I'm a bit confused about the warm-bloodedness thing. I was trying to look this up actually because I've just been kind of confused generally as to what the distinction is supposed to be between endotherms and homeotherms-but-not-endotherms. (I am not any sort of biologist, if that wasn't clear.) Like, OK, poikilothermy seems obviously distinct, but once you're within homeotherms, I was like, I don't really see an obvious distinction between these different methods of maintaining temperature that should distinguish some of them as "endo"? Especially considering that like a main method of thermogenesis is shivering, which -- as a muscle-based mechanism -- seems to be getting pretty close to a behavioral mechanism, you know?
It's the shivering thing that kicked off me looking this up, really. Because I keep reading that non-shivering thermogenesis, which is based on this uncoupling, happens only in brown fat cells, and that adults don't have much of these?? So they get all their heat from shivering?? And that's just like... that can't be right.
IDK, I am basically clicking around on Wikipedia here, so some parts contradict other parts. Like, oh, maybe all cells have UCP1, just brown fat cells have *more* of it. But other parts say no it's only brown fat cells. Or maybe adults have more brown fat cells than thought. I am confused!
Because like yeah generating heat from uncoupling, that's pretty distinct, much more so than shivering! And it also, y'know, matches everyday experience, where you don't start shivering the instant you're colder than is comfortable. But it's really confusing to keep reading that adult humans don't have much in the way of non-shivering thermogenesis going on. Like, huh? What's up with that statement? Where does that come from? Or is there some way I'm missing that it could actually be true??
In theory, making friends online should be easy. Instead of luck and circumstance of the physical world, the virtual world should give us access to the few most compatible friend-candidates out of billions.
And yet, I still default to the physical world for finding new friends.
Question 1: where, online, have you found "true" friendship and how?
Question 2: I know that some have tried (and failed) to create a social network for the non-masses. Do you think there is opportunity for a social network for people with long attention spans that rewards the building of deep relationships? If yes, do you think it should be an open network (like Reddit), or more akin to a dating/matching app that filters the billions down to the most compatible? Ex. If love of Nietzsche is non-negotiable, would be easier to filter by that first.
Re: comments. Just do what you're already doing, Scott. I think they add something special to the Substack.
The addition of "don't highlight" is more than enough to guarantee anyone who doesn't want their comments seen (on a public forum!) won't be surprised. You could ask Substack to put a small text under "Discussion" with a disclaimer, if you want to make absolutely sure.
I've seen a few people saying that they see subscriber only posts in the rss feed, even though they aren't subscribed. While it seems that substack doesn't provide a non-subscriber feed, I've taken a shot at filling that void here: https://pycea.tk/acxfeed . It's a direct clone of the official feed, just with sub only posts removed.
On comments, this seems like a rare social problem with a technical solution.
You've already asked Substack for a lot of changes to make these comments more like WordPress. It seems to me this is another feature you'd like: either a checkbox to mark a comment as highlightable / non-highlightable, or a checkbox at site-level per user, whichever's easier. Then on the admin side, comments can get different background colors depending on whether they should be highlighted or not (or something).
Might take a while to implement depending on how many other higher priorities there are, but it would save you from having to manually maintain lists of people who want comments highlighted.
Maybe it's just me, but the simulation hypothesis (that we're essentially living in a simulation by some superior intelligence) seems completely ludicrous, seeing how our univere very convincingly appears to show behavior spanning 30 orders of magnitude in time, 30 orders of magnitude in space, and a sheer degree of size, scale and detail that is completely absurd if you were trying to study something specific in a simulation. Am I missing something?
Is there maybe a more general phenomenon where conspiracy theorism, or extreme politics in general, comes from a kind of frustrated religious instinct?
On DNP: I feel like the survey missed a "It made me consider DNP but it did not make it look worth the risk". After all, watching video of BASE jumpers part of my brain wants to try that, but that does not mean I consider that a good idea.
This is a little silly, but I'd like to add my novel to the Top Web Fiction website, and it looks like it's invitation-only. Is there anyway you could send me an invitation?
I'm assuming that somebody here will have a more informed take than my own, so I'm writing this in the hope that I get put in my place. Here are my thoughts:
First of all, it's already interesting (although I assume well-known) that facial expressions + head pose (it looks like these are correlated) do so well. Obviously these are choices, and choices reflect the culture of one's peers, but I'm not sure I'd have guessed that the bias is so strong. Relatedly, I definitely would not have guessed that "sunglasses" would be so poor of a predictor relative to, well, anything else.
I was personally surprised that the algorithmic accuracy was greater than human accuracy at the same task. After reading a bit, I think that I just haven't kept up with the SOTA of facial recognition --- apparently algorithms now beat humans in general. This is surprising to me, since I learned that our brains are pretty specialized for facial recognition, but I guess this is just my ignorance speaking. Fine.
Anyway, for the big picture: the attributes against which facial recognition was compared were all *voluntary*. Users had the ability to chose whether to look up or down in their profile photo; whether to smile or frown or wear sunglasses. There are many such choices, and including more of them (like more subtle aspect of the facial expression) might have yielded an even more accurate prediction. Facial recognition itself, though, will include demographic characteristics (*partially* controlled for by the study), but also things like "has this person had plastic surgery". What I really want to know is, how much of the predictive power lies with these uncontrollable characteristics, and how much lies with aspects of the facial expression that can be faked?
I now wait for somebody to tell me that the entire study is flawed because it didn't control for whether the pictures were taken at day or at night.
The problem with agreeing that sticking DNH = do not highlight or something like that on the end of comments is that someone can now search specifically for those tags, and ... if they have really no morals, write a bot that automatically reposts these comments as highlights on sneerclub or something.
I use an online alias here that's not linked to my offline identity or any other online ones, and so should you.
A possible solution for the comment highlight issue would be to make these posts subscriber-only. The obvious downside is that non-subscribers will lose some interesting content, and it kind of goes against your original commitment to make most of your posts free. On the other hand, this puts highlighted comments out of reach for search engines and the internet at large - so the threat to commenters' privacy goes way down.
I was under the impression that you were going to publish the best 5 book reviews and invite additional comments to help decide on prizes, kudos and glory. Are you now going to publish them all? All 100?
Ok, a) that’s a lot b) you’ll publish mine. Have you thought this through?
The most politically influential person you never heard of? Is it possible an unassuming children's books author shaped the world we live in?
A strange thing happened a few days ago. My wife and I were at the rumpus room reading Who's Bashing Whom? Trade Conflict in High Technology Industries ( http://cup.columbia.edu/book/a/9780881321067 ) when our oldest son told us he had been assigned for reading a book by British writer Aldous Huxley. My wife and I had never heard of him before. What happened to Beatrix Potter, C. S. Lewis, Roald Dahl, R. L. Stine, Horace Greeley, Lemony Snicket, Lewis Carroll and G. A . Henty? Has the great state of California cancelled them? Anyway, I decided to ask my own questions and do some research on this fellow. Apparently, his books are required reading in many Democratic-controlled school districts, are very strongly opposed by concerned parents due to anti-religion and anti-family themes and sexual content ( https://en.m.wikipedia.org/wiki/List_of_most_commonly_challenged_books_in_the_United_States ) and (surprise, surprise! ) were widely read in the communist Soviet Union ( https://www.jstor.org/stable/3831583 ). The book my son was assigned to read proposes the abolishing families, replacing religion with orgies and banning the Bible and the works of William Shakespeare.
But who was this Huxley guy and how did he shape the world we live in? He was a scion of the Huxley family, founded by Victorian scientist Thomas Huxley, who was nicknamed Darwin's Bulldog due to his rabid championing Darwin's ideas. More than any other single individual, Thomas Huxley was responsible for the triumph of Darwinism, which led, in the 20th Century, to the Holocaust and the Gulag. Aldous Huxley, besides writing books, also studied Oriental religions such as Buddhism and Hinduism and became widely familiar with the substances used in such religions to create altered conscience states known as trances. On these matters, he wrote a book called The Doors of Perception, which kicked off the psychedelic movement. The famous rock n' roll band The Doors was named after the book. Huxley's eloquence and charism made him a kind of intellectual patron saint of conscience altering drugs, the consequences of which America and the world at large experience to this very day. Here he can be watched being interviewed by famous journalist Mike Wallace https://m.youtube.com/watch?v=alasBxZsb40 .
The rabbit hole, however, goes much deeper. Huxley wrote a biography of famous Catholic priest Urbain Grandier, who was burned alive by his own correligionists for practising withcraft. Is it to take things too far to suppose that, while maintaining his façade as a beloved children's books author, Huxley was able to combined his studies on Hinduist devil-worship and his research on Grandier's Medieval withchraft to make demonic forces do his bidding and assure his success at Hollywood, where he worked as a scriptwriter? If you don't believe so, maybe you should read this hair-raising report of the visit by an Evangelical author who originally did not know who Huxley was to the Buddhist Monastery in which founding he was instrumental and decide for yourself if you choose the blue pill or the red pill. https://midwestoutreach.org/2019/05/18/thomas-merton-the-contemplative-dark-thread/
Don't highlight is kinda funny because it announces to the world that you'd rather not announce to the world this opinion. I think we just have to accept the nature of social media. Anything you post can go viral at any time after you post it. Could be getting highlighted, could be ten years down the road at a job interview. That's how it works, highlighting or not. We should keep highlighting as is because it keeps us in touch with the reality of how online forums work.
My comments at the end of the survey, because I think people not Scott need to hear them.
1) IANAL, and I think you should get actual legal advice on this because the idea that "pointing out a potential investment opportunity" is illegal doesn't pass the smell test with me. But hey, I think growing my own corn and feeding it to my own pigs isn't "interstate commerce" so what do I know.
2) There is, in my mind, more ethical issues with promoting your buddy's "non profit" (from which the buddy draws a 6 figure salary, or is using to build points for their kid to get into Harvard, or what ever) than in promoting a stranger's business. I think the glorification of "non-profits" completely misses how modern society/economics treats these as wealth & prestige builders for the individuals employed by them, and is another casualty of the demonization of normal business and investing.
3) In any case, you should say if (known to you) a friend/relative/ex/employer is involved in any such promotion. To me, that's the only ethical hurdle beyond 'don't take money for promoting things unless you disclose that' and 'def don't promote things you don't think should be promoted, even if they pay you for the promo.'
4) I like the highlights reels, as they often bring out more conversations. They are not a complete substitute for reading the whole comments, which is occasionally enlightening as to what *doesn't* get promoted.
5) You should never promote something in a malicious way, or in order to cause issues for someone.
6) What we say is our own responsibility. Period. You are not the boss a' me, and you don't get to police what I say. I am a grown adult and I do not yield you that power.
The covid infection and hospital rates in the US have both dropped steeply over that past two months. The time correlates with when we started vaccinating people. But to (non expert) me, it seems like it has dropped way faster than we have been vaccinating. I haven't found any news stories that say whether this drop was due to vaccinations or maybe due to post-holiday social distancing or maybe even due to some amount of herd-immunity.
What's the consensus here? Why have cases/hospitalizations dropped so precipitously in the US over the past two months?
People should just adopt some quick signal they can indicate if they don't want their posts highlighted, something like an asterisk or DNH or whatever.
> I’m a little worried about the Comment Highlights posts, because they broadcast things people said down in the comments where nobody would ever read them out to the entire Internet.
Just wanted to chime in and say 1. I don't think there's anything wrong with how you've handled it so far, 2. I do think this is a fair/nontrivial concern which is worth considering, and 3. I don't see any easy solutions, so the answer might just have to be to accept it?
