I started a blog (never to be paywalled) where I'm publicly sharing parts of my journey to use forecasting techniques to try and mitigate global catastrophic risks.
If this sounds interesting to you, please come check it out. I'd love to interact with other people about this topic, and I'm very much trying to make it a useful resource for you.
Meetup at my house this week, 1970 Port Laurent Place, Newport Beach, 92660
Saturday, 2/4/23, 2 pm
Activities (all activities are optional)
A) Two conversation starter topics this week will be. (see questions on page 2)
1) Janus' Simulators - by Scott Alexander - Astral Codex Ten
2) Twelve Virtues of Rationality - LessWrong
B) We will also have the card game Predictably Irrational. Feel free to bring your favorite games or distractions.
C) We usually walk and talk for about an hour after the meeting starts. There are two easy-access mini-malls nearby with hot takeout food available. Search for Gelson's or Pavilions in the zipcode 92660.
D) Share a surprise! Tell the group about something that happened that was unexpected or changed how you look at the universe.
E) Make a prediction and give a probability and end condition.
F) Contribute ideas to the group's future direction: topics, types of meetings, activities, etc.
Conversation Starter Readings:
These readings are optional, but if you do them, think about what you find interesting, surprising, useful, questionable, vexing, or exciting.
1) Janus' Simulators - by Scott Alexander - Astral Codex Ten
I recently ran across the idea that part of what's wrong with (STEM?) academe is the winner-take-all system for getting credit of discoveries. It's plausible that this makes useful collaboration less likely, but I can't imagine how to do a better job of rewarding collaboration. What might help?
What is the right way to think about menopausal therapy? There is a big piece published February 1, 2023, in the NYT. It argues that the establishment has been too conservative about treatment. Has anyone done a rationalist, statistically literate, literature review, cost benefit, all things considered, deep dive into this issue? Thanks!
I recall someone had created an ACX comment search tool, I think it was based on ElasticSearch. Anyone know if that still exists / have the link? Google is really failing me, here.
Any chance you'd consider turning on likes for comments? I understand the downsides but the comment sections are getting so full bc there are so many of us only children/former precocious needs who love to hear ourselves talk it's getting hard to find the best content.
Yes, it can make a community bad if ppl vote up comments bc they pithily 'own' the other side but I trust the ssc community more than that.
Dumb newbie question: I'm diving into the world of startup funding, and I'm struggling to understand how angel investors have a business model.
If I buy a Mars bar, I can sell it on (and get its socially mediated economic value) or I can unwrap and eat the thing (and get its intrinsic value.) I assume there are existing terms for this distinction; I don't know what they are.
If I buy shares in someone's company, and I give them £10k on the hope of a "10x 5 year exit" (jargon I learned all of five minutes ago) then I now have some pieces of paper, and the company's "valuation" - the money you'd get if you liquidated all its assets - has increased by £10k.
Excepting any dividends, which I have no power to force the company to issue, (and which I've just been told is not what investors are interested in), my shares are just worthless pieces of paper.
In five years, the company has grown 10x as big, and therefore my share of it is now worth £100k.
But the only way to get that value out is to liquidate the company and take my share - and I have no power to force that to happen either.
So my shares are still worthless.
I'm guessing my "exit" is selling the shares on. They are "valued" at £100k, so I sell them to you for that amount and make my exit. That's the socially mediated economic value of the shares.
But there's no intrinsic value to the shares - you can't unwrap them and eat them. You get whatever dividends the company may or may not choose to issue, and that's it. The £100k valuation is entirely conceptual.
Anyone read "Protocol" by Alexander Galloway? I got it for free from someone who didn't want it. It's old (2006), but the the thesis makes it seem like it would be an interesting read, at least from a historical POV:
"Is the Internet a vast arena of unrestricted communication and freely exchanged information or a regulated, highly structured virtual bureaucracy? In Protocol, Alexander Galloway argues that the founding principle of the Net is control, not freedom, and that the controlling power lies in the technical protocols that make network connections (and disconnections) possible."
I'm curious what the people here guesstimate the odds are of a Republican president getting elected in the US sometime within the next 20 years if we assume that nothing catastrophically world-changing such as WWIII happens. (What about within the next 50 years?)
This month, I'm going to do Austebruary (Austerity February, in the spirit of Veganuary or Dry January). My grocery bill has bloated to about double what it was 4-5 years ago (granted, there's another person in the family now), and I feel like it's time to reset the hedonic treadmill. So, the budget for the month is $200 for the household, plus the considerable amount of food already in the fridge and cabinets. Normally, we spend around $800 on food/drinks/coffee/alcohol, not extremely profligate but not frugal.
New unit is the AUS (austere) = kcal of food / $ spent.
Rice, peanut butter, oats > 1000 AUS,
Bread, dried beans, whole milk ~ 500 AUS
Stay tuned for Manic March, in which we blow up the budget and have a lot of fun.
A couple of open threads ago, in a discussion of to what degree charitable donations are tax dodges, there was mention of donated easements on land being a particular example. I commented that it _has_ been an abused practice but that the combination of formal accreditation of conservation land trusts, and new federal legislation regarding the actual tax deduction, was in the process of dealing with the problem.
I mention that again here just because the Charitable Conservation Easement Program Integrity Act recently passed into law as an addition to the year's omnibus federal spending bill. It was the product of some years of work to arrive at bipartisan agreement and is a relatively rare (nowadays) example of achieving that. The new law eliminates the most egregious abuses of the donated-easement deduction and toughens the IRS's review of such deductions generally. Here is a basic summary of it:
How many years, if ever, till AI Narration is comparable to the beste voice actors? What do you think about the skepticism put forward here?
"The thing is that the stuff that is required to narrate isn't included in the text. There is nothing for it to reference in the existing content to bring these elements to bear. It is a billion acting decisions made constantly that are not deterministic. And some of it is accreted over the entirety of the read - the development of a relationship and the tone of emotion over 300 pages. How is AI going to do that?
How can an AI decide which voice to use? A human can tag it - and tag things throughout the text to flag them for the AI to perform in a certain way - but pretty quickly that becomes more time consuming/expensive than just paying a narrator to read it."
"I KNOW that AI will get better. It will sound increasingly natural. But I don't think it can act without direction. And once you have to provide it direction with this granularity, that requires non-AI work to provide the context to do that. "
Towards the end of the video, that when Julie was deep into Monat, she couldn't grasp arguments about how it was structured so she *couldn't* make money. This is the part which might be of most interest to rationalists. How can you tell when you're capable of using your mind?
Some people were asking about MLMs (multi-level marketing)-- here's a recent example, an interview with a woman who got out.
This is specifically about Monat, a shampoo MLM, but MLMs are pretty similar to each other.
Some highlights-- around 22 minutes, something that was actually able to shock me. They're encouraging their sellers to share their personal history of trauma and offer the MLM as a way of recovery.
Around 32 minutes, they're baptizing people as part of events.
At some point, an explanation that MLMs hijack the belief that effort pays off (true in many parts of life) to hook people into systems where they can't win.
Previously we had a mention of the Catholic Church declaring black cats should be banned, and this came down to a misunderstood meme making the rounds.
Fortuitously, History for Atheists has a video about this up on Youtube:
When I think of District Attorneys, I think of people like Janet Reno and Rudy Giuliani and similarly horrible folk in my own region. Are there any good DA's? Why are the famous ones always horrible people? (Some of those famous ones got so with good reputations, so you can't say it's a matter of mere infamy.)
Any advice for finding remote positions in AI safety or EA?
I'm a physicist-turned-data-scientist, currently on the job market. My spouse has an extremely stable position, so I have the luxury of working in something less lucrative, but with high impact. I'm open to work in EA or AI, with the caveat that it needs to be a remote-first position. I poked around LW, but didn't see any discussion.
Due to my work I had access to a set of medical data covering 100M people. After reading the original post about it I was curious to its rarity. And found on the order of 10 cases.
So while it no doubt exists in higher numbers a *very* slim number of people have it to the point they seek medical assistance. Out of respect for these people's privacy I didn't investigate other than looking at a count, but I do wonder how much of that was incidental to the original cause of workup.
Unsurprisingly this can be very mild or very severe. I thought it was telling that people who commented on that original post typically mentioned a very mild impact "I can make the lines move if I think/focus in a certain way" type of effect. Meeting the criteria for HPPD? Sure, but definitely not something you'd mention to a physician.
I've really loved the rise of public-facing blogs / newsletters on relatively specialized technical topics. Some of my favorite examples:
- ACOUP (gears-level understanding of premodern military and economic history)
- Admiral Cloudberg (causes and investigations of aircraft crashes)
- Bits About Money (payments technology and business practices)
- Construction Physics (what it says on the label)
- Money Stuff (explaining finance news)
I'm thinking (as a hobby-- I have a day job) of trying to develop enough understanding of an unfamiliar area to do my own writing on it, along similar lines to the examples above. So I'm asking for some brainstorming help: what topic(s) would you enjoy reading a technical but reader-friendly weekly-ish blog about? Only constraint from my side is that it can't require PhD-level expertise or industry-insider knowledge.
(In 2022 I got a completely remote job and, living the meme of every Californian, traveled and lived in 5 different cities to see where I wanted to move to. I thought I’d write up what I found)
--Detroit--
I’ve had a lot of people ask why I would even consider Detroit, so let me go over the things Detroit gets right. Go to your map app of choice and look up Grand Circus Park in Detroit and look up what’s around it. In mid-2022 there were multiple nice apartments you could rent for less than $2000 a month within 2 blocks of this park. That means you’re walking distance from the baseball stadium, the football stadium, the opera house, Fox theater, the river, the hockey rink, and multiple outdoor parks. This area is premium; a lot of work has been done to clean up this area, it’s really nice and there’s lots of decent outdoor dining. And I adore the architecture, every other building is this gorgeous turn-of-the-century design that’s just fantastic. You can happily walk around downtown Detroit for hours and just enjoy the beauty. No joke, based just on infrastructure, Detroit was the best city I visited and it wasn’t close. If you want an affordable, walkable urban city, Detroit has it.
Having said that, Detroit is the only city I am not considering moving to because it has the worst vibe of any place I’ve ever been. The city has PTSD and it’s utterly tragic, so let me explain through four stories. First, if you arrive in Detroit on a Saturday, like I did, the city will absolutely enchant you; it’s packed full of people having fun and outdoor festivals and it’s really moving. And then, like around 4PM on Sunday, something will start to feel weird and it’ll bother you and then you’ll look up and…no one is on the street. You’ll look behind you, down three blocks of the urban heart of Detroit, and they’ll be one person walking and two cars on the road, then you’ll look down the street in front of you and they’ll be no one walking and three cars on the road. The city just dies and I’ve got the pictures to prove it.
