I'm curious why you care about the read:unread ratio. Is it just to avoid having shelves and shelves of unread books, or do you actually want to conserve a certain number of unread books on your shelves?
Ways that I keep my library in check:
1. Avoiding charity shops unless I have a non-book reason to go in, because I almost always leave with one or more books since they're so cheap second-hand - Easier now they're all closed due to Covid!
2. Donating books I don't plan on rereading to charity shops/family/friends - harder now due to Covid!
3. Buying most books I want new on Kindle, which doesn't take up physical space - of course, this means my digital library has... 175 titles? Most of those are free Gutenberg editions of classic texts though, and a lot of those are unread.
There is something wonderfully decadent about the personal library. It calls to mind A Series o Unfortunate Events, where, each time the siblings moved, the personal library of their new caretaker was the symbol that they had arrived home. The word alone is sufficient to conjure up images of homeliness, coziness, comfort, and knowledge. It's old-fashioned, but if I had a permanent residence there would definitely be a reading room with some comfortable chairs, a lamp with a green glass lampshade, and wall-to-wall bookshelves.
Thinking about the fantasy of the personal library as safe space, it comes to mind that although I assume a certain feminine nature to the fetishization of libraries, Beauty and the Beast style, libraries are pretty agender. They combine both domesticity and academia, a kind of nurturing growth and the professionalism of a lawyer or doctor. Class-wise, the idea of a personal library strikes me as an 'old money' kind of thing, but perhaps it universalizes to educated people in general. I bet someone has made a decent term paper about the sociology or psychology of libraries - and if they haven't, someone sure could.
I know what you mean about the difficulty of keeping too many books, I left most of my childhood collection to my younger brothers and have only recently started filling a bookshelf of my own. I'm not sure if I'd ever dedicate an entire room to a library, I'd be much more likely to use my entire house as a library, assigning different books to different rooms for decoration and categorisation.
I currently have a rather eclectic collection, ranging from Dante's Inferno to the 99% Invisible City, but since most of it was bought from charity shops there's a lot of thriller/mysteries that I'll realistically only read once.
It's always worth having unread books though, my internet went down at the start of lockdown and I'd never have made it through the next few days without plenty of books to keep me company!
My strategy is to buy special editions where possible, it turns book shopping into a fun adventure, makes sure I don't buy too frequently, and also, I have really cool copies of a lot of books! I mean signed copies, really old copies, or any collections or anything that seem fun.
If anyone has recommendations on how I would go about getting Discworld in _any_ form, I'd love to hear it.
I've managed to read most of the Discworld series in a random order by borrowing it from libraries. Not the most efficient way to do it, but definitely the cheapest and most interesting.
A quick Googling suggests that eBay would be a good way to buy most of the books secondhand, which would be cheaper than buying them new:
Obviously there's postage fees, but for that price I'd hope they'd send it to me in a replica of the Luggage, legs and all.
It's kind of sad how that site still describes their offering as "The Complete Collection... So Far!" but I guess the original author being dead doesn't always stop them from releasing additional work.
Jesus, that's a lot! I might save up to buy these all one day, but I'll more likely try to get used copies from little bookshops so that they all have more individual value to me.
Just finished watching the Amazon Prime adaptation of "Good Omens" and I was very sad that Pratchett was missing, it really felt like his touch was missing from a lot of things.
My strategy is to download the Kindle preview as a reminder that I wanted to read this book, and not to buy the book until I get to the end of the preview. But it doesn't work for books that haven't been released yet.
I do something similar with video games, but follow the inverse ratio. (I can "earn" a new game by getting 100% completion in 2 old games.) Your strategy makes sense for books, since you're increasing the visual impressiveness of your library while maintaining the same risk of embarrassment if someone grills you on a book you haven't read. Obviously, that doesn't apply to games at all.
I have a similar problem but a different approach.
Any 'shiny object' book I might want to read at some point goes on an Amazon wish list, and I'll occasionally check which ones are deeply discounted. I managed to accumulate ~200 unread Kindle books, many of which are anthologies or collections, and now I'm trying to bring that down to something more reasonable. I'll still buy books, but only if I plan to read them in the next month or if they're free, with a monthly cap of around $10 to spend.
This way I can still finish long series (Dune, most recently) where the later books are full price, but I also have incentive to finish books I already own and read more classics that are out of copyright.
Current known concerns about Substack comments (if you have more concerns, please post them as replies to this comment, one concern per reply; if you agree with someone else's concern, heart the comment):
1. Need to press "load more comments" to load more comments and "1 more reply" to see reply
2. Hearts need to be removed. SSC commentariat previously voted that likes/upvotes distort commenting behavior too much and they don't want them.
2a. People are concerned that you can't subscribe to comment replies without getting an email every time someone hearts it, but removing hearts should solve this.
3. Cannot currently edit comments.
4. Cannot currently report comments.
5. Cannot currently see which comments are new since last visit (as in old SSC where new comments would have a green square around them).
6. Cannot currently comment without it reloading the page, no easy way to get back to your old location.
7. Cannot currently without getting popup telling you to subscribe
8. Cannot currently link to some comments (link just goes to top of page)
> 2a. People are concerned that you can't subscribe to comment replies without getting an email every time someone hearts it, but removing hearts should solve this.
Unfortunately the "Send emails for Likes and Replies to my comments" setting is site-wide, not per-Substack, so that does not solve the problem: if someone comments on any other Substacks, and those do have hearts, they can't have this setting on even though it would work fine for their comments here.
For what it's worth, I just made a gmail filter that deletes any email from substack with the text "liked your comment". I guess notifications about this comment might get caught up in it since it has that text, but I figure it should pretty well handle me getting reply notifications and not like notifications. And I do really like the reply notifications, please keep that!
I actually really like the lack of reply notifications from SSC. I believe it produces higher quality discussion.
For example, in reddit I get an inbox of replies, listed without context, and the option to quickly respond. This encourages addictive behavior (in the usual dopamine-optimization ways), and causes me to reflexively defend myself against any response that is (perceived as) even slightly critical. I end up talking to the one person I'm in a back-and-forth with, not pausing to see if my comments are in any way useful to the broader thread (which I literally can't see while responding). By contrast, on SSC I had to consciously look for responses, thus ensuring I only comment when I'm actually invested in the idea. I had the opportunity to see what had been said, and often someone else had made my point better already. I took the time to actually wonder what my response was adding.
This analysis might be specific to my personal psychology (I find reddit replies to be one of my worst addictions, it's really quite bad), but that's at least one argument against reply notifications
Are you using RES for Reddit? I always click on the "Context" link for comment reply notifications in my inbox because replying as you describe is horrible for all of the reasons you mention.
I strongly disagree. It often involves a lot of back-and-forth replies to fully clarify our positions, and understand that of the other person. The latency between replies is one of the reasons I often find it harder to fully understand each other in a comment section than in a spoken conversation (others are that it's slower to type than to speak, and that we can't interrupt each other to ask for clarification). Reply notifications make this somewhat better.
A back-and-forth with one person shouldn't be a major problem in a threaded comment section (especially if it were possible to collapse threads).
(Btw there actually was a third-party reply notification system on SSC, though I didn't find out about it until more than a year into commenting on SSC.)
Reply and like notifications come from different addresses, so you can set up a more reliable filter than that. Reply notifications come from forum@mg1.substack.com while like notifications come from reactions@mg1.substack.com.
I would prefer it if, instead of sending an email when someone replies to my comment, it notifies me by changing the colour of a mail icon at the top of the page (the same as Reddit does). I would like substack to add this feature -- maybe make it optional on a per-user basis.
I very much liked the hearts feature change, as it enables me to find "better" comments when I don't have the time to read the whole comment section. Can someone tell me why and how voting is distortionary ?
Similar to the “like” button. People are conditioned to write comments to get likes/hearts rather than to say what they want to say regardless of the popularity.
Wouldn't that make the ranking of your comment be the metric you're chasing? I can't see how many hearts I got, but I can see I got more than X person I care about doing better than.
As someone wrote in a comment below, sorting by chronological order introduces it's own distortions to the comments. I don't really know how to evaluate what's the better option.
There already appears to be some randomness in the sorting (or at least, top first doesn't seem to just sort all comments in descending order of number of hearts).
Thanks, I guess I get that, though I don't know if I consider that drawback sufficient for getting rid of the feature as a whole.
Might be worth thinking about better ways of implementing some kind of rating system (maybe as likes only showing some time after the comment was posted, so as not to enable some instant gratification effect, and to prevent early comments from getting a head-start).
LessWrong has the upvote/downvote feature, which I'm sure has all sorts of downsides too, but at least beats the hearts feature in e.g. that it doesn't reward too controversial comments
I feel like upvote/downvote rewards controversial comments more than likes, because people often take their downvotes as a badge of honor. (I know I certainly do when I say something sensible in the wrong subreddit and get it downvoted to -20 or whatever.)
Ars Technica is even worse - downvoted comments are collapsed for subscribers and invisible to non-subscribers. I don't think they fully brigade yet, but increasingly the echo chamber is mindlessly defended.
True, but in my opinion having an imperfect classifier is still better than no classifier at all. I don't have enough time to read every single comment, so I find the heart-counts to be helpful even if there are some false positives and false negatives as far as identifying insightful comments.
Maybe I don't have a threshold, and just want to view comments in descending order of hearts. Or, say I do have some threshold, say it's 10, then what?
I almost think “likes” with no “dislikes” are even more distortionary. On Reddit I will often “counter vote” posts that seem like they are getting too many upvotes or downvotes because they are red meat to one side of the commentariat. Can’t do that here - posts that elicit a strong positive reaction will get the reflexive “like”, optimizing for (one-sided) heat rather than light.
Having both upvotes and downvotes is even more prone to circlejerks. With upvotes only, at least it's likely that high-quality comments expressing a variety of views will be highly upvoted. With both upvotes and downvotes, even high-quality comments expressing minority views on controversial issues are likely to be downvoted on the net, unless people are *very* disciplined about not downvoting comments that are well-argued just because they disagree.
That's only a solution if people don't actually judge the likely quality of a comment by its votes, and commenters don't optimize for votes. But then what's the point of votes?
Actually, if a forum were to have both upvotes and downvotes, it may be less bad if they are displayed separately, rather than displaying the net score. A comment that has -4 points may have 30 upvotes and 34 downvotes, or it may have no upvote and 4 downvotes. These indicate very different things, and this information is lost when only the net score is displayed. One is likely to be an important argument for a controversial position, the other a poor-quality comment that few people have read.
Presence or absence of like or heart reacts influences behavior of the commentariat. "Chasing likes" is a thing.
Now, this is not necessarily a bad thing. But virtually all relevant discussion boards on the web have like buttons or their equivalent. In the interest of diversity, discussion spaces without such features should exist. SSC was just that space and imho it worked great.
otoh, i would often wonder how good arguments were, and without likes I could only assume 50/50. If I find one side of an argument utterly ridiculous, and the likes confirm that, I can tilt my view one way. If the likes don't confirm that, it informs me that maybe *I* am the one misunderstanding something.
yeah, and i would tend to agree.. but I also often found myself going insane at some of the arguments at ssc. and while it is not a good way to form opinions, it is useful information to know how much support an opinion has in the population.
i don't know if 'likes' is a great solution, and it may bring more problems than benefits, but there is at least some tradeoff to consider.
The optimal way to form opinions is to be right about everything.
But since an individual can't easily implement that, it's usually relevant to take into account the opinion of others, particularly others that one expects to be well-informed, and that one knows something about their reasoning processes.
I think in a lot of subreddits, upvotes or downvotes are a really poor proxy for anything, because they just reflect the biases of that community. But a community like the one here is likely to be somewhat better, particularly depending on the type of comment one is looking at. (It's bad for ones that depend on empirical knowledge, probably much less bad for ones that depend on conceptual analysis.)
> i would often wonder how good arguments were, and without likes I could only assume 50/50.
With likes, you still do not know, because people commonly upvote comments they agree with, rather than comments that are well written.
This is why likes create echo chambers, because good heterodox writers typically get pissed off at getting few likes (and thus few readers) and leave, while poor orthodox writers who are first to make a popular argument, get a lot of likes, so they get encouraged.
