaged by mom or dad sitting in the living room drunk or high all day suddenly.
I'm looking to set up (play money, for now) prediction markets within my company on private company topics...which means I can't use third party applications, but can (and deeply desire to) use third-party source code to get off the ground. Anyone know of a recommended codebase for something like this? The closer to vanilla use of common languages like html/js and python, the better.
With an estimated 35-45 years to live as per every life expectancy calculator I can find and a net worth of approximately $2.5M, how would you devote the remainder of your life to fighting wokeness and/or feminism with maximum effectiveness? If I keep working at my (quite woke) job for the next 5 years, I will probably accumulate another $1-1.5M. Assume no family.
Some metrics seem much easier to Goodhart than others. For example
-easy to manipulate: "In response to Labour targets in the 1990s, one force recorded a 27 per cent reduction in the number that had a target (‘theft from a motor vehicle’) alongside a 407 per cent increase in the number that had no target (‘vehicle interference’)." (https://www.lrb.co.uk/the-paper/v45/n18/john-lanchester/get-a-rabbit)
-presumably much harder to manipulate: homicide versus suicide or accidental death.
Assuming I'm not mistaken, what are some not-too-technical sources discussing this further? For general interest, not AI alignment.
I'm looking for help. Mostly in the form of advice, I guess, though I wouldn't say no to other forms (not money, I still have some to keep me afloat for a while).
There are three aspects to my problems: work, a place to live, and psychology.
1. I'm a scientist from Russia. Not a particularly good one, with a candidate's degree (the lower level in two-level PhD system) and over 25 published papers but my h-index is not even 5. I'm not really keeping up to date with current research in my area, and it's been quite a while since I tried to tackle some serious problem. I can do middle-level work if given a problem to work on, get some results and write those papers. This might have been enough if I was 30 but I'm ~40 now.
My work also involves programming (and I do it for my own entertainment/personal projects), but I'm not up to date on a lot of tools and concepts. It's a bit better in my own area (machine learning) but there aren't many job offers. I think with enough effort I could get myself in shape to be employable in a wider area, but it's hard to put in the effort with uncertain prospects.
Still, I am hoping to find a job I'm somewhat qualified to do.
2. I wasn't a fan of Russia since at least 00s, but never found enough motivation to do something about it. Eventually, I fled to Armenia in late 2022 when they started mobilization (not that my risk of being drafted was high, but I decided to take a safer route). It's nice here but I'm not sure it works as a long-term solution. For one thing, Armenia still has strong ties with Russia and I'm not sure what happens to people like me if Russia passes laws making us criminals. So while finding a decently paying job here would be a nice start, eventually I hope to find a job that would allow me to move elsewhere.
Which is a huge problem because I have zero experience with looking jobs abroad (and negligible experience doing it in my own country, for that matter).
Armenia is also problematic in that it's hard to find a decently paying job here if I decide to switch to something I'm less skilled with just to break the pattern.
3. I don't know what it is with me, but things do not look good. I lack motivation and willpower. With my old low-requirement job it was mostly enough, though I had a tendency to almost miss deadlines whenever things I didn't really like doing (e.g. preparing papers) came up. The actual work of finding things out was more interesting so less problematic. It seems to be exacerbated by me working from home (though I wouldn't give a scout's promise to become a highly performing worker if I had to work from the office, but it should help if I'm at least somewhat interested and can't switch off immediately whenever I encounter something I don't feel like doing). Lately it got worse with the necessity to look for a new job, which I'm scared of. Maybe also because of the overall shitty situation, too (shame and guilt over what my country does with overt support by a lot of people I knew and respected, having to leave my home unprepared and leave behind friends and relatives in what seems to be turning into a totalitarian state). But the end result is that I do basically nothing to advance further towards this goal of a new job - mostly just wasting time, sometimes doing something that is still meaningful and can qualify as "work" but is an obvious way to avoid doing something I want even less. I'm not even sure where writing this post falls, but I'll try to commit to following at least some of the advice I get if there is any.
Is it depression? I'm not sure, it doesn't feel like I imagine depression should. Though lately there have been days when I just wanted to stay in bed, but that was because none of the things I had planned were enjoyable, and I was trying to stop allowing myself doing anything besides them. (Also there's growing sense of urgency so even the distracting things I normally like are less enjoyable. But that's more like a stress.)
So, I'll appreciate any thoughts and suggestions. In particular:
Re 1. Is there any kind of job you think I should focus on looking for? I still have time and money to maybe get additional training (or prepare myself) in some math or programming related topic if it would let me land a job with high probability.
(I probably won't say no to a direct offer too, if everything I wrote hasn't discouraged you.)
Re 2. Particularly if I can get this job in any livable country that wouldn't kick me out to Russia on short notice, for example if Russia decides that missing an appointment with a military registration office is now a crime. Although I feel like I'm just not good enough to look for a job abroad (anyone willing to expand their search this much can probably find better candidates), is there anything I can do about it?
Re 3. I am close to recognizing that I need professional psychological/psychiatric help, but if that is so, will the help help? Do my problems look like something where talking it out or taking some pills is likely to result in a large improvement? Again, I'm constrained with the resources I could spend on this but I can afford giving it a try if the expected payoff is good. Or are there things I could just try by myself?
What are the best free sites that use AI to summarize long articles (such as the book review context stuff)? So far the free GPT's that I have tried do a bad job.
I tried https://tldrthis.com/ , signing in through google to get me an AI (human-like) summary that is detailed/section-wise and asked for summarizing this article:
and it gave a decent summary, but they only allow 10 articles for free.
Is it too silly to hope that something like a wikipedia could be created for slow-reading low attention span folks (like me) leveraging AI to summarize several books/articles? Sorry if I am being too stupid/naive.
Scott, do you have any thoughts on Data Falsificada[1]? The summary is in the first paragraph: "[A] series of posts detailing evidence of fraud in four academic papers co-authored by Harvard Business School Professor Francesca Gino."