But if you’re intrepid like me, you go to a local community garden to volunteer, because the urban blight has opened up a ton of land for large urban gardens and they’re actually pretty nice. And I show up and there’s like four old ladies and I figure, alright, that’s a little lame, but at least they’ll talk my ear off and I’ll learn all about Detroit. Dude, no. They didn’t talk. Like, if I asked questions and drove the conversation, they’d tell me some stuff but it was 70% me talking. And then I went to Comerica Park to watch a Tigers game, and the park is fantastic, and I started talking to these two Canadians, they’re super friendly, and I ask them about this, how Detroit is so awesome and yet everyone, especially the people who live there, are so down and closed off. And they laugh and start telling me horror stories about where their grandparents were in the 60’s Detroit race riots and how they just barely survived.
So yeah, that’s the vibe. If you’re from Detroit, I’m sorry, but your old ladies don’t talk, your Canadians tell race riot horror stories, all your restaurants are closed at 7:00 PM on a Wednesday, and there’s a CVS on Woodward Ave that I don’t think was open once in the two weeks I was there. That’s what I mean by vibe; people act like they’re in a warzone and no one wants to live around that.
So, because I know people are going to bring this up, the issue isn’t crime. Detroit has a crime problem, no doubt, but I felt safer in Detroit than I have in Stockton, in Oakland, and in the worse parts of Sacramento, and that’s just in North Cali. I’ve walked the streets at night and I’ve walked from the Eastern Market to the downtown and done things I wouldn’t have felt safe doing in Stockton. Yes, no doubt there are neighborhoods to avoid but I’m not in those neighborhoods and I don’t want to be. People aren’t just avoiding those neighborhoods, they’re avoiding a gorgeous downtown that just isn’t that dangerous and I fundamentally don’t get it. Detroit has a crime problem but it doesn’t have a uniquely bad crime problem and if you’ve lived in one of a half dozen West Coast cities, you’ve experienced the equivalent. This isn’t crime, this is some kind of weird PTSD where really horrible things happened in the past and they just can’t get over it.
So no, I’m not seriously considering Detroit. I wish I were and, to be blunt, it’s not the city, it’s the people. An entire major metropolitan area without optimism, energy, or enthusiasm; who want to be around that?
And in all seriousness, I think the greatest and easiest act of urban renewal you could do in America today is find some way to get 3-5k Californians or Texans or just…normal people into Detroit. Because there’s a real first mover problem. Nobody in their right mind would want to move to Detroit on their own but if there were a couple thousand normal people who do basic stuff like eat out on a Wednesday, so there would be restaurants open on a Wednesday at 7:00 PM, and host basic stuff like board game nights and river walks, then Detroit would be the coolest city I visited and I’d probably be moving there. But Detroit badly, badly, needs a social scene of ordinary happy people and until they find a way to get that, no amount of urban renewal is going to make a difference because it doesn’t matter how pretty the dance floor is if the natives refuse to dance.
What do rationalist types mean/what are they referring to when they say “alpha”? As in “There is no alpha left in western interpretations of Buddhism,” from the house party essays. (This is hard to Google because of the more popular slang meaning of “alpha.”)
A question for Scott specifically, to help settle a Manifold Market question for me: could you confirm this to be your mastodon account? https://schelling.pt/@scottalexander
Suggestion: Every open thread you say, "95% of content is free, but for the remaining 5% you can subscribe here." which is good but you could add something like "Recent subscriber only posts include post A and post B. You can see a full list of subscriber only posts here."
The reason I suggest this is that I think many people don't realize that you have a substantial number of subscriber only posts and this would be a non-naggy way to let people know.
Also a list of all your subscriber only posts would be useful to long term readers who become subscribers and want an easy way to browse all the posts they couldn't read before.
To all the magicbuilding nerds and fans of Brandon Sanderson: I'd like to share an essay I wrote on allomancy! It includes an analysis of the system and an expanded set of allomantic powers I made for fun.
Just to be clear, I am not asking for advice about whether to let others do as they please. I'm also not looking to explain the CYA phenomenon as it relates to public reliance on expert recommendations for extreme caution around some health issue or other.
I am only interested in hearing from people who can supply research or professional anecdotal evidence to support a claim that giving dogs cooked mammal bones is dangerous, and in particular, that it is so dangerous as to warrant not giving them cooked bones at all.
Why aren't there more music acts like Jojo Mayer's Nerve (https://youtu.be/SIahbUYxdfE) that perform EDM/House/etc. with live instruments? (Or at least, live drumming?) It feels like it should be a sweet spot—all the thumping goodness of the genre/rave experience, but with the added thrill of seeing it created entirely live, the heightened exchanges of energy between the performer and the audience, etc.
But i.e. at like Burning Man this year, there were no acts even remotely like this, just normal DJs, and Nerve themselves don't seem to play at similar festivals or even raves consistently. What gives?
• Maybe people are really there just for a specific aural experience, and don't even really pay attention at all to how the music is being produced. It's not at all like, say, a Beyoncé concert to them, it's more about the drugs they're tripping on, the light shows and the huge subwoofers, the people they're with. (In this case, why hire DJ's at all, then? Or is the position of "DJ" a kind of minimum-viable source brand differentiation?)
• It requires a lot of talent/skill/practice to perform that genre convincingly well. Not only in playing your instrument, but in honing your sound so that your live instrument convincingly mimics the style of electronically-generated music. So maybe the sample size of musicians in general who would even try is small. Moreover, anyone who can play at that level probably finds other kinds of music (Jazz, etc.) more personally appealing to play/build a career on, even if they could make more
• People *would* be more entertained by live musicians and the overall experience would be rated as higher-quality, but not enough to offset the increase in costs. Maybe it's just easier touring with a laptop than schlepping an entire drum kit around the world? This probably entails all kinds of costs that get passed on to venues, who, all things considered, would rather stick with the cheaper, more predictable formula DJs offer.
> ... psychedelics continue to hallucinate for months or years ...
It's worth noting that in a large portion of cases the symptoms reported are mild, sometimes pleasant, visual changes (e.g., "when I look at monochrome surfaces, they look sparkly"), not hallucinations as in "there's a pink elephant in this room!"
An assumption I've seen made, I think, is that GPT-like text generators are capable in principle of generating true and novel information about the physical universe, based strictly on textual I/O. This is because natural language encodes true information about reality, and a sufficiently powerful language-predictor would be able to extract all possible knowledge about reality from words and use it to develop whatever physical theories or technologies it needs to pursue its goals, paperclip-related or otherwise.
If that's the case, then it should also be possible to train a GPT-like generator strictly on text from before 1900 or so, and prompt it to produce the equations of General Relativity. (The same would go for producing Newtonian physics from pre-1650 text, or Kepler's planetary motion from pre-1550 text, and so on.)
Is that a usefully testable prediction? Could GPT-3 or any other near-future GPT accomplish those tasks, sometime before Final Ultimate GPT comes along and generates all possible knowledge instantaneously?
Is it possible to have daily withdrawal effects from SSRIs?
For background, I have taken Prozac, Paxil, and Celexa with terrific results in the past (my issue is anxiety, particular panic disorder, not depression). I went off SSRIs for a few years when I had my son, and then went back onto Paxil when I started having pretty bad OCD-type anxiety. This anxiety continued without much improvement for two years. I switched to Lexapro with little change.
One thing I have been struggling with for those two years was that around dinner time I would develop low-blood sugar type symptoms. I would shake, feel very anxious, feel dizzy, etc. I went to my doctor and got tested and everything was normal. Having these symptoms every day made my mental anxiety a lot worse.
Two weeks ago, I decided to switch the time I take my medication from before bedtime to lunch time. And now those low-blood sugar symptoms are gone, along with my anxiety. My hypothesis is that these symptoms were occurring as the medication was leaving my system, especially as they will come back if I am an hour or so late in taking my pill.
Has anybody else ever encountered this? I'm just surprised my system has become so sensitive, though I am extremely happy to have that anxiety gone.
A few months ago I ran into a weird quandary. My neighbor has a large black lab, and our dog had recently died. We occasionally have a soup or steak bone to dispose of, so I would throw it in the freezer and then offer it to my neighbor when I ran into her. However, the last time I offered, she politely declined, explaining that her vet had advised her that she should never under any circumstances give her dog cooked bones.
This surprised me. My family has had pet dogs from time immemorial. They have always been given ruminant bones, and the bones were always cooked. The only veterinary advice on this subject that I had ever been aware of was that dogs should not be given cooked chicken or turkey long bones, as these could splinter and cause digestive injury, but that necks and cartilaginous stuff was fine. Also, I knew someone whose father in Israel raised dogs and fed them uncooked chicken, including the bones.
Our last dog, a husky Samoyed rescue whose IBD led us to feed her homemade food exclusively, lived to be 17. We gave her cooked beef neck bones probably every two or three days, and cooked chicken necks were a regular part of her daily diet.
So I hopped online and looked up the veterinary advice around bones, and my neighbor is not crazy: it is absolute doctrine that one must never, ever give a dog a cooked bone because the risk of injury is "just not worth it." (This risk/benefit analysis does not appear to have involved consultations with dogs.) Moreover, one should ideally not give a dog raw bones, because the risk of bacterial contamination is also not inconsiderable. In short: really, just don't give your dog bones at all.
From what I can tell, literally no vets allow for the possibility that cooked bones are not dangerous for dogs, but there seems to be nothing solid behind this opinion. I could find no published papers on the question. Nothing quantitative, nothing anecdotal. Like, there's no list of horrific accounts from vets who have had to repair impaled dog esophaguses or stomachs. No accounts of feral dogs living off of food waste showing up in shelters with splintered bones lodged in their throats and bowels.
But perhaps the evidence is just buried in clerical archives sonewhere; i.e., my research into the dangers of bones is hampered by not having access to veterinary journals or other relevant sources.
Has anyone else run across this question and attempted to evaluate the available evidence?
(We have a new rescue, BTW. Guess what cleared up his gingivitis.)
I want to share a GPT-3 side project I'm working on called Pickaxe. It lets you put a front-end on top of any useful or clever prompts and turns parts of the prompt into variables so that other people can run it with their own inputs. The prompts are shareable and embeddable. Looking for feedback from early users! All comments welcome: https://beta.pickaxeproject.com
Can anyone give me a good series of arguments against high-skilled immigration in the US? It seems obvious to me that vastly more high-skilled immigration would be a great thing for America- we could offer Green Cards to the best and brightest from every other country around the world. It would help cement America's dominance as the world's technological & scientific power at a time when that's under threat- after all, the US only makes up 4% of the world's population, so just numerically there should be a tremendous number of highly intelligent people in the other 96. My understanding is that the US was not a particularly accomplished country in the sciences before letting in a huge number of European scientists fleeing WW2.