I think we all agree and accept there are pitfalls with the like button. But I do not agree that there is necessarily zero information in the likes of a community like ssc. Given the heterodox tilt of ssc, I am not even convinced that our commentariat wouldn't give more likes to a 'good heterodox' writer versus a 'poor orthodox' writer.
I understand the risk of making an echo chamber, but they also would serve as a reality check to the heterodox ideas; is this an idea with merit, or is it one of those "devil's advocate, heterodox for the sake of being contrarian, but doesn't really hold up" ideas?
Incentivising good content, while imperfectly, seems like a good thing, in particular as people's time is limited it gives people the ability to find the best stuff. If people prefer not to browse that way they can always switch to chronological
Imho "gets a lot of likes" and "good content" are not remotely identical categories.
E.g. short controversial political statements might get tons od likes from the ingroup, while longform thoughful and important posts on, say, ways to improve vaccine production, might languish, um, "unloved", since few people understand technical problems involved.
Of course they're not identical. But I know of no better way to filter good content to the top than by a like system. Every other approach is far worse.
This is true. Forums with "likes" or "hearts" tend to become echo-chambers, over time -- unless they're very large, and then they fracture into a number of smaller echo-chambers, e.g. Twitter.
"Hearts" are totally inappropriate for a forum where unpopular opinions, and adversarial arguments, may on occasion be put forth. The lack of hearts leads to better discussion, more discussion, and freer thought.
"More hearts" doesn't necessarily mean "better" comments. Sometimes (and it's a big temptation that I've fallen for myself) you'll clicky that heart because it's an automatic "yay my side!" rather than "your reasoned argument has won my conviction". Sometimes you'll clicky that heart because Joe and Bill are tussling in the comment thread and you'll backing Bill against Joe (who is just plain flat out *wrong*). They do distort reactions and I'd prefer if they're gone.
It might work to have hearts and brains, so that people can indicate both what they like and what they think are good arguments or valuable added information.
I propose a dummy heart button that doesn't do anything (like some close-door buttons on elevators) and a second "this is a well-reasoned argument button" that does the actual sorting.
I'm only partially joking. I think a problem with simply imploring people to upvote for good reasons is that sometimes you have an emotion and just want to find some way to express it. A placebo button seems an interesting way to combat that impulse.
Adjective tags in slashdot were once a useful idea. The "funny" crowd could read the comments they like while the "thought provoking" crowd can pick out the ones they like.
Tags don't have to be pro/con, they could be more descriptive.
The heart button can allow us to appease our limbic system while the second "well-reasoned argument button" can allow us to flex our forebrain.
I would even go so far as to say that the latter button should be extended to good will arguments, even if they were weak arguments and were subsequently bested.
Problem is, ANY kind of rating system would cause commenters to optimize for rating (not necessarily consciously), and since any rater is just as good as the next one, it would optimize for the lowest common denominator, which, while arguably higher here than in most other internet fora, is still the lowest.
I'm not sure why one would think it's specifically the *lowest* common denominator. I would assume it would be the *most frequent* common denominator. Which isn't necessarily lowest, depending on the topic and the community.
Voting systems aren't perfect, but they are better than a system which rewards people who post the fastest and in greatest volume, which is defacto what a chronological system becomes. And is very easily dominated by a small number of people with fringe views who post a lot.
I'm definitely a big fan of having some way to sort the comments other than chronologically. I want to see a rundown of the most interesting comments by some kind of metric!
At least from my own experience, the old SSC system incentivised comments that resulted in more replies to them, especially replies that resulted in even more replies, and that was great, since it optimized for polite and engaging discussion, while a system with likes would optimize for likes.
This is something that the internet has learned over the past 10 years, very much to everybody's detriment. "Likes" distort communities and make discussions more superficial and facile.
The least toxic of the big social media sites, by far, is Reddit. Which is the only one which not only allows upvotes, but also has downvotes, and uses both of them heavily in their ranking algorithm. This is not a coincidence.
Often though, it optimized for being contrarian. A bulletproof argument will get someone to agree, but there is no reason to reply to it... except if you are willing to make a flawed counter argument.
Those deep reply chains were maybe half good discussion, but half were one of the posters just ignoring someone's key claim.
If placing value on more replies is a way to incentivize polite discussion, then you're already dealing with a community so unusual that no Guardian link is going to accurately describe our social dynamics.
Note that comments incentivizing more replies are known as "hate-clicks" in media, and "getting ratioed" on Twitter. This is how normies act when incentivized to get more replies. At best, it'll be scissor statements. If we're like that, I want to incentivize fewer replies. And while we're weird, I don't think we're that weird - either we're ignoring the incentive, or we're responding to it in toxic ways. In the first case, hearts are harmless. In the second case, they are vastly superior, because optimizing for them can't be nearly so toxic.
right, in the old SSC, more often the only response you got to a comment was 'here is why you are wrong'. there was no way to calibrate my thought process.
Well, there's the "good, necessary or kind" rule against that. So either that rule, or we are that weird, or a little bit of both, but it resulted in the old SSC having possibly the best comment section I've ever encountered - without any noticeable external incentive other than getting replies (and, ok, not getting banned by Scott).
I think the “small number of people with fringe views who post a lot” are also the ones most likely to spam the like button on posts they agree with, so it might not help much.
I understand the downsides of likes that others have mentioned, but I want to second the sentiment that likes are useful to quickly see the highest-quality comments. Suppose you've read the post and want to spend 5 minutes reading the comments. If the comments are not ranked, you won't likely read the most interesting comments.
One compromise could be to use hearts for "soft" ranking, i.e. surface some newer comments above liked ones. From what it looks, Substack might already be doing something like that.
I think that's one of the problems:short term people love upvote mechanics. But it tends to have long term downsides on the average quality of forums.
The comparison that comes to mind is the egg weight thing: people love big eggs, select chickens for big eggs and you get bigger eggs... until after a while you get chickens that fight constantly.
Because you're also selecting for other things as well.
Select comments a little and you make it easier to find good comments... but a little while later you have Facebook-quality comments and everyones fighting
Hearts/likes/upvotes favor earlier comments more than better comments in my experience. An expert can comment 12 hours later with a high quality comment and just get buried at the bottom due to the first comments already having been substantially raised via a low quality but popular/agreeable comment. (...said the man 23 hours after the original post...)
It does value both, and an earlier comment of similar quality will beat a later one. But later comments that are substantially better usually rise to the top pretty quickly.
FWIW I agree that voting is distortionary, but IMO it's the lesser of two evils—I very much like being able to quickly find the well-liked comments, and I'm willing to accept some amount of distortion to get that.
'off-stack': I've only just joined and got my 1st mail thread-notification. Could you color the link-title please, so I remember where to click? Took me a little while just now.. and yes, I'm stupid.
The links in the text of the post are a different color, but the link that goes to this article is just the title of the post. Since the tile is the same color as the normal text, it's not obvious it's a link.
Ah, I didn't know that clicking "show more comments" would eventually jump me there! I just assumed that if it was behind that link it was humanly un-findable.
At least for me, it doesn't (OS X Chrome). I have a bookmarklet to work around this -- executing "document.location.href = document.location.href" will cause the page to jump to the anchor in the URL. So I just click my bookmarket after hitting 'load more' a hundred times or whatever.
A way to hide comment threads would be really nice. Bonus if it remembers that choice through reloads. And a way to quickly navigate (like old SSC having a "parent" link) would be helpful.
Oh gosh yes please dump the hearts, they really confuse me as they make me go "Twitter? Not Twitter? Where am I?". Some way to edit and format comments would be great also.
This is a biggy (for me at least). Substack, please add Markdown support. Also, enable highlighting text in a yellow background -- as if it was highlighted with a yellow marker pen -- using the <mark> tag.
Found out by accident that if you click on the gray line in the left margin of a comment thread it collapses it. But the view doesn't seem to be moved accordingly: instead of, say, scrolling you to where the collapsed comment is, it looks like it just keeps you X many screens down from the top, which may be considerably further south from where you were.
Also, in the process of writing this comment I realize Substack doesn't currently support markdown. I see that user ersatz has already raised this issue.
Yes, and if you are on a touch screen (like a tablet), collapsing is FAR to easy to do by accident). The mobile interface is seems nigh useless for comments, as any length comment is nearly unreadable. Not exactly sure why, but I think it is something about the text sizing being too big just makes it very hard to process.
You'd have to load all of the comments you want to search first (in batches of whatever size Substack uses), but then you should be able to use something like 'Find in page' in whatever browser you're using.
I am strongly in favor of a voting system. Despite having followed your blogs since the early lesswrong days I almost never went into the comment section on SSC because it was always so long as to be unreadable, and the lack of a voting feature meant that discussions were easily dominated by a small number of people with fringe views who posted a lot. (I'll avoid specifics so as not to derail, but I'm sure anyone who was on it frequently can think of a few examples.) Voting systems are not perfect but they are better than a system that rewards whoever posts fastest and in greatest volume, which is what a chronological system defacto is.
"Voting systems are not perfect but they are better than a system that rewards whoever posts fastest and in greatest volume, which is what a chronological system defacto is."
I agree this was a problem with the SSC commentariat. But instead of intruducing a voting system, I suggest tackling the problem directly by limiting number of comments one user might post over a time period.
That is a vastly worse solution. None of them were spammers - they were honestly engaging. Hard-blocking that is a major overreaction. Far better to let them keep talking, but let it fall to the bottom, where only interested parties will see it.
Yes, the comments were often very long. But I found it relatively easy to navigate (but I did also spend hours sometimes reading one post's comments):
1. Collapse comment 'sub-trees' that seem to be uninteresting – that can prune a LOT of overall comments.
2. Ignore comments from commenters that I found to be consistently uninteresting.
One nice thing tho is that there are many 'venues' to discuss a post, not just directly 'in the comments' (on the main site). That's probably necessary to some extent if the number of commenters continues to increase.
With upvotes, you mostly see the most popular opinion on any given issue (which may or may not be correct). If you mostly see comments by the most prolific commenters, at least it's likely that a variety of views have prolific exponents.
Re the voting system. It seems like even if you swap to "top first" it doesn't fully sort all the comments in order of votes. Not sure if thats a bug or its using something like the reddit "best" algorithm that tries to normalize the for how long a post has been up, to reduce the benefits of being teh first to reply to an article
I've been getting into photography lately and I like sidequests so if anyone wants a high resolution, debatably high quality photo of anything within a reasonable radius of Madison, Wisconsin, @ me with your request.
An optional but highly rewarding sidequest to natural light photography is... lit photography! Artificially lit photography? Photographing with strobes! :D
It allegedly appears this far south but I've never seen it personally, but when it's next supposed to be down here I will probably give it a try unless it's cloudy
Our planet is chock-full of agents with wildly varying degrees of intelligence that are able to mentally adapt to their environment within an individual lifespan, as opposed to evolutionary timespans. This can't be coincidence. There has to be a shared learning mechanism, and it has to be unaffected by these paradoxes.
As for AI, the "only" trick is to understand and imitate that learning mechanism to a sufficient degree, and then apply it to a machine. "Sufficient" meaning, to make the machine fit for its intended task. Then, if that machine is able to sense, learn, and act with sufficient similarity to an animal, then we could teach it to do anything that an animal with the same sensorium, mental capacity, and actuators could be taught to do.
The mode of learning might even turn out to be completely different from what we find in nature, as long as it gets the job done - after all, we optimize airplanes for their greatest utility to us, not for their similarity to birds. Though if we found that, indeed, bird flight is the only way to ever achieve flight at all, then bird flight it is.
The only way in which you could equate AI paradoxes with the speed of light is if we could regularly observe light-speed travel but just not know how to do it ourselves yet.
When you say, "I've yet to read a statement that made sense of why Polanyi's Paradox and Moravec's Paradox will not cripple the effort" I think you would do better to decouple the criticism with the projected end result. It's one thing to say that you think there are relevant considerations that you think receive too little attention, but when you staple on a "and thus this endeavor is doomed to failure!" rider, you're making a stronger statement with much weaker support.