I posted this in the subreddit, and haven't gotten any response, so hopefully someone here will know something:
Just came across a supposed image pulled from a rabbit's retina in the late 1800's by Dr. W. Kühne, supposedly creating the field of "Optography". While he himself claimed that it wasn't good enough to be of use in forensic science, I'm skeptical that he even managed to do it at all.
Supposedly it's based on the fact that Rhodopsin, a pigment in the retina, bleaches in the presense of bright lights, and can afterwards be fixed through chemical processes. The fixed Rhodopsin can then, according to Kühne, be turned into a recognizeable image (although not detailed enough to be of much use).
I buy the physical claims about Rhodopsin. I'm sure it bleaches as described, and I even buy that it can be fixed in that bleached/unbleached state. I do not believe that any images, useful or not, were actually produced with this method.
Unfortunately, some google scholar searching and a little bit of internet research have failed to turn up very much. It's mostly focusing on how widespread it got in victorian science fiction.
I found one citation from the 70's called "Optography of the retina. Resumption of Küne's studies", but it's just a citation, I can't actually find the article. Wikipedia has a link to an article that claims it was replicated (based on the title, I'm guessing it's not itself a replication, just a report of someone else doing it) but the link goes nowhere and the title doesn't even return anything on Google Scholar.
This has all the hallmarks of being a complete load of bullshit. But I can't find enough evidence to actually prove it.
If anyone has any actually reliable information on this, I would be greatly appreciative.
I recently realised that a spookily-consistent mathematical rule has described the previous political "re-alignments" in the US, and that this rule says the next re-alignment will be in 2028.
The rule is that each Party System lasts for four years longer than the previous one. I'm defining the precise dates of the party systems by my own reasoning, which differs very slightly by how historians may define them, but I think my definitions are justified (and I discovered the rule only after settling on these dates):
1776: The Revolution, and the framing of the constitution, and the largely uncompetitive, pre-party presidencies of Washington and Adams.
1800: The First Party System begins with the election of Jefferson, entirely dominated by what's now called the Democratic-Republican Party.
1828: The Second Party System begins with the election of Jackson, establishing the Democratic Party.
1860: The Third Party System begins with the election of Lincoln, establishing the Republican Party.
1896: The Fourth Party System begins with the campaign (and loss) of William Jennings Bryan, and the election of McKinley.
1936: The Fifth Party system begins with the re-election of FDR, firmly establishing the New Deal Coalition.
1980: The Sixth Party System begins with the election of Reagan, establishing the modern conservative movement.
In brief: the US established; 24 years later a party system established; 28 years later a *competitive* (two real options) party system established; 32 years later the *current* two-option party system established; 36 years later the current parties with their recognisable ideologies (D=left, R=right) established; 40 years later the modern "flavour" of Democratic ideology (big government, unions+minorities) established; 44 years later the modern "flavour" of Republican ideology (evangelicals+business+national security establishment) established. 48 years after 1980 will be 2028.
The only part of the above that seems to differ from how the party systems are normally defined is using 1936 for the start of the Fifth Party System, instead of 1932 when FDR was first elected. But I think this makes sense, because 1932 was not really an endorsement of a new ideology like 1828 and 1860 and 1980 were (or a repudiation of a new ideology like 1896 was) but a rejection of the incumbent over the depression. Also, I think FDR downplayed his big-government plans in his first campaign. Only his re-election really created a new ideological alignment. The rest of the dates seem uncontroversial to me.
So, why would this rule hold? It's too perfect to be a coincidence, I think. But what's causing it? I don't know exactly, but there are many phenomena that have very complicated causal explanations but can still be modeled by a fairly simple rule: that's the way the world often works, and the complex causes end up interacting with each other and canceling each other out to produce (when you zoom out) a simple, straightforward trend. Most likely, there's a general (mostly unconscious) awareness that each party system is coming to an end, and each time enough political will to squeeze one further election out of the old system before it collapses. (Incidentallly, you can even extend the rule back before the Revolution--20 years--to the beginning of the Seven Years- War in 1756, which was the catalyst for the causes of the revolution).
Finally, how will this most likely play out in 2028? I see three posibilities:
1. Desantis (or someone else who isn't Trump) wins the Republican primary and wins the general election. I know this isn't looking good at the moment, but I think it's more likely than people think (though I won't elaborate here) and it gives the clearest path to a political re-alignment. Say it's Desantis, and he returns to a more traditional GOP governance. The rural working class is sick of this and they permanently split from the GOP to form their own Trumpian populist party. The Democrats go into meltdown like they did in 2016, and the socialist faction again blame the woke elite faction. This time there's an actual split: the socialists leave the Democratic Party. Is it possible the Trump populists and the Sanders socialists find common ground and unite to become the Populist Party? I don't know, but it seems perfectly possible. The Trump supporters can now embrace class warfare and economic leftism along with right-wing nationalism to become a truly anti-elite force: Trump can outright adopt a restributionist agenda, truly drain the swamp, no longer have to play nice with the old guard of capitalism. Meanwhile, the socialists no longer have to gently criticise the woke faction: they can go all-out anti-woke, condemn identity politics as viciously as they condemn capitalism, educated elites as viciously as business elites. With the Populist Party established as a powerful force, the fight of 2028 will be between the two old parties, to decide which of them survives as the elite opposition to the Populists, and which one dies. That could go either way.
2. Trump wins the primary and the general. It's less clear how a split happens here, but maybe he decides to this time govern in a more economically leftist way, and the traditional Republicans leave their own party. It's impossible to see them joining the elite Democrats, and hard to see Dem socialists joining the Republican Party. But maybe these things happen in some weird way?
3. Biden is re-elected. In this case, he surely moves further to the left on economic matters, and that keeps the socialists and wokes in the same party for now. The GOP probably permanently splits, but the significance of this is less clear. Maybe a new populist Trump party challenges Biden/Harris from the left in 2028?
What do people think of this? Please don't dismiss the rule I described as a mere coincidence without explaining why it seems to so perfectly describe the past.
What happened to the convenient link in the email, that let a reader click to go direct to opening the page online? Today's message has a "subscribe here for more" link where it at various times had "view in browser" (one link) or "open in app or online" (2 links).