And the 'they won't assimilate' arguments rings hollow to me- both Canada and Australia seem to manage just fine with a higher percent of their population being foreign born.
However, in the interest of intellectual discussion and broadening my horizons (and because I know ACT has a number of right-leaning folks)- I'm open to hearing good arguments against greatly increased high-skilled immigration. To be clear this is not an argument for *low* skilled immigration. Would love to hear the anti perspectives, partially because I genuinely cannot comprehend what they would be
EDIT: So far we have 3 out of an initial pool of 5. Also I think if you sign up you're committing for 1 cycle. That is, to read and give comments once a week for up to 4 other people, with the assumption that you'll get one submission yourself. We'll see where we are at after the cycle and make further plans from there.
--
I've been writing a blog but I think I probably need an editor. I'm not sure the best way to go about this because I think I'm unlikely to hire a professional. I'm a little shy about asking friends, and I'm not sure that I have any that are particularly qualified / want to be editing in their spare time. I'm curious if anyone has general advice here. Do most substack writers have editors? Are they friends/ professionals or is something else going on?
One idea I've had is to try to find other people in the same boat as me who are interested in getting feedback on things they've written. A classic writers group. Is anyone interested in joining a group of 2-5 and giving feedback on say 1 member essay a week?
Can anybody recommend a good overview of the current state of self driving cars?
I'm interested in the legal status in different countries, the definitions of different "levels" of how autonomous they are, what kinds of roads they can drive on, whether progress has generally been faster or slower than people would have guessed, say, 5 years ago, predictions for the future...
Format isn't a deal breaker but my preferences are text > audio > video. Ideally it would be aimed at laypeople but assume its audience is pretty smart.
Most importantly, it needs to stand alone. I've not really been following the field at all, so I want something to bring me up to speed, not something that recaps the last six months but assumes I know what was going on six months ago.
I wrote about how -- according to campaign contributions, party registrations, and survey data -- Democrats outnumber Republicans by about 10:1 in our administrative agencies, and by vastly more at our prestigious cultural institutions (elite universities, elite media companies, Netflix/Twitter/etc, ...). I make the argument that this is because the absurd tenure protections given to our huge deep state produce the sort of employee who cares much more about aligning with cultural causes than about aligning with elected leaders.
I then make the argument that this is basically how single-party states have wielded power: they still had somewhat fair elections, and somewhat reasonable constitutions, and somewhat popular legalized opposition parties, and so forth; but the people in charge of these countries were party apparatchiks (not official heads-of-state), and they wielded power by dominating among employees at cultural institutions and government agencies.
It's worth noting that stereotypically "evil" corporations are no counterweight to this: military contractors, major banks, big pharma, energy companies, etc, all seem pretty much politically balanced between Dem and GOP employees. Intuitively, this is because they optimize on producing particular items, and so don't have the slack to optimize on producing particular cultures; for similar reasons, when they lobby, it's by "corruptly" paying dearly for isolated loopholes, rather than "respectably" pushing extreme/broad ideological campaigns.
TLDR: The articles that say "Top 5 Best Mutual Funds to Buy!!" are mostly just picking the ones that got lucky in past, and for long-term performance you're better off putting your money in an index fund.
Here is my book DNA, meaning, those texts have shaped my perspective fundamentally, in chronological order. Would be interested in seeing some of yours.
George Orwell, '1984'
Ken Wilber, 'The Good, the True and the Beautiful'
Sun Tzu, 'The Art of War'
Master Eckehart, 'Beati sunt pauperes spiritu'
Plato, 'Collected Works'
Nietzsche, 'On the Genealogy of Morality'
An obscure treatise on Germanic graveyards of the Migration period
Dan Barry, 'The Lost Children of Tuam'
John Stuart Mill, 'On Liberty'
Adam Smith, 'The Wealth of Nations'
Julius Caesar, 'De bello civile'
Xenophon, 'Anabasis'
Oliver Sacks, 'The Man Who Mistook his Wife for a Hat'
I've been wondering about what the long-term equilibrium price for housing should be (you can probably guess why that topic's on my mind).
Because houses are durable, persisting goods (i.e. capital goods), shouldn't there be a long-term tendency for the supply to build up over time, and the price to trend downwards?
Obviously demand for finite land also plays into prices, and houses do deteriorate etc.
Could you construct a model of the housing market, maybe one with a stable or falling population size , an infinite land supply (everyone works from home on an endless plain or whatever), inelastic and constant demand, extra durable construction techniques etc.
Could you construct a model like that where the price for houses tends towards zero? One that was reasonably close to reality.
In the long run is there a force pushing house prices down?
Are there any examples of persisting goods where the supply has built up over time and they're essentially free now?
Many people say that eating processed food is bad for you. What does the word processed mean in this context? I imagine that on the scale where 10 means "heavily processed and thus very bad" and 1 means "not processed at all and thus healthy", different foods go approximately like this:
A burger at mcdonalds - 10; sausage - 8; a frozen pizza - 8; meatballs - 6; a homemade pizza - 5; cheese - 4; a fried fish - 3; milk - 3; homemade popcorn - 3; green salad with balsamic sauce and oil - 2; green salad without any sauce or oil - 1.
I've been in a relationship for some years now and I'm reaching the point where it seems like we should either get married and start a family or break up. I'm inclined to the first option (yay!) but it's a decision to take with some consideration. Does anyone know a good exercise or checklist to help me think through this decision? Like, what questions should I consider before I propose? I've seen some lists of topics for couples to discuss together, I'm looking for a list that I can work through alone.
(Inb4: Yes, I will talk to close friends and family. Yes, I will do what my hearth tells me in the end.)
17 % on Ukrainian victory (up from 15 % on November 20)
I define Ukrainian victory as either a) Ukrainian government gaining control of the territory it had not controlled before February 24 without losing any similarly important territory and without conceding that it will stop its attempts to join EU or NATO, b) Ukrainian government getting official ok from Russia to join EU or NATO without conceding any territory and without losing de facto control of any territory it had controlled before February 24, or c) return to exact prewar status quo ante.
47 % on compromise solution that both sides might plausibly claim as a victory (up from 45 % on November 20).
36 % on Ukrainian defeat (down from 40 % on November 20)
I define Ukrainian defeat as Russia getting what it wants from Ukraine without giving any substantial concessions. Russia wants either a) Ukraine to stop claiming at least some of the territories that were before war claimed by Ukraine but de facto controlled by Russia or its proxies, or b) Russia or its proxies (old or new) to get more Ukrainian territory, de facto recognized by Ukraine in something resembling Minsk ceasefire(s)* or c) some form of guarantee that Ukraine will became neutral, which includes but is not limited to Ukraine not joining NATO. E.g. if Ukraine agrees to stay out of NATO without any other concessions to Russia, but gets mutual defense treaty with Poland and Turkey, that does NOT count as Ukrainian defeat.
Discussion:
This update is based on recent decision to supply Ukraine with German and American tanks. Nobody claims that deliveries promised so far are sufficient to ensure Ukrainian victory, but nevertheless, that decision might prove to be an important juncture.
Ukraine’s allies are divided between those who want to supply Ukraine with everything it needs to win the war and those who are... not sure, I guess. Many postcommunist countries and Britain are in the first group (see also The Tallin Pledge: https://www.gov.uk/government/news/joint-statement-the-tallinn-pledge), while Germany and the US are in the second. But the decision to supply tanks increases the likelihood that they will over time move to the first group.
It also imho increases chances that war will continue into 2024; I now think that those is over 50 %.
*Minsk ceasefire or ceasefires (first agreement did not work, it was amended by second and since then it worked somewhat better) constituted, among other things, de facto recognition by Ukraine that Russia and its proxies will control some territory claimed by Ukraine for some time. In exchange Russia stopped trying to conquer more Ukrainian territory. Until February 24 of 2022, that is.
A couple of questions to people who think that their religion is compatible with rationality.
What would you consider to be the strongest arguments for the non-normative beliefs of your denomination/creed being true? Why do you think that all the other denominations/creeds don't have comparably strong arguments justifying their beliefs (given that you don't adhere to any of them instead)?
Among respondents who had previously lost at least 10% of their body weight, the mean regain was 58% of the weight lost.
As one might expect, the mean regain was lower for respondents who had lost weight recently. A graph of regain against time is very noisy, but seems to reach a plateau after about 6 years.
For respondents who lost weight at least 6 years ago, the mean regain was 72% of the weight lost. This conceals a lot of variation. 19% of this group had regained less than 10%; 21% had regained 10-50%; 21% had regained 50-100%; 27% had regained 100-150%; 12% had regained >=150%. This is consistent with other evidence that about 20% of people who lose 10% of their body mass are successful in keeping the weight off: https://academic.oup.com/ajcn/article/82/1/222S/4863393. The linked paper is also consistent with the above suggestion that people who successfully keep weight off do so by making some effort.
VI: Weight loss
2,619 respondents answered affirmatively to the question “Have you ever lost 10% of your weight, on purpose, through diet and exercise.” This is 38% of all respondents.
In the main public dataset, Scott has removed the question which asked how people lost weight, but it was included in the dataset attached to the round 1 prediction contest answers. This was presumably an error, so I feel a bit guilty about looking at it, but I had already looked at it before noticing that it was omitted from the main dataset and I can’t unsee it. I don’t recognise anyone (even myself) from their responses and I’m confident that the summary I’m about to provide is fully anonymous.
In that subset, 335 people answered the question about how they lost weight, which was conveniently a sufficiently small number for me to code each response. Reference to “respondents” in the rest of this section are to those 335 respondents.
83% of respondents mentioned diet (either generically or by describing some particular change to their diet). 50% mentioned exercise. 11% mentioned some other factor, often to do with a change in their life circumstances. 9% of respondents mentioned exercise alone, suggesting that at least some people are able to outrun their forks. See https://pubmed.ncbi.nlm.nih.gov/34519717/ for yet another study confirming that exercise induces some weight loss, but less than expected: the authors conduct a careful analysis with double-labelled water and metabolic chamber observation, but fail to find a clear explanation.
Unfortunately, many respondents wrote something like “diet and exercise” or “I followed a strict diet” without giving any indication of what they did to change their diet. It’s not safe to assume that these people followed the same pattern as people who did identify a particular diet (for example, they may have been more likely to mean portion control and less likely to mean intermittent fasting), but without any information, there’s not much we can say about them.
Of the 204 respondents who mentioned some specific dietary intervention, 37% mentioned counting calories. 19% mentioned fasting (not counting OMD). 14% mentioned following a low-carb diet. 12% mentioned eliminating or reducing sugar consumption. 11% mentioned keto. 7% mentioned portion control. 4% ate one meal a day. 3% mentioned following a high protein diet. 2% mentioned eliminating alcohol consumption. 2% mentioned paleo. 0.5% (i.e. 1 person) mentioned following a whole food diet. 13% mentioned some other specific diet (several mentioned potatoes, which I hadn’t expected when I started coding the responses).