On the last comment highlights I was in this *exact* position. I wrote a hasty/unedited post about Josh Hawley (that basically amounted to a rant) under my real name, and was slightly surprised to see it featured more prominently. Of course, I stand by the general idea, but it was written kinda impulsively out of annoyance, so if I had thought more than a couple of people would read it, I would have written it very differently (and been careful about the argument), and I wouldn't have "chosen" to have it featured.
To be clear, I don't personally mind, and if this had actually troubled me I could have tried to message you (to omit the name, or remove the comment). But I can at least sympathize with how it might take someone by surprise, even though in retrospect it seems fairly obvious (it's an open discussion thread, the blog generates discussion by highlighting certain comments, if you don't want to be associated with something you post, don't put a name on it?).
TLDR: the norms seem pretty obvious and straightforward (it's an open/public blog, of course any comments might be featured to generate further discussion), but I can still see how someone would be surprised.
On the other hand, I just don't see any easy solutions! If you omit the name of the poster by default, I assume some folks would feel like they weren't getting "credit" for thoughtful comments they made. Nor do I think it's feasible to individually ask posters for permission.
I think the easiest improvement is just occasional mentions of these norms. For instance, on each of these "Comment Highlights" posts, I'd add a quick note explaining 1. this blog generates further discussion by highlighting comments, 2. if you want a featured comment removed, message you, and 3. if you don't want a comment featured in the future, include a note in that comment (or maybe it could be in someone's profile).
Obviously, you will still get occasional "surprises" by someone who is unfamiliar with these norms. But I don't think that edge case is a particularly big deal, it seems overall perfectly reasonable to me for a blog to highlight comments that are written on that blog. And as long as you include that occasional note to the "Comment Highlights", it would only ever be a worry for someone who was new to the blog (and growing pains like that are inevitable for any community norms).
Regarding the DNP company, I think it's a trust issue. Theoretically, there's nothing wrong with saying "this company looks cool and you can invest in it here", but in practice nine out of ten times someone with a platform says that kind of thing it's because they have some financial stake in doing so rather than because they legitimately think it's cool. I believe Scott when he says he just think it's cool, but for anyone without a strong prior about Scott's trustworthiness, the reasonable thing to think upon seeing products shilled is "Oh, one of *those* writers."
I'm not a gamer, but I'm looking for action video game recommendations (for PlayStation 4, Nintendo Switch, or Steam) that I can use for brain-training my middle-aged brain. Ideally, the game should dynamically increase in difficulty as I get better. If such a thing doesn't exist, I would like something that can be set at a super-easy level (something that would insult a five-year-old) and then ramped up manually. I tried "Call of Duty: Black Ops", but it was too difficult even at the easiest level. I don't have great eye-hand coordination, and my reflexes aren't the best. I'm not interested in walking simulators, puzzlers, or similar.
The problem in #6 seems like something you could automate pretty easily. Get a friend with programming skills to make a quick program for you. You put all the names into a list, then when you want to highlight a comment, have a search algorithm check if the name's in the list.
RE: Book reviews, I'm good with posting them sans names for blinding purposes; does that mean if mine should make the cut I shouldn't link friends to it from outside the blog until after they've been unblinded?
I read the Vox piece and got pretty much the same impression from it that I got from the original malaria-vaccine story. When I posted it elsewhere, my prefatory comment was "this is really preliminary, only in mouse studies, but it looks interesting."
A lot of people around here care about housing and urbanism and related economic issues, and he's probably the best economist working on this stuff. He's got the same "just plainly explain economic issues in a simple, non-moralizing technocratic way" that I enjoy about ACX or Zvi's blog. I used to recommend his book to people, but he manages to get pretty much everything in there across in a 30-60 minute youtube lecture.
I can't believe they canceled a book about sitting by a pool and maybe catching something big. Besides it's dedicated to his Dad! (I have happy memories of fishing with my dad.) Maybe it's the fishing the woke left objects to? (cry or make a joke.. my two options.) From what can tell the book was pulled because it has Eskimo fish. And Eskimo is no longer used. (This is news to me.) They would now be called Inuit fish. Is there some harm done by having Inuit children learn they were once called Eskimo? And what joy will be lost without the rhymes and illustrations. Can anyone make an argument for pulling the book? I wanted to close this rant with his words from “On Beyond Zebra” (also pulled)
“The places I took him!
I tried hard to tell
Young Conrad Cornelius o'Donald o'Dell
A few brand-new wonderful words he might spell.
I led him around and I tried hard to show
There are things beyond Z that most people don't know.
I took him past Zebra. As far as I could.
And I think, perhaps, maybe I did him some good...”
I take two grams of Metformin a day for anti-aging reasons. From the Washington Post: "researchers noticed that diabetics who took [metformin] outlived non-diabetics who did not. Moreover, metformin had shown an effect in separate studies against each of the three diseases [dementia, heart disease and cancer]"
Scott, please consider doing a metformin, much more than you wanted to know post. I think taking metformin has the potential to be a big win for rationalists.
There are ways for book reviews to be self-deanonymizing to various degrees ("As a person named John Doe, I found Mary Roach's "Stiff" highly upsetting" etc.), so watch out for this if you end up going ahead with the "publish everything anonymously" plan.
On DNP ethics: I briefly considered going into more detail on safe-ish DNP dosing strategy and decided not to, primarily due to the fact that comments aren’t editable so if I accidentally say something wrong I can’t cleanly go back and fix it.
[Also this topic seems like an ACTUAL Copenhagen Interpretation of Ethics situation. If I give advice that makes an unsafe thing seem less dangerous so people do more of that unsafe thing, whose fault IS it when they hurt themselves doing it?]
I can’t imagine giving the same amount of attention to 100 book reviews. It seems likely that they’ll get attention based on superficial interest or the order they’re posted.
One possibility would be to post them all at once and ask for each volunteer judge to read them in a different order, based on a randomly generated list. You can keep going down the list until you get bored and stop judging.
The interactive fiction competition (ifcomp) seems like good prior work for an Internet competition based on judging.
Re: the mass of book reviews - suppose there was a call for volunteers to read some subset of the entries and respond with yes or no to these questions: (1) Does this seem like a good fit for SSC/ACX? and (2) Is the writing of tolerable quality?
Would this meaningfully whittle down the number of entries to review in depth? I would volunteer to read a dozen or so with those questions in mind.
I am looking to collect databases from real businesses and business-like entities, including those that have failed or otherwise become "past-tense". Read on if you or someone you know might have access to such things.
I'm a software engineer, specializing in data systems (i.e. a data engineer), with about 16 years in the industry under my belt. Something that's always frustrated me about the way that we design and build systems, is the way that knowledge fails to diffuse through the industry, because we don't _study_ what we do, and especially we don't study our failures.
As an example, the 2010s witnessed the full hype cycle (rise and fall) of "NoSQL" databases, such as MongoDB, Cassandra, DynamoDB, Riak, Aerospike, and many others. Did they turn out to be any good? Individually, in local circumstances, some engineers know the answer, or at least _an_ answer. Collectively, we have no idea. This knowledge only spreads as the primary sources write blog posts (mostly terrible), or move on to new jobs and tell stories (distorted by all sorts of biases). What we *should* be doing is studying what was actually built, out in the open, where everyone can see it if they're interested.
Additionally, I find it very difficult to teach other engineers about data systems, in a scalable way, without open example material. There are many online courses in SQL and things of that nature, but they always deal with trivially small, trivially clean data sets, without any of the richness or messiness of Real World Data. Many years ago, my own skill in dealing with data grew by leaps and bounds the instant I was exposed to real business data and asked to solve real business problems with it.
To these ends, I am looking to collect real business data sets. I use the term "business" loosely, in the same sense that engineers often say "business logic". Non-profits, community efforts, personal side projects, these things all count. The key thing I'm after are custom-built databases, meaning they either started from a blank MySQL/Postgres/MongoDB/etc, or heavily customized an off-the-shelf system like Wordpress or Salesforce.
I recognize there are thorny issues here with respect to intellectual property and personal data privacy. I do not expect anyone to just hand over a database and wish me well. We would have to work something out, whether that's an NDA, or thorough anonymization, or whatever.
In any event, if you possess a data set like this, and *might* be willing to share it for research purposes, please reply here and we can figure out how to connect and discuss.
(This is the third time I've posted this. My plan is to re-post it periodically. If that runs afoul of any rules, written or unwritten, let me know and I can adjust.)
Addendum:
I recently became aware of this book https://fightchurnwithdata.com/ which might seem unrelated, but I mention because it contains specific details about how to model measurements of subscription customers and their behavior. This is the _kind_ of thing I'd like to be able to produce from my research: specific guidance on how to model specific business concepts and processes.
One of the commenters in this thread mentioned something rather illuminating (for me):
> This sort of false sense of urgency is very common amongst conspiracy theories - the idea that the aliens are coming, that the Illuminati have control and are about to enslave us all, that the Storm is coming, that all police are secretly racist, that we're going to run out of food and all starve.
Does anyone have a good recommendation for books on the pre-American history of slavery? I've read several books on American slavery, but am looking to expand. It seems like every single book out there is about American slavery. I'm looking for something that is more apolitical, just the facts.
I think a lot about how to create effective organizations. Let's say that we want to run a big organization, a cooperation, non-profit or government. There are a couple of failure modes.
One is lack of structure, where every decision is taken willy-nilly and nothing is written down. This trends towards nepotism and corruption, where decisions are based on flattery on who-knows-who in the old boys gang.
Another is bureaucratization and procedural-ism, where everyone does things right but no-one does the right thing. This seems to be something that comes creeping into an organization and is very hard to stamp out without burning things to the ground and rebuilding. Ossified companies fall to creative destruction, governmental agencies can seemingly live on forever. One reason that is often used to explain bureaucratization is that organizations respond to violations and accidents by adding more procedure to avoid repeats in the future, inadvertently stiffening themselves. Another is the middle-management and bureaucrats that were hired to perform important work tend to invent tasks for themselves to do during downtime, and these tasks don't go away when things gets busier. I guess there are many possible just-so stories about the origins of bureaucratization. Does anyone know of more rigorous takes on this phenomena?
A third failure mode is to have both: Strict regulations that can be skirted if you grease the right palm or know the management. This is akin to organizational anarcho-tyranny.
A fourth failure mode is to have a little bit of both: Just enough regulations to create extra work, just enough personal touch that the extra work doesn't matter and the management gets to decide in the end anyway.
What are the success modes? One seems to be to hire the right people and then leave them alone to do the job. Presumably they will create the level of bureaucracy needed to get the task done and no more. I'm not sure this is the secret sauce though. Is anyone aware of an example when "hire good people and leave them alone" failed?
Another success mode might be "meritocracy": Simple, relevant and transparent rules/checklists/procedure/criteria where it's applicable, human judgement at the lowest level everywhere else. Reward results, but beware of Goodhart's law. (This is what I personally think is the best police for governmental agencies.)
I would like to plot this on a two-by-two but it seems like one access becomes "good-vs-bad" which isn't that useful.
Slack is probably another important factor: lots of slack can create amazing results (Bell Labs?) but can also create an adult daycare. Little slack leaves little room for experimentation, but it's also my impression of how some very successful companies (e.g. Tesla?) are run.
Thoughts and comments on effective organizations? Was "Good to Great" correct all along? Am I missing something important?
My answer on the investment question was "It was completely appropriate for you to include that the company wanted investors and to provide an address where people could get in touch with them".
I think it would be very bad for EA (or society in general) to add more obstacles in the way of starting altruistic for-profit projects, signal-boosting and discussing those projects, investing in those projects, or otherwise trying to get them to succeed.