This is a good general bit of epistemic hygiene but it's also highly relevant here, specifically. Consider that the former of these "paradoxes" has been reborn in this new age of narrow AI. We have networks that can solve protein structures damn near as well as we can actually measure the structures. We just don't... actually really know how they do it. Many people are very upset that these neural networks have such black box character, but it's ultimately not very different from our understanding of human cognition. I don't understand why you think this is a bottleneck; we've demonstrated pretty comprehensively that we can build an intelligence with learning ability even if we can't teach it exactly how to accomplish a task.
As to the latter, Moravek's paradox is definitely a part of the mainstream consciousness of robotics researchers. Everyone knows that it's hard to make robots which can do complex physical tasks but easy to make ones that can do well in a game of checkers (assuming they can move the pieces, at least). It's not especially surprising, since we're comparing our work to that of bio-robots strongly optimized for locomotion and perception and weakly optimized for checkers. That doesn't mean the effort is doomed to fail; it just means that the problem is hard.
An interest read. The author is vastly underselling the main characteristic feature of modern machine learning, which is generalization. You don’t need to create ten million precise computer models corresponding to every possible coffee mug that has ever or will ever exist. A relatively simply system can learn to identify a coffee mug with relatively good accuracy, given a modest training set of coffee mugs. A manipulator arm can manipulate a new, never-before-seen coffee mug by performing a minor inference over the space of mug-like objects it has played with before.
So all the numerical estimates of cost and timing are 10-100x too high, IMO.
It’s the hooking together of the different systems into a multi-modal ML system which can seamlessly transition into a robot body with sensors that hasn’t been done yet. Everything *individual* part is a solved problem, not something requiring billions more in investment.
If your situation gets any more odd, it might turn out that there is an enormous demand for you to refer prospective patients and psychiatrists to each other, and that this takes up so much time and is so lucrative, that it shifts the entire business model of your practice. What a world.
Maybe, but I'm guessing most of the people signing up don't just want a psychiatrist, they want Scott as their psychiatrist, so unless he can duplicate himself, those people will probably lose interest if referred elsewhere.
I read the Cult of Smart a month or so ago. I know a lot of people here have read it too. Overall I wasn't impressed with the conclusions of the book. But I can't get over the premise: intelligence is mostly inherited and school won't likely increase your IQ.
That model of the world -- where a significant amount of your intelligence comes from your parents; and there is a meritocracy; and there is some sort of self-sorting -- is a different model than the world I am used to. It really makes a lot of policies obsolete.
It kinda scares me to think about how we could govern in that world. All the same, if that is reality, I'm not sure it makes sense to ignore it.
"I think your fear misses the mark, because merit isn't based on iq alone."
For sure. But for a lot of things, particularly in the advanced society we have now, IQ becomes a bit of a ceiling. So, sure, smart people won't all be successful. But not-smart people are a lot less likely to.
1. I'm going to post a review of Cult of Smart soon
2. We need to be really careful about how we define this. It's not exactly true that school "won't increase your IQ" - there are some studies showing it does, although the exact interpretation is complicated and the magnitude isn't as much as some people would like. I think de Boer's main point is that however much school increases IQ, by the end of a given amount of school, genetically-more-gifted people will still do better than genetically-less-gifted people. Which actually doesn't go much beyond common sense - everyone knows that even within school classes some kids do better than others.
3. Curious what policies you think it makes obsolete. I'm worried some people might over-update on this. My impression is that since everyone already knows that some people do better than others even after a lot of schooling, most people have an intuitive idea that genetic giftedness exists and it already factors into most policy discussions, even if we're bad at talking about it explicitly.
If a country is reasonably meritocratic, people "sort" and intelligence (talent) is inherited then you have a lot of problems that will *naturally* arise:
1. Poor people will cluster in poor areas. The schools in those areas will be bad (if ranked based on the scores of the students). Rich people will cluster in rich areas and those schools will be good (if ranked based on the scores of students). Changing tax policies or adding funding to the bad schools won't help (above a baseline).
2. Ethnic groups whose immigration correlates with their talent-level will be, as a group, advantaged/disadvantaged and that will stick with those groups until they "blend" in (though sorting could prevent blending.). For example if immigration policies tend toward high-talent from some countries (i.e. Asia) and lower (or average) talent from others (i.e. Guatemala), you would expect Asian students to do better than Guatemalan, *in aggregate* kinda for the long term. That sorta scuttles a lot of the diversity plans we have today.
3. With the greater ability to meet people (via internet and dating apps), sorting will become more and more efficient. That means a growing gap between haves and have-nots. Even in a meritocratic country, the gini coefficient would continue to grow.
My perspective on these things is to ask what counts as a negative outcome in itself, and what counts as a negative outcome because it indicates something else. For example, on point 2, if you use diversity as a measure of fairness (in an equality-of-opportunity sense) then lack of diversity is indeed very distressing, but in the hypothetical you propose, is lack of diversity really such a bad thing anymore?
Of course there are some inherent negative consequences to points 2 and 3 which are worthy of discussion, concern, and efforts to mitigate. I think they're not as bad as they initially sound though (subject to the debatable premise that they are caused by immutable differences in population genetics, and not discriminatory policy). In other words, a consequence of updating only half of your beliefs in response to new data.
"if you use diversity as a measure of fairness (in an equality-of-opportunity sense) then lack of diversity is indeed very distressing"
I guess that's the thing that is bothering me. As a society we have decided that that is the case. Not every individual feels that way, but as a group that's what we've decided. So we've implemented policies to try to fix it. But those policies don't seem to be working. So we "crank things up" to make sure we *really* fix things. What's the next step?
I don't see our society suddenly deciding that diversity and "equity" aren't important. So where do we go from here?
Saying that IQ is genetic doesn't mean that two mediocre people can't produce a brilliant child, nor that a brilliant, successful couple won't have a below-average child who's successful (by some definition) because they were "born on third base."
A corollary of your point #2 is that "affirmative action", as it's usually understood, is doomed to fail.
With respect to "diversity plans", a lot of research shows that having employees from different backgrounds and life experiences helps companies succeed. Given your point #2, it's worth moving heaven and earth to find those brilliant children from communities that are underrepresented in the historical picture of "success".
There are diversity plans and "diversity plans". In the real world, "diversity" is a reference to <i>Baake</i> and pretty much never means "diversity of backgrounds and experiences".
These aren't really problems unless your goal is radical egalitarian economics accomplished through white collar brain jobs. Low income families in wealthy societies are better off than median income families in a low-income country, and those "bad" schools may not be so unbearable if we expel the trouble-makers and still give all the kids discount laptops.
The families from Guatemala are almost certainly happier in the US than in their native country: they wouldn't have left otherwise!
Problems arise when the educated classes raise their ladders and effectively eliminate meritocracy, or otherwise police the rest of society with bad policies because they are too divorced from reality to effectively address problems. It's not encouraging to let my kids fall to lower income status if they descend into a mess of divorce, obesity, and meth.
check out "coming apart" on how far the sorting has gone, but really probably just "the bell curve" on the basic issues around iq heritability.
very curious if people know of good books engaging with policy implications (I'll try reading CoS, hear (from Tyler) good things about thing being an honest try, but given the guy's ideology is pry unpalatable to me not rly expecting much..)
Sorry if this is a weird question, but can I ask why you wanted to write a review of The Cult of Smart? (I'll be a little vague here because I want to ask about the meta-level issue without wading into the object-level issues, but hopefully what I mean is still clear.)
I ask because it seemed to me, as someone with what I think are broadly similar views on the relevant subjects, on the basis of interviews and articles about it, that The Cult of Smart basically reiterated what some other books have said about the relevant science, but consciously tried to make it acceptable to a broader range of people on the political spectrum. In doing so, de Boer was taking a stand that was somewhat controversial among his political tribe. And I think that was a really salutary contribution to The Discourse that de Boer deserves praise for.
However, 1) It didn't really seem to me that the book's core thesis would be that novel/interesting to people who had already been following/discussing these issues, as SSC has been doing for some time and, more importantly, 2) It seemed to me that an impartial analysis of the book's arguments on their own terms, with no reference to the surrounding intellectual ecosystem, would likely produce some fairly harsh criticism---but criticism that could be better aimed, considering said ecosystem, at different targets.
So I tried to avoid discussing it when it came up in various fora, because the stuff I agreed with I thought was obvious/well-established in these circles at this point and I felt it would have been tactless/counterproductive to criticize the stuff I disagreed with for not being even more controversial in the mainstream. Consequently, I'm wondering if there's something I'm missing that made it a good choice for an ACX book review.
(Probably the answer is that it's because it sounded like an interesting book, and I'm insane and trapped inside a Straussian 4-D chess match against myself. But I still had to ask.)
You are right that deBoer's fight is with the ogressive-pray eft-lay (odd-numbered thread), but I think the book's focus on educational policy in particular makes it more than a re-hash of ideas previously discussed.
"But I can't get over the premise: intelligence is mostly inherited and school won't likely increase your IQ."
I think this is mostly true? You could have kept me at school 24 hours a day with regular beatings on the hour every hour and you would never have driven either mathematics or music into me (I couldn't carry a tune in a bucket, as they say, and crows sound like Jessye Norman by comparison with my singing voice). So "more school" would never have pushed me into STEM.
On the other hand, education is no burden (as another saying has it) and it's better to go to school and get a taste of everything and some basic knowledge, and a direction to go in for more knowledge, than try to re-invent the wheel yourself. Very smart, self-directed, disciplined kids with supportive parents and access to sources of knowledge may do great without school, but the mass of us ordinary types need the structure and discipline. Not that modern schooling can't be improved upon, but no need to throw the entire thing out.
So in conclusion: school won't turn a sow's ear into a silk purse, but it's important to give everyone a basic level of education.
"Access to sources of knowledge" is pretty much free now for anyone with an internet connection. It's the supportive parents, meaning parents who both want to help and can help, that is the scarce resource. I wonder to what extent one could have a private non-school substitute for that, people in the business of helping kids who are unschooling find the knowledge they want — probably doing it online.
I can't help wondering whether more thoughtful and attentive teaching would have done you some good-- there's more to education than quantity and harshness.
There are probably hard limits on how much people can learn because of talent and interest, but I can also believe that most education doesn't get near those limits.
I vehemently disagree with many of the book's policy prescriptions, but it's hard-hitting and worthy of respect.
My main takeaway was this: we actually know what it takes to increase measured school performance - aggressive filtering of students by ability.
I doubt DeBoer would conclude: "OK, let's do that!" But he does (a) recommend allowing students to drop out earlier, and (b) state "the data shows that what really matters is the academic performance of the top 5 percent of students."
“Education” as it is currently instantiated is just a stacked sequence of testing and sorting gates, so we should expect to find that the primary function of this system has been to sort people according to native ability.
True education has never been tried, because it is much harder, many times harder, than just sifting for natural high achievers.
What exactly would true education be and how would it differ from education as it exists now? I would assume that elements of true education exist in the current framework; concretely, I know how to do calculus and write essays, which I didn't before. Certainly there is a lot of sorting by ability in the current system, but I don't see how there could be a substantially different system that better achieves the goal of imparting knowledge onto willing and capable recipients.
True education in this framing would involve successfully training people who might, by default, never learn how to write essays or do calculus.
As you say, a calculus classroom might not look much different, but the overriding objectives will be completely different. It's as simple as changing from an attitude of "we're going to go over this material once, then grade everyone on their performance" to an attitude of "every individual student is going to be trained in this material until their performance is adequate". This might require something like individual AI tutors. Like I said, it's very hard.
There's got to be some intermediate case where people are allowed to drop subjects where there seems to be no or little hope of improvement.
Also, part of what's needed is thoughtfulness by the teacher and/or teaching system. Not just going over the material again and again, but working on figuring out what the student needs to learn and how they can best learn it.
I don't think AI tutors is what's required. There are computer programs that teach math to students by letting them learn at their own pace, but not letting them advance until they prove that they've mastered a topic. All that's required is getting schools to use these programs. And perhaps some testing to make sure they work better than current systems.