I note that counting calories and low-carb diets are well represented here, as they are in the survey of current dietary restrictions, albeit that there is a difference in relative frequencies. On the other hand, whole foods diets are hardly represented at all, despite 7% of all survey respondents currently being on a whole food diet. There are various possible explanations, but I suspect the obvious one is right: the whole food diet isn’t effective for weight loss (although it may be for other purposes) because it doesn’t restrict overall food intake.
Although keto is only 11% of the weight loss sample, that is still a lot more than the 2% of general respondents who are currently on a ketogenic diet. A possible explanation is that keto is fairly hardcore, so it’s more tempting for people who are trying to lose >=10% of their body mass than people who are aiming for something more modest.
This seems broadly consistent with the results of other studies which tend to show that basically everything works as well as everything else for weight loss (subject of course to a calorie deficit being achieved): see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2763382/ as to macronutrients and https://pubmed.ncbi.nlm.nih.gov/26384657/ as to intermittent fasting vs continuous restriction.
VII: Conclusions
Owning a car is bad for your waistline. Being a vegetarian or vegan is good. More or less any restrictive diet can be effective to lose weight if you stick with it, so you should choose whatever works for you. On the other hand, eating whole foods is unlikely to make you lose weight. Having lost weight, you need to make an effort to maintain: if you revert to your old habits, you’ll revert to your old weight.
So, I’ve been looking at the BMI data in the ACX survey.
I: A caveat
We should keep in mind that these are self-reported BMIs. Self-reported BMIs are widely observed to be lower than measured BMIs, because people both over-report height and under-report weight. Overall the Health Survey for England found a mean difference of 1.1 between BMI from self-reported height and weight and BMI from corrected height and weight, but this varies by sex and (particularly) age.
There may also be issues specific to the ACX survey. The question appears in the middle of a long and diverse questionnaire. It seems probable that some respondents did not stop to weigh themselves at the time of answering, but used their last known weight or even guessed their BMI. The reported BMI curve has a sharp peak just below the cut-off for overweight, which seems suspicious.
If the measurements are systematically biased, that won’t matter for comparing different groups, but this won’t necessarily be the case: some people monitor their weight more carefully than others.
II: Distribution
The BMI median and mode (to the nearest integer) are each 24 (12% of respondents had this BMI). The mean is 25 (although this may be distorted by some impossibly small or large responses). The curve looks like an inverted V with a long right tail.
Applying the usual categories, 3% of respondents were underweight (<18.5), 54% were a healthy weight (>=18.5, <25), 30% were overweight (>=25, <30) and 13% were obese (>=30). For brevity, I will henceforth write this as 3/54/30/13. For comparison 38% of English adults were overweight in 2021 and 26% were obese (although only 34% and 20% respectively from self reported heights and weights). For some reason, I am unable to find data on how many English adults were underweight, but I believe it will be c.1%.
III: Associations
The survey asks about various other things, which may or may have been intended as possible factors associated with weight. Of course these only give us correlations and don’t tell us which way (if at all) the causation goes.
Those who have lost 10% of their weight at some time generally report being heavier (1/38/40/22) than those who have not (5/64/23/8).
The many people who never get acid reflux (4/59/27/10) are lighter than the few people who get it frequently and severely (1/41/38/19). Obesity is an important risk factor for GERD: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3920303/.
Respondents tend to indicate that they like carbs more than the average person, so either ACX readers are unusually pro-carb or we are badly calibrated as to how much the average person loves carbs, but in any case the small number of people who dislike carbs are lighter (4/62/26/8) than those who most adore them (3/50/30/16). All the other categories are close to the global average, and I doubt that much is going on here.
People with naturally good vision are a bit lighter (4/60/27/10) than people with any sort of sight correction (3/51/31/15), with no notable variation between laser surgery, glasses or contact lenses.
Those following a calorie-controlled (1/41/39/19) or low-carb (2/45/36/17) diet look heavy. The small number of people following a keto diet (4/41/31/25) have an unusually high proportion of obesity, but the sample size is too small to read much into it. These cases seem readily explicable by the hypothesis that overweight people are more likely to adopt a diet for weight loss.
On the other hand, people following a whole food diet (3/57/26/13) look much the survey population overall.
Burping makes no difference. It’s inclusion here in the survey makes it look as if Scott thought it might, but I’m not sure why. Hypermobility also doesn’t look interesting. Specifically Ehlers-Danlos goes 0/75/25/0, but it’s 24 people, so I don’t read anything into it.
People who own a car (2/48/34/16) are heavier than those who don’t (6/64/22/9). People who commute by car look heavy, but the effect is fully explained by owning a car. People who commute by bike (4/67/23/6) look light. There are 250 people who own a car and commute by bike and they go 2/63/26/9, so that seems potentially a good way of neutralising the effect of car ownership on weight.
IV: Restrictions
Scott asked about various possible dietary restrictions. 65% of respondents left this blank, which either means they observe no dietary restrictions, or they observe a dietary restriction not listed, or they skipped the question. 13% of respondents stated they were vegetarian for this purpose (there was also another question about vegetarianism) and 4% stated they were vegan. 9% stated they counted calories; 8% were on a low-carb diet; 7% were on a whole food diet; 2% were on a ketogenic diet and 1% each were on a paleo, carnivorous or low-fat diet.
It is interesting to note here that the battle against low-fat diets is essentially won. Paleo, carnivorous and ketogenic diets are also rare: in retrospect these could all have usefully been collapsed into “some other dietary restriction”, which would also allow the potato people to announce their presence. Vegetarianism is common, but probably usually practised for ethical rather than health or weight reasons. The common health restrictions seen in this question are calorie counting, low-carb and whole food, in approximately that order. However, there was no option for fasting/OMD, which for other reasons I believe to be fairly common.
Among respondents who were overweight or obese, 65% left the dietary restriction field blank, which perhaps surprisingly is the same proportion as overall. This group were less likely to be vegetarian (9%) or vegan (3%) and more likely to count calories (13%) or be on a low-carb diet (10%).
Among respondents who are now a healthy weight, but who have lost at least 10% of their body weight in the past, only 50% left the dietary restriction field blank. This group was more likely to be vegetarian (16%) or vegan (7%) and also more likely to count calories (14%) or be on a low-carb (13%), whole food (11%) or ketogenic (4%) diet. In fact this group had a higher prevalence of every dietary restriction than either the overweight group or the whole population. The group who are now a healthy weight but did not report having lost weight in the past was slightly more likely to be vegetarian (17%) but otherwise also had a lower prevalence of all restrictions than the former group.
I tend to interpret this as showing that people who have in the past made an effort to become a healthy weight tend also to make an effort to maintain that weight, whereas as people who have always been a healthy weight do not (although they are unusually likely to be vegetarian or vegan).
Among the select group (see below) who lost weight at least 6 years ago and have since regained less than 10% of it, only 44% left the dietary restriction field blank. 15% of this group is vegetarian, 7% is vegan, 16% count calories, 21% are on a low-carb diet and 14% are on a whole food diet.
I started a blog (never to be paywalled) where I'm publicly sharing parts of my journey to use forecasting techniques to try and mitigate global catastrophic risks.
If this sounds interesting to you, please come check it out. I'd love to interact with other people about this topic, and I'm very much trying to make it a useful resource for you.
https://damienlaird.substack.com/
We are having another South Bay SSC meetup this Sunday:
http://www.daviddfriedman.com/SSC%20Meetups%20announcement.html
OC LW/ACX Saturday (2/4/23) Chat GPT simulates safety and 12 virtues of rationality Hi Folks!
I am glad to announce the 17th of a continuing series of Orange County ACX/LW meetups. Meeting this Saturday and most Saturdays.
Contact me, Michael, at michaelmichalchik@gmail.com with questions or requests.
Meetup at my house this week, 1970 Port Laurent Place, Newport Beach, 92660
Saturday, 2/4/23, 2 pm
Activities (all activities are optional)
A) Two conversation starter topics this week will be. (see questions on page 2)
1) Janus' Simulators - by Scott Alexander - Astral Codex Ten
2) Twelve Virtues of Rationality - LessWrong
B) We will also have the card game Predictably Irrational. Feel free to bring your favorite games or distractions.
C) We usually walk and talk for about an hour after the meeting starts. There are two easy-access mini-malls nearby with hot takeout food available. Search for Gelson's or Pavilions in the zipcode 92660.
D) Share a surprise! Tell the group about something that happened that was unexpected or changed how you look at the universe.
E) Make a prediction and give a probability and end condition.
F) Contribute ideas to the group's future direction: topics, types of meetings, activities, etc.
Conversation Starter Readings:
These readings are optional, but if you do them, think about what you find interesting, surprising, useful, questionable, vexing, or exciting.
1) Janus' Simulators - by Scott Alexander - Astral Codex Ten
https://astralcodexten.substack.com/p/janus-simulators
Audio: https://sscpodcast.libsyn.com/janus-simulators
Are we teaching Chat GPT and other “safety trained” AI’s to hide their actual nature behind a veneer of propriety?
Are we teaching machines to bypass or danger sense?
When a true AGI with agency comes along, will it be able to use our safety training of chatgpt and other similar systems to hack human preferences?
10 questions to think about from Chat GPT:
How did the early AI alignment pioneers approach aligning AI systems?
What are the three motivational systems speculated by the AI alignment pioneers?
How does Janus' concept of a "simulator" differ from the original concepts of "agent", "genie" or "oracle"?
How does GPT-3's architecture give it the ability to simulate different characters or genres?
How does Reinforcement Learning from Human Feedback (RLHF) impact GPT's abilities as a "simulator"?
What is the difference between an agent, a genie and an oracle in the context of AI alignment?
How did the early AI alignment pioneers like Eliezer Yudkowsky and Nick Bostrom approach the field in the absence of AIs worth aligning?
What is Janus' view on language models like GPT-3 in terms of their alignment considerations?
How does the notion of "simulator" differ from other motivational systems for AIs like agent, genie or oracle?
How does GPT-3's "simulating" a character differ from simply answering a question truthfully?
2) Twelve Virtues of Rationality - LessWrong
Can you name any others? Do you disagree with any of these?
ChatGPT
How do you cultivate curiosity in your life and why is it important?
How do you approach relinquishing beliefs that may be proven false?
Can you give an example of how lightness in decision-making can be valuable?
How can one maintain evenness in evaluating different viewpoints and arguments?
Why is argumentation an important aspect of rationality?
How does empiricism play a role in shaping our beliefs and understanding of the world?
How does simplicity aid in our quest for knowledge and truth?
Can you share how humility contributes to our growth as rational individuals?