Also, adding norms is just really costly in general, in the same way that it's costly to keep accumulating more and more laws. Law #1 of the Internet: people generally need to relax way more on the Internet and treat fewer things as norm violations. Otherwise we risk slowly choking the life out of cool things by demanding those things follow more and more norms.
If there are no legal issues, and people aren't lying, then I think people should just talk about investment opportunities however they want.
I suspect many people with highlighted comments don't just feel ok with it but proud to provide insight and thankful to get share that insight with more people.
Survey taken. I will add my experiences of the weight loss drug, the company shilling, and the comment-highlights posts here.
1) On the weight-loss drug. I recall reading the first paragraph of your post, and feeling excited (reliable weight-loss?)and beginning to plot to maybe obtain some of this drug. Then I read the next two paragraphs and felt almost stung by the side-effects (get cataracts and then literally burn to death from the inside?) and was far less likely to ever take the drug than had I not read the post. I think you are not a responsible adult if you read that post and then want to take the drug.
2) On shilling the company, I didn’t read the version of the post with it in. Nonetheless I’ll add my two cents: I think it’s cool if you do that thing once in a while, but I think if you did it regularly it would be less fun, and also your life would get worse because people would be asking you for stuff ALL THE TIME AND NEVER STOP.
3) The comment highlights posts are great! Otherwise, I do not read the comment sections. (More specifically, I do often search for the threads you Scott are in, and read only those.) The comment highlights are always great, way better than what I get from reading the comments myself, and if you didn’t post them I still wouldn’t read the comments, it’s too much noise to signal. Please don’t take them away from me.
(My random guess is that the class one was unusually sore for a user and they were kinda mad at you? But that you’ve been doing this for years without much pushback? If this overly specific story is true I suggest maybe just being a bit more sensitive on very political comments/posts and then otherwise going about business as usual.)
Is the following statement about mRNA vaccines at all correct, or did I go way of base somehow:
"mRNA vaccine technology is a breakthrough because the same delivery system can be used for almost any RNA sequence, which means mRNA vaccine candidates can be developed almost a soon as a pathogen's DNA is sequenced. Traditional vaccines are hard to develop because making a pathogen that is too damaged to be infectious, but intact enough to provoke an immune response, is a hard problem that has to be re-solved for each and every pathogen."
(Now that I think about it, why bother with mRNA at all? Why not cut out the middleman and inject the spike proteins directly? Does the immune system not panic over random proteins in the bloodstream, where it does panic when those same proteins are being synthesized by foreign mRNA?)
Now that the book review contest is closed, can there be a master list of titles/mass reply email to confirm the ones that were received? (Or, selfishly/personally, can I ask if mine was received? I sent it in a week or two before the March 1st deadline. But I think there was a discussion of this in a previous OT, so it might still be a broader concern.)
Also, while I don't have a problem with the temporary anonymity suggestion, one thing to consider is how it might affect post authors' commenting on the reviews. I presume that at least some reviewers would want to join in the debates/discussions in the comments before the final judgement.
There was a really interesting NYT article on the Vertical Club, an upscale 1980s NYC gym, a couple weeks ago. (https://www.nytimes.com/2021/02/24/style/vertical-club.html) Relevantly here, it provides additional evidence for Fussell's argument that fashion moves from the lower to the upper classes:
"Aside from those office gyms, New York fitness spaces were often considered “grimy” bodybuilding spots or cruising venues for gay men, stigmatized during the AIDS crisis. The Vertical Club, its name suggesting “the up and up,” brought in the light.
The obvious glamour of the club tinctured virtue with vice, “bolstering that image that this was sexy, this was a singles club, this was where people went to meet each other,” said Natalia Mehlman Petrzela, another fitness historian. “But at the same time they had to sanitize that image, because they wanted to attract this white-collar clientele.”"
I miss the Report button. Since it doesn't look like that's getting implemented anytime soon, is there any way to contact Scott or a moderator about a particular comment?
We've recently been modeling and forecasting elections in Australia, and our latest one, on the Western Australian state election this week, seems to have been well received in the local psephology community. We put up two articles on how the modeling works and how we intend to judge it after Saturday.
Would this be of interest to any Aussies curious about politics about statistics over here? We'd love any feedback or engagement that people were willing to offer.
My conviction is that all political posturing that doesn't rely on material interests is just larping, and people who "get into" a specific kind of politics or "grow out of" it just need to get a real, more fulfilling hobby that doesn't damage their mental health.
Why the clause about material interests? Because you cannot "grow out" of having a shitty wage or a shitty job or harsher treatment by the justice system, it's something that's imposed on you until your material conditions are improved one way or another. In this case alone political posturing and organizing is useful.
An interesting case is people whose material conditions *are* dependent on their political posturing, i.e. "thought leaders", op-ed writers, professional opinion havers, blog monetizers, or in other words grifters. It is useless to engage such people with the purpose of changing their mind, since their very posturing is part of their brand and they're not going to threaten their livelihood to change it. (It can still be useful to confront them if they are spreading damaging lies.)
I've been thinking a lot about obesity lately, and some of your statements on the post about DNP got me thinking about it again. In particular about how in 1938 obesity was thought by most to be primarily a cosmetic problem.
In 1999, in article published in JAMA* concluded that obesity caused approximately three hundred thousand deaths per year.
According to the CDC** between 1999 and 2018, prevalence of obesity went from from 30.5% to 42.4%, and the US population went up by nearly fifty million as well.
Doing some back of the envelope math (and assuming that change in prevalence is not a consequence of changing definitions or better detection), we're
looking at an additional 40-54 million obese adults in the US. If the deaths track with the increasing population and obesity rates, then in 2020, between 450 and 550 thousand people died of obesity related causes.
Or roughly the same total mortality as the COVID-19 pandemic.
There are some pretty sizable error bars on this estimate, but even if I take the lowest reasonable guess and put the deaths at 250000 per year, consistent since 1999, that's still enough to make it the third highest cause of death in the US, behind heart disease and cancer (both of which obesity either increases the likelihood of, or massively complicates treatment).
I don't have any good solutions to this (fatshaming clearly hasn't worked), but I think that if we took the problem a third as seriously as we have taken the pandemic, we would get a lot more QALYs out of it for the effort.
Failed/failing ad agency, consultant, tech entrepreneur = least sympathy.
Is this purely a subconscious reaction to the perceived risk/reward ratio of each business owner? (Which would make this dynamic fair in a way). Is it a class thing? Or is it arbitrary?
Is this dynamic simply because we have more face to face interactions with analog small businesses?
Will I ever be able to start a Go Fund Me for my digital consultancy should I need to, or will it always be, “try again, loser”?
I've been checking at ebay, and the Seuss books are still available, though at high prices. I don't think ebay is actually taking the book offers down.
Anyone have predictions of whether the deprecated Seuss books are even vaguely sensible investments?
Also, if you just want to see what's in the books, there ae podcasts of people reading them on youtube.
Bret and Heather made McElligot's Pool sound pretty good, so I watched it. I'm surprised there weren't vegetarians going after it. It has some Seuss charm, but the lesson seems kind of stupid if taken literally-- imagining entertaining outcomes doesn't redeem an unworkable project. On the other hand, imagining entertaining outcomes is a fine leisure activity, especially if you're Dr. Seuss.
On yet another hand, I could read it as exploring your subconscious vs. someone who tells you there's nothing to the world but what you can see.
If this is true, I wonder whether suspected post-viral syndromes such as chronic fatigue syndrome, chronic Lyme disease etc might be able to be treated in the same way using vaccines for these diseases.
Given that Moderna was able to design its COVID vaccine in two days, it seems plausible that we could have vaccine candidates for such diseases very quickly. If we had to run a normal clinical trial for an EBV vaccine that would take a long time and be expensive (which is probably why such diseases have languished without vaccines). But if it's being trialed as a therapy for CFS, that would be a relatively simple trial to run (just randomize 500 CFS patients into placebo and vaccine).
Speculating more, is it possible that the widespread availability of flu vaccine reduces that infection's association with apparent post-viral syndromes (i.e. people end up getting treated for their postviral illness every year anyway).
I've never thought that wokeness would all "just pass." I've always thought that unchecked, its core assumptions would have a tremendously corrosive effect on society. Yet when I read Scott, it seems that he lives in a world of blissful ignorance. Maybe it's because I live in a city that's become a pigsty, with gibbering crazy people on every street corner, and I see its downstream effects every day - I don't know.
Am I alone? Am I the only one who thinks we're living in a form of End Times (I use that religious phrase in a completely secular way).
Here's an idea - perhaps not novel, but an idea - to help with the quote-problem. Why not just make opt-in? Most of these comments are made by people with special interest or experience in the fields of choice. At the end of your post, add a blurb to the effect of, "If you share something insightful and are okay with being commented in a comments post, please note it at the top of your post."
On warm-bloodedness: because of reptiles vs mammals and birds, I've associated warm-bloodedness with high baselines activity levels. Think of a lizard just sitting there versus a mouse scurrying about. But insects seem to be extremely active, think of the mosquito that flies around you constantly. Insects are ectothermic aka cold-blooded if only they had blood [1]. And the general square-cubed law suggests smaller animals like insects are generally more cold-blooded-like. So how do insects maintain their high activity levels? Or are they not as energetic as I think (maybe I only see them when they're in the mood to bug)?
[1] Okay, mosquito's do have blood, it's just not rightfully theirs!
Am I the only one to whom the name of this blog in all-lowercase-no-space (as in the URL) reads like "astral cod exten"? (I guess that's because the string "ex" is way more common at the beginning than at the end of words)
You're a talented director in your 30s, and a major Hollywood studio has given you $200 million to make a new "Alien" sequel. Ridley Scott and James Cameron have no power over the project. Your only mandate from the studio execs is that your movie be so iconic and memorable that fans forget about the Colonial Marines/pulse rifle/LV-426/Ripley/Hicks paradigm for good. The whole visual and technological aesthetic of "Aliens" needs to be overshadowed by the uniquely different one in your film. Moreover, you can't copy the style of any other Alien movies or use any of their characters.
What do you do? What kinds of characters do you pick? What distinctive technologies, weapons, equipment and space ships do your characters have? What choices do you make with respect to cinematography?
Just had my first solo experience with the American healthcare system, and it was weird. I was vomiting constantly for 4 hours, and thought I might need to see a doctor. I went on my health insurance website, and they offered telemedicine visits for $50 or I could talk with a nurse for free. I decided to talk with the nurse, he was friendly and said I needed to go to an urgent care center after asking some questions about my symptoms. I asked about the telemedicine, but he said for my case in-person care would be more beneficial, and he sent me a list of in-network urgent care care centers. I go to the nearest one, and it is a generic office building with no signage indicating a urgent care center. I go in, and the security guard at the front desk says this is the correct address, but there is nothing like that there. Then I go to another center on the list that I know exists, sit in a waiting room, sit in an exam room, then get my vitals taken by a nurse. I had to explain that I probably did not need a strep test for my sore throat that appeared immediately after vomiting. Then the main provider, who is an NP, comes in, asks me some questions, is very interested in when my last poop was, pokes various places in my stomach, then says it is probably a virus and gives me a prescription. Turns out she gave me Ondansetron, which apparently costs around $10 a pill but my insurance negotiated it down to $0.75. Still don't know how much the office visit will cost me. I know things like this are old news to most of you, and I knew the US healthcare system was broken before, but it still strikes me as odd how totally unlike other capitalist transactions this is.