Widespread literacy and basic arithmetic would suggest we actually can educate. The difference between calculus and the three 'Rs is that you actually need to do the three 'Rs to succeed, so we put a great deal of effort in establishing some baseline.
Perhaps we should do more, but I would rather see better reading comprehension and intuitive algebra before trying to get everyone to jump on to the calculus bandwagon.
From the employer side, education filters serve as a damn fine way to stratify talent. So still useful from this side.
The standard caveats about what IQ is and isn't all apply here, and drastically lessen the impact of this model in my mind. While you may not be able to 'train' someone into a higher IQ, you can train them in mental discipline, ways and habits of thinking, types of problem solving modalities, information and procedures, etc etc etc.
There are some specific jobs and activities that you'll never be good at without a high IQ, but they're pretty rare - although they are probably high-visibility and high-prestige among the people reading this sentence, but that's a selection bias and not representative of most of the world. In many normal jobs and activities, a well-trained and disciplined average IQ person can still outperform an undisciplined and untrained high-IQ person. It still makes a lot of sense to focus education and child-rearing on those things.
Basically, even if everyone does have a 'ceiling' to their abilities related to IQ, very very few people in the world are actually operating at their ceiling, so most have room to improve.
check out "coming apart" on how far the sorting has gone, but really probably just "the bell curve" on the basic issues around iq heritability.
very curious if people know of good books engaging with policy implications (I'll try reading CoS, hear (from Tyler) good things about thing being an honest try, but given the guy's ideology is pry unpalatable to me not rly expecting much..)
I would like to contribute, but feel that the current subscription price is a bit higher than my willingness to pay. I would gladly pay the discounted price, but definitely do not qualify for it. Is there some other way to contribute? (I understand it's a trade-off - if you make it easier for people like me to contribute, you may get less from people who would otherwise contribute the full amount under the current setup).
I'm always surprised by how many sites insist on the recurring-charge subscription model exclusively, and offer no way at all to make one-time donations. I can't be the only one who likes to support content providers, but doesn't fancy riding herd on dozens of automatically repeating credit-card charges.
An interesting model, one alternate to recurring subscription, is to hold articles ransom. No one can read the article until the price target on it's head has been paid, anyone can contribute however much to this target and once it is paid everyone can view it. This has been tried (and I think pioneered somewhat) by Vitalik Buterin of Ethereum.
I took a 1 year subscription and then immediately canceled. It told me I will still have the benefits of being a subscriber until next year, when I will reconsider.
It’s been a disappointing winter. I’ve been looking forward to some solid sheets of snow but mostly we’ve been met with sparse paddings of sludge and ice. I’m hoping for one last good snowfall before things get warm again, or for the snow to be done completely as the dreary grey and green isn’t doing it for me.
Couldn't agree more. It has so far been a Winterless Winter here in Southern Ontario. Growing up it would snow in late November or early December, and persist till March... This year has been alarmingly devoid of snow, and eerily warm... It's hard not to be anything but concerned. And with lockdown and no hockey allowed on the rinks... Hardly a Canadian winter at all.
I'm on the south sides of the lakes, close enough to Erie to see it if I just peak over the Horizon. It's usual for us to have a buffer due to the lake, but every year I feel like the snow recedes farther into my memory. We only got two full coatings this year, one of which that was kind enough decimate some sick trees on our property with a heavy coating of ice!
We got five inches of snow in College Station, TX, two weeks ago! Most snow we've had in the past seven years! But otherwise, the problem of the winter has mainly been dealing with allergies from whatever trees are enjoying the warm weather before and after that snowfall.
I'm so sorry you're missing out on one of the guaranteed joys of winter. No tree allergies! Luckily, up here, all the Bradford Pears haven't started blooming yet. They're beautiful over-pollinating stinky trees and just make spring its own kind of splendid!
Minor heads up: by default, gmail was filtering the Astral Codex Ten substack emails under promotions. I found out by accident - I'm not sure if any keyword accidentally triggered it.
I've been on Matthew Yglesias's substack for a few weeks now, and I've noticed that about a quarter of the e-mails from his go into my promotions tab, while the rest go into the main inbox. I don't know how it decides which, because I think it's happened to a similar proportion of the three types of e-mails (notifications of new posts, notifications of replies to my comments, and notifications of likes on my comments).
Parenting (and holding down a job while parenting) during extended quarantines/lockdowns may be one of the most mundanely interesting things I've done... EVER. Aside from that... re-learning sleight of hand card magic, which I used to do much more of as a teenager.
Hm, it turns out there are a lot of contenders for this title. Parenting for sure-- notably, overhearing my son's interactions in distance learning classes, drawing conclusions from that about his academic and social growth paths, figuring out what school activities were actually intrinsically motivating to participate in with him (turns out father-son PE class is super underrated even for non-athlete nerds like me!) and teaching him fun things to complement the classes.
Other than that, learning a lot about how local institutions work and how they fail is definitely on the list. I am going to have Many New Opinions about San Francisco when this is done, hopefully non-redundant ones that challenge everyone's preconceptions. Suffice it to say for now that between virtual civic participation and hiking every single pretty path I can find within walking distance, I feel way more rooted in the place now than a year ago.
Learning what I sound like when I record myself is another top candidate. I had intended to do a lot of music learning and music making in 2020 anyway (quit my job for unrelated reasons at the beginning of March, which was... interestingly fortuitous and terrible at the same time) but would not have done nearly as much self-recording if I'd been able to make music with others in person, and it's a useful enough tool that I wish I'd been motivated to do more of it earlier.
Mine is getting a second home and moving to a city, because the quarantine has freed me up from being stuck in the exurban town where the university I teach is based. It's weird being in Austin and not doing half of the things that I would normally come to Austin to do in the past. But it's really nice being here and doing the other half of the things (mainly involving food trucks, bike paths, and a bit of outdoor socializing with a few friends that live here).
There is something very weird about an economic crisis that leaves a substantial fraction (perhaps even a majority) of households with *greater* amounts of disposable income than they had in the before times.
Yes it's true. But it's also true that household savings has increased in nearly all income brackets, even among people that don't describe their financial situation as having improved.
My house rehearsed a musical every weekday for three months and then livestreamed a performance. Only about half of us had any prior music or acting experience, but everyone was involved and everyone got cast in a major role. We had choreography and costumes and (some) live instrumentals. Still kind of weird to me that that actually happened.
I drank a lot, ate a ton of food, and got to spend more time with daughter by eliminating the commute.
Also, played COD: Warzone with my buddies every night for about 3 months straight. That was pretty sweet. We haven't gamed since college, but since it's the only way to really spend time together, we have done a lot.
I also made the entire annual budget by brute-forcing Excel. Again. Someday we really need to get out of Excel. That experience along with a number of dropped balls from other departments has resulted in a +10 to grumpy.
Buying a car, putting all of our stuff into storage, and going Airbnb-hopping for 6 months as a way to visit other parts of the US and get a feel for what it's like to live outside of big cities. Of course, since we're staying home except for grocery shopping, it's not quite like getting the full experience, but it's been a cool way to re-evaluate my preferences for, e.g. density vs quiet, etc.
I taught myself Python, since all the non-Covid science labs were closed and so becoming proficient in bioinformatics was the best use of my time. I'm actually quite happy with my abilities, Python is just so much better reformatting and graphically displaying data than Microsoft Excel, and I wish I'd bothered sooner.
I probably won't pursue a career in coding though, I suffer from "techno-rage" even though I know that shouting at computers doesn't make my hacked-together code miraculously function - only Stack Exchange has that power.
Nice! Programming can be humbling if you deeply internalize that the computer is a good listener, sometimes the makers have just told it to listen to the wrong things (hardware makers, language designers, other programmers, you!). It can be very relaxing or very stressful, I'm glad you sampled it either way.
I don't know if it's interesting or not that I have spent pretty much the whole year failing to launch a coliving/coworking/makerspace/etc project that would be very interesting if it launched.
I got a Virtual Reality kit and accidentally got into gamified exercise. Lost 30lbs, although I put a bit of that back on over Christmas. I just wanted to play the new Half Life and instead I've done 500 hours of sweaty Beat Saber.
Staying consistent with my 1k calories per day of VR cardio has been a real confidence boost, and so has seeing the results in the mirror. I'm in the best shape of my life - or, to be less grandiose about it, I'm in shape for the first time in my life. I also stream 1 hour of my workout each day, which has been a fun experience and really motivational.
I think it accounts for the majority of the difference between me last year who wanted to learn piano and me now who's two weeks into piano lessons.
Once I have a six pack and grade 8 piano, I'll have completed my mindset training and be prepared to start working on the real goal: getting good at Super Smash Bros Melee.
Is your characterization of your column and apology as "badly" and "bad job" intended to mean that you believe it was possible to do those things in a way that did not produce the negative responses that you go? I often encounter that sort of thinking in places I am relatively certain it is not accurate, but I can't tell in the context of this post. Saying/thinking these things when there was no less-bad approach available leads to faulty conclusions, in my experience.
This is of course a matter of taste, but I generally preferred the aesthetic of the previous blog. I don't know how much control you have with Substack, but experimenting with different fonts and formatting options might be worthwhile.
On desktop, you can get pretty close to the previous aesthetic by following the instructions in this post, you just have to install a Chrome or Firefox extension: https://applieddivinitystudies.com/slatestarsubstack/
Hopefully this does not breach the 'no-politics' nature of this thread but... Has anyone read "Cynical Theories: How Activist Scholarship Made Evyerhting about Race, Gender, and Identity..."? Does anyone have any opinions on how it treats the subject of Postmodernism? Does it get at the crux of it? I am only a few chapters in, but I am recalling that Jordan Peterson was rightly pilloried for his lack of nuance in understanding Postmodernism. These authors, however, seem to have done their homework. So far, it's an excellent, balanced book.
Like on so many academic topics, may be worth starting with the Very Short Introduction series. It tends to be well researched, and adopts a sort of detached and comprehensive account of whatever topic that can lend a certain credibility (though I can't vouch for this title in particular, haven't read it personally. Many of the others I've read are very good.).
Thanks for recommending Cynical Theories! I am also interested in this subject and second your request for any more book recommendations on the subject of postmodernism.
It's a really interesting book. It is Academic, in the sense that it is NOT polemical and is well researched, but it is Popular in the sense that it is clearly written and obviously written for a non-specialist audience. There are not too many books out there that could fall under the heading of 'Popular Philosophy', and even less that seem to do such a good job of it. Even more to the credit of the authors is that they offer tangible solutions to the problems of Critical Theory, and are sensitive to the complaints of what they oppose.
I think there's a major problem with the word "postmodernism" which is that no one can agree on what they mean by it. It's almost as bad as the term "neoliberalism" has become, but with the advantage that most of the fighting about "postmodernism" is probably a few decades ago.
If you want a steelman of postmodernism, you could do worse than Michel Foucault's side of the debate he had with Noam Chomsky on the subject of "human nature" on Dutch television: https://www.youtube.com/watch?v=3wfNl2L0Gf8
This is a classic. I'd seen it several years ago but forgot about it until now, so thanks for reminding me of this. Not quite as exciting as Norman Mailer vs. Feminists in Town Bloody Hall ( https://www.youtube.com/watch?v=gGYmyou0sKM ) - but still… Actually, the Foucault/Chomsky debate incredibly rewarding. Despite what postmodernism has spawned, Foucault has some very resonant ideas.
But your right. In the end, like all ideological terms, ‘postmodernism’ AS A WORD just dissolves into space. It’s fuzzy. However, it is a theory that is alive and thriving in Academia especially, as is often pointed out. I think it behooves us to understand it fully since its tendrils are still growing/evolving, and enter popular ideology in important ways.
The general consensus I’ve seen from philosophers in my orbit is that it doesn’t - it’s not as bad as Peterson, but the scholarship is only superficially good. Sam Hoadley Brill has a review in Liberal Currents that takes a narrow look at what he considers the most egregious chapter, and that seemed to match the opinion of most of the people I talked to (I’ve not read the book myself, I’m just conveying the reactions of the folks I know who have)
Thanks for drawing my attention to Brill's article (https://www.liberalcurrents.com/the-cynical-theorists-behind-cynical-theories/). Seems to counter a lot of the points made in the 'Cynical Theories book, and points to some important misreadings of the authors they cite to build their points. Looking forward to checking this out further. Cheers!