The intermediate state, with "view in browser" at the upper right, had a button with "read in app" a bit lower down. That button is still there, but "view in browser" is completely gone.
Obviously I got to the post in the browser, in spite of the user interface dis-improvement. You haven't made the blog completely inaccessible to those who never do anything on a cell phone they can do with a real computer, let alone to those who simply don't have a cell phone.
But the email interface to this blog is now making a pretty clear statement - we only offer convenience to app-lovers here. Most likely, that statement comes straight from substack, and is perhaps motivated by some idea they have that app users are more profitable. (Perhaps the app violates privacy more than the web site, giving them more to sell?)
You might want to restore the old appearance and links, if substack will let you do so.
Were the Crusades bad? I mean, schooling has led me to believe they were bad, but the older I get the more I’m convinced my schooling was a bunch of lies and prejudicial assertions.
On the one hand, one can say that we can’t judge the past at all because their values were too different from ours. OTOH, does that mean we can’t judge Hitler? If we can judge Hitler, then why can’t we judge the Pope Urban II?
Don’t we arrive at our present values precisly by judging the past, and saying no, yes, no, no,yes, no?
If in the year 2023 we can confidently say Hitler was bad then we shold be able to, with some level of confidence, say whether Pope Urban II was good or bad. No?
I say Hitler was bad, and the Crusades were good. Christian culture thrived after the Crusades.
All my priors say something the author is hiding a lie in plain site and my gut says it has something to do with their over/loose use of "average" and "typical". But I just can't place it.
I mean, my god, of course soldiers in war are going to eat 6k calories. And of course pedestrians will need more calories than drivers. But I don't think those answers are really enough to call it the "solution" to the puzzle. Any suggestions on where I can stand to make this all line up?
If I think of an AI as metaphorically a different species than ones that have come before, then it scares me to think that this new species would be more intelligent than humans. Because of course a much more intelligent species is going to makes mince meat out of a less intelligent one.
But if I think of AI as an extension of human intelligence in the transhumanist sense it doesn’t scare me so much.
I think AI is much more likely to be like the latter than the former. What are the arguments against that?
I started to wonder how common it is for people to embrace the idea of there being "systems" in the political context, where actions/policy choices have downstream consequences, which could be evaluated and selected for using a consequentialist framework (ie in order to decide if we should pursue x vs y policy, we should consider the anticipated systems-wide aggregate outcomes).
After speaking to a handful of generic liberals on this (ie civil servants, policy professionals, grad school educated pmcs types), I was shocked at how reprehensible everyone I spoke with found this idea. Discussing this concept felt as if I were broaching a highly taboo and inappropriate subject. The primary response I heard was that "you simply can't look at the consequences to understand what we should do" (this response, almost word for word, came from someone who self-advertises as being interested in "systems thinking"!)
I used to believe that many people disliked markets due to issues of exploitation and inequality, rather than objecting to the concept of a market itself. I also understood that some people had reservations about philosophical consequentialism, but I never thought this would extend to political contexts, where it could be considered merely one factor among many to consider.
I'm now very confused and was wondering if anyone here knows of any writing on this topic — specifically any psychology books/articles on why some people do or do not embrace the idea of systems and consequentialist reasoning.
Does anyone have a good argument for an average person buying their house? All the good ones I've heard seems to go against it(undiversified, inefficient, expected returns no better than renting), and yet it remains so popular.
This Arts & Culture feature from Google actually has a ton of interesting pages. The page for Care of Magical Creatures shows a paradoxical two-horned unicorn.
Given that I'm just a lowly commentator, I don't exactly need to post an erratum, but I wanted to admit that I was at least partially wrong about Joe Biden a month or so ago when I said that there wouldn't be any evidence that he was aware of how much more-than-baseline corrupt the Hunter business was. I'd still insist that we don't really know what the baseline level of corruption in Washington is, but my actual claim seems wrong in retrospect.
South Bay Meetup 9/23 cancelled.
We had plumbers in this week, they said they would be finished Wednesday, now say Friday. The meetup is scheduled for Saturday and we don't think we can get things sufficiently back in order to do it. I have sent the cancellation out on my mailing list, asked Scott to post it, and am putting this up as well.
When I visit https://astralcodexten.com, I'm taken to a sort of generic business-looking website. When I visit https://www.astralcodexten.com, I'm taken to the normal blog. Is this intentional (as some sort of ARG or defense against the public internet?) or accidental? Seems strange to me.
What traits does the aesthetic known as "classiness" entail? What exactly does it look like / feel like to have a classy home, a classy outfit, etc?
Most importantly: Is the defining feature of classiness simply that it signals status? Is classiness actually 100% a matter of social signalling, or does it have aesthetic qualities independent of signalling ability, at all?
Hi everyone! EA for Jews is running another round of the EA and Judaism Intro Fellowship this October (https://eaforjews.org/take-action/fellowship/). I thought it might be of interest to some folks here. Please help spread the word to any Jewish communities you are part of!
How many personalities are there? I don't like the Bryer/Miggs stuff because they ignore so much. Seems to me personaltiies are at least:
1) Hedonists
2) Infovors
3) Status-mongers
4)Money-mongers
5) Scientists
6) Artists
7) Family-mongers
8) Moralists
9( Ethicists
10) Futurists
I am going somewhere with this. The idea is everyone is at east one of these, but often more than one.
Feel free to add categories.
EDIT: 11: Builders
I added that to include people like Elon Musk. I guess it says a lot that category was an oversight.
Today I was thinking: coupons and loyalty programs basically use pointless makework as a form of costly signalling so that companies can get information to more effectively price-discriminate among their customers. But their existence is probably bad for economic efficiency overall because it distorts price signals and diverts labor into pointless makework. The cost of providing the product doesn't actually depend on who's buying it in the vast majority of these cases of price discrimination. Price discrimination transfers some of the surplus from consumers to producers but never increases the total surplus and it might create some deadweight loss if done imperfectly so banning price discrimination would probably not do any net harm.