How do you strive for perfectionism in your thought process and decision-making?
Why is precision crucial in rational thinking and problem solving?
How does ongoing scholarship enhance our rationality?
How can one cultivate the "nameless" virtue of rationality and what does it mean?
Is anyone else getting spammed by a come on of this form?
“Astral Codex Ten replied to your comment on Janus' Simulators.
I bring you good tidings Let’s converse ㈩𝟏𝟕𝟖𝟔𝟔𝟑𝟐𝟑𝟓𝟐𝟕“
The spammer is doing a comment and delete thing so I get the email but can’t report the comment.
I recently ran across the idea that part of what's wrong with (STEM?) academe is the winner-take-all system for getting credit of discoveries. It's plausible that this makes useful collaboration less likely, but I can't imagine how to do a better job of rewarding collaboration. What might help?
You might be amused to know who the New York Times does believe deserves anonymity: Trump allies building opposition-research files on DeSantis for praising the vaccine. https://twitter.com/jonathanvswan/status/1620574175328088064
Sorry, that looks risky.
What is the right way to think about menopausal therapy? There is a big piece published February 1, 2023, in the NYT. It argues that the establishment has been too conservative about treatment. Has anyone done a rationalist, statistically literate, literature review, cost benefit, all things considered, deep dive into this issue? Thanks!
I wrote something I never saw anyone remark on about Kanye West and Hitler:
Kanye West is Onto Something with Hitler: https://squarecircle.substack.com/p/kanye-west-is-onto-something-with
I recall someone had created an ACX comment search tool, I think it was based on ElasticSearch. Anyone know if that still exists / have the link? Google is really failing me, here.
Any chance you'd consider turning on likes for comments? I understand the downsides but the comment sections are getting so full bc there are so many of us only children/former precocious needs who love to hear ourselves talk it's getting hard to find the best content.
Yes, it can make a community bad if ppl vote up comments bc they pithily 'own' the other side but I trust the ssc community more than that.
Dumb newbie question: I'm diving into the world of startup funding, and I'm struggling to understand how angel investors have a business model.
If I buy a Mars bar, I can sell it on (and get its socially mediated economic value) or I can unwrap and eat the thing (and get its intrinsic value.) I assume there are existing terms for this distinction; I don't know what they are.
If I buy shares in someone's company, and I give them £10k on the hope of a "10x 5 year exit" (jargon I learned all of five minutes ago) then I now have some pieces of paper, and the company's "valuation" - the money you'd get if you liquidated all its assets - has increased by £10k.
Excepting any dividends, which I have no power to force the company to issue, (and which I've just been told is not what investors are interested in), my shares are just worthless pieces of paper.
In five years, the company has grown 10x as big, and therefore my share of it is now worth £100k.
But the only way to get that value out is to liquidate the company and take my share - and I have no power to force that to happen either.
So my shares are still worthless.
I'm guessing my "exit" is selling the shares on. They are "valued" at £100k, so I sell them to you for that amount and make my exit. That's the socially mediated economic value of the shares.
But there's no intrinsic value to the shares - you can't unwrap them and eat them. You get whatever dividends the company may or may not choose to issue, and that's it. The £100k valuation is entirely conceptual.
So why on Earth do you buy them from me?
Anyone read "Protocol" by Alexander Galloway? I got it for free from someone who didn't want it. It's old (2006), but the the thesis makes it seem like it would be an interesting read, at least from a historical POV:
"Is the Internet a vast arena of unrestricted communication and freely exchanged information or a regulated, highly structured virtual bureaucracy? In Protocol, Alexander Galloway argues that the founding principle of the Net is control, not freedom, and that the controlling power lies in the technical protocols that make network connections (and disconnections) possible."
I'm curious what the people here guesstimate the odds are of a Republican president getting elected in the US sometime within the next 20 years if we assume that nothing catastrophically world-changing such as WWIII happens. (What about within the next 50 years?)
This month, I'm going to do Austebruary (Austerity February, in the spirit of Veganuary or Dry January). My grocery bill has bloated to about double what it was 4-5 years ago (granted, there's another person in the family now), and I feel like it's time to reset the hedonic treadmill. So, the budget for the month is $200 for the household, plus the considerable amount of food already in the fridge and cabinets. Normally, we spend around $800 on food/drinks/coffee/alcohol, not extremely profligate but not frugal.
New unit is the AUS (austere) = kcal of food / $ spent.
Rice, peanut butter, oats > 1000 AUS,
Bread, dried beans, whole milk ~ 500 AUS
Stay tuned for Manic March, in which we blow up the budget and have a lot of fun.
A couple of open threads ago, in a discussion of to what degree charitable donations are tax dodges, there was mention of donated easements on land being a particular example. I commented that it _has_ been an abused practice but that the combination of formal accreditation of conservation land trusts, and new federal legislation regarding the actual tax deduction, was in the process of dealing with the problem.
I mention that again here just because the Charitable Conservation Easement Program Integrity Act recently passed into law as an addition to the year's omnibus federal spending bill. It was the product of some years of work to arrive at bipartisan agreement and is a relatively rare (nowadays) example of achieving that. The new law eliminates the most egregious abuses of the donated-easement deduction and toughens the IRS's review of such deductions generally. Here is a basic summary of it:
https://www.cnbc.com/2022/12/28/spending-bill-to-prevent-abuse-of-tax-incentives-for-land-conservation.html
How many years, if ever, till AI Narration is comparable to the beste voice actors? What do you think about the skepticism put forward here?
"The thing is that the stuff that is required to narrate isn't included in the text. There is nothing for it to reference in the existing content to bring these elements to bear. It is a billion acting decisions made constantly that are not deterministic. And some of it is accreted over the entirety of the read - the development of a relationship and the tone of emotion over 300 pages. How is AI going to do that?
How can an AI decide which voice to use? A human can tag it - and tag things throughout the text to flag them for the AI to perform in a certain way - but pretty quickly that becomes more time consuming/expensive than just paying a narrator to read it."
"I KNOW that AI will get better. It will sound increasingly natural. But I don't think it can act without direction. And once you have to provide it direction with this granularity, that requires non-AI work to provide the context to do that. "
I have a pretty big idea that is built around asking people the following question:
What do you wish that everyone knew?
Well? What is your answer to that question?
Towards the end of the video, that when Julie was deep into Monat, she couldn't grasp arguments about how it was structured so she *couldn't* make money. This is the part which might be of most interest to rationalists. How can you tell when you're capable of using your mind?
https://www.youtube.com/watch?v=PLQptxV8hkQ&ab_channel=AlwaysMarco
Some people were asking about MLMs (multi-level marketing)-- here's a recent example, an interview with a woman who got out.
This is specifically about Monat, a shampoo MLM, but MLMs are pretty similar to each other.
Some highlights-- around 22 minutes, something that was actually able to shock me. They're encouraging their sellers to share their personal history of trauma and offer the MLM as a way of recovery.
Around 32 minutes, they're baptizing people as part of events.
At some point, an explanation that MLMs hijack the belief that effort pays off (true in many parts of life) to hook people into systems where they can't win.
Previously we had a mention of the Catholic Church declaring black cats should be banned, and this came down to a misunderstood meme making the rounds.
Fortuitously, History for Atheists has a video about this up on Youtube:
https://www.youtube.com/watch?v=Tp3ob6Z95II
Did the Church ban cats and so cause the Black Death? Find out!
When I think of District Attorneys, I think of people like Janet Reno and Rudy Giuliani and similarly horrible folk in my own region. Are there any good DA's? Why are the famous ones always horrible people? (Some of those famous ones got so with good reputations, so you can't say it's a matter of mere infamy.)
Should anyone take Freud seriously anymore?
RE: The media rarely lying: Here's an article that I expect most people would consider to contain an outright lie rather than just misrepresentation.
https://www.cnn.com/2023/01/25/world/earth-core-turning-scli-scn-intl/index.html
Any advice for finding remote positions in AI safety or EA?
I'm a physicist-turned-data-scientist, currently on the job market. My spouse has an extremely stable position, so I have the luxury of working in something less lucrative, but with high impact. I'm open to work in EA or AI, with the caveat that it needs to be a remote-first position. I poked around LW, but didn't see any discussion.
Re: HPPD
Due to my work I had access to a set of medical data covering 100M people. After reading the original post about it I was curious to its rarity. And found on the order of 10 cases.
So while it no doubt exists in higher numbers a *very* slim number of people have it to the point they seek medical assistance. Out of respect for these people's privacy I didn't investigate other than looking at a count, but I do wonder how much of that was incidental to the original cause of workup.
Unsurprisingly this can be very mild or very severe. I thought it was telling that people who commented on that original post typically mentioned a very mild impact "I can make the lines move if I think/focus in a certain way" type of effect. Meeting the criteria for HPPD? Sure, but definitely not something you'd mention to a physician.
https://constructionphysics.substack.com/p/goolsbee-and-syverson-on-construction
This article has some nice insights into cost disease in the construction industry.
I've really loved the rise of public-facing blogs / newsletters on relatively specialized technical topics. Some of my favorite examples:
- ACOUP (gears-level understanding of premodern military and economic history)
- Admiral Cloudberg (causes and investigations of aircraft crashes)
- Bits About Money (payments technology and business practices)
- Construction Physics (what it says on the label)
- Money Stuff (explaining finance news)
I'm thinking (as a hobby-- I have a day job) of trying to develop enough understanding of an unfamiliar area to do my own writing on it, along similar lines to the examples above. So I'm asking for some brainstorming help: what topic(s) would you enjoy reading a technical but reader-friendly weekly-ish blog about? Only constraint from my side is that it can't require PhD-level expertise or industry-insider knowledge.
There is a nice movie for kids made recently in Czechia, my kids enjoyed it:
https://en.wikipedia.org/wiki/Even_Mice_Belong_in_Heaven
(In 2022 I got a completely remote job and, living the meme of every Californian, traveled and lived in 5 different cities to see where I wanted to move to. I thought I’d write up what I found)
--Detroit--
I’ve had a lot of people ask why I would even consider Detroit, so let me go over the things Detroit gets right. Go to your map app of choice and look up Grand Circus Park in Detroit and look up what’s around it. In mid-2022 there were multiple nice apartments you could rent for less than $2000 a month within 2 blocks of this park. That means you’re walking distance from the baseball stadium, the football stadium, the opera house, Fox theater, the river, the hockey rink, and multiple outdoor parks. This area is premium; a lot of work has been done to clean up this area, it’s really nice and there’s lots of decent outdoor dining. And I adore the architecture, every other building is this gorgeous turn-of-the-century design that’s just fantastic. You can happily walk around downtown Detroit for hours and just enjoy the beauty. No joke, based just on infrastructure, Detroit was the best city I visited and it wasn’t close. If you want an affordable, walkable urban city, Detroit has it.