Is anyone here of non-binary gender? If so, I'd like to get your impressions of this Steven Crowder debate with a transgender person: https://youtu.be/w0zv62EBfMU
To be clear, I know little about the trans rights movement and have never spoken with a trans person about them. What little I do hear about it tends to come from news articles and social media posts about things like biological males competing in female high school sports and people being punished for mis-gendering trans people. These news reports have never impassioned me, and I've had the building suspicion that the trans rights cause is being misrepresented (at least in the information channels I frequent most often), and made to look bad by a small number of its most extreme supporters.
In the Louder with Crowder video, I thought "Madison" was very intelligent and well-spoken, and kept their cool better than Crowder did. They got hung up on the issue of whether using words to deliberately misgender a person counts as "violence," and I thought that debate over that was frustrating.
If you are trans, this is an earnest invitation to help me understand your outlook. The Crowder video is (unfortunately?) the best place I can think of to start.
I really like the link back to the comment in the Highlights, since then I can get context and see if others disagreed. I wouldn't mind seeing names removed, but I would be sad to lose the whole link (I realize people can just click the link to get the name, but I'd expect 90% of people just read the summary)
Could someone please click on my profile and see if my bio appears? I don't see anyone else's, but I'm not sure if that's because I'm the only one who wrote one, or if its because ACX/substack is just not displaying it.
Is this a thing: rather than rule-out for psychiatric diagnosis, probabilistic diagnosis? The whole idea of medicine is to diagnose and treat, but what if some things are fuzzy (but can still be dealt with)? Elaborating makes it sound stupid, but I've wondered for a while if this isn't closer to what psychiatrists do. Some medical things are much more clearcut - there is a cartilage tear or there isn't - but psychology/psychiatry seems to have room for simultaneous states.
Question for people with some experience in leftist spaces.
I've been frequenting some online leftist communities more and more recently. I'll find myself agreeing pretty strongly with the overall sentiment and mostly positive about their proposed policy but then I'll get whiplashed right the heck out again when someone mentions obviously how the US intelligence apparatus secretly controls everything, that various well known figures are all FBI (or whomever) plants and that everyone who doesn't entirely agree with them is controlled opposition. Instead of being laughed out of the room, this idea is almost always taken seriously.
I find this idea dumb enough to make me distrust the overall movement even though I'm largely sympathetic to its stated aims and goals. Yes - I've heard of COINTELPRO, no its effects and ramifications aren't some all-consuming feature of modern politics.
Have I just gotten unlucky, or is this kind of stuff really just a feature of modern-day leftism?
I'm a regular reader who normally doesn't read comments in general. Here there are so many that they intimidate me so I tend to ignore them. I hope you are able to keep doing the comment highlights because they are pretty great!
(I think this is the third time I've tried to post this; I hope the other two aren't annoying you in some kind of comment queue.)
I propose that AstralCodexTen should be abbreviated as ACT, and SlateStarCodex be respelled as SlAtesTarcodex and abbreviated SAT, in honor of the two most prominent standardized academic tests. (This originated as a joke at an earlier SSC meetup that I attended when we briefly discussed calling it something different because of the blog name change, but I now endorse it semi-unironically.)
Being faced on a daily basis with the awful Covid19 death figures, I wondered what the total daily death figures were. I also wondered why the daily figures of all deaths was never mentioned - wouldn't it give better context to the Covid deaths? Not being a epidemiologist or a statistician, I searched online and found that the figures for Italy's deaths over the last few years were available at https://www.macrotrends.net/countries/ITA/italy/death-rate?q=England+Death+Rate+1950-2021 I knew that Italy has been badly affected by Covid19. The daily death rate for Italy for the past 3 years is 10.4 to 10.7 deaths per day per million of population - that is, about 640 deaths per day. The rise from 10.4 to 10.7 may well be Covid related - if so, it means an additional 24 deaths per day are the result of the pandemic. The increase doesn't seem so much - and may be a statistical variation - and yet Italy is regarded as being hit hard by Covid19.
It really seems like substack could solve this "don't quote me" problem for you. Maybe they allow people to click a config in their profile to indicate that they don't want highlights, and it would show up as an ❌ next to their name. You could imagine such software also allowing users to choose gender, or to choose background and get a 🧬 or 🦿, idk. They have product people to scope this stuff.
I say that as a software dev manager - your use cases are good data for them, since you have developed a successful community before. This sounds pretty easy for them to solve. People seem to like emojis next to their name, so even if you don't have any option on your blog other than ❌, other communities might use them. Sports blogs with team mascots and all that.
Just pointing out an option other than "I solve it" or "my customers solve it". You have customer support now!
I surprised myself and got ACX-Simple (a rewrite of the Substack user interface from scratch, without any of the Substack stuff aside from the CSS) to dynamically load comments. All the pieces just fit together.
So I think this is of interest as indicating what a national government thinks about AI and its implementation. On the one hand, they are not considering the kinds of disruptive existential threat that discussions about AI here and in previous fora like to talk about - the Fairy Godmother AI that will bring forth post-scarcity Fully Automated Luxury Gay Space Communism, or the Bad AI that will turn us all into paperclips. They seem to imagine it as yet another new technology that will provide jobs for humans, provided they have 'learned to code' as it were; certainly they are not envisaging AI as replacing humans altogether in this area :
"17. Most of the AI training programmes worldwide are based on the fundamentals of data
science, machine/deep learning, and ethical aspects using generic use-cases and examples.
While they build the fundamental skills of the AI workforce, a significant gap continues to
exist in the immediate employability of the trainees and their ability to apply their AI skills
directly on real-world research and business problems in various domains such as
environmental sciences, financial services, biotechnology, material sciences and
autonomous systems. Thus, there is a need to customise the curricula of training
programmes to address the specific AI skills required for such specific domains, particularly
focusing on relevant datasets, problems, AI algorithms, and implications on privacy, ethics
and security aspects.
18. However, a similar concern arises with regard to the need for specialist curricula in
technical fields. For instance, the methods and tools for dataset preparation, AI algorithms,
AI solution deployment and ethical/privacy implications for areas such as environmental
sciences, healthcare, material sciences are radically different. Thus, developing AI skills for
each of these domains needs customisation of the curriculum of the training programmes to
address the requirements and specificities of the problems in each domain."
On the other hand, they do have some intimations of Fairy Godmother AI, although again - the AI will improve diagnostics and treatment so much, nobody will be able to claim disability benefit because we will all be deemed fit to work - there is nothing about "nobody, able-bodied or not, will *need* to work because AI will do it all!":
Right to Life, Liberty, and Security of Person - AI-based diagnostic systems enhance
the enjoyment of the right to life by making accurate, high-quality diagnostic services
more widely available.
• Right to Desirable Work - The improved health outcomes that AI-based diagnostic
systems are likely to produce will reduce the number of people who are excluded
from the dignity of work for medical reasons.
• Right to an Adequate Standard of Living - By detecting diseases earlier and more
accurately, AI-based diagnostic systems will improve living standards and quality of
life.
• Right to Education - Should AI-based diagnostic systems deliver on their promise,
fewer people will be excluded from the enjoyment of the right to education due to illhealth.
So there you go - no more skiving off contributing to society and leeching on others by whining that you are physically or mentally disabled and so unable to work, you slackers!
Hey! Sorry if this is too much noise, this is mainly a test/greeting post to familiarize myself with the commenting system here… interface is pretty different from what I'm used to. Let's see what happens when I hit the button…
If you were writing the Substack comment interface from scratch, how would you want the web page to tell you about new comments? What kind of "workflow" do you want?
Here are ideas to brainstorm. Some are mutually exclusive, some build on each other.
* The title changes and/or the tab thumbnail gets a big blue dot on it to attract your attention when there are new comments.
* You press a button to manually check for new comments instead of them loading in the background.
* You are told that there are new comments, but new comments are not dynamically put up on the page until you press a button that populates them all.
* Placeholders are put on the page in the place where new comments go (I think the default Substack behavior) for you to explicitly load, one at a time.
* No background loading of comments. You reload the page and new comments get the "∼New∼" tag, like on the old SSC.
* New comments are automatically loaded and put in place, with some tag like "∼New∼" so you search for them.
* * You have a way of clearing all those "new" tags. (But how do you know that you have read everything if background loading is happening?)
Slight bonus points if what you suggest works natively with ACX-Tweaks, but don't lock yourself into that. Dream big.
can the options for why it was inappropriate be multiple-choice (checkbox) instead of single-choice (radio button)? i have multiple reasons why i think it was inappropriate
I don't know whether it would be worth the trouble for you, but you could ping people about specific comments you want to quote.
I'm a bit confused about the warm-bloodedness thing. I was trying to look this up actually because I've just been kind of confused generally as to what the distinction is supposed to be between endotherms and homeotherms-but-not-endotherms. (I am not any sort of biologist, if that wasn't clear.) Like, OK, poikilothermy seems obviously distinct, but once you're within homeotherms, I was like, I don't really see an obvious distinction between these different methods of maintaining temperature that should distinguish some of them as "endo"? Especially considering that like a main method of thermogenesis is shivering, which -- as a muscle-based mechanism -- seems to be getting pretty close to a behavioral mechanism, you know?
It's the shivering thing that kicked off me looking this up, really. Because I keep reading that non-shivering thermogenesis, which is based on this uncoupling, happens only in brown fat cells, and that adults don't have much of these?? So they get all their heat from shivering?? And that's just like... that can't be right.
IDK, I am basically clicking around on Wikipedia here, so some parts contradict other parts. Like, oh, maybe all cells have UCP1, just brown fat cells have *more* of it. But other parts say no it's only brown fat cells. Or maybe adults have more brown fat cells than thought. I am confused!
Because like yeah generating heat from uncoupling, that's pretty distinct, much more so than shivering! And it also, y'know, matches everyday experience, where you don't start shivering the instant you're colder than is comfortable. But it's really confusing to keep reading that adult humans don't have much in the way of non-shivering thermogenesis going on. Like, huh? What's up with that statement? Where does that come from? Or is there some way I'm missing that it could actually be true??
In theory, making friends online should be easy. Instead of luck and circumstance of the physical world, the virtual world should give us access to the few most compatible friend-candidates out of billions.
And yet, I still default to the physical world for finding new friends.
Question 1: where, online, have you found "true" friendship and how?
Question 2: I know that some have tried (and failed) to create a social network for the non-masses. Do you think there is opportunity for a social network for people with long attention spans that rewards the building of deep relationships? If yes, do you think it should be an open network (like Reddit), or more akin to a dating/matching app that filters the billions down to the most compatible? Ex. If love of Nietzsche is non-negotiable, would be easier to filter by that first.
Re: comments. Just do what you're already doing, Scott. I think they add something special to the Substack.
The addition of "don't highlight" is more than enough to guarantee anyone who doesn't want their comments seen (on a public forum!) won't be surprised. You could ask Substack to put a small text under "Discussion" with a disclaimer, if you want to make absolutely sure.
I've seen a few people saying that they see subscriber only posts in the rss feed, even though they aren't subscribed. While it seems that substack doesn't provide a non-subscriber feed, I've taken a shot at filling that void here: https://pycea.tk/acxfeed . It's a direct clone of the official feed, just with sub only posts removed.
On comments, this seems like a rare social problem with a technical solution.
You've already asked Substack for a lot of changes to make these comments more like WordPress. It seems to me this is another feature you'd like: either a checkbox to mark a comment as highlightable / non-highlightable, or a checkbox at site-level per user, whichever's easier. Then on the admin side, comments can get different background colors depending on whether they should be highlighted or not (or something).
Might take a while to implement depending on how many other higher priorities there are, but it would save you from having to manually maintain lists of people who want comments highlighted.