There are two reasons I find that review really valuable - the first is that it doesn’t focus on Cynical Theories’ (mis)interpretations of the original postmodernists, but in the part where they talk about the modern academics they’re critiquing. I saw so many people talking about Pluckrose & Lindsay’s reading of postmodernists like Foucault and Derrida, most of which sounded really bad but it also doesn’t seem to be the central argument of the book? It’s not a great sign if their account of the history is really wrong but their core claim seems to be that modern scholars are *mis*applying the original postmodernism anyway.
The other thing I appreciated was the highlighting of the lax scholarship - minimal citations, references to entire books or even just authors for contentious claims, that sort of thing. It helped confirm that I wasn’t interested in reading it
My browser just ate my comment, so here is a short version: I think Cynical Theories is very good, but you have to see that is solely about the reception about post-structuralism(= the works of Foucault, Derrida, etc.), and less about post-structuralism itself. Pluckrose and Lindsay show how a very playful philosopy interested in off-beat readings of classic texts became an orthodox school of thought obsessed with reading oppression into everything.
They do this by thoroughly highlighting the works of thinkers like Kimberlé Crenshaw, who were able to turn post-structuralism into a highly effective tool for activists. The cost was, of course, that this created a version of post-structuralism with the sophistry turned up to eleven.
In my opinion, there are some things they could expand upon, especially on the more philosophical and historical side (and they have done so partly through articles at Areo Magazine and New Discourses). They more or less admit in the chapter about Critical Race Theory that it takes as much from the Frankfurt School as from Post-structuralism. This is very interesting, because the Frankfurt School was very activist, and in fact had a big influence on radical political activism in the 60s. In my opinion, if you squint your eyes, Critical Social Justice is the language and arguments of Post-structuralism fused with the activism of the Frankfurt School.
The other thing that could be expanded upon is the American ingredients in Critical Social Justice. Critical Race theory is, of all things, an outgrowth of American legal theory. Generally, the American university system plays a big role in why Critical Social Justice and Applied Postmodernism is how it is. American universities are obsessed with teaching a very specific Canon (see all the Great Books courses for example), which is why Applied Postmodernism hates Canons as a reaction. Critical Social Justice then simply replaces the Canon with another. Continental universities, by comparison, are much more random, you can go from a professor who is a hardcore deconstructivist to a professor who takes 19th century dance theater at face value in an instant.
I think for reading recommendations, since you are reading a work of secondary literature at the moment, it would be interesting to read some actual works by post-structuralists. Foucault is quite ok to read if you are careful not to be tricked by his rhetoric, I can recommend Discipline and Punish. If you have a masochistic streak, read Gender Trouble by Judith Butler afterwards, and experience how she builds upon Foucault in some... interesting ways.
By the way, I read the article about Cynical Theories in Liberal Currents, I found it very nit-picky, and not in any way relevant to the core arguments of Cynical Theories.
I read it about a month ago and I feel like it treated the concept fairly. It is well-researched and it shows when they steelman the philosophical arguments used in the various postmodern and critical theories.
On the other hand, it is clear that the authors have a view (correct, in my opinion), but it's pretty obvious when they are giving it.
So what should we call this blog? Obvious candidates that I've seen around are ACX and ACT. I strongly prefer ACX for some reason. Or will folks just stick with SSC?
ACT is already Acceptance and Committment Therapy and Assertive Community Treatment in psychiatry, probably other things in other fields. I think ACX is free.
As long as we're okay with being _constantly_ confused for the Audiobook Creation Exchange and Access Control Experts, we can probably manage. It'll be tough, but I think we can make it through.
1) Has anyone else had issues with the RSS feed? I didn't get a notification for this post despite subscribing.
2) There's a lot of talk over which cities are most livable by various metrics - housing supply, jobs, transit/pedestrian friendliness. I'm curious about anecdotal subjective experience - people who have lived in multiple cities, which one was your favourite overall? (And I mean including every factor - e.g. "NYC was dirty and badly run but I liked it overall because I had an interesting highly-paid job" should still count, since job market is somewhat location-dependant).
Hm, it just showed up for me (about an hour after it was posted). So I guess it's not a general RSS issue, my reader just only updates every few hours.
I'm interested if there's any truth to whether different cities have actual network effects that matters for things like innovation. I read a blog post recently that said you can't walk around NYC without thinking about how much money you have and ... I don't know, NYC is very big. I guess it depends where you live. But I wonder if that's an accurate reflection of density in mixed income neighborhoods?
I will say that NYC is difficult with little kids and no car, but it's far more pedestrian friendly than other cities in the US that I've been in because of public transit. The MTA may be a failing system with unsustainable costs and maddening inefficiencies, but it also comes close to being a truly comprehensive system for the city in a way that e.g. Pittsburgh, Boston, Chicago's transit systems do not. Likewise the CitiBike system is maddening but it's well on the way to being a hugely impactful program, even as the city seems to do everything in its power to keep biking dangerous.
When I visited London it felt very close to NYC to me, with the exception of the tube which was eerily clean and well functioning.
Not sure if this is helpful, but the RSS feed is working. (Notifications depend on how you’re subscribed to the feed. I manually checked that the feed now contains this post.)
SF is currently undergoing a real estate and tech job implosion that is going to reset a lot of markets here 10-20 years. I think it will be a very popular place to move to next year.
With the new blog having just launched there are probably a lot of new and returning readers and commenters that just want to say hi or introduce themselves. I’ll start.
I found SSC a bit over two years ago and binged the archives like mad. I had come across overcoming bias back in 2009 and read those archives up to that point but didn’t really stick around or engage with the community. I just lurked on Scott’s blog as well, but I have really missed it since it has been down and now that it is back I want to engage more; So Hello everyone!
I will avoid mentioning who I work for to preserve anonymity, but I am an aerospace engineer specializing in rocket propulsion and will be happy to comment on or answer questions regarding rocketry technology or space exploration. Other intellectual interests include history, and psychology especially as it relates to leadership and management.
In terms of cost disease, is rocketry more like tech (where things get exponentially cheaper over time), or transit/healthcare (where things get inexplicably more expensive and become less affordable over time)?
Oh man, if I had my own blog I could probably write an entire multi-thousand word post on this. The shortest version I can manage is that it is hard to tell. Up until very recently most rocket development was done by big aerospace companies working closely with their biggest customers: NASA and the military. This tended to result in a cost disease type progression for a variety of reasons. Now with companies like SpaceX and Blue Origin doing more of their own things we are starting to see prices for orbital launch fall a lot, but it is hard to tell if this will be sustained. Personally, I think it will be, at least for a while, and we will continue to see falling costs at the same time as increasing capabilities. Definitely a cool time to be in the field.
For more boring types of rocket propulsion systems (i.e. not launch vehicles) it is clearer cut. The size and cost of things like attitude control thrusters for satellites have been following the ‘tech’ model and dropping steadily even as performance improves, largely due to advances in electric propulsion. Small military rocket propulsion systems for things like anti-tank missiles, air to air missiles etc. are usually pretty simple and cheap, with the insane cost of the overall system driven more by the targeting abilities and the need to make the entire thing as close to indestructible and idiot-proof as possible than by the propulsion system.
My unsupported guess about transit/healthcare/education is that the things aren't getting more expensive, instead they are the only things that aren't getting cheaper. I think its been shown that if you break down where the money is going in a cost disease ridden industry, there isn't any one particular subsection that is getting more expensive. I'm suspicious that if you do the the same thing for an industry without cost disease, you will find that there are subsections getting cheaper.
This is a nice idea! And perfectly aligns with my goal for this year, to become more involved in the community.
I don't remember how long ago I found SSC - the answer to this is lost in the hazy days of my post-highschool years. I've been reading HPMOR (and was a major contributor to one of the translations) as it was being released, then stopped midway, then continued after the hiatus (I think there was a hiatus?), then the sequences somewhere in there. I've always been a lurker, but maybe this will change from now on :)
I'm about to finish my MSc in physics (quantum optics), and very excited to be starting soon my PhD in cognitive science. Most of what I know about it is from reading Scott's book review of Surfing Uncertainty, so I can't answer any questions about that yet, but I'm always happy to chat about physics if anyone's interested.
Apparently physicists are seen as "generally smart people who can do well even without domain-specific knowledge". I don't know if that's true, but I definitely took advantage of that. It also helped that the PI I'm going to work with is a physicist, and was looking for some like-minded students :)
I found the SSC blog while Scott was on his little hiatus, so I was unable to participate. I had previously been binge reading LessWrong. (I really should "hunker down" and read the Sequences, in order, and HP:MOR.) I find it hard to engage in the rationality community (though I guess SSC is technically not part of that) since everything they say sort of goes over my head. I'm a senior in high school. I like math and science "generally", but I don't really know what I want to do in college and whatnot.
I only recently managed to finish the Sequences (after knowing about them for, jeez, nearly ten years), but I've read HPMOR like four times, no hunkering down required! It's so fun. And I first read it when I was in high school, so definitely accessible.
I also had the problem of people saying things that went over my head for a long time. I spent a couple years reading like crazy (I read the entire SSC archive while in college) and listening a bunch, and now I'm at a point where I feel comfortable contributing most of the time – although not in every single domain, e.g. I don't have much to contribute when it comes to technical AI safety... but that's okay! I guess what I mean to say is that a lot of the things are only hard to understand because they're high-context, not because you're not smart :)
Thanks for the encouragement. I guess, I still can't really participate in the conversation. But helps to know that I CAN learn these things. Like, me wanting to learn these things is independent of whether I can actually participate, so the existence of these conversations are something to be grateful for (that once I learn, I have somewhere people will listen), rather than be jealous of.
I also have the same feelings about contributing - Even with stuff where I have some subject-matter experience like Math/Programming/Finance, we have a whole bunch of actual experts who know way more than me (well, unless for some reason we ever get to talking about random integer matrices, which was the specific subject of my PhD), so even then I don't feel like I can add much.
👋 I've been reading SSC since ~2014, but have never contributed online - I went to a few EA/Rationalist meetups but didn't attend consistently. I'm a software engineer working at a biotech startup, with a previous life in art conservation. I'm looking to improve my writing and get more involved with the community (man, does quarantine make you reassess where your social skill points are being spent!). I also do a lot of crafts and am happy to talk about any of the above.
>I will avoid mentioning who I work for to preserve anonymity, but I am an aerospace engineer specializing in rocket propulsion and will be happy to comment on or answer questions regarding rocketry technology or space exploration.
Ah. Another member of the SSC/ASX aerospace fraternity. I'm also an aerospace engineer, specialized in rockets in college, currently doing military aviation stuff. And there's John Schilling, who there's a decent chance you know professionally. A couple of others, too. Hopefully gbdub makes it back.
I'm here, but still trying to figure out how to make efficient use of this comment system. Short term, you may have better luck finding me at Data Secrets Lox. And if you know a source for good, cheap attitude control thrusters, I'd like to hear about it. In my world, the good ones aren't cheap and the cheap ones don't always work.
I missed Scott's insightfulness and creativity. I found out about Scott through a chain of HPMOR -> Eliezer -> SSC, but another chain from HPMOR is my interest in Fan Fiction, and now I'm writing my second quest on Sufficient Velocity. I'm doing one that I call a Let's Quest (like a Let's Play but User vote controlled).
I was born in a tiny town, then I lived in a big city and now a small town again, and SSC has been a community to stay in touch with communities that just don't exist locally now that I'm back in a small town.
I've been married for 5 years and now becoming a first time parent in May.
Hi! I found SSC almost 2 years ago I think? I think like many people I had the experience of finding some post I really liked, and then finding another, and another, and eventually giving up and just started reading anything I could find. I got to reading LW and overcoming bias, and HPMOR, and I found everything extremely interesting, especially as I'm currently in a master programme AI, and it has a way of relating back to almost everything.