Advertising wastes people's time conducting emotional manipulations that have no connection to the quality of the product. The information value of advertising is negligible except for obscure products, but most advertising is brand advertising for things we've seen before. Customers can gossip amongst themselves or use trusted third parties like consumer reports to gain much better information than advertising can provide. Most advertisers are incurring a lot of expense in an arms race that provides no value to anyone except zero-sum market share gains.
I think a high tax on advertising and coupon/loyalty programs would probably be a good thing even though I am generally way out on the right on economic issues.
There was a article years ago - 10 proposals for the US that make sense but will never pass or something like that. One being self valuing your house to determine property taxes but can be bought for that price by anyone at anytime. Have been searching in vain for a while - any leads?
Well, I got my covid shot today. My flu shot was last week, so hopefully I'll be good for this season.
Here is a thing that I can't say to my friends, because it codes as anti-vax: I sometimes wonder if vaccines might be bad for you. Not in a "vaccines cause autism" way, but more like this: the covid vaccine will make me feel exhausted and feverish tomorrow, as my immune system kicks into overdrive and frantically converts all available resources into antibodies. I feel like that probably counts as a high-stress event? How many times can my immune system do that before it starts to get worn out?
I got the covid shot anyway, because covid is alarmingly common in my friend group, and because it's all sorts of bad for you, and I really don't want to get it. But I kind of resent that they made it this strong. I resent that nobody has researched whether a one-quarter-strength covid vaccine would work approximately as well, and be less of a strain on people's bodies.
This paper argues that all the different cognitive biases identified by psychologists can be reduced to a small number of fundamental beliefs coupled with confirmation bias. Of course, as rationalists, we're all aware of our cognitive biases and thus we are able to set our biases aside when looking at the evidence. ;-)
https://journals.sagepub.com/doi/10.1177/17456916221148147
I'm writing a post on The Adventures of Huckleberry Finn, just would like to hear what some of you, any of you, think about the novel. An informal survey but your quick thoughts would be appreciated.
How do you all handle diversity statements? For context, I'm a STEM undergraduate applying to graduate programs, and this part of the process really has me stumped. I struggle somewhat with writing to begin with, so statements of interest are hard enough, but these truly challenge me, not least because I worry that these are designed as filters to keep out people with more milquetoast liberal views on diversity like myself. Not to mention, I have a pretty privileged background, all things considered. Have others been able to even find programs that don't require these? Have commenters here found success with any particular strategy in writing diversity statements?
https://www.radioiowa.com/2023/09/18/desantis-says-trumps-abortion-remarks-show-the-danger-in-reelecting-him/
I am no fan of Donald Trump. I am certainly no fan of Ron DeSantis. When I read this, I had the feeling I needed to choose between an opportunist or an ideologue. When Trump casually says that he can cut a deal about one of the most profound schisms between right and left. It’s very tempting to believe him. I think overall I prefer opportunism to ideology.
I mean, I know they are all playing games with each other but still…
I remember reading about an author's fictional characters being expressions of 'personified mental constructs'[my words, probably]. Not sure how to do a search for this - it's like misspelling a word so badly spell check shrugs.
Does anyone have any solid starting point I can work off of? You would have my gratitude and a gift of 1000 points!
"Why does every school teach history but hardly any have courses on the future?"
Futures thinking seems to be a burgeoning fad in K12 education. What do rationalists think?
The idea is to teach young people that their voice matters, and that they can change the future.
Is anyone doing analysis of the efficiency of the incentives being given for the CO2 emissions avoided or atmospheric CO2 removed. I've seen some high numbers but from sources that are opposed to any incentive, so I distrust them, but nothing from the loyal opposition. In principle I know that the form of the subsidy -- for investment in instead of output from the target technologies -- is wrong, but principles do not give magnitudes.
Some organizations in Europe signal their wokeness by having a token Black person at an important position. Other organizations signal their wokeness by having a token Gypsy person at an important position. I haven't seen an organization that would have *both*.
The general pattern seems to be that commercial companies prefer Black people, while non-profits prefer Gypsy people.
Is this true in general, or just in my bubble?
LLMs are basically trying to do to human writing what the book review submitters were trying to do to Mr. Alexander's writing. For LLMs, the metric is how much their writing can seem like that of a generally skilled human. For book reviews, the metric is whether an audience selected for liking the Astral Codex Ten house style likes your kind of writing.
It's been established that Mr. Alexander is the best at writing like himself (1st place, plus a good-but-not-great finish for the second entry). Currently, humans are also the best at writing like humans more generally, although it seems like the difference is getting small for some more prosaic tests.
This suggests to me that it's going to be very, very hard to get LLMs to write better than humans. Obviously it's not open-and-shut, but I would presume some of the competitors spent much longer on their entries than Scott spent on the actual Njal's saga review (minus the gimmicky parts that I utterly ignored), or otherwise had more resources at their disposal.
In particular, I think this is a sign that we shouldn't be particularly worried about LLMs ripping off specific authors and being successful at doing so. For one, this demonstrates that it's a pretty tough job to outdo the literal benchmark on a task. But also, a lot of people are complaining about various aspects of Scott's style showing up too much in the book reviews; I'd expect complaints about there being too much Jane Austen in some hypothetical JaneAustenBot's works. At least for me, books by a specific author have clearly diminishing marginal returns.
I wouldn't take a random contest on a blog too seriously, but it's an interesting antidote to some of the more extreme forms of concern about LLMs rapidly becoming superhuman.
Help a friend out:
I'm looking for a forum to discuss philosophy/politics/culture. I already know of DSL ... and not them. Can anyone suggest a forum?
I'm trying to understand partner dances. I was forced to learn some foxtrot back in middle school and I've picked up some country dance as a happy amateur now when I've matured. I'm trying to branch out into some Latin dances now as well. But I don't feel like I "get it".
1. Most (well-known) dances seem to have a "social" scene and a "show"/"competition" scene. To me as a casual observer, the "show" dance seems very far divorced from the "social" dance. I feel like if there's this sport called "golf" and when I try to learn it my teachers tell me to practice hitting this white ball with this club to make it go into a hole, but when I watch a golf competition on TV they're playing soccer. As a beginner, I try to learn the steps, and I try to learn the "feel" and posture of the dance. But are those things even important? Can I show up at a shag dance and dance the whole night without doing any shag basics steps?