Maybe it's just me, but the simulation hypothesis (that we're essentially living in a simulation by some superior intelligence) seems completely ludicrous, seeing how our univere very convincingly appears to show behavior spanning 30 orders of magnitude in time, 30 orders of magnitude in space, and a sheer degree of size, scale and detail that is completely absurd if you were trying to study something specific in a simulation. Am I missing something?
Since we're talking about politics, I found this 538 article about the link between qanon and white evangelicals interesting. https://fivethirtyeight.com/features/why-qanon-has-attracted-so-many-white-evangelicals/
Is there maybe a more general phenomenon where conspiracy theorism, or extreme politics in general, comes from a kind of frustrated religious instinct?
What measures do people take to obscure their identity online? Beyond using a pseudonym
On DNP: I feel like the survey missed a "It made me consider DNP but it did not make it look worth the risk". After all, watching video of BASE jumpers part of my brain wants to try that, but that does not mean I consider that a good idea.
Non-fungible tokens (explainer here https://www.theverge.com/22310188/nft-explainer-what-is-blockchain-crypto-art-faq). Fad or next big thing? I don't really understand why anyone is buying them
This is a little silly, but I'd like to add my novel to the Top Web Fiction website, and it looks like it's invitation-only. Is there anyway you could send me an invitation?
Claim: poltiical orientation can be determined (unreliably, but statistically different from random) via facial recognition. Here's the link: https://www.nature.com/articles/s41598-020-79310-1.
I'm assuming that somebody here will have a more informed take than my own, so I'm writing this in the hope that I get put in my place. Here are my thoughts:
First of all, it's already interesting (although I assume well-known) that facial expressions + head pose (it looks like these are correlated) do so well. Obviously these are choices, and choices reflect the culture of one's peers, but I'm not sure I'd have guessed that the bias is so strong. Relatedly, I definitely would not have guessed that "sunglasses" would be so poor of a predictor relative to, well, anything else.
I was personally surprised that the algorithmic accuracy was greater than human accuracy at the same task. After reading a bit, I think that I just haven't kept up with the SOTA of facial recognition --- apparently algorithms now beat humans in general. This is surprising to me, since I learned that our brains are pretty specialized for facial recognition, but I guess this is just my ignorance speaking. Fine.
Anyway, for the big picture: the attributes against which facial recognition was compared were all *voluntary*. Users had the ability to chose whether to look up or down in their profile photo; whether to smile or frown or wear sunglasses. There are many such choices, and including more of them (like more subtle aspect of the facial expression) might have yielded an even more accurate prediction. Facial recognition itself, though, will include demographic characteristics (*partially* controlled for by the study), but also things like "has this person had plastic surgery". What I really want to know is, how much of the predictive power lies with these uncontrollable characteristics, and how much lies with aspects of the facial expression that can be faked?
I now wait for somebody to tell me that the entire study is flawed because it didn't control for whether the pictures were taken at day or at night.
The problem with agreeing that sticking DNH = do not highlight or something like that on the end of comments is that someone can now search specifically for those tags, and ... if they have really no morals, write a bot that automatically reposts these comments as highlights on sneerclub or something.
I use an online alias here that's not linked to my offline identity or any other online ones, and so should you.
So ACX is becoming a competitor to Goodreads?
A possible solution for the comment highlight issue would be to make these posts subscriber-only. The obvious downside is that non-subscribers will lose some interesting content, and it kind of goes against your original commitment to make most of your posts free. On the other hand, this puts highlighted comments out of reach for search engines and the internet at large - so the threat to commenters' privacy goes way down.
I was under the impression that you were going to publish the best 5 book reviews and invite additional comments to help decide on prizes, kudos and glory. Are you now going to publish them all? All 100?
Ok, a) that’s a lot b) you’ll publish mine. Have you thought this through?
I'm worried that 100 book reviews are too many and will dilute the blog with boring content.
Maybe post 10/week as part of links posts, have people fill out weekly surveys, and then post the top 5 as their own posts?
Because honestly, I didn't come here to read a book review at the quality that I could write one ;)
The most politically influential person you never heard of? Is it possible an unassuming children's books author shaped the world we live in?
A strange thing happened a few days ago. My wife and I were at the rumpus room reading Who's Bashing Whom? Trade Conflict in High Technology Industries ( http://cup.columbia.edu/book/a/9780881321067 ) when our oldest son told us he had been assigned for reading a book by British writer Aldous Huxley. My wife and I had never heard of him before. What happened to Beatrix Potter, C. S. Lewis, Roald Dahl, R. L. Stine, Horace Greeley, Lemony Snicket, Lewis Carroll and G. A . Henty? Has the great state of California cancelled them? Anyway, I decided to ask my own questions and do some research on this fellow. Apparently, his books are required reading in many Democratic-controlled school districts, are very strongly opposed by concerned parents due to anti-religion and anti-family themes and sexual content ( https://en.m.wikipedia.org/wiki/List_of_most_commonly_challenged_books_in_the_United_States ) and (surprise, surprise! ) were widely read in the communist Soviet Union ( https://www.jstor.org/stable/3831583 ). The book my son was assigned to read proposes the abolishing families, replacing religion with orgies and banning the Bible and the works of William Shakespeare.
But who was this Huxley guy and how did he shape the world we live in? He was a scion of the Huxley family, founded by Victorian scientist Thomas Huxley, who was nicknamed Darwin's Bulldog due to his rabid championing Darwin's ideas. More than any other single individual, Thomas Huxley was responsible for the triumph of Darwinism, which led, in the 20th Century, to the Holocaust and the Gulag. Aldous Huxley, besides writing books, also studied Oriental religions such as Buddhism and Hinduism and became widely familiar with the substances used in such religions to create altered conscience states known as trances. On these matters, he wrote a book called The Doors of Perception, which kicked off the psychedelic movement. The famous rock n' roll band The Doors was named after the book. Huxley's eloquence and charism made him a kind of intellectual patron saint of conscience altering drugs, the consequences of which America and the world at large experience to this very day. Here he can be watched being interviewed by famous journalist Mike Wallace https://m.youtube.com/watch?v=alasBxZsb40 .
The rabbit hole, however, goes much deeper. Huxley wrote a biography of famous Catholic priest Urbain Grandier, who was burned alive by his own correligionists for practising withcraft. Is it to take things too far to suppose that, while maintaining his façade as a beloved children's books author, Huxley was able to combined his studies on Hinduist devil-worship and his research on Grandier's Medieval withchraft to make demonic forces do his bidding and assure his success at Hollywood, where he worked as a scriptwriter? If you don't believe so, maybe you should read this hair-raising report of the visit by an Evangelical author who originally did not know who Huxley was to the Buddhist Monastery in which founding he was instrumental and decide for yourself if you choose the blue pill or the red pill. https://midwestoutreach.org/2019/05/18/thomas-merton-the-contemplative-dark-thread/
Don't highlight is kinda funny because it announces to the world that you'd rather not announce to the world this opinion. I think we just have to accept the nature of social media. Anything you post can go viral at any time after you post it. Could be getting highlighted, could be ten years down the road at a job interview. That's how it works, highlighting or not. We should keep highlighting as is because it keeps us in touch with the reality of how online forums work.
My comments at the end of the survey, because I think people not Scott need to hear them.
1) IANAL, and I think you should get actual legal advice on this because the idea that "pointing out a potential investment opportunity" is illegal doesn't pass the smell test with me. But hey, I think growing my own corn and feeding it to my own pigs isn't "interstate commerce" so what do I know.
2) There is, in my mind, more ethical issues with promoting your buddy's "non profit" (from which the buddy draws a 6 figure salary, or is using to build points for their kid to get into Harvard, or what ever) than in promoting a stranger's business. I think the glorification of "non-profits" completely misses how modern society/economics treats these as wealth & prestige builders for the individuals employed by them, and is another casualty of the demonization of normal business and investing.
3) In any case, you should say if (known to you) a friend/relative/ex/employer is involved in any such promotion. To me, that's the only ethical hurdle beyond 'don't take money for promoting things unless you disclose that' and 'def don't promote things you don't think should be promoted, even if they pay you for the promo.'
4) I like the highlights reels, as they often bring out more conversations. They are not a complete substitute for reading the whole comments, which is occasionally enlightening as to what *doesn't* get promoted.
5) You should never promote something in a malicious way, or in order to cause issues for someone.
6) What we say is our own responsibility. Period. You are not the boss a' me, and you don't get to police what I say. I am a grown adult and I do not yield you that power.
7) Damn, it's good to have you back, Scott.
Were it the upper class people who did not want their comments highlighted? It would be typical for their secretive subculture as described by Fussel.
The covid infection and hospital rates in the US have both dropped steeply over that past two months. The time correlates with when we started vaccinating people. But to (non expert) me, it seems like it has dropped way faster than we have been vaccinating. I haven't found any news stories that say whether this drop was due to vaccinations or maybe due to post-holiday social distancing or maybe even due to some amount of herd-immunity.
What's the consensus here? Why have cases/hospitalizations dropped so precipitously in the US over the past two months?
People should just adopt some quick signal they can indicate if they don't want their posts highlighted, something like an asterisk or DNH or whatever.
Reminder that the monthly DSL effortpost contest closes in about 24 hours. Lots of good stuff this month.
https://www.datasecretslox.com/index.php/topic,2812.0.html
> I’m a little worried about the Comment Highlights posts, because they broadcast things people said down in the comments where nobody would ever read them out to the entire Internet.
Just wanted to chime in and say 1. I don't think there's anything wrong with how you've handled it so far, 2. I do think this is a fair/nontrivial concern which is worth considering, and 3. I don't see any easy solutions, so the answer might just have to be to accept it?
On the last comment highlights I was in this *exact* position. I wrote a hasty/unedited post about Josh Hawley (that basically amounted to a rant) under my real name, and was slightly surprised to see it featured more prominently. Of course, I stand by the general idea, but it was written kinda impulsively out of annoyance, so if I had thought more than a couple of people would read it, I would have written it very differently (and been careful about the argument), and I wouldn't have "chosen" to have it featured.
To be clear, I don't personally mind, and if this had actually troubled me I could have tried to message you (to omit the name, or remove the comment). But I can at least sympathize with how it might take someone by surprise, even though in retrospect it seems fairly obvious (it's an open discussion thread, the blog generates discussion by highlighting certain comments, if you don't want to be associated with something you post, don't put a name on it?).
TLDR: the norms seem pretty obvious and straightforward (it's an open/public blog, of course any comments might be featured to generate further discussion), but I can still see how someone would be surprised.
On the other hand, I just don't see any easy solutions! If you omit the name of the poster by default, I assume some folks would feel like they weren't getting "credit" for thoughtful comments they made. Nor do I think it's feasible to individually ask posters for permission.
I think the easiest improvement is just occasional mentions of these norms. For instance, on each of these "Comment Highlights" posts, I'd add a quick note explaining 1. this blog generates further discussion by highlighting comments, 2. if you want a featured comment removed, message you, and 3. if you don't want a comment featured in the future, include a note in that comment (or maybe it could be in someone's profile).
Obviously, you will still get occasional "surprises" by someone who is unfamiliar with these norms. But I don't think that edge case is a particularly big deal, it seems overall perfectly reasonable to me for a blog to highlight comments that are written on that blog. And as long as you include that occasional note to the "Comment Highlights", it would only ever be a worry for someone who was new to the blog (and growing pains like that are inevitable for any community norms).
Regarding the DNP company, I think it's a trust issue. Theoretically, there's nothing wrong with saying "this company looks cool and you can invest in it here", but in practice nine out of ten times someone with a platform says that kind of thing it's because they have some financial stake in doing so rather than because they legitimately think it's cool. I believe Scott when he says he just think it's cool, but for anyone without a strong prior about Scott's trustworthiness, the reasonable thing to think upon seeing products shilled is "Oh, one of *those* writers."