I'd like to get more involved as well, it seems I had just gotten into the habit of following every post (and every link in every post, Scott has a way of filling up my reading time!) when the blog got deleted. Hello everyone!
Hello and welcome everyone! I go by Aristides, and have been reading SSC since the Last Pychiatrist stopped posting, and I searched for other writers, about 7 years ago. Despite Scott's humbleness, I consider Scott's influence on my thoughts to have been much more positive, so it turns out that TLP leaving, worked in my favor. I have been an occasional commenter for 2 or 3 years, and look forward to be more involved in this site.
I am a law grad that works HR for a government hospital. I am also Eastern Orthodox and moderately conservative. Hopefully others will find my perspective interesting.
There's a massive Greek Orthodox church going up in our small city, and I've been curious about the perspective of that group. I have family members that converted to Catholic and I've become protestant non-denominational after growing up without a church background, so it would be an interesting discussion.
I work with our HR director and accountants and provide IT support and reporting as needed., especially now with processing end of year stuff. Are you busy at end of year as well. I was just writing a 5500 census report, and waiting for it to finish loading data to verify it looked correct.
I used to be a non-denominational protestant as well, and converted a few years ago. I am hardly an expert in it, since a large percentage of sermons I have attended have been in a language I do not speak, and I stopped attending once COVID hit. Overall, I love the atmosphere, and after studying the theology, I believe it has the highest percent chance of being correct. Catholicism has had too much corruption in its history for the doctrine of papal infallibility to make sense to me, and all Protestant denominations made so many changes to original Christianity, that if any of them are actually the truth, they discovered it through sheer luck. That's my thoughts anyways.
My end of calendar year is the least busy time, since the fiscal year starts at Oct 1 here, and that's when any of the annoying reports are. No one wants to discipline or fire employees around the holidays either, so that helps keep work light, too. That said, I think work will pick up soon with the administration change.
Hi! I've read SSC for at least 5 years now, but had always avoided the comments section due to lack of interest. But in the recent SSC hiatus I decided that was silly and from now on I'll try to participate more.
I'm a cryptographer and am also in the process of learning some stuff about law (in the US). Vaguely EA too but more human-centric than most of their discourse seems to be these days.
(Is anyone else experiencing weird lag when typing their comments, or is it just me?)
Followed from Overcoming Bias to LessWrong. Enjoyed many of Scotts posts there but didn't know they were by the same person until it came up in the post about governing fictional worlds.
Never really engaged in the comments on either site, might make more of an effort to do so this time around.
Giddy with excitment to see the blog up and running again. I got a notification that there's a new post while I was writing this!
What's the state of research in rocket engines? Is it mostly a solved problem? Do people still innovate in say, the basic geometry of combustion chambers?
It definitely is not a solved problem, there are several big open questions in rocketry, and advances are being made if not as fast as we might like. To give a quick rundown of some problem areas and places where new things are happening:
Combustion instability: this is probably the biggest single unsolved problem in rocketry. The short, simple explanation is that combustion instabilities are pressure oscillations in the combustion chamber that can grow via various feedback mechanisms until they are strong enough to destroy the engine. This almost killed the entire Apollo program, as the F1 engines for the Saturn V initially had extreme combustion instability problems. The various mechanisms involved, and the interactions between them are complex enough that they still cannot be modeled effectively with computer simulations, so even today it is not possible to predict in advance whether a given engine design will be stable or not. Larger engines are more prone to instability problems, so small scale testing doesn’t really help, the F1 engine on the Saturn V was a scaled-up version of an earlier missile engine that was very reliable. Combustion instability is therefore generally considered to be the single biggest risk in any new engine development project, and holds back innovation in any area that might affect it as companies try and stick to iterative improvement on existing designs that are known to be stable.
Pressure-gain combustion: The details are complex, but in layman’s terms if you design the engine such that the propellants explode (detonate) rather than burn the thermodynamics allow you to extract more useful work from a given amount of propellant… in theory. In practice it remains to be seen if a practical engine design can realize enough of the theoretical gains to be worth the other tradeoffs involved. Many designs have been proposed in this area, and it is a very active field of research. Rotating-Detonation engines (https://en.wikipedia.org/wiki/Rotating_detonation_engine) seem to be the lead contender, but it will likely be some time before one is built as anything other than a experiment.
In terms of more immediately applied progress being made, SpaceX’s Raptor engine will be the first full-flow staged combustion engine to see flight service; and the Raptor and the Blue Origin BE4 will finally bring methane fuel into widespread use. Electric propulsion is still improving rapidly, with novel designs still being invented regularly.
Some other areas: Solid propulsion is so simple there are not a lot of areas to improve on, but there is still active research on improved propellant formulations. New propellants and additives are also being actively investigated for hybrid rockets, monopropellant thrusters, and hypergolic thrusters. In particular there is a lot of interest in replacing hydrazine as a monopropellant, and the monomethyl hydrazine (MMH) and nitrogen tetraoxide (NTO) combination for hypergolic propellants, as these are highly toxic and sufficiently volatile that they require extensive safety measures during handling to avoid breathing the fumes. NASA’s Green Propellant Infusion Mission (https://en.wikipedia.org/wiki/Green_Propellant_Infusion_Mission) was a test flight of one potential replacement for hydrazine in monopropellant applications, and ended successfully quite recently.
So while we might wish for faster progress, there is definitely a lot of exciting work going on in the field, and this is just the tip of the iceberg, there are plenty of other areas of research interest besides these.
Ignition! is a fantastic read and very accessible, it is also freely available as a pdf (legally) here: https://library.sciencemadness.org/library/books/ignition.pdf it is more focused on the propellant chemistry than the rocket engineering but still well worth it.
after that if you are willing to spend a little money and don't just want pop-sci books go for one of the following textbooks, both will be at least somewhat accessible to anyone with a hard science/engineering background.
"rocket propulsion elements" by George Sutton or
"rocket propulsion" (cambridge aerospace series) by Heister, Anderson, Pourpoint, and Cassady.
I was introduced to SSC/HPMOR at around the same time, 2014 or so. Definitely remember trying to puzzle out the end of HPMOR in time for the last chapters. Loved SSC, Scott's writing style, and the ideas behind it all. Read through Eliezer's Sequences, and then mostly just lurked on SSC until 2020 happened. Started in on more LessWrong after that to bridge over until recently, when ACX emerged!
Studying at a German technical university in my bachelor's, my interests include space/rocket technologies, long walks in uninhabited places, and since March (thank you COVID-19) programming with Python, followed closely by HTML, CSS, and JS to facilitate getting functioning Heroku Apps going.
Matt Levine’s posts have made me want to invest in a SPAC upfront (i.e. for $10 with the warrant). Is this one of those things where you have to be a VHNWI and have a relationship with a bank?
So you're saying you *don't* want to buy shares+warrants of a SPAC that already exists, like PSTH or something? To do it "upfront" means investing in the IPO of the SPAC- like any other IPO, you'd need to participate in the opening book, and that means either being a bank and participating in the road show, or having a big broker and a good relationship so that you can get some of the few chunks allocated to retail brokers.
Basically what Andrew said. I don't follow Levine's newsletter anymore (a little too snarky for my taste) but my impression is SPACs are a lot more interesting for the sellside and bankers than anyone on the buyside.
It's interesting that SPACs have become fashionable just when direct listings have become possible.
Brokers make their livings from their contacts. They earn commission from finding buyers for blocks of shares, for instance.
In the IPO process the chosen brokers place blocks of Pre IPO shares with their friends, making sure that the company value for those pre IPO shares is below fair price. On IPO day, the fair price is discovered by the market, the shares "pop", the friends make a handsome profit, and of course the people who owned the company before the IPO process make less than fair market price.
I'm showing my age, but the lack of liquidity discount for unlisted shares has traditionally been about 40%. It's lower now.
The below is at least true for Sweden.
Valuations before public listings: VC's and private owners (VC funds are often partly pension money and therefore IMO not entirely private money) have been holding off longer before listing successful companies for the last couple of decades. They can fund the company themselves, so access to public market money isn't necessary for the company. The incumbent investors get to keep more of the market cap growth during late exponential scaling to themselves.
Those successful companies have *no* trouble finding money. A story in my vicinity unfolding now: A promising company closed a financing round in late December 2020. There wasn't space in that round to admit a new well known investor; the incumbent investors took the entire round. That new investor returned days after the previous round was closed, offering a 25% higher valuation to invest money the company doesn't need.
For the vast majority of less attractive companies, capital is accessible. It's been stimulus all the way down for decades now and money is looking for assets to buy.
- Small company without much traction / generational transition: brokers sell to private investors. This is a lively market. Lots of brokers, LOTS of companies, lots of investors, some of them building conglomerates / adding value to their favorite company flavor. Brokers deserve their 7-10% commission; it's hard work.
- Easy to understand product, can't deal with the overhead of listing, still want access to some public capital: try Alternativa Aktiemarknaden. Report your numbers quarterly, trade is organized one week/ month. There are various alternatives for crowdfunded shares, passion driven shares like teams or bands, etc. It works pretty well. https://www.pepins.com/market/listing#show
- Small company, can handle the administrative overhead of listing, worth at least 50MUSD. List on the NASDAQ/OMX.
so, direct listings are now mainstream and it shouldn't be a big deal for any company to use them, yet a lot of companies just turned to SPACs which are more expensive and fee-ridden than conventional IPOs.
Direct listings are not yet mainstream; brokers are fighting its adoption. Brokers and their best clients do lots of listings while founders do at most two, so founders rely on brokers expertise in the listing process. Brokers press their informational advantage and recommend the traditional IPO process since it generates large income streams for them and for their best clients.
For the VC side, try the Bill Gurley podcast above.
I don't know what SPACs have suddenly become popular; the below is speculation. I think it's a bit similar to the ICO frenzy a few years ago.
- Brokers and well known players want to make money on their contact nets and ability to hype companies.
- Companies want high valuations and good visibility quickly.
- Sarbanes - Oxley made listing onerous, so there are fewer listed companies.
- There is a lot of private capital around, so companies don't have to list to get financing in late exponential growth, reducing the number of publicly listed companies.
- Private equity buys companies / divisions off the public markets, reducing the number of publicly listed companies.
- Investors who don't mind a lot of risk don't have much choice, especially if they're young and haven't yet found their way onto broker's call lists. The success of Robinhood demonstrates Millennial risk appetite.
-Creative and clever talent flows to the finance sector.
Not many investment vehicles, lots of money around, insatiable risk appetite and tough regulations to reduce fraud plus a surfeit of creative clever talent
- Regulatory arbitrage / less onerous paths to public capital - like ICOs a few years ago - are going to be popular in this environment.
This afternoon, I heard of another SPAC being set up by stellar investors. They believe they can manage more money than the billion USD or so they manage now. They're thinking of setting up a SPAC, and will allow new customers to invest with them. For a fee.
They are skilled. Decades of experience with generally excellent returns and no terrible years. Up 98% in 2020, while investing exclusively in listed shares. (!)
I can understand why people might be interested in investing in *their* SPAC, even if the fees will be high.
Maybe that is the attraction of SPACs. Customers get a chance to ride with the stellar investors for a fee.
Glad that the blog is back! I'm gonna make immediate use of it by asking this: I'm starting a PhD in Psychology in the Netherlands. My PhD friends have told me Twitter is essential for keeping up with relevant publications, events, trends and development and that I should absolutely make an account.
A) What does everyone think about this assessment and B) if I do make a Twitter account, what are good ways to filter out all of the bullshit?
Science PhD student here. I heard something similar and got on twitter for this reason. I've come to find science twitter to be undewhelming, very heavy wokeness-signaling (I say as someone much more woke-aligned than a typical gray triber), and fairly little actual science. The actual science does tend to be good though, and I've seen some *very* cool times where scientific collaborations actually formed on twitter and produced really cool stuff. Overall science twitter has more than 0 value but is also pretty annoying. Note though that your field may be totally different.
That said, after getting on science twitter I ended up discovering rationalist/postrat/whatever twitter and I've been really enjoying that little subculture. I ended up just splitting it up between two different twitter accounts and I find I spent more time in the rat-sphere twitter than science twitter.