2. Related: Often when I watch a video from a dance show, I don't see the dancers do any of the basic moves that I learn as a beginner. Often I wouldn't know what dance they were dancing unless the video title told me. How do you identify dances? Can you do it from the music alone ("this is a bachata song so they're dancing bachata")? Is there's some good flashcard-style way to learn to identify different dances? Like, I get a ten second clip of people dancing and I've to identify it. I guess I should hack it together myself using YouTube... Do dances have signature moves that makes them recognizable (but "advanced" dancers don't do the basic moves per 1., and also couldn't I do the signature move of one dance while dancing another dance just because I'm cool?)? How do I spot what dance someone is dancing?
3. Are there a music theory but for dance? Are there a natural limit to the types and styles of partner dances? Could I make up my own partner dance right now from scratch by just deciding on a couple of steps to a 4x4 beat? Or would what I make just turn out to be rumba with a backstep or something? Most (all?) partner dances seem to belong to a "tradition": do dances naturally split and involve like this? Do they fuse? Are new dances invented from scratch sometimes?
4. Related: Can dances be wildly combined? Could someone just smash together rumba and foxtrot and have a new fusion dance? Would it be recognizable as a rumba+foxtrot fusion? Could such a fusion be done in many different ways? Must the dance I'm dancing match my partner? What happens if I dance shag but my partners dance rumba? What happens if I dance lindy hop to bachata music?
(I know I could get an answer to many of these questions myself by just trying it myself, but I'm interested in the deeper answer that I don't really have the terminology to ask about.)
I know Scott has posted stuff about Buddhism in the past, so I assume there are some "secular Buddhists" here I could ask this question to. I know the Buddhist concept of rebirth is different from the traditional conception of reincarnation, that it's only the karma that gets reborn etc., however, suppose that I am a skeptical materialist that denies even that. If we believe that death doesn't bring rebirth into Samsara, but is a full extinguisment of everything one was, does that make it equivalent to Nirvana? Essentially, if we accept the first noble truth that life is Dukkha, and desire to extinguish that Dukkha, if we don't believe death results in rebirth into this world, then wouldn't suicide be a much quicker and more expedient means than the long and difficult road to achieving Nirvana in this lifetime? Or is there something to Nirvana that is more than just extinguishment, making it still preferable to death even in a world with no rebirth? Please note that I'm not actually contemplating suicide based on this line of reasoning, this is more of an academic exercise and something I'm genuinely curious about, whether Buddhism reduces to implying suicide if the belief in life as suffering is maintained but the belief in rebirth isn't. Essentially, death to an atheist materialist looks a lot like what Buddhists describe as the ultimate goal of paranirvana, i.e. permanently dying and not coming back, so if you think regular death brings that by default anyway, why not just die?
What is the general impression of AI generated art? It seems Aella takes flak for using it in her posts, but I generally like them, and I wouldn't even say her pictures are that effective. Are AI-generated images seen as not classy or something? Is it a generational thing?
How's everyone's experience with meditation? Do people here meditate regularly? Do you get a lot out of it? I did some every day during Covid, and eventually had a kinda-transcendental experience after drinking a beer on a beach where I was extremely dialed in to my surroundings- not, off in my own head thinking about something else. I was just very 'there' and not elsewhere for a little while. Despite that, I got away from regular meditation until recently, when I started up again. I'm now averaging probably 4 days a week, every morning after I get out of the shower and before I have coffee. It's a slow but steady learning curve to get better, though I haven't experienced any benefits outside of the actual meditation yet.
Are other people meditating regularly? Seeing benefits?
Piano players: I want to learn a few pieces nearly perfectly. By this I mean that I'd like to be able to play them in front of a crowd with no stumbles/startovers and *maybe* miss a note here and there. It seems that as I practice, my accuracy asymptotically approaches 90% or 95% and takes a long time to even get close to that.
I am convinced that really good pianists still have to practice a lot but they don't have to deal with asymptotic progress and they can progress fairly linearly up to 100%. At piano recitals, I marvel at the way performers will play multiple pieces without a single mistake.
How do I get there? Is there a way to shoot for 110% so that I hit 100% with less effort? For instance, if I want to lift 100 lbs 6 times, I'd work on being able to lift 120 lbs 8 times. At that point, lifting 100x6 would be almost effortless. What's the analogous approach for piano? Learning to play it in the dark?
Thanks!
Ukrainian forces are slowly breaking through Russia's three concentric lines of defenses in southwestern Ukraine. They've now reached the second line.
Why doesn't Russia respond by building a fourth line farther back? Ukraine's men are moving so slowly that Russia probably has enough time to build it before the Ukrainians reach it.
And if the slow pace continues, what's to stop Russia from building fifth, sixth and seventh defensive lines in time?
Anybody here play Eclipse Phase? I've been looking to join a game and I've been having a devilishly hard time finding anyone interested.
For the fans of forecasting: Main takeways from FORPOL - Czech project exploring the use of forecasting in policy: https://forum.effectivealtruism.org/posts/ZL5MqHLDNP5qTxwvw/forecasting-for-policy-forpol-main-takeaways-practical
Being Human
Contents:
Instinct to Social Morality............... 2
Advancing Humanness.................... 3
Steps to Betterment........................... 4 [TBD]
https://drive.google.com/file/d/1q7l72cRgfZ41545e-xJP0OxzWETySuHK/view?usp=sharing
What would the effect on the economy be if factories were 100% automated and they could shut down without ill consequence during recessions? No furloughed human workers, no increase in unemployment and welfare claims, no families damaged by mom or dad sitting in the living room drunk or high all day suddenly.