Feel free to quote with full attribution any comment I make here.
I'm not a gamer, but I'm looking for action video game recommendations (for PlayStation 4, Nintendo Switch, or Steam) that I can use for brain-training my middle-aged brain. Ideally, the game should dynamically increase in difficulty as I get better. If such a thing doesn't exist, I would like something that can be set at a super-easy level (something that would insult a five-year-old) and then ramped up manually. I tried "Call of Duty: Black Ops", but it was too difficult even at the easiest level. I don't have great eye-hand coordination, and my reflexes aren't the best. I'm not interested in walking simulators, puzzlers, or similar.
How about asking people in a reply to their comment whether they're willing to have their comment highlighted, and only highlight if they say yes?
This does little to publicize the comment, and gives people a chance to cancel the comment if, after thought, they think they shouldn't have said it.
It's opt-in, which is safer than opt-out. It will lose some potential highlight-worthy comments, but I hope not a lot of them.
I thought the approach of "please don't comment here if you don't want people to read your comment" made a lot of sense.
The problem in #6 seems like something you could automate pretty easily. Get a friend with programming skills to make a quick program for you. You put all the names into a list, then when you want to highlight a comment, have a search algorithm check if the name's in the list.
RE: Book reviews, I'm good with posting them sans names for blinding purposes; does that mean if mine should make the cut I shouldn't link friends to it from outside the blog until after they've been unblinded?
I'm selfishly disappointed that this means you won't be doing a Conversation with Tyler.
What's the policy on discussing treatment for medical diagnoses?
I read the Vox piece and got pretty much the same impression from it that I got from the original malaria-vaccine story. When I posted it elsewhere, my prefatory comment was "this is really preliminary, only in mouse studies, but it looks interesting."
Content recommendation: Ed Glaeser's online lectures (e.g. https://www.youtube.com/watch?v=6WEusDeJfXI ).
A lot of people around here care about housing and urbanism and related economic issues, and he's probably the best economist working on this stuff. He's got the same "just plainly explain economic issues in a simple, non-moralizing technocratic way" that I enjoy about ACX or Zvi's blog. I used to recommend his book to people, but he manages to get pretty much everything in there across in a 30-60 minute youtube lecture.
McElligot's Pool (by Dr. Seuss)
“This book is dedicated to
T.R. Geisel of Springfield Mass.,
The Worlds Greatest Authority
on Blackfish, Fiddler Crabs and Deegel Trout”
Searching for 'Deegel trout' I found this nice blog post,
https://seussblog.wordpress.com/2012/11/18/mcelligots-pool/
I can't believe they canceled a book about sitting by a pool and maybe catching something big. Besides it's dedicated to his Dad! (I have happy memories of fishing with my dad.) Maybe it's the fishing the woke left objects to? (cry or make a joke.. my two options.) From what can tell the book was pulled because it has Eskimo fish. And Eskimo is no longer used. (This is news to me.) They would now be called Inuit fish. Is there some harm done by having Inuit children learn they were once called Eskimo? And what joy will be lost without the rhymes and illustrations. Can anyone make an argument for pulling the book? I wanted to close this rant with his words from “On Beyond Zebra” (also pulled)
“The places I took him!
I tried hard to tell
Young Conrad Cornelius o'Donald o'Dell
A few brand-new wonderful words he might spell.
I led him around and I tried hard to show
There are things beyond Z that most people don't know.
I took him past Zebra. As far as I could.
And I think, perhaps, maybe I did him some good...”
How far back do you go when trying to keep up with ACX threads?
I tried this in the hidden OT but here's the question again. What are some new examples of patterns that should not have been crystallized?
https://slatestarcodex.com/2014/03/15/can-it-be-wrong-to-crystallize-patterns/
Was Hellen Keller a hoax? Has any other deaf mute reached her level of sophistication?
I take two grams of Metformin a day for anti-aging reasons. From the Washington Post: "researchers noticed that diabetics who took [metformin] outlived non-diabetics who did not. Moreover, metformin had shown an effect in separate studies against each of the three diseases [dementia, heart disease and cancer]"
https://www.washingtonpost.com/health/anti-aging-drugs-metformin-study/2021/03/05/d9de870a-7882-11eb-9537-496158cc5fd9_story.html
Scott, please consider doing a metformin, much more than you wanted to know post. I think taking metformin has the potential to be a big win for rationalists.
Gwern on metformin: https://www.gwern.net/Longevity#metformin
"
There are ways for book reviews to be self-deanonymizing to various degrees ("As a person named John Doe, I found Mary Roach's "Stiff" highly upsetting" etc.), so watch out for this if you end up going ahead with the "publish everything anonymously" plan.
I enjoyed that survey much more than I typically enjoy surveys. Well done I guess?
On DNP ethics: I briefly considered going into more detail on safe-ish DNP dosing strategy and decided not to, primarily due to the fact that comments aren’t editable so if I accidentally say something wrong I can’t cleanly go back and fix it.
[Also this topic seems like an ACTUAL Copenhagen Interpretation of Ethics situation. If I give advice that makes an unsafe thing seem less dangerous so people do more of that unsafe thing, whose fault IS it when they hurt themselves doing it?]
I can’t imagine giving the same amount of attention to 100 book reviews. It seems likely that they’ll get attention based on superficial interest or the order they’re posted.
One possibility would be to post them all at once and ask for each volunteer judge to read them in a different order, based on a randomly generated list. You can keep going down the list until you get bored and stop judging.
The interactive fiction competition (ifcomp) seems like good prior work for an Internet competition based on judging.
Re: the mass of book reviews - suppose there was a call for volunteers to read some subset of the entries and respond with yes or no to these questions: (1) Does this seem like a good fit for SSC/ACX? and (2) Is the writing of tolerable quality?
Would this meaningfully whittle down the number of entries to review in depth? I would volunteer to read a dozen or so with those questions in mind.
Hello ACX commenters,
I am looking to collect databases from real businesses and business-like entities, including those that have failed or otherwise become "past-tense". Read on if you or someone you know might have access to such things.
(previous post generated a bit of good discussion... https://astralcodexten.substack.com/p/hidden-open-thread-1595#comment-1235343)
Background:
I'm a software engineer, specializing in data systems (i.e. a data engineer), with about 16 years in the industry under my belt. Something that's always frustrated me about the way that we design and build systems, is the way that knowledge fails to diffuse through the industry, because we don't _study_ what we do, and especially we don't study our failures.
As an example, the 2010s witnessed the full hype cycle (rise and fall) of "NoSQL" databases, such as MongoDB, Cassandra, DynamoDB, Riak, Aerospike, and many others. Did they turn out to be any good? Individually, in local circumstances, some engineers know the answer, or at least _an_ answer. Collectively, we have no idea. This knowledge only spreads as the primary sources write blog posts (mostly terrible), or move on to new jobs and tell stories (distorted by all sorts of biases). What we *should* be doing is studying what was actually built, out in the open, where everyone can see it if they're interested.
Additionally, I find it very difficult to teach other engineers about data systems, in a scalable way, without open example material. There are many online courses in SQL and things of that nature, but they always deal with trivially small, trivially clean data sets, without any of the richness or messiness of Real World Data. Many years ago, my own skill in dealing with data grew by leaps and bounds the instant I was exposed to real business data and asked to solve real business problems with it.
To these ends, I am looking to collect real business data sets. I use the term "business" loosely, in the same sense that engineers often say "business logic". Non-profits, community efforts, personal side projects, these things all count. The key thing I'm after are custom-built databases, meaning they either started from a blank MySQL/Postgres/MongoDB/etc, or heavily customized an off-the-shelf system like Wordpress or Salesforce.
I recognize there are thorny issues here with respect to intellectual property and personal data privacy. I do not expect anyone to just hand over a database and wish me well. We would have to work something out, whether that's an NDA, or thorough anonymization, or whatever.
In any event, if you possess a data set like this, and *might* be willing to share it for research purposes, please reply here and we can figure out how to connect and discuss.
(This is the third time I've posted this. My plan is to re-post it periodically. If that runs afoul of any rules, written or unwritten, let me know and I can adjust.)
Addendum:
I recently became aware of this book https://fightchurnwithdata.com/ which might seem unrelated, but I mention because it contains specific details about how to model measurements of subscription customers and their behavior. This is the _kind_ of thing I'd like to be able to produce from my research: specific guidance on how to model specific business concepts and processes.
Can you get substack to implement a "don't highlight" option when someone creates a post that only you can see?
One of the commenters in this thread mentioned something rather illuminating (for me):
> This sort of false sense of urgency is very common amongst conspiracy theories - the idea that the aliens are coming, that the Illuminati have control and are about to enslave us all, that the Storm is coming, that all police are secretly racist, that we're going to run out of food and all starve.
I would love to see Scott weigh in on that.
Does anyone have a good recommendation for books on the pre-American history of slavery? I've read several books on American slavery, but am looking to expand. It seems like every single book out there is about American slavery. I'm looking for something that is more apolitical, just the facts.
https://www.datasecretslox.com/index.php/topic,2863.msg85806.html#msg85806
I found an interesting photo poster of a Girl with a dog.
It was behind my mirror - it's almost a century old.
I think a lot about how to create effective organizations. Let's say that we want to run a big organization, a cooperation, non-profit or government. There are a couple of failure modes.
One is lack of structure, where every decision is taken willy-nilly and nothing is written down. This trends towards nepotism and corruption, where decisions are based on flattery on who-knows-who in the old boys gang.
Another is bureaucratization and procedural-ism, where everyone does things right but no-one does the right thing. This seems to be something that comes creeping into an organization and is very hard to stamp out without burning things to the ground and rebuilding. Ossified companies fall to creative destruction, governmental agencies can seemingly live on forever. One reason that is often used to explain bureaucratization is that organizations respond to violations and accidents by adding more procedure to avoid repeats in the future, inadvertently stiffening themselves. Another is the middle-management and bureaucrats that were hired to perform important work tend to invent tasks for themselves to do during downtime, and these tasks don't go away when things gets busier. I guess there are many possible just-so stories about the origins of bureaucratization. Does anyone know of more rigorous takes on this phenomena?
A third failure mode is to have both: Strict regulations that can be skirted if you grease the right palm or know the management. This is akin to organizational anarcho-tyranny.
A fourth failure mode is to have a little bit of both: Just enough regulations to create extra work, just enough personal touch that the extra work doesn't matter and the management gets to decide in the end anyway.
What are the success modes? One seems to be to hire the right people and then leave them alone to do the job. Presumably they will create the level of bureaucracy needed to get the task done and no more. I'm not sure this is the secret sauce though. Is anyone aware of an example when "hire good people and leave them alone" failed?
Another success mode might be "meritocracy": Simple, relevant and transparent rules/checklists/procedure/criteria where it's applicable, human judgement at the lowest level everywhere else. Reward results, but beware of Goodhart's law. (This is what I personally think is the best police for governmental agencies.)
I would like to plot this on a two-by-two but it seems like one access becomes "good-vs-bad" which isn't that useful.
Slack is probably another important factor: lots of slack can create amazing results (Bell Labs?) but can also create an adult daycare. Little slack leaves little room for experimentation, but it's also my impression of how some very successful companies (e.g. Tesla?) are run.
Thoughts and comments on effective organizations? Was "Good to Great" correct all along? Am I missing something important?
I get no notification when someone replies to a comment on any substack, so if I don’t go back to look, I don’t see the reply.
Is that true for everyone? Is there some way to turn it on?