Math professor here. It may be different in psychology, but it's hard for me to imagine how Twitter can be useful. https://psyarxiv.com/ is probably a better way to keep up with what's new. At the same time, when some truly important papers in your area come out you will hear about them from your advisor or other students. Your advisor is a much better source of information on what's in trend than twitter.
Personally I am most productive when I change all my social media passwords and hide them somewhere far away. The most important thing to succeed in academia is to be able to keep your concentration for a long time, patiently struggle through long and intense periods of boredom and confusion, sometimes literally banging your head against the wall.
It's really hard to do if you didn't hide your Twitter password in a good place.
Thank you, the productivity issue definitely crossed my mind as I already have a hard time resisting the temptation to check fora, messages and feeds when I'm supposed to work. Definitely going to keep it off my phone.
Not in academia, so no opinion about A, but for B, I make *heavy* use of muted words (e.g. "trump", "election", "woke", etc) and it makes Twitter tolerable - even enjoyable sometimes!
Senior PhD friends telling you is good evidence, ask more students and profs to be even more certain.
A: very much field dependent. I finished PhD in pure math a few years back and never used twitter. from my ex soon to be a bio professor I know twitter is essential tool in biology or at least a range of fields around hers: many/most profs are there, people use it to announce their papers, all major conferences have hashtags around which the conversation evolves etc..
B: you seem to not understand what twitter is. it's a social network. you follow a bunch of people, and what is shown in your feed is mostly what they post. if you follow researchers in your field and twitter is indeed actively used as a professional tool in that field, I'd expect most of their tweets being professional.
Not having twitter, I don't know how it works, true. I suspected it might do a thing similar to facebook and 'curate' the tweets of the people you follow according to shareability or engagement or whatever. If I can just pick a list and not worry about it being filled with the outrage du jour, that would be nice.
I don't know about academia specifically, but my general impression is that, if something's important, it will move from Twitter to slower-paced media/fora. If you read the relevant books, journals, etc. and talk to your professors and classmates, my guess is that you'll get most of what you need, and possibly more than you would get by spending the equivalent time on Twitter. But, again, this is based on speculative conjecture, not informed experience.
And I think it's worth noting, as other commenters have mentioned, that there's a difference between how a perfectly serene and rational truth-seeker would ideally use Twitter and how you, personally, are likely to actually use Twitter. At least for me, I've found that only using the app to send DMs/contact people about specific things has been a huge improvement over trying to use it well in its core functions. It's sort of a point of pride for me at this point that I now often hear about whatever the latest stupid Internet controversy is for the first time when I read an article about it in The New York Times. But YMMV.
I'd be interested if people have specific recommendations for people to follow on Twitter for psychology publications/events/etc., both global and in the Netherlands. Do people have some tips?
Political scientist here. In political science, Twitter is very useful for keeping up with publications and current debates. The most useful thing you get on twitter than you don't get in other places is real time debates about research. This gives you a much better sense of where the discipline is right now, as opposed to 3 years ago when the papers that are being published were written.
It's hard to filter out the bullshit not least because some of the bullshit is actually useful information about the norms and practices of people in the discipline. My best suggestions would be to just scroll past the tiresome tweets, mute anyone who's raising your blood pressure too often and mute keywords that come up in repeated discussions or memes that don't add anything.
Academic Twitter is mostly good, but at random times once or twice a year it becomes incredibly stupid. On bio Twitter recently:
- an offhand joke about a certain species of worm turned into an enormous fight in which people who made fun of the worm were somehow accused of being bigoted.
- an online funeral was organized (by some fairly high-profile biologists) for a Native American scientist who supposedly had died of COVID, but in fact turned out to be the fake identity of a white scientist.
Pretty much any academic community has these bizarre dramas and if you get caught up in them, they might totally blow up your career. So, don't do that.
Actionable advice: you should make a Twitter account. Follow people in your field. Don't post much, never argue with anyone. If you notice someone who is very argumentative or seems to spend a lot of their time mocking other Twitter users, mute them. Do not install the app on your phone.
AFAIK, there are several possible 'varieties', or possibly several distinct ideas with the same name. My personal favorite is David Chapman's 'meta-rationality'. His site "Meaningness" is great, though perpetually unfinished.
Irrelevant suggestion - bring back the punny titles for open threads. They brought some levity and variety and I don't think the stated reason for getting rid of them (too confusing for newbies) ever made that much sense.
I'm not just saying this because you never got around to Openguin Thread... okay I kinda am.
This made me wonder if GPT-3 is any good at puns. Alas, according to Gwern:
"Better [than GPT-2], but not as much better as one would expect given the leap on many other capabilities. Trying to generate puns or rhymes, it seems like GPT-3 know extremely well what they are on an abstract level, and will appropriately manipulate words and attempt to make puns or rhymes (see the shoggoth-cat dialogue below for a particularly striking example), but the words it chooses just aren’t right on a phonetic basis."
The mentioned dialogue has puns like:
"Well, the best pun for me was the one he searched for the third time: “You didn’t eat all my fish, did you?” You see, the word “fish” can be replaced with the word “cats” to make the sentence read “Did you eat all my cats?”"
So I think we're still quite a bit away from GPT-3 generated thread title puns.
Health question: this last year, unexplained and unplanned I've lost about 30 pounds of weight, and about five inches off my waist and I now have to use suspenders to hold my pants up,
Has this happened to anyone else?
Unfortunately in this time of covid-19 speaking to a physician or nurse practitioner is really hard, I called for six months with lung symptoms before finally being seen, an x-ray and my weight loss prompted my doctor to say "80% lung cancer", but a biopsy found a fungal infection instead. I'm guessing at least a year until pre-2020 style medical care is available again, weight loss remains a mystery.
Lost a similar amount of weight, maybe a little less; but I'm fairly certain it's just due to malnutrition. Too depressed to make proper meals and feed myself most of the time, making up the difference with coffee flavoured with milk and sugar.
As an already-skinny person, I once lost 10-15% of my body weight over a summer because I was playing computer games all day. I lost muscle from the inactivity, but I also lost fat, just because being so inactive and so absorbed in playing games suppressed my appetite. (I do *not* recommend this as a weight-loss method; I very much doubt it would work for anyone who *wanted* that result!) I gained the weight back once I started being active again. Has Covid led you to be sitting in the same place all day a lot more often than you used to?
* There was even at the time a lot of government and institutional support for prominent charter networks
* Closer investigation into what's going on in those schools should have made it clearer that they were focused on test prep in a way that wouldn't lead to sustainable long-term benefits (KIPP has had a hard time showing increased college enrollment and graduation, though I think after changing to focus on that they now have a modest impact on that)
* Research was very limited at the time on impact of no excuses charters
I'm having a hard time separating disturbing things I've heard from people who have done stints at KIPP (and Success Academy etc) from what was understood in 2011. I think this should be seen as a failure of Effective Altruism/Givewell but I'm interested in defenses.
Was KIPP really the best place to give money in the US in 2011?
One of the ideas behind Brave (brave.com) is to have privacy by default. It comes with a built-in tracker-, add-, script-, and fingerprint-blocker. You don't need to get add-ons at all (I have none). So as far as I can tell it is basically as convinent as possible.
I didn't really thoroughly research existing browsers, though; no comment on how it compares to other options.
I've been using Brave for a while now, and I'm pretty happy with it. The only changes I would make out of the box are to set tracker/ad blocking to aggressive, and fingerprint blocking to strict. I haven't encountered any broken sites with that, except for Google Maps which doesn't load some of the JS (one could argue that if you really want privacy, you shouldn't be using Google in the first place). It's also really simple to disable the shields (blocking) on an individual site.
If you're up for it, you can add uBlock Origin and Privacy Badger, which are pretty good without any config changes. uBlock is good for getting anything that isn't blocked by Brave, as well as the ability to add custom block lists. It also makes it easy to zap specific elements of a site. Privacy Badger also blocks some things that the others don't get, like Disqus comments. Both extensions can be disabled on an individual basis.
Downloading Brave and adding uBlock and Privacy Badger should be about 3 minutes of work, so I'd recommend that as a good starting point.
(Bonus advice: get a password manager while you're at it. Helps against password reuse as well as phishing attacks. 1Password is a good option.)
Scott's difficulties setting up his psychiatry practice remind me of Paul Morphy (https://en.wikipedia.org/wiki/Paul_Morphy) who, after becoming widely recognized as the world chess champion at 21, retired from competitive play to open a law practice. His law practice failed because everyone who came to him just wanted to talk about chess. Scott, the widely recognized world blogging champion, may be facing a similar predicament.
Now that my living room no longer looks like I've been fighting Terminators, I am happy to report that I have found a practical use for RGB lighting! To my utter shock, my new PC rig booted on the first try, but one of my memory modules didn't light up. What do you know, Windows is only detecting one module. Fortunately, it just needed to be reseated, but who knows how long I would have gone with only 16GB before I noticed the problem? Since the RGB not only alerted me to the issue, but obviously indicated which module had the problem, I retract my prior opinion that RGB memory modules were frivolous nonsense. (RGB fans, on the other hand...)
Any good books/resources on soft skills? I suddenly found myself in a position in work where being a bit better at persuasion, negotiation and de-escalation would be much appreciated.
I'm finding it hard to find quality resources and easy to find crap.
It feels so "basic" to recommend, and I've heard some good criticisms of it that I forget at the moment; but if you haven't read it, it's definitely worth reading.
It's not an instructional book and doesn't have anything about business negotiation, but it is a great example of what talking about emotionally loaded topics in an emotionally intelligent way looks like in action.
Assuming you're after the emotional (not logical) component of negotiation and de-escalation.
The Harvard Negotiation Project has some really good books on the subject. I would definitely recommend "Difficult Conversations" and "Getting to Yes".
The Power Paradox: How We Gain and Lose Influence, was the most useful one for me, coming into HR from a law background. I probably should have read more, but that was the most engaging for me.
I had a shower thought about what will happen to Western diets if meat (and only meat) suddenly increases in price, as has been predicted extensively. I haven’t been able to find examples of this happening to a society before and I was curious if it’s more likely we’ll expand the window of acceptable animals to eat, move to insect proteins, or move to mass vegetarianism? I haven’t been able to find much via Google, so I thought I’d reach out here.
Given how much meat substitutes have improved in quality and prevalence over the last few years, I'm assuming that those will become much more popular. Insect protein doesn't have nearly as much of a head start as Impossible Burgers and Beyond Burgers etc. so I don't see that catching on, as much as I would like to. Eating other animals seems similarly unlikely because they're not factory farmed, and the meat produced at the largest scale is obviously the cheapest, so I don't see how e.g. horse meat would compete.
That last question is the biggest one to me. it just seems unlikely; there is so much land that is unsuitable for growing vegetables or cereals; but is suitable for grazing livestock;
Rising incomes in Africa/Asia causing the people there to purchase more meat?
By limiting demand to a smaller number of people, namely those who can afford it, we might expect that the quality of meat being sold will increase. What has happened to American beef — the best in the world 30 years ago — is a sin and a shame, and for meat eaters, a tragedy. So, as a member of that increasingly decreasing class of people, I would look forward to it.
Do you have a link, so that a young cow addict can read about what happened to American beef? I always thought it tasted fine...
Separately, the economic question of "cost of manufacture goes up --- does quality rise or fall?" is not obvious to me. I actually had the opposite guess (more pressure to cut costs causes quality to decline), but after reflection I'm no longer sure at all. Should the answer be obvious?
Not much, unless the price increase is something on the order of doubling or more. Maybe you'd see increased adoption of extended meat blends, which are basically only used in institutional settings right now. Fast food would either adopt them or try and shift more upscale. Most people who can afford it will pay more at the grocery store rather than modify their diets.
I'm am not a historian, but remember that for much of pre medieval history mainly the rich ate meat on a regular basis. Perhaps this will recur, with "true meat" costing hundreds of dollars a kilo once "lab meat" reduces in price.