I'm looking to set up (play money, for now) prediction markets within my company on private company topics...which means I can't use third party applications, but can (and deeply desire to) use third-party source code to get off the ground. Anyone know of a recommended codebase for something like this? The closer to vanilla use of common languages like html/js and python, the better.
https://www.youtube.com/watch?v=KOJkZbgGIvE&ab_channel=JessicaKellgren-Fozard
A claim that LGBTQ+ people have a higher than average rate of disability. Might be a worthwhile survey question.
With an estimated 35-45 years to live as per every life expectancy calculator I can find and a net worth of approximately $2.5M, how would you devote the remainder of your life to fighting wokeness and/or feminism with maximum effectiveness? If I keep working at my (quite woke) job for the next 5 years, I will probably accumulate another $1-1.5M. Assume no family.
L*bgen being sued: https://arstechnica.com/tech-policy/2023/09/most-notorious-illegal-shadow-library-sued-by-textbook-publishers/
School in canada removes all books published before 2008 because they're not woke enough (paging FdB) - https://www.cbc.ca/news/canada/toronto/peel-school-board-library-book-weeding-1.6964332
More confirmation that Scott was correct about the Ferguson effect - https://www.unz.com/isteve/approaching-35000-incremental-homicides-and-traffic-fatalities-in-2020-2022-compared-to-2018-19-average/
Some metrics seem much easier to Goodhart than others. For example
-easy to manipulate: "In response to Labour targets in the 1990s, one force recorded a 27 per cent reduction in the number that had a target (‘theft from a motor vehicle’) alongside a 407 per cent increase in the number that had no target (‘vehicle interference’)." (https://www.lrb.co.uk/the-paper/v45/n18/john-lanchester/get-a-rabbit)
-presumably much harder to manipulate: homicide versus suicide or accidental death.
Assuming I'm not mistaken, what are some not-too-technical sources discussing this further? For general interest, not AI alignment.
I'm looking for help. Mostly in the form of advice, I guess, though I wouldn't say no to other forms (not money, I still have some to keep me afloat for a while).
There are three aspects to my problems: work, a place to live, and psychology.
1. I'm a scientist from Russia. Not a particularly good one, with a candidate's degree (the lower level in two-level PhD system) and over 25 published papers but my h-index is not even 5. I'm not really keeping up to date with current research in my area, and it's been quite a while since I tried to tackle some serious problem. I can do middle-level work if given a problem to work on, get some results and write those papers. This might have been enough if I was 30 but I'm ~40 now.
My work also involves programming (and I do it for my own entertainment/personal projects), but I'm not up to date on a lot of tools and concepts. It's a bit better in my own area (machine learning) but there aren't many job offers. I think with enough effort I could get myself in shape to be employable in a wider area, but it's hard to put in the effort with uncertain prospects.
Still, I am hoping to find a job I'm somewhat qualified to do.
2. I wasn't a fan of Russia since at least 00s, but never found enough motivation to do something about it. Eventually, I fled to Armenia in late 2022 when they started mobilization (not that my risk of being drafted was high, but I decided to take a safer route). It's nice here but I'm not sure it works as a long-term solution. For one thing, Armenia still has strong ties with Russia and I'm not sure what happens to people like me if Russia passes laws making us criminals. So while finding a decently paying job here would be a nice start, eventually I hope to find a job that would allow me to move elsewhere.
Which is a huge problem because I have zero experience with looking jobs abroad (and negligible experience doing it in my own country, for that matter).
Armenia is also problematic in that it's hard to find a decently paying job here if I decide to switch to something I'm less skilled with just to break the pattern.
3. I don't know what it is with me, but things do not look good. I lack motivation and willpower. With my old low-requirement job it was mostly enough, though I had a tendency to almost miss deadlines whenever things I didn't really like doing (e.g. preparing papers) came up. The actual work of finding things out was more interesting so less problematic. It seems to be exacerbated by me working from home (though I wouldn't give a scout's promise to become a highly performing worker if I had to work from the office, but it should help if I'm at least somewhat interested and can't switch off immediately whenever I encounter something I don't feel like doing). Lately it got worse with the necessity to look for a new job, which I'm scared of. Maybe also because of the overall shitty situation, too (shame and guilt over what my country does with overt support by a lot of people I knew and respected, having to leave my home unprepared and leave behind friends and relatives in what seems to be turning into a totalitarian state). But the end result is that I do basically nothing to advance further towards this goal of a new job - mostly just wasting time, sometimes doing something that is still meaningful and can qualify as "work" but is an obvious way to avoid doing something I want even less. I'm not even sure where writing this post falls, but I'll try to commit to following at least some of the advice I get if there is any.
Is it depression? I'm not sure, it doesn't feel like I imagine depression should. Though lately there have been days when I just wanted to stay in bed, but that was because none of the things I had planned were enjoyable, and I was trying to stop allowing myself doing anything besides them. (Also there's growing sense of urgency so even the distracting things I normally like are less enjoyable. But that's more like a stress.)
So, I'll appreciate any thoughts and suggestions. In particular:
Re 1. Is there any kind of job you think I should focus on looking for? I still have time and money to maybe get additional training (or prepare myself) in some math or programming related topic if it would let me land a job with high probability.
(I probably won't say no to a direct offer too, if everything I wrote hasn't discouraged you.)
Re 2. Particularly if I can get this job in any livable country that wouldn't kick me out to Russia on short notice, for example if Russia decides that missing an appointment with a military registration office is now a crime. Although I feel like I'm just not good enough to look for a job abroad (anyone willing to expand their search this much can probably find better candidates), is there anything I can do about it?
Re 3. I am close to recognizing that I need professional psychological/psychiatric help, but if that is so, will the help help? Do my problems look like something where talking it out or taking some pills is likely to result in a large improvement? Again, I'm constrained with the resources I could spend on this but I can afford giving it a try if the expected payoff is good. Or are there things I could just try by myself?
I am wondering: What book should you never have read?
What are the best free sites that use AI to summarize long articles (such as the book review context stuff)? So far the free GPT's that I have tried do a bad job.
I tried https://tldrthis.com/ , signing in through google to get me an AI (human-like) summary that is detailed/section-wise and asked for summarizing this article:
https://worksinprogress.co/issue/why-we-didnt-get-a-malaria-vaccine-sooner
and it gave a decent summary, but they only allow 10 articles for free.