(I guess I’ll have to check back on this one.)
My answer on the investment question was "It was completely appropriate for you to include that the company wanted investors and to provide an address where people could get in touch with them".
I think it would be very bad for EA (or society in general) to add more obstacles in the way of starting altruistic for-profit projects, signal-boosting and discussing those projects, investing in those projects, or otherwise trying to get them to succeed.
Also, adding norms is just really costly in general, in the same way that it's costly to keep accumulating more and more laws. Law #1 of the Internet: people generally need to relax way more on the Internet and treat fewer things as norm violations. Otherwise we risk slowly choking the life out of cool things by demanding those things follow more and more norms.
If there are no legal issues, and people aren't lying, then I think people should just talk about investment opportunities however they want.
I suspect many people with highlighted comments don't just feel ok with it but proud to provide insight and thankful to get share that insight with more people.
Survey taken. I will add my experiences of the weight loss drug, the company shilling, and the comment-highlights posts here.
1) On the weight-loss drug. I recall reading the first paragraph of your post, and feeling excited (reliable weight-loss?)and beginning to plot to maybe obtain some of this drug. Then I read the next two paragraphs and felt almost stung by the side-effects (get cataracts and then literally burn to death from the inside?) and was far less likely to ever take the drug than had I not read the post. I think you are not a responsible adult if you read that post and then want to take the drug.
2) On shilling the company, I didn’t read the version of the post with it in. Nonetheless I’ll add my two cents: I think it’s cool if you do that thing once in a while, but I think if you did it regularly it would be less fun, and also your life would get worse because people would be asking you for stuff ALL THE TIME AND NEVER STOP.
3) The comment highlights posts are great! Otherwise, I do not read the comment sections. (More specifically, I do often search for the threads you Scott are in, and read only those.) The comment highlights are always great, way better than what I get from reading the comments myself, and if you didn’t post them I still wouldn’t read the comments, it’s too much noise to signal. Please don’t take them away from me.
(My random guess is that the class one was unusually sore for a user and they were kinda mad at you? But that you’ve been doing this for years without much pushback? If this overly specific story is true I suggest maybe just being a bit more sensitive on very political comments/posts and then otherwise going about business as usual.)
PSA: if you're an over 55 or so in the UK you can get a vaccine now
Has anyone read "The Integral Trees"? Could the solar system depicted in it actually exist?
Is the following statement about mRNA vaccines at all correct, or did I go way of base somehow:
"mRNA vaccine technology is a breakthrough because the same delivery system can be used for almost any RNA sequence, which means mRNA vaccine candidates can be developed almost a soon as a pathogen's DNA is sequenced. Traditional vaccines are hard to develop because making a pathogen that is too damaged to be infectious, but intact enough to provoke an immune response, is a hard problem that has to be re-solved for each and every pathogen."
(Now that I think about it, why bother with mRNA at all? Why not cut out the middleman and inject the spike proteins directly? Does the immune system not panic over random proteins in the bloodstream, where it does panic when those same proteins are being synthesized by foreign mRNA?)
The question about whether not having highlights posts would be a worse outcome is easy to misread.
What would be some good blogs etc. (like Zvi) for vaccine stuff centered on Canada?
Now that the book review contest is closed, can there be a master list of titles/mass reply email to confirm the ones that were received? (Or, selfishly/personally, can I ask if mine was received? I sent it in a week or two before the March 1st deadline. But I think there was a discussion of this in a previous OT, so it might still be a broader concern.)
Also, while I don't have a problem with the temporary anonymity suggestion, one thing to consider is how it might affect post authors' commenting on the reviews. I presume that at least some reviewers would want to join in the debates/discussions in the comments before the final judgement.
There was a really interesting NYT article on the Vertical Club, an upscale 1980s NYC gym, a couple weeks ago. (https://www.nytimes.com/2021/02/24/style/vertical-club.html) Relevantly here, it provides additional evidence for Fussell's argument that fashion moves from the lower to the upper classes:
"Aside from those office gyms, New York fitness spaces were often considered “grimy” bodybuilding spots or cruising venues for gay men, stigmatized during the AIDS crisis. The Vertical Club, its name suggesting “the up and up,” brought in the light.
The obvious glamour of the club tinctured virtue with vice, “bolstering that image that this was sexy, this was a singles club, this was where people went to meet each other,” said Natalia Mehlman Petrzela, another fitness historian. “But at the same time they had to sanitize that image, because they wanted to attract this white-collar clientele.”"
I miss the Report button. Since it doesn't look like that's getting implemented anytime soon, is there any way to contact Scott or a moderator about a particular comment?
We've recently been modeling and forecasting elections in Australia, and our latest one, on the Western Australian state election this week, seems to have been well received in the local psephology community. We put up two articles on how the modeling works and how we intend to judge it after Saturday.
Would this be of interest to any Aussies curious about politics about statistics over here? We'd love any feedback or engagement that people were willing to offer.
Forecast:
https://armariuminterreta.site/projects/2021-wa-state-election-forecast/
Method:
https://armariuminterreta.site/2021/02/17/how-meridiem-works-2021/
How we intend to judge it:
https://armariuminterreta.site/2021/03/08/how-we-intend-to-judge-our-wa-2021-forecast
My conviction is that all political posturing that doesn't rely on material interests is just larping, and people who "get into" a specific kind of politics or "grow out of" it just need to get a real, more fulfilling hobby that doesn't damage their mental health.
Why the clause about material interests? Because you cannot "grow out" of having a shitty wage or a shitty job or harsher treatment by the justice system, it's something that's imposed on you until your material conditions are improved one way or another. In this case alone political posturing and organizing is useful.
An interesting case is people whose material conditions *are* dependent on their political posturing, i.e. "thought leaders", op-ed writers, professional opinion havers, blog monetizers, or in other words grifters. It is useless to engage such people with the purpose of changing their mind, since their very posturing is part of their brand and they're not going to threaten their livelihood to change it. (It can still be useful to confront them if they are spreading damaging lies.)
I've been thinking a lot about obesity lately, and some of your statements on the post about DNP got me thinking about it again. In particular about how in 1938 obesity was thought by most to be primarily a cosmetic problem.
In 1999, in article published in JAMA* concluded that obesity caused approximately three hundred thousand deaths per year.
According to the CDC** between 1999 and 2018, prevalence of obesity went from from 30.5% to 42.4%, and the US population went up by nearly fifty million as well.
Doing some back of the envelope math (and assuming that change in prevalence is not a consequence of changing definitions or better detection), we're
looking at an additional 40-54 million obese adults in the US. If the deaths track with the increasing population and obesity rates, then in 2020, between 450 and 550 thousand people died of obesity related causes.
Or roughly the same total mortality as the COVID-19 pandemic.
There are some pretty sizable error bars on this estimate, but even if I take the lowest reasonable guess and put the deaths at 250000 per year, consistent since 1999, that's still enough to make it the third highest cause of death in the US, behind heart disease and cancer (both of which obesity either increases the likelihood of, or massively complicates treatment).
I don't have any good solutions to this (fatshaming clearly hasn't worked), but I think that if we took the problem a third as seriously as we have taken the pandemic, we would get a lot more QALYs out of it for the effort.
*https://jamanetwork.com/journals/jama/fullarticle/192032
**https://www.cdc.gov/obesity/data/adult.html
How do we determine what failing/struggling businesses are deserving of public sympathy and support, often in the form of crowdfunding and the like?
It’s interesting how “struggling” applies to the business for which we give emotional energy, while “failed” more often applies to those we don’t.
Struggling restaurant, local book shop, martial arts studio = most sympathy
Struggling landscaper, electrician, roofer = middle sympathy.
Failed/failing ad agency, consultant, tech entrepreneur = least sympathy.
Is this purely a subconscious reaction to the perceived risk/reward ratio of each business owner? (Which would make this dynamic fair in a way). Is it a class thing? Or is it arbitrary?
Is this dynamic simply because we have more face to face interactions with analog small businesses?
Will I ever be able to start a Go Fund Me for my digital consultancy should I need to, or will it always be, “try again, loser”?
I've been checking at ebay, and the Seuss books are still available, though at high prices. I don't think ebay is actually taking the book offers down.
Anyone have predictions of whether the deprecated Seuss books are even vaguely sensible investments?
Also, if you just want to see what's in the books, there ae podcasts of people reading them on youtube.
Bret and Heather made McElligot's Pool sound pretty good, so I watched it. I'm surprised there weren't vegetarians going after it. It has some Seuss charm, but the lesson seems kind of stupid if taken literally-- imagining entertaining outcomes doesn't redeem an unworkable project. On the other hand, imagining entertaining outcomes is a fine leisure activity, especially if you're Dr. Seuss.
On yet another hand, I could read it as exploring your subconscious vs. someone who tells you there's nothing to the world but what you can see.
There's a saying that you can't manage what you can't measure. This is false, you can't manage what you can't perceive.
And even that is incomplete, because you need to be thoughtful about what you need to manage and what you need to let take care of itself.
There's some interesting anecdotes emerging that long COVID symptoms are alleviated after receiving the vaccine (see https://www.theverge.com/2021/3/2/22308965/covid-vaccine-shots-symptoms-improve-chronic-long-haulers).
If this is true, I wonder whether suspected post-viral syndromes such as chronic fatigue syndrome, chronic Lyme disease etc might be able to be treated in the same way using vaccines for these diseases.
Given that Moderna was able to design its COVID vaccine in two days, it seems plausible that we could have vaccine candidates for such diseases very quickly. If we had to run a normal clinical trial for an EBV vaccine that would take a long time and be expensive (which is probably why such diseases have languished without vaccines). But if it's being trialed as a therapy for CFS, that would be a relatively simple trial to run (just randomize 500 CFS patients into placebo and vaccine).
Speculating more, is it possible that the widespread availability of flu vaccine reduces that infection's association with apparent post-viral syndromes (i.e. people end up getting treated for their postviral illness every year anyway).
I've never thought that wokeness would all "just pass." I've always thought that unchecked, its core assumptions would have a tremendously corrosive effect on society. Yet when I read Scott, it seems that he lives in a world of blissful ignorance. Maybe it's because I live in a city that's become a pigsty, with gibbering crazy people on every street corner, and I see its downstream effects every day - I don't know.
Am I alone? Am I the only one who thinks we're living in a form of End Times (I use that religious phrase in a completely secular way).
In the realm of nominative determinism:
Terry Boot has replaced Peter Foot as Shoe Zone's financial director.
Here's an idea - perhaps not novel, but an idea - to help with the quote-problem. Why not just make opt-in? Most of these comments are made by people with special interest or experience in the fields of choice. At the end of your post, add a blurb to the effect of, "If you share something insightful and are okay with being commented in a comments post, please note it at the top of your post."
On warm-bloodedness: because of reptiles vs mammals and birds, I've associated warm-bloodedness with high baselines activity levels. Think of a lizard just sitting there versus a mouse scurrying about. But insects seem to be extremely active, think of the mosquito that flies around you constantly. Insects are ectothermic aka cold-blooded if only they had blood [1]. And the general square-cubed law suggests smaller animals like insects are generally more cold-blooded-like. So how do insects maintain their high activity levels? Or are they not as energetic as I think (maybe I only see them when they're in the mood to bug)?
[1] Okay, mosquito's do have blood, it's just not rightfully theirs!