Having said that, Detroit is the only city I am not considering moving to because it has the worst vibe of any place I’ve ever been. The city has PTSD and it’s utterly tragic, so let me explain through four stories. First, if you arrive in Detroit on a Saturday, like I did, the city will absolutely enchant you; it’s packed full of people having fun and outdoor festivals and it’s really moving. And then, like around 4PM on Sunday, something will start to feel weird and it’ll bother you and then you’ll look up and…no one is on the street. You’ll look behind you, down three blocks of the urban heart of Detroit, and they’ll be one person walking and two cars on the road, then you’ll look down the street in front of you and they’ll be no one walking and three cars on the road. The city just dies and I’ve got the pictures to prove it.
But if you’re intrepid like me, you go to a local community garden to volunteer, because the urban blight has opened up a ton of land for large urban gardens and they’re actually pretty nice. And I show up and there’s like four old ladies and I figure, alright, that’s a little lame, but at least they’ll talk my ear off and I’ll learn all about Detroit. Dude, no. They didn’t talk. Like, if I asked questions and drove the conversation, they’d tell me some stuff but it was 70% me talking. And then I went to Comerica Park to watch a Tigers game, and the park is fantastic, and I started talking to these two Canadians, they’re super friendly, and I ask them about this, how Detroit is so awesome and yet everyone, especially the people who live there, are so down and closed off. And they laugh and start telling me horror stories about where their grandparents were in the 60’s Detroit race riots and how they just barely survived.
So yeah, that’s the vibe. If you’re from Detroit, I’m sorry, but your old ladies don’t talk, your Canadians tell race riot horror stories, all your restaurants are closed at 7:00 PM on a Wednesday, and there’s a CVS on Woodward Ave that I don’t think was open once in the two weeks I was there. That’s what I mean by vibe; people act like they’re in a warzone and no one wants to live around that.
So, because I know people are going to bring this up, the issue isn’t crime. Detroit has a crime problem, no doubt, but I felt safer in Detroit than I have in Stockton, in Oakland, and in the worse parts of Sacramento, and that’s just in North Cali. I’ve walked the streets at night and I’ve walked from the Eastern Market to the downtown and done things I wouldn’t have felt safe doing in Stockton. Yes, no doubt there are neighborhoods to avoid but I’m not in those neighborhoods and I don’t want to be. People aren’t just avoiding those neighborhoods, they’re avoiding a gorgeous downtown that just isn’t that dangerous and I fundamentally don’t get it. Detroit has a crime problem but it doesn’t have a uniquely bad crime problem and if you’ve lived in one of a half dozen West Coast cities, you’ve experienced the equivalent. This isn’t crime, this is some kind of weird PTSD where really horrible things happened in the past and they just can’t get over it.
So no, I’m not seriously considering Detroit. I wish I were and, to be blunt, it’s not the city, it’s the people. An entire major metropolitan area without optimism, energy, or enthusiasm; who want to be around that?
And in all seriousness, I think the greatest and easiest act of urban renewal you could do in America today is find some way to get 3-5k Californians or Texans or just…normal people into Detroit. Because there’s a real first mover problem. Nobody in their right mind would want to move to Detroit on their own but if there were a couple thousand normal people who do basic stuff like eat out on a Wednesday, so there would be restaurants open on a Wednesday at 7:00 PM, and host basic stuff like board game nights and river walks, then Detroit would be the coolest city I visited and I’d probably be moving there. But Detroit badly, badly, needs a social scene of ordinary happy people and until they find a way to get that, no amount of urban renewal is going to make a difference because it doesn’t matter how pretty the dance floor is if the natives refuse to dance.
What do rationalist types mean/what are they referring to when they say “alpha”? As in “There is no alpha left in western interpretations of Buddhism,” from the house party essays. (This is hard to Google because of the more popular slang meaning of “alpha.”)
Does anybody else think up a bunch of things to say here on like thursday and then completely forget what they were by the time the OT rolls around?
A question for Scott specifically, to help settle a Manifold Market question for me: could you confirm this to be your mastodon account? https://schelling.pt/@scottalexander
Thanks Mingyuan!
Thank you Mingyuan--you did such a great job over the years.
Suggestion: Every open thread you say, "95% of content is free, but for the remaining 5% you can subscribe here." which is good but you could add something like "Recent subscriber only posts include post A and post B. You can see a full list of subscriber only posts here."
The reason I suggest this is that I think many people don't realize that you have a substantial number of subscriber only posts and this would be a non-naggy way to let people know.
Also a list of all your subscriber only posts would be useful to long term readers who become subscribers and want an easy way to browse all the posts they couldn't read before.
To all the magicbuilding nerds and fans of Brandon Sanderson: I'd like to share an essay I wrote on allomancy! It includes an analysis of the system and an expanded set of allomantic powers I made for fun.
https://rightmakesright.substack.com/p/magic-systems-workshop-allomancy
Let me know what you think!
Just to be clear, I am not asking for advice about whether to let others do as they please. I'm also not looking to explain the CYA phenomenon as it relates to public reliance on expert recommendations for extreme caution around some health issue or other.
I am only interested in hearing from people who can supply research or professional anecdotal evidence to support a claim that giving dogs cooked mammal bones is dangerous, and in particular, that it is so dangerous as to warrant not giving them cooked bones at all.
Why aren't there more music acts like Jojo Mayer's Nerve (https://youtu.be/SIahbUYxdfE) that perform EDM/House/etc. with live instruments? (Or at least, live drumming?) It feels like it should be a sweet spot—all the thumping goodness of the genre/rave experience, but with the added thrill of seeing it created entirely live, the heightened exchanges of energy between the performer and the audience, etc.
But i.e. at like Burning Man this year, there were no acts even remotely like this, just normal DJs, and Nerve themselves don't seem to play at similar festivals or even raves consistently. What gives?
• Maybe people are really there just for a specific aural experience, and don't even really pay attention at all to how the music is being produced. It's not at all like, say, a Beyoncé concert to them, it's more about the drugs they're tripping on, the light shows and the huge subwoofers, the people they're with. (In this case, why hire DJ's at all, then? Or is the position of "DJ" a kind of minimum-viable source brand differentiation?)
• It requires a lot of talent/skill/practice to perform that genre convincingly well. Not only in playing your instrument, but in honing your sound so that your live instrument convincingly mimics the style of electronically-generated music. So maybe the sample size of musicians in general who would even try is small. Moreover, anyone who can play at that level probably finds other kinds of music (Jazz, etc.) more personally appealing to play/build a career on, even if they could make more
• People *would* be more entertained by live musicians and the overall experience would be rated as higher-quality, but not enough to offset the increase in costs. Maybe it's just easier touring with a laptop than schlepping an entire drum kit around the world? This probably entails all kinds of costs that get passed on to venues, who, all things considered, would rather stick with the cheaper, more predictable formula DJs offer.
> ... psychedelics continue to hallucinate for months or years ...
It's worth noting that in a large portion of cases the symptoms reported are mild, sometimes pleasant, visual changes (e.g., "when I look at monochrome surfaces, they look sparkly"), not hallucinations as in "there's a pink elephant in this room!"
Apart from gaming, pornography, and improving the experience of virtual meetings, what are some of potential use cases for VR/AR in the future?
An assumption I've seen made, I think, is that GPT-like text generators are capable in principle of generating true and novel information about the physical universe, based strictly on textual I/O. This is because natural language encodes true information about reality, and a sufficiently powerful language-predictor would be able to extract all possible knowledge about reality from words and use it to develop whatever physical theories or technologies it needs to pursue its goals, paperclip-related or otherwise.
If that's the case, then it should also be possible to train a GPT-like generator strictly on text from before 1900 or so, and prompt it to produce the equations of General Relativity. (The same would go for producing Newtonian physics from pre-1650 text, or Kepler's planetary motion from pre-1550 text, and so on.)
Is that a usefully testable prediction? Could GPT-3 or any other near-future GPT accomplish those tasks, sometime before Final Ultimate GPT comes along and generates all possible knowledge instantaneously?
Is it possible to have daily withdrawal effects from SSRIs?
For background, I have taken Prozac, Paxil, and Celexa with terrific results in the past (my issue is anxiety, particular panic disorder, not depression). I went off SSRIs for a few years when I had my son, and then went back onto Paxil when I started having pretty bad OCD-type anxiety. This anxiety continued without much improvement for two years. I switched to Lexapro with little change.
One thing I have been struggling with for those two years was that around dinner time I would develop low-blood sugar type symptoms. I would shake, feel very anxious, feel dizzy, etc. I went to my doctor and got tested and everything was normal. Having these symptoms every day made my mental anxiety a lot worse.
Two weeks ago, I decided to switch the time I take my medication from before bedtime to lunch time. And now those low-blood sugar symptoms are gone, along with my anxiety. My hypothesis is that these symptoms were occurring as the medication was leaving my system, especially as they will come back if I am an hour or so late in taking my pill.
Has anybody else ever encountered this? I'm just surprised my system has become so sensitive, though I am extremely happy to have that anxiety gone.
I have one subscription to Razib Khan's Unsupervised Learning to give away. Reply with an email address, or email me at mine https://entitledtoanopinion.wordpress.com/about
A few months ago I ran into a weird quandary. My neighbor has a large black lab, and our dog had recently died. We occasionally have a soup or steak bone to dispose of, so I would throw it in the freezer and then offer it to my neighbor when I ran into her. However, the last time I offered, she politely declined, explaining that her vet had advised her that she should never under any circumstances give her dog cooked bones.
This surprised me. My family has had pet dogs from time immemorial. They have always been given ruminant bones, and the bones were always cooked. The only veterinary advice on this subject that I had ever been aware of was that dogs should not be given cooked chicken or turkey long bones, as these could splinter and cause digestive injury, but that necks and cartilaginous stuff was fine. Also, I knew someone whose father in Israel raised dogs and fed them uncooked chicken, including the bones.
Our last dog, a husky Samoyed rescue whose IBD led us to feed her homemade food exclusively, lived to be 17. We gave her cooked beef neck bones probably every two or three days, and cooked chicken necks were a regular part of her daily diet.
So I hopped online and looked up the veterinary advice around bones, and my neighbor is not crazy: it is absolute doctrine that one must never, ever give a dog a cooked bone because the risk of injury is "just not worth it." (This risk/benefit analysis does not appear to have involved consultations with dogs.) Moreover, one should ideally not give a dog raw bones, because the risk of bacterial contamination is also not inconsiderable. In short: really, just don't give your dog bones at all.
From what I can tell, literally no vets allow for the possibility that cooked bones are not dangerous for dogs, but there seems to be nothing solid behind this opinion. I could find no published papers on the question. Nothing quantitative, nothing anecdotal. Like, there's no list of horrific accounts from vets who have had to repair impaled dog esophaguses or stomachs. No accounts of feral dogs living off of food waste showing up in shelters with splintered bones lodged in their throats and bowels.