Is it too silly to hope that something like a wikipedia could be created for slow-reading low attention span folks (like me) leveraging AI to summarize several books/articles? Sorry if I am being too stupid/naive.
Does anyone know a good TLDR or summary of some of Erik Hoel's ideas on neuroscience and consciousness? Any thoughts on it?
Scott, do you have any thoughts on Data Falsificada[1]? The summary is in the first paragraph: "[A] series of posts detailing evidence of fraud in four academic papers co-authored by Harvard Business School Professor Francesca Gino."
[1] https://datacolada.org/109
I posted this in the subreddit, and haven't gotten any response, so hopefully someone here will know something:
Just came across a supposed image pulled from a rabbit's retina in the late 1800's by Dr. W. Kühne, supposedly creating the field of "Optography". While he himself claimed that it wasn't good enough to be of use in forensic science, I'm skeptical that he even managed to do it at all.
Supposedly it's based on the fact that Rhodopsin, a pigment in the retina, bleaches in the presense of bright lights, and can afterwards be fixed through chemical processes. The fixed Rhodopsin can then, according to Kühne, be turned into a recognizeable image (although not detailed enough to be of much use).
I buy the physical claims about Rhodopsin. I'm sure it bleaches as described, and I even buy that it can be fixed in that bleached/unbleached state. I do not believe that any images, useful or not, were actually produced with this method.
Unfortunately, some google scholar searching and a little bit of internet research have failed to turn up very much. It's mostly focusing on how widespread it got in victorian science fiction.
I found one citation from the 70's called "Optography of the retina. Resumption of Küne's studies", but it's just a citation, I can't actually find the article. Wikipedia has a link to an article that claims it was replicated (based on the title, I'm guessing it's not itself a replication, just a report of someone else doing it) but the link goes nowhere and the title doesn't even return anything on Google Scholar.
This has all the hallmarks of being a complete load of bullshit. But I can't find enough evidence to actually prove it.
If anyone has any actually reliable information on this, I would be greatly appreciative.
I recently realised that a spookily-consistent mathematical rule has described the previous political "re-alignments" in the US, and that this rule says the next re-alignment will be in 2028.
The rule is that each Party System lasts for four years longer than the previous one. I'm defining the precise dates of the party systems by my own reasoning, which differs very slightly by how historians may define them, but I think my definitions are justified (and I discovered the rule only after settling on these dates):
1776: The Revolution, and the framing of the constitution, and the largely uncompetitive, pre-party presidencies of Washington and Adams.
1800: The First Party System begins with the election of Jefferson, entirely dominated by what's now called the Democratic-Republican Party.
1828: The Second Party System begins with the election of Jackson, establishing the Democratic Party.
1860: The Third Party System begins with the election of Lincoln, establishing the Republican Party.
1896: The Fourth Party System begins with the campaign (and loss) of William Jennings Bryan, and the election of McKinley.
1936: The Fifth Party system begins with the re-election of FDR, firmly establishing the New Deal Coalition.
1980: The Sixth Party System begins with the election of Reagan, establishing the modern conservative movement.
In brief: the US established; 24 years later a party system established; 28 years later a *competitive* (two real options) party system established; 32 years later the *current* two-option party system established; 36 years later the current parties with their recognisable ideologies (D=left, R=right) established; 40 years later the modern "flavour" of Democratic ideology (big government, unions+minorities) established; 44 years later the modern "flavour" of Republican ideology (evangelicals+business+national security establishment) established. 48 years after 1980 will be 2028.
The only part of the above that seems to differ from how the party systems are normally defined is using 1936 for the start of the Fifth Party System, instead of 1932 when FDR was first elected. But I think this makes sense, because 1932 was not really an endorsement of a new ideology like 1828 and 1860 and 1980 were (or a repudiation of a new ideology like 1896 was) but a rejection of the incumbent over the depression. Also, I think FDR downplayed his big-government plans in his first campaign. Only his re-election really created a new ideological alignment. The rest of the dates seem uncontroversial to me.
So, why would this rule hold? It's too perfect to be a coincidence, I think. But what's causing it? I don't know exactly, but there are many phenomena that have very complicated causal explanations but can still be modeled by a fairly simple rule: that's the way the world often works, and the complex causes end up interacting with each other and canceling each other out to produce (when you zoom out) a simple, straightforward trend. Most likely, there's a general (mostly unconscious) awareness that each party system is coming to an end, and each time enough political will to squeeze one further election out of the old system before it collapses. (Incidentallly, you can even extend the rule back before the Revolution--20 years--to the beginning of the Seven Years- War in 1756, which was the catalyst for the causes of the revolution).
Finally, how will this most likely play out in 2028? I see three posibilities:
1. Desantis (or someone else who isn't Trump) wins the Republican primary and wins the general election. I know this isn't looking good at the moment, but I think it's more likely than people think (though I won't elaborate here) and it gives the clearest path to a political re-alignment. Say it's Desantis, and he returns to a more traditional GOP governance. The rural working class is sick of this and they permanently split from the GOP to form their own Trumpian populist party. The Democrats go into meltdown like they did in 2016, and the socialist faction again blame the woke elite faction. This time there's an actual split: the socialists leave the Democratic Party. Is it possible the Trump populists and the Sanders socialists find common ground and unite to become the Populist Party? I don't know, but it seems perfectly possible. The Trump supporters can now embrace class warfare and economic leftism along with right-wing nationalism to become a truly anti-elite force: Trump can outright adopt a restributionist agenda, truly drain the swamp, no longer have to play nice with the old guard of capitalism. Meanwhile, the socialists no longer have to gently criticise the woke faction: they can go all-out anti-woke, condemn identity politics as viciously as they condemn capitalism, educated elites as viciously as business elites. With the Populist Party established as a powerful force, the fight of 2028 will be between the two old parties, to decide which of them survives as the elite opposition to the Populists, and which one dies. That could go either way.