Am I the only one to whom the name of this blog in all-lowercase-no-space (as in the URL) reads like "astral cod exten"? (I guess that's because the string "ex" is way more common at the beginning than at the end of words)
You're a talented director in your 30s, and a major Hollywood studio has given you $200 million to make a new "Alien" sequel. Ridley Scott and James Cameron have no power over the project. Your only mandate from the studio execs is that your movie be so iconic and memorable that fans forget about the Colonial Marines/pulse rifle/LV-426/Ripley/Hicks paradigm for good. The whole visual and technological aesthetic of "Aliens" needs to be overshadowed by the uniquely different one in your film. Moreover, you can't copy the style of any other Alien movies or use any of their characters.
What do you do? What kinds of characters do you pick? What distinctive technologies, weapons, equipment and space ships do your characters have? What choices do you make with respect to cinematography?
Just had my first solo experience with the American healthcare system, and it was weird. I was vomiting constantly for 4 hours, and thought I might need to see a doctor. I went on my health insurance website, and they offered telemedicine visits for $50 or I could talk with a nurse for free. I decided to talk with the nurse, he was friendly and said I needed to go to an urgent care center after asking some questions about my symptoms. I asked about the telemedicine, but he said for my case in-person care would be more beneficial, and he sent me a list of in-network urgent care care centers. I go to the nearest one, and it is a generic office building with no signage indicating a urgent care center. I go in, and the security guard at the front desk says this is the correct address, but there is nothing like that there. Then I go to another center on the list that I know exists, sit in a waiting room, sit in an exam room, then get my vitals taken by a nurse. I had to explain that I probably did not need a strep test for my sore throat that appeared immediately after vomiting. Then the main provider, who is an NP, comes in, asks me some questions, is very interested in when my last poop was, pokes various places in my stomach, then says it is probably a virus and gives me a prescription. Turns out she gave me Ondansetron, which apparently costs around $10 a pill but my insurance negotiated it down to $0.75. Still don't know how much the office visit will cost me. I know things like this are old news to most of you, and I knew the US healthcare system was broken before, but it still strikes me as odd how totally unlike other capitalist transactions this is.
If a blogger doesn't appear on tim pool or joe rogan does he even exist?
also lex fridman
Is anyone here of non-binary gender? If so, I'd like to get your impressions of this Steven Crowder debate with a transgender person: https://youtu.be/w0zv62EBfMU
To be clear, I know little about the trans rights movement and have never spoken with a trans person about them. What little I do hear about it tends to come from news articles and social media posts about things like biological males competing in female high school sports and people being punished for mis-gendering trans people. These news reports have never impassioned me, and I've had the building suspicion that the trans rights cause is being misrepresented (at least in the information channels I frequent most often), and made to look bad by a small number of its most extreme supporters.
In the Louder with Crowder video, I thought "Madison" was very intelligent and well-spoken, and kept their cool better than Crowder did. They got hung up on the issue of whether using words to deliberately misgender a person counts as "violence," and I thought that debate over that was frustrating.
If you are trans, this is an earnest invitation to help me understand your outlook. The Crowder video is (unfortunately?) the best place I can think of to start.
I really like the link back to the comment in the Highlights, since then I can get context and see if others disagreed. I wouldn't mind seeing names removed, but I would be sad to lose the whole link (I realize people can just click the link to get the name, but I'd expect 90% of people just read the summary)
Unofficial book review survey: https://forms.gle/1aU6BdcAt5ZKit7n7
I'll send the results to Scott.
Could someone please click on my profile and see if my bio appears? I don't see anyone else's, but I'm not sure if that's because I'm the only one who wrote one, or if its because ACX/substack is just not displaying it.
Is this a thing: rather than rule-out for psychiatric diagnosis, probabilistic diagnosis? The whole idea of medicine is to diagnose and treat, but what if some things are fuzzy (but can still be dealt with)? Elaborating makes it sound stupid, but I've wondered for a while if this isn't closer to what psychiatrists do. Some medical things are much more clearcut - there is a cartilage tear or there isn't - but psychology/psychiatry seems to have room for simultaneous states.
Everything you think you know is wrong. Discuss.
Question for people with some experience in leftist spaces.
I've been frequenting some online leftist communities more and more recently. I'll find myself agreeing pretty strongly with the overall sentiment and mostly positive about their proposed policy but then I'll get whiplashed right the heck out again when someone mentions obviously how the US intelligence apparatus secretly controls everything, that various well known figures are all FBI (or whomever) plants and that everyone who doesn't entirely agree with them is controlled opposition. Instead of being laughed out of the room, this idea is almost always taken seriously.
I find this idea dumb enough to make me distrust the overall movement even though I'm largely sympathetic to its stated aims and goals. Yes - I've heard of COINTELPRO, no its effects and ramifications aren't some all-consuming feature of modern politics.
Have I just gotten unlucky, or is this kind of stuff really just a feature of modern-day leftism?
I'm a regular reader who normally doesn't read comments in general. Here there are so many that they intimidate me so I tend to ignore them. I hope you are able to keep doing the comment highlights because they are pretty great!
(I think this is the third time I've tried to post this; I hope the other two aren't annoying you in some kind of comment queue.)
I propose that AstralCodexTen should be abbreviated as ACT, and SlateStarCodex be respelled as SlAtesTarcodex and abbreviated SAT, in honor of the two most prominent standardized academic tests. (This originated as a joke at an earlier SSC meetup that I attended when we briefly discussed calling it something different because of the blog name change, but I now endorse it semi-unironically.)
Does anyone know what's supposed to be wrong with Seuss's Scrambled Eggs Super!?
Changing the comment order from "Chronological" to anything else causes no effect, and that dropdown and other Javascript on the page stops working.
Being faced on a daily basis with the awful Covid19 death figures, I wondered what the total daily death figures were. I also wondered why the daily figures of all deaths was never mentioned - wouldn't it give better context to the Covid deaths? Not being a epidemiologist or a statistician, I searched online and found that the figures for Italy's deaths over the last few years were available at https://www.macrotrends.net/countries/ITA/italy/death-rate?q=England+Death+Rate+1950-2021 I knew that Italy has been badly affected by Covid19. The daily death rate for Italy for the past 3 years is 10.4 to 10.7 deaths per day per million of population - that is, about 640 deaths per day. The rise from 10.4 to 10.7 may well be Covid related - if so, it means an additional 24 deaths per day are the result of the pandemic. The increase doesn't seem so much - and may be a statistical variation - and yet Italy is regarded as being hit hard by Covid19.
It really seems like substack could solve this "don't quote me" problem for you. Maybe they allow people to click a config in their profile to indicate that they don't want highlights, and it would show up as an ❌ next to their name. You could imagine such software also allowing users to choose gender, or to choose background and get a 🧬 or 🦿, idk. They have product people to scope this stuff.
I say that as a software dev manager - your use cases are good data for them, since you have developed a successful community before. This sounds pretty easy for them to solve. People seem to like emojis next to their name, so even if you don't have any option on your blog other than ❌, other communities might use them. Sports blogs with team mascots and all that.
Just pointing out an option other than "I solve it" or "my customers solve it". You have customer support now!
I surprised myself and got ACX-Simple (a rewrite of the Substack user interface from scratch, without any of the Substack stuff aside from the CSS) to dynamically load comments. All the pieces just fit together.
https://github.com/EdwardScizorhands/ACX-simple/#how-to-install-and-use if you want to try it.
Ireland will be publishing a National Artificial Intelligence (AI) Strategy sometime this year (it was supposed to happen last year, but pandemic). https://www.gov.ie/en/press-release/51e5f-minister-troy-hosts-roundtable-on-the-national-ai-strategy/
We have also contributed a submission to the EU White Paper on AI: https://enterprise.gov.ie/en/Publications/Publication-files/National-Submission-EU-White-Paper-on-AI.pdf
So I think this is of interest as indicating what a national government thinks about AI and its implementation. On the one hand, they are not considering the kinds of disruptive existential threat that discussions about AI here and in previous fora like to talk about - the Fairy Godmother AI that will bring forth post-scarcity Fully Automated Luxury Gay Space Communism, or the Bad AI that will turn us all into paperclips. They seem to imagine it as yet another new technology that will provide jobs for humans, provided they have 'learned to code' as it were; certainly they are not envisaging AI as replacing humans altogether in this area :
"17. Most of the AI training programmes worldwide are based on the fundamentals of data
science, machine/deep learning, and ethical aspects using generic use-cases and examples.
While they build the fundamental skills of the AI workforce, a significant gap continues to
exist in the immediate employability of the trainees and their ability to apply their AI skills
directly on real-world research and business problems in various domains such as
environmental sciences, financial services, biotechnology, material sciences and
autonomous systems. Thus, there is a need to customise the curricula of training
programmes to address the specific AI skills required for such specific domains, particularly
focusing on relevant datasets, problems, AI algorithms, and implications on privacy, ethics
and security aspects.
18. However, a similar concern arises with regard to the need for specialist curricula in
technical fields. For instance, the methods and tools for dataset preparation, AI algorithms,
AI solution deployment and ethical/privacy implications for areas such as environmental
sciences, healthcare, material sciences are radically different. Thus, developing AI skills for
each of these domains needs customisation of the curriculum of the training programmes to
address the requirements and specificities of the problems in each domain."
On the other hand, they do have some intimations of Fairy Godmother AI, although again - the AI will improve diagnostics and treatment so much, nobody will be able to claim disability benefit because we will all be deemed fit to work - there is nothing about "nobody, able-bodied or not, will *need* to work because AI will do it all!":
Right to Life, Liberty, and Security of Person - AI-based diagnostic systems enhance
the enjoyment of the right to life by making accurate, high-quality diagnostic services
more widely available.
• Right to Desirable Work - The improved health outcomes that AI-based diagnostic
systems are likely to produce will reduce the number of people who are excluded
from the dignity of work for medical reasons.
• Right to an Adequate Standard of Living - By detecting diseases earlier and more
accurately, AI-based diagnostic systems will improve living standards and quality of
life.
• Right to Education - Should AI-based diagnostic systems deliver on their promise,
fewer people will be excluded from the enjoyment of the right to education due to illhealth.
So there you go - no more skiving off contributing to society and leeching on others by whining that you are physically or mentally disabled and so unable to work, you slackers!
I'd be interested in comments on this.
Hey! Sorry if this is too much noise, this is mainly a test/greeting post to familiarize myself with the commenting system here… interface is pretty different from what I'm used to. Let's see what happens when I hit the button…
Did they get rid of likes on comments? For fuck's sake, people. This makes the comments substantially less appealing.
If you were writing the Substack comment interface from scratch, how would you want the web page to tell you about new comments? What kind of "workflow" do you want?
Here are ideas to brainstorm. Some are mutually exclusive, some build on each other.
* The title changes and/or the tab thumbnail gets a big blue dot on it to attract your attention when there are new comments.
* You press a button to manually check for new comments instead of them loading in the background.
* You are told that there are new comments, but new comments are not dynamically put up on the page until you press a button that populates them all.
* Placeholders are put on the page in the place where new comments go (I think the default Substack behavior) for you to explicitly load, one at a time.
* No background loading of comments. You reload the page and new comments get the "∼New∼" tag, like on the old SSC.
* New comments are automatically loaded and put in place, with some tag like "∼New∼" so you search for them.
* * You have a way of clearing all those "new" tags. (But how do you know that you have read everything if background loading is happening?)
Slight bonus points if what you suggest works natively with ACX-Tweaks, but don't lock yourself into that. Dream big.
I wrote a short piece about the Dutch elections and how we have 37 parties we can vote for, including one based on age, one on animal rights and one on partying: https://considerthenl.blogspot.com/2021/03/37-party-system-dutch-parliamentary.html
can the options for why it was inappropriate be multiple-choice (checkbox) instead of single-choice (radio button)? i have multiple reasons why i think it was inappropriate