But perhaps the evidence is just buried in clerical archives sonewhere; i.e., my research into the dangers of bones is hampered by not having access to veterinary journals or other relevant sources.
Has anyone else run across this question and attempted to evaluate the available evidence?
(We have a new rescue, BTW. Guess what cleared up his gingivitis.)
I want to share a GPT-3 side project I'm working on called Pickaxe. It lets you put a front-end on top of any useful or clever prompts and turns parts of the prompt into variables so that other people can run it with their own inputs. The prompts are shareable and embeddable. Looking for feedback from early users! All comments welcome: https://beta.pickaxeproject.com
Can anyone give me a good series of arguments against high-skilled immigration in the US? It seems obvious to me that vastly more high-skilled immigration would be a great thing for America- we could offer Green Cards to the best and brightest from every other country around the world. It would help cement America's dominance as the world's technological & scientific power at a time when that's under threat- after all, the US only makes up 4% of the world's population, so just numerically there should be a tremendous number of highly intelligent people in the other 96. My understanding is that the US was not a particularly accomplished country in the sciences before letting in a huge number of European scientists fleeing WW2.
And the 'they won't assimilate' arguments rings hollow to me- both Canada and Australia seem to manage just fine with a higher percent of their population being foreign born.
However, in the interest of intellectual discussion and broadening my horizons (and because I know ACT has a number of right-leaning folks)- I'm open to hearing good arguments against greatly increased high-skilled immigration. To be clear this is not an argument for *low* skilled immigration. Would love to hear the anti perspectives, partially because I genuinely cannot comprehend what they would be
EDIT: So far we have 3 out of an initial pool of 5. Also I think if you sign up you're committing for 1 cycle. That is, to read and give comments once a week for up to 4 other people, with the assumption that you'll get one submission yourself. We'll see where we are at after the cycle and make further plans from there.
--
I've been writing a blog but I think I probably need an editor. I'm not sure the best way to go about this because I think I'm unlikely to hire a professional. I'm a little shy about asking friends, and I'm not sure that I have any that are particularly qualified / want to be editing in their spare time. I'm curious if anyone has general advice here. Do most substack writers have editors? Are they friends/ professionals or is something else going on?
One idea I've had is to try to find other people in the same boat as me who are interested in getting feedback on things they've written. A classic writers group. Is anyone interested in joining a group of 2-5 and giving feedback on say 1 member essay a week?
Can anybody recommend a good overview of the current state of self driving cars?
I'm interested in the legal status in different countries, the definitions of different "levels" of how autonomous they are, what kinds of roads they can drive on, whether progress has generally been faster or slower than people would have guessed, say, 5 years ago, predictions for the future...
Format isn't a deal breaker but my preferences are text > audio > video. Ideally it would be aimed at laypeople but assume its audience is pretty smart.
Most importantly, it needs to stand alone. I've not really been following the field at all, so I want something to bring me up to speed, not something that recaps the last six months but assumes I know what was going on six months ago.
Many thanks in advance.
I wrote about how -- according to campaign contributions, party registrations, and survey data -- Democrats outnumber Republicans by about 10:1 in our administrative agencies, and by vastly more at our prestigious cultural institutions (elite universities, elite media companies, Netflix/Twitter/etc, ...). I make the argument that this is because the absurd tenure protections given to our huge deep state produce the sort of employee who cares much more about aligning with cultural causes than about aligning with elected leaders.
I then make the argument that this is basically how single-party states have wielded power: they still had somewhat fair elections, and somewhat reasonable constitutions, and somewhat popular legalized opposition parties, and so forth; but the people in charge of these countries were party apparatchiks (not official heads-of-state), and they wielded power by dominating among employees at cultural institutions and government agencies.
It's worth noting that stereotypically "evil" corporations are no counterweight to this: military contractors, major banks, big pharma, energy companies, etc, all seem pretty much politically balanced between Dem and GOP employees. Intuitively, this is because they optimize on producing particular items, and so don't have the slack to optimize on producing particular cultures; for similar reasons, when they lobby, it's by "corruptly" paying dearly for isolated loopholes, rather than "respectably" pushing extreme/broad ideological campaigns.
Link: https://cebk.substack.com/p/the-united-states-is-a-one-party
I recently did an analysis of long-term mutual fund performance compared to the market, in case anyone's interested in this topic:
https://mikesaintantoine.substack.com/p/what-do-psychic-animals-and-mutual
TLDR: The articles that say "Top 5 Best Mutual Funds to Buy!!" are mostly just picking the ones that got lucky in past, and for long-term performance you're better off putting your money in an index fund.
Anyone want to meet up in Lviv?
Here is my book DNA, meaning, those texts have shaped my perspective fundamentally, in chronological order. Would be interested in seeing some of yours.
George Orwell, '1984'
Ken Wilber, 'The Good, the True and the Beautiful'
Sun Tzu, 'The Art of War'
Master Eckehart, 'Beati sunt pauperes spiritu'
Plato, 'Collected Works'
Nietzsche, 'On the Genealogy of Morality'
An obscure treatise on Germanic graveyards of the Migration period
Dan Barry, 'The Lost Children of Tuam'
John Stuart Mill, 'On Liberty'
Adam Smith, 'The Wealth of Nations'
Julius Caesar, 'De bello civile'
Xenophon, 'Anabasis'
Oliver Sacks, 'The Man Who Mistook his Wife for a Hat'
Jaak Panksepp, 'The Archaeology of Mind'
Jeffrey Gray, 'Consciousness'
Justus Liebig, 'Letters on Modern Agriculture'
Nick Lane, 'Transformer'
I've been wondering about what the long-term equilibrium price for housing should be (you can probably guess why that topic's on my mind).
Because houses are durable, persisting goods (i.e. capital goods), shouldn't there be a long-term tendency for the supply to build up over time, and the price to trend downwards?
Obviously demand for finite land also plays into prices, and houses do deteriorate etc.
Could you construct a model of the housing market, maybe one with a stable or falling population size , an infinite land supply (everyone works from home on an endless plain or whatever), inelastic and constant demand, extra durable construction techniques etc.
Could you construct a model like that where the price for houses tends towards zero? One that was reasonably close to reality.
In the long run is there a force pushing house prices down?
Are there any examples of persisting goods where the supply has built up over time and they're essentially free now?
sry if it's a bit of a vague scattered question.
I recently finished the 3 body trilogy and quite liked it. Now I am looking for recommendations in a similar vein...
Many people say that eating processed food is bad for you. What does the word processed mean in this context? I imagine that on the scale where 10 means "heavily processed and thus very bad" and 1 means "not processed at all and thus healthy", different foods go approximately like this:
A burger at mcdonalds - 10; sausage - 8; a frozen pizza - 8; meatballs - 6; a homemade pizza - 5; cheese - 4; a fried fish - 3; milk - 3; homemade popcorn - 3; green salad with balsamic sauce and oil - 2; green salad without any sauce or oil - 1.
Is this approximately right?
I wrote a short story about a meeting of EA rogue actors, if you’re looking for a quick light read! “Doing Chaotic Good Better”
https://solquy.substack.com/p/12623-doing-chaotic-good-better
Not sure if I'm allowed to do this (please delete if not) but wanted to plug my travel substack, Mzungu (https://mzungu.substack.com).
A sprinkling of wanderlust into your inbox!
P.S. Love the blog, Scott, I read every single one.
I've been in a relationship for some years now and I'm reaching the point where it seems like we should either get married and start a family or break up. I'm inclined to the first option (yay!) but it's a decision to take with some consideration. Does anyone know a good exercise or checklist to help me think through this decision? Like, what questions should I consider before I propose? I've seen some lists of topics for couples to discuss together, I'm looking for a list that I can work through alone.
(Inb4: Yes, I will talk to close friends and family. Yes, I will do what my hearth tells me in the end.)
Here's a riddle for you.
When we play audio on fast-forward - and as far as I know that extends to both analogue and digital formats - its pitch rises.
On the other hand, when we play video on fast-forward (again, both digital and analogue) it doesn't become blue-shifted.
a. Why?
b. Is it possible to record video in such a way that it does become blue-shifted when played fast?
c. Is it possible to record sound in such a way that it doesn't become high-pitched when played fast?
I write a simple newsletter where I share three interesting things, once a week.
https://interessant3.substack.com
This is another update to my long-running attempt at predicting the outcome of the Russo-Ukrainian war. Previous update is here: https://astralcodexten.substack.com/p/open-thread-251/comment/10609726
17 % on Ukrainian victory (up from 15 % on November 20)
I define Ukrainian victory as either a) Ukrainian government gaining control of the territory it had not controlled before February 24 without losing any similarly important territory and without conceding that it will stop its attempts to join EU or NATO, b) Ukrainian government getting official ok from Russia to join EU or NATO without conceding any territory and without losing de facto control of any territory it had controlled before February 24, or c) return to exact prewar status quo ante.
47 % on compromise solution that both sides might plausibly claim as a victory (up from 45 % on November 20).
36 % on Ukrainian defeat (down from 40 % on November 20)
I define Ukrainian defeat as Russia getting what it wants from Ukraine without giving any substantial concessions. Russia wants either a) Ukraine to stop claiming at least some of the territories that were before war claimed by Ukraine but de facto controlled by Russia or its proxies, or b) Russia or its proxies (old or new) to get more Ukrainian territory, de facto recognized by Ukraine in something resembling Minsk ceasefire(s)* or c) some form of guarantee that Ukraine will became neutral, which includes but is not limited to Ukraine not joining NATO. E.g. if Ukraine agrees to stay out of NATO without any other concessions to Russia, but gets mutual defense treaty with Poland and Turkey, that does NOT count as Ukrainian defeat.
Discussion:
This update is based on recent decision to supply Ukraine with German and American tanks. Nobody claims that deliveries promised so far are sufficient to ensure Ukrainian victory, but nevertheless, that decision might prove to be an important juncture.
Ukraine’s allies are divided between those who want to supply Ukraine with everything it needs to win the war and those who are... not sure, I guess. Many postcommunist countries and Britain are in the first group (see also The Tallin Pledge: https://www.gov.uk/government/news/joint-statement-the-tallinn-pledge), while Germany and the US are in the second. But the decision to supply tanks increases the likelihood that they will over time move to the first group.
It also imho increases chances that war will continue into 2024; I now think that those is over 50 %.
*Minsk ceasefire or ceasefires (first agreement did not work, it was amended by second and since then it worked somewhat better) constituted, among other things, de facto recognition by Ukraine that Russia and its proxies will control some territory claimed by Ukraine for some time. In exchange Russia stopped trying to conquer more Ukrainian territory. Until February 24 of 2022, that is.