2. Trump wins the primary and the general. It's less clear how a split happens here, but maybe he decides to this time govern in a more economically leftist way, and the traditional Republicans leave their own party. It's impossible to see them joining the elite Democrats, and hard to see Dem socialists joining the Republican Party. But maybe these things happen in some weird way?
3. Biden is re-elected. In this case, he surely moves further to the left on economic matters, and that keeps the socialists and wokes in the same party for now. The GOP probably permanently splits, but the significance of this is less clear. Maybe a new populist Trump party challenges Biden/Harris from the left in 2028?
What do people think of this? Please don't dismiss the rule I described as a mere coincidence without explaining why it seems to so perfectly describe the past.
What happened to the convenient link in the email, that let a reader click to go direct to opening the page online? Today's message has a "subscribe here for more" link where it at various times had "view in browser" (one link) or "open in app or online" (2 links).
The intermediate state, with "view in browser" at the upper right, had a button with "read in app" a bit lower down. That button is still there, but "view in browser" is completely gone.
Obviously I got to the post in the browser, in spite of the user interface dis-improvement. You haven't made the blog completely inaccessible to those who never do anything on a cell phone they can do with a real computer, let alone to those who simply don't have a cell phone.
But the email interface to this blog is now making a pretty clear statement - we only offer convenience to app-lovers here. Most likely, that statement comes straight from substack, and is perhaps motivated by some idea they have that app users are more profitable. (Perhaps the app violates privacy more than the web site, giving them more to sell?)
You might want to restore the old appearance and links, if substack will let you do so.
If anyone's interested in AI stuff, we might have our first report of Gemini (Google's upcoming model that's supposed to compete with OpenAI's GPT-4).
https://twitter.com/BrianRoemmele/status/1702671064239833415?s=20
"It is equivalent to ChatGPT-4 but with newly up to the second knowledge base. This saves it from some hallucinations."
Plug for meat-eaters who want to find more humanely raised meat: https://findhumane.com/
Were the Crusades bad? I mean, schooling has led me to believe they were bad, but the older I get the more I’m convinced my schooling was a bunch of lies and prejudicial assertions.
On the one hand, one can say that we can’t judge the past at all because their values were too different from ours. OTOH, does that mean we can’t judge Hitler? If we can judge Hitler, then why can’t we judge the Pope Urban II?
Don’t we arrive at our present values precisly by judging the past, and saying no, yes, no, no,yes, no?
If in the year 2023 we can confidently say Hitler was bad then we shold be able to, with some level of confidence, say whether Pope Urban II was good or bad. No?
I say Hitler was bad, and the Crusades were good. Christian culture thrived after the Crusades.
Okay, I'm not usually one to ask for spoilers, but I could use some help with this puzzle.
Jonathan Blow doublecrossed this thread:
https://nitter.net/Outdoctrination/status/1694842310625010144#m
All my priors say something the author is hiding a lie in plain site and my gut says it has something to do with their over/loose use of "average" and "typical". But I just can't place it.
I mean, my god, of course soldiers in war are going to eat 6k calories. And of course pedestrians will need more calories than drivers. But I don't think those answers are really enough to call it the "solution" to the puzzle. Any suggestions on where I can stand to make this all line up?
If I think of an AI as metaphorically a different species than ones that have come before, then it scares me to think that this new species would be more intelligent than humans. Because of course a much more intelligent species is going to makes mince meat out of a less intelligent one.
But if I think of AI as an extension of human intelligence in the transhumanist sense it doesn’t scare me so much.
I think AI is much more likely to be like the latter than the former. What are the arguments against that?
After writing this post in the subreddit last week https://old.reddit.com/r/slatestarcodex/comments/16chk0g/understanding_the_general_apathy_towards_systemic/
I started to wonder how common it is for people to embrace the idea of there being "systems" in the political context, where actions/policy choices have downstream consequences, which could be evaluated and selected for using a consequentialist framework (ie in order to decide if we should pursue x vs y policy, we should consider the anticipated systems-wide aggregate outcomes).
After speaking to a handful of generic liberals on this (ie civil servants, policy professionals, grad school educated pmcs types), I was shocked at how reprehensible everyone I spoke with found this idea. Discussing this concept felt as if I were broaching a highly taboo and inappropriate subject. The primary response I heard was that "you simply can't look at the consequences to understand what we should do" (this response, almost word for word, came from someone who self-advertises as being interested in "systems thinking"!)
I used to believe that many people disliked markets due to issues of exploitation and inequality, rather than objecting to the concept of a market itself. I also understood that some people had reservations about philosophical consequentialism, but I never thought this would extend to political contexts, where it could be considered merely one factor among many to consider.
I'm now very confused and was wondering if anyone here knows of any writing on this topic — specifically any psychology books/articles on why some people do or do not embrace the idea of systems and consequentialist reasoning.
Does anyone have a good argument for an average person buying their house? All the good ones I've heard seems to go against it(undiversified, inefficient, expected returns no better than renting), and yet it remains so popular.
This week I made two magical discoveries and I offer them up in exchange for other whimsical or magical items which you might consider worth sharing:
1. Codex Seraphinianus - an encyclopedia of an imaginary world. (https://en.wikipedia.org/wiki/Codex_Seraphinianus)
2. The Ripley Scroll - a Medieval step-by-step guide for making the Philosopher's Stone. (https://artsandculture.google.com/story/exploring-the-ripley-scroll/sgLCNEuJqIJzLg)
This Arts & Culture feature from Google actually has a ton of interesting pages. The page for Care of Magical Creatures shows a paradoxical two-horned unicorn.
Given that I'm just a lowly commentator, I don't exactly need to post an erratum, but I wanted to admit that I was at least partially wrong about Joe Biden a month or so ago when I said that there wouldn't be any evidence that he was aware of how much more-than-baseline corrupt the Hunter business was. I'd still insist that we don't really know what the baseline level of corruption in Washington is, but my actual claim seems wrong in retrospect.
My wife wrote "It Only Takes 200 Joules To Restart A Heart:" https://bessstillman.substack.com/p/it-only-takes-200-joules-to-restart, which I think is great. I would, though.