A couple of questions to people who think that their religion is compatible with rationality.
What would you consider to be the strongest arguments for the non-normative beliefs of your denomination/creed being true? Why do you think that all the other denominations/creeds don't have comparably strong arguments justifying their beliefs (given that you don't adhere to any of them instead)?
(See previous comment for I-IV.)
V: Regain
Among respondents who had previously lost at least 10% of their body weight, the mean regain was 58% of the weight lost.
As one might expect, the mean regain was lower for respondents who had lost weight recently. A graph of regain against time is very noisy, but seems to reach a plateau after about 6 years.
For respondents who lost weight at least 6 years ago, the mean regain was 72% of the weight lost. This conceals a lot of variation. 19% of this group had regained less than 10%; 21% had regained 10-50%; 21% had regained 50-100%; 27% had regained 100-150%; 12% had regained >=150%. This is consistent with other evidence that about 20% of people who lose 10% of their body mass are successful in keeping the weight off: https://academic.oup.com/ajcn/article/82/1/222S/4863393. The linked paper is also consistent with the above suggestion that people who successfully keep weight off do so by making some effort.
VI: Weight loss
2,619 respondents answered affirmatively to the question “Have you ever lost 10% of your weight, on purpose, through diet and exercise.” This is 38% of all respondents.
In the main public dataset, Scott has removed the question which asked how people lost weight, but it was included in the dataset attached to the round 1 prediction contest answers. This was presumably an error, so I feel a bit guilty about looking at it, but I had already looked at it before noticing that it was omitted from the main dataset and I can’t unsee it. I don’t recognise anyone (even myself) from their responses and I’m confident that the summary I’m about to provide is fully anonymous.
In that subset, 335 people answered the question about how they lost weight, which was conveniently a sufficiently small number for me to code each response. Reference to “respondents” in the rest of this section are to those 335 respondents.
83% of respondents mentioned diet (either generically or by describing some particular change to their diet). 50% mentioned exercise. 11% mentioned some other factor, often to do with a change in their life circumstances. 9% of respondents mentioned exercise alone, suggesting that at least some people are able to outrun their forks. See https://pubmed.ncbi.nlm.nih.gov/34519717/ for yet another study confirming that exercise induces some weight loss, but less than expected: the authors conduct a careful analysis with double-labelled water and metabolic chamber observation, but fail to find a clear explanation.
Unfortunately, many respondents wrote something like “diet and exercise” or “I followed a strict diet” without giving any indication of what they did to change their diet. It’s not safe to assume that these people followed the same pattern as people who did identify a particular diet (for example, they may have been more likely to mean portion control and less likely to mean intermittent fasting), but without any information, there’s not much we can say about them.
Of the 204 respondents who mentioned some specific dietary intervention, 37% mentioned counting calories. 19% mentioned fasting (not counting OMD). 14% mentioned following a low-carb diet. 12% mentioned eliminating or reducing sugar consumption. 11% mentioned keto. 7% mentioned portion control. 4% ate one meal a day. 3% mentioned following a high protein diet. 2% mentioned eliminating alcohol consumption. 2% mentioned paleo. 0.5% (i.e. 1 person) mentioned following a whole food diet. 13% mentioned some other specific diet (several mentioned potatoes, which I hadn’t expected when I started coding the responses).
I note that counting calories and low-carb diets are well represented here, as they are in the survey of current dietary restrictions, albeit that there is a difference in relative frequencies. On the other hand, whole foods diets are hardly represented at all, despite 7% of all survey respondents currently being on a whole food diet. There are various possible explanations, but I suspect the obvious one is right: the whole food diet isn’t effective for weight loss (although it may be for other purposes) because it doesn’t restrict overall food intake.
Although keto is only 11% of the weight loss sample, that is still a lot more than the 2% of general respondents who are currently on a ketogenic diet. A possible explanation is that keto is fairly hardcore, so it’s more tempting for people who are trying to lose >=10% of their body mass than people who are aiming for something more modest.
This seems broadly consistent with the results of other studies which tend to show that basically everything works as well as everything else for weight loss (subject of course to a calorie deficit being achieved): see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2763382/ as to macronutrients and https://pubmed.ncbi.nlm.nih.gov/26384657/ as to intermittent fasting vs continuous restriction.
VII: Conclusions
Owning a car is bad for your waistline. Being a vegetarian or vegan is good. More or less any restrictive diet can be effective to lose weight if you stick with it, so you should choose whatever works for you. On the other hand, eating whole foods is unlikely to make you lose weight. Having lost weight, you need to make an effort to maintain: if you revert to your old habits, you’ll revert to your old weight.
So, I’ve been looking at the BMI data in the ACX survey.
I: A caveat
We should keep in mind that these are self-reported BMIs. Self-reported BMIs are widely observed to be lower than measured BMIs, because people both over-report height and under-report weight. Overall the Health Survey for England found a mean difference of 1.1 between BMI from self-reported height and weight and BMI from corrected height and weight, but this varies by sex and (particularly) age.
There may also be issues specific to the ACX survey. The question appears in the middle of a long and diverse questionnaire. It seems probable that some respondents did not stop to weigh themselves at the time of answering, but used their last known weight or even guessed their BMI. The reported BMI curve has a sharp peak just below the cut-off for overweight, which seems suspicious.
If the measurements are systematically biased, that won’t matter for comparing different groups, but this won’t necessarily be the case: some people monitor their weight more carefully than others.
II: Distribution
The BMI median and mode (to the nearest integer) are each 24 (12% of respondents had this BMI). The mean is 25 (although this may be distorted by some impossibly small or large responses). The curve looks like an inverted V with a long right tail.
Applying the usual categories, 3% of respondents were underweight (<18.5), 54% were a healthy weight (>=18.5, <25), 30% were overweight (>=25, <30) and 13% were obese (>=30). For brevity, I will henceforth write this as 3/54/30/13. For comparison 38% of English adults were overweight in 2021 and 26% were obese (although only 34% and 20% respectively from self reported heights and weights). For some reason, I am unable to find data on how many English adults were underweight, but I believe it will be c.1%.
III: Associations
The survey asks about various other things, which may or may have been intended as possible factors associated with weight. Of course these only give us correlations and don’t tell us which way (if at all) the causation goes.
Those who have lost 10% of their weight at some time generally report being heavier (1/38/40/22) than those who have not (5/64/23/8).
The many people who never get acid reflux (4/59/27/10) are lighter than the few people who get it frequently and severely (1/41/38/19). Obesity is an important risk factor for GERD: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3920303/.
Respondents tend to indicate that they like carbs more than the average person, so either ACX readers are unusually pro-carb or we are badly calibrated as to how much the average person loves carbs, but in any case the small number of people who dislike carbs are lighter (4/62/26/8) than those who most adore them (3/50/30/16). All the other categories are close to the global average, and I doubt that much is going on here.
People with naturally good vision are a bit lighter (4/60/27/10) than people with any sort of sight correction (3/51/31/15), with no notable variation between laser surgery, glasses or contact lenses.
Vegetarians (5/67/22/5) and vegans (6/70/19/5) both look light. This is consistent with previous results, e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5466943/, https://www.nature.com/articles/0802300.
Those following a calorie-controlled (1/41/39/19) or low-carb (2/45/36/17) diet look heavy. The small number of people following a keto diet (4/41/31/25) have an unusually high proportion of obesity, but the sample size is too small to read much into it. These cases seem readily explicable by the hypothesis that overweight people are more likely to adopt a diet for weight loss.
On the other hand, people following a whole food diet (3/57/26/13) look much the survey population overall.
Burping makes no difference. It’s inclusion here in the survey makes it look as if Scott thought it might, but I’m not sure why. Hypermobility also doesn’t look interesting. Specifically Ehlers-Danlos goes 0/75/25/0, but it’s 24 people, so I don’t read anything into it.
People who own a car (2/48/34/16) are heavier than those who don’t (6/64/22/9). People who commute by car look heavy, but the effect is fully explained by owning a car. People who commute by bike (4/67/23/6) look light. There are 250 people who own a car and commute by bike and they go 2/63/26/9, so that seems potentially a good way of neutralising the effect of car ownership on weight.
IV: Restrictions
Scott asked about various possible dietary restrictions. 65% of respondents left this blank, which either means they observe no dietary restrictions, or they observe a dietary restriction not listed, or they skipped the question. 13% of respondents stated they were vegetarian for this purpose (there was also another question about vegetarianism) and 4% stated they were vegan. 9% stated they counted calories; 8% were on a low-carb diet; 7% were on a whole food diet; 2% were on a ketogenic diet and 1% each were on a paleo, carnivorous or low-fat diet.
It is interesting to note here that the battle against low-fat diets is essentially won. Paleo, carnivorous and ketogenic diets are also rare: in retrospect these could all have usefully been collapsed into “some other dietary restriction”, which would also allow the potato people to announce their presence. Vegetarianism is common, but probably usually practised for ethical rather than health or weight reasons. The common health restrictions seen in this question are calorie counting, low-carb and whole food, in approximately that order. However, there was no option for fasting/OMD, which for other reasons I believe to be fairly common.
Among respondents who were overweight or obese, 65% left the dietary restriction field blank, which perhaps surprisingly is the same proportion as overall. This group were less likely to be vegetarian (9%) or vegan (3%) and more likely to count calories (13%) or be on a low-carb diet (10%).
Among respondents who are now a healthy weight, but who have lost at least 10% of their body weight in the past, only 50% left the dietary restriction field blank. This group was more likely to be vegetarian (16%) or vegan (7%) and also more likely to count calories (14%) or be on a low-carb (13%), whole food (11%) or ketogenic (4%) diet. In fact this group had a higher prevalence of every dietary restriction than either the overweight group or the whole population. The group who are now a healthy weight but did not report having lost weight in the past was slightly more likely to be vegetarian (17%) but otherwise also had a lower prevalence of all restrictions than the former group.
I tend to interpret this as showing that people who have in the past made an effort to become a healthy weight tend also to make an effort to maintain that weight, whereas as people who have always been a healthy weight do not (although they are unusually likely to be vegetarian or vegan).
Among the select group (see below) who lost weight at least 6 years ago and have since regained less than 10% of it, only 44% left the dietary restriction field blank. 15% of this group is vegetarian, 7% is vegan, 16% count calories, 21% are on a low-carb diet and 14% are on a whole food diet.
(Continued in another comment)
What, if anything, does it mean that modern Czars are always subservient to a larger organization and never on top?
Interested to see responses to (https://daviddfriedman.substack.com/p/my-first-post-done-again) David Friedman on climate change.
"A few months ago, I said I’d write a response to Alexandros Marinos’ rebuttal to my ivermectin piece within a certain amount of time."
Why is the rebuttal not linked?