I could see factory-farmed meat increasing in price relative to lower-intensity farming, but small farmers producing free-run chicken and pork and rotationally grazed grass-fed beef to sell directly to customers has never been all that expensive and isn't dependent on volatile inputs and layers of middle-men. Buying whole frozen chickens and half-cows from farmers is more work, but it's not really more expensive, and it tastes better, it's better for the animals, and it's much better for the environment.
The economies of scale involved are interesting. Just letting cows eat grass is really quite cheap as long as land is cheap and it rains enough. Some breeds are optimized for getting really swole on expensive industrial inputs, and others (e.g. Black Angus) are smaller, hardier, and able to take care of themselves with a little supplemental hay in winter. The smaller breeds also have an easier time calving, reducing vet bills.
Of course, I'm in Canada where we have lots of land and we don't do the kind of intensive feedlot beef production that's common south of the border. Chicken has never been dirt-cheap either. The factory farming situation isn't great but it isn't US levels of bad either.
The change I expect to see is both a shift from the crappiest, least-meaty, most-environmentally-destructive meat products to plant-based analogues, along with a surge in decentralized consumer-direct animal agriculture. The latter is already happening quite rapidly where I am.
for most people, the list of famous psychiatrists begins and ends with Jung and Freud, who for cultural reasons catered to rich people.
the real art of being a good psychiatric counselor has NOTHING to do with being like Jung and Freud in that respect. The good results from any patient are equivalent to the good results from any other patient, there are no bonus points for being there for people who are richer, who have more friends, who are more gifted, who are more interesting, who are more special in any way. All of God's creatures are uniquely spectacular.
I hate fake Tables of Contents. Most books have them. You know the ones, the book is divided into five chapters or so, and each chapter has many subsections that are not even listed in the Table of Contents, so when you want to find something again, you can't, or when you want to remind yourself of the general cut and thrust of the book, you can't do it near so well, because instead all you get from the TOC is a handful of scarcely meaningful chapter titles. It's as though you want to see the anatomy of the book, and the book just keeps repeating in a sing-song, "Head, Shoulders, Knees and Toes." And you reply, "Damn you, I want to see the whole skeleton!"
Do you ever have this experience?
I am thinking about solving it, but is there demand?
I've most often been annoyed by this in PDFs of RPG sourcebooks. For some reason, there seems to be a trend of releasing them with just the major chapters bookmarked and not any of the useful subheadings.
What are some books similar to Joseph Henrich's The WEIRDest People in the World? (In that they discuss the same/related issues, but from a different angle of some sort.) I ask because I've found Tyler Cowen's advice to read in clusters very useful, so I'd appreciate at least a couple more to round out a cluster including Henrich's book.
Two examples of possibilities here: The Open Society and its Enemies, by Karl Popper, and Fairness and Freedom, by David Hackett Fischer. The former is a work of philosophy criticizing, among other things, a thread of collectivism running through the works of Plato, Hegel, and Marx, as opposed to individualism, and the latter is a comparative history of the US and New Zealand considering their individualist/free market/liberal etc. roots. So they cover what seems like closely related conceptual ground in terms of Western individualism, but from different perspectives: Henrich discusses anthropology, Popper discusses philosophy, and Fischer discusses history. (Note, I've already read The Open Society.) Hopefully that illuminates what I meant about books that form a cluster. (They can also be books on the subject within the same discipline but with different intellectual loci.)
I'm really looking forward to reading Guns, Germs, and Steel, but it seems like it fits into a slightly different cluster to me: namely, regarding the rise of agriculture/Eurasian civilizations/the state etc., along with e.g. Azar Gat's War in Human Civilization and Ian Morris' Why the Rest Rules. (By contrast, I understood Henrich's book to be about Dark Ages and beyond developments.) Although now that I think about I guess there is overlap in terms of the "what made the West unique?" angle.
Democracy Now! is always great.
Thanks for that survey on religion and economics, it looks quite pertinent.
I've recently been watching Criminal Minds on Netflix, as I'm a sucker for police procedurals. It's a fun show, but I have to wonder, how legit is FBI profiling? Seems like one of those things that could easily just be all smoke and mirrors.
This comment has solidified my dislike of the "like" feature. It has 6 likes as of when I post this. That's probably 6 people who want to see the Naval Gazing links back, but I can't be sure, and it's frustrating compared to getting actual information.
Naval Gazing is great, and I like the idea of links in at least the first few OTs of ACX at a minimum to get some mind share on the new site.
Likes are bad if they funge against higher-content replies, but IMO better than no feedback at all - a comment with no responses but a relatively high number of likes is useful information. Still would prefer an opt-in system though.
You have a solid point. With how much effort it takes for you to write a Naval Gazing posts, you would hope that your fans would take the time to make a full comment, not just a quick like.
For what it's worth, while I've always been impressed by how extensive your Naval Gazing posts are, I've honestly only read 3 of them, since it is outside of my area of interest. Hopefully the people that regularly read your Naval Gazing will actually comment, since I remember it being fairly popular.
Naval Gazing has done fine during the hiatus, and I'm not starving for comments there. But those people don't need links posts to remind them of things. I would ask this occasionally on SSC, and usually got several people saying that they enjoyed the posts because they were irregular readers and liked seeing what I was up to.
are you certain those six would have replied otherwise? perhaps that is six new positive responses that you would not have seen at allin the old system
For some reason the RSS on Naval Gazing hasn't been working with my RSS reader (NewsBlur) lately, so I selfishly wouldn't mind the occasional reminder that you exist and have new posts to read.
IIRC, most of the actual donors ended up not being Nobel laureates, because Nobels tend to be awarded for work from previous decades and thus to older recipients. (Also, hilariously, because customers complained that many of the Nobel laureates were too short.) So they ended up being younger, modestly accomplished all around the board types instead.
Have another unsourced theory: there are two kinds of genius. One is related to general good health and possibly height. The other is related to weird shit and possibly related to physical problems.
Can you freeze sperm indefinitely? If so, we should be collecting samples from up-and-coming thinkers and then giving them out decades later when we know which ones will get Nobels. (idea may or may not be stolen from rat-tumblr post about housecat genetics)
Color is mostly #aaa on #111, which is great on my monitor, but YMMV. Has an issue with the skeleton dummy posts that show up before content loads (I can't find their HTML, so they're still glaring white), but I'm not sure how annoying that will actually be. Seems to work otherwise. I'll try to keep this updated when site changes start breaking it, but feel free to fork/do whatever with it.
If all they have about you is something you wrote in college, and where the only issue was some miscommunication... I wouldn't worry too much. There are many worse people and stories. I hope the NYT doesn't publish, but if they do... Remember that in a few years it will just be a painful but distant memory.
That being said, I feel an urgent need for some way for new comments to be easy to find.
I've recommended having brains as well as hearts so that we can mark intellectually satisfying comments as well as just liking them. I don't have enough experience with assorted fora to have an opinion about rating systems. Part of my motivation is to get more nuanced ratings, but I also want to see those cute little brains to click on.
It might also be good (if feasible) to have an "I want to remember this" marker. For that matter, the want-to-remember marker could be part of a rating system. It might be more honest than mere likes.
- a "+1" or "This!" (no need to clutter the comment list just for this)
- a "-1" or "Nope!" (ditto)
The first pair would put a number on the famous light/heat ratio and constitute the SSC equivalent to up- and down-votes, and the second would be a nanoscale survey on the comment's position.
"I also want to see those cute little brains to click on."
Indeed, we should consider servicing the needs of those of our fellow-readers/subscribers who are members of the zombie community (or perhaps, in deference to their sensibilities, I should say the "alternatively-alive community") 😀
As for new comments, it seems like keeping the tab open should work; it automatically shows "1 new reply"/"n new replies" buttons that you can use to load them. You can search for "new repl" to find them. Of course this is suboptimal in that you lose track if you close the tab (or the browser) for any reason.
Gee, several people responded to my comment, so I went here to reply, and I can't even find my comment! Even searching the page (either for my name or for a keyword I used) doesn't turn it up. Very disappointing.
I believe that the current comment structure may suit many Substack users fairly well, but that the SSC crowd needs more of a forum for conversation around the pieces: it would be nice if Substack could offer publishers the choice between forum-style and current style comments.
If you go to Data Secrets Lox, it is a forum with many of the old SSC commenters posting regularly. Personally, I dislike the forum system, so I'll be staying here, but it has a big, active community.
I feel like I very recently saw a paper/blog post/something arguing that the IAT could not, in principle, measure *implicit* bias because it requires the person taking the test to make *explicit* decisions about whatever social category is the focus of the test. Has anyone else seen this or similar arguments? If so, could you point me to it?
I thought the IAT was measuring reaction time, and the point was that unconscious bias is equivalent to strong associations, which in turn influence reaction time?
That said, when I took the IAT, I came out having a more positive association with black people than white people, and I'm white. But I was very conscious of the reason, which was that I've got a learned response to most of the classic associations between "black people" and "[negative stereotype]", which is to stop and step back and take another look at what's going on. (Very useful when trying to assess the level of racial bias in something written 100+ years ago.) So it really wasn't testing my internalized racism, so much as my ability to examine a text to discover how racist it is.
I suppose I could be tested more accurately if they showed me negative stereotypes that were new to me, or that I didn't know were negative. But hopefully that's too Orwellian to pass muster, at least for a while.
I'm actually kind of curious whether anyone's flipped the IAT around, and tried to see what sort of associations cause a shorter or longer reaction time. If one accepts the hypothesis that faster reaction time implies stronger associations, which imply bias, then this could actually give hard data on claims that "[something] is racist". But I'd still suspect that it would be confounded by an inability to distinguish "X is associated to Y in my mind, because I think it's true" and "X is associated to Y in my mind, because bad people think it's true, and I need to keep an eye out for them".
The IAT does make use of reaction times. More specifically, it based on the difference in mean RTs between particular conditions. So, e.g., for the white/black race IAT, in one condition you have to push one button when you see either a white face or a positive-valence word and a different button when you see either a black face or a negative-valence word. Then in another condition, you push one button for white faces or a negative-valence word and a different button for black faces or a positive-valence word. The idea is that if you are slower in the second condition (with white/negative and black/positive pairings) than in the first (white/positive and black/negative), this reflects implicit bias.
The thing I saw was, I think, making an argument that because you make explicit decisions about race in this (version of the) task, it isn't measuring implicit anything. I didn't read the argument in detail, and maybe it was a bad argument, but I'm curious to find it and read it.
- And been looking at the wikipedia article on https://en.wikipedia.org/wiki/Toxin-antitoxin_system , which serves to select for cells that inherit the antitoxin (as otherwise the toxin kills them). There's a few papers that it references that I may look over as well.
Seeing this new blog open prompted me to finally start reading Unsong, and it's great trippy fun! (I'm currently on the last couple chapters of Genesis.) I also keep catching myself trying to Kabbalistically analyze random names and words.
History is fun! More snippets from MacCulloch's book on the Reformation.
Got those pesky marital troubles? Pope won't cough up with an annulment? Never fear, the Reformers are here!
Philipp Melanchthon [big name German reformer if you didn't know https://en.wikipedia.org/wiki/Philip_Melanchthon] ...in 1531 he suggested to Henry VIII that he should solve the Aragon marriage impasse by getting the Pope to provide a dispensation for bigamy.
Uhhhh - thanks, Philipp, but *two* queens at the same time? Maybe not workable?
Henry wasn't the only one struggling with the problem of "I got this hot new squeeze who's demanding I put a ring on it, problem is I'm already tied with the old ball and chain, what do?"
"Against this background of attempted accommodations and adjustments, a moment of opportunity arose in 1540–1 that Charles V was determined to seize. He had a bizarre stroke of luck – a severe dent to the prestige of his most truculent Protestant opponent within the Empire, Philipp of Hesse, which also damaged the reputation of the leading German reformers. It was revealed that Philipp had recently committed bigamy, with the express if reluctant written agreement of Luther, Bucer [another Big Name in the German Reformation] and Melanchthon. Their action was the ultimate in Protestant wooing of the magistrate. The signatories’ logic (apart from the unspoken acknowledgement of how much they owed to Philipp) was that bigamy was better than the adultery which had characterized Philipp’s chaotic private life over the previous decade; Melanchthon and Bucer in particular were only being consistent with the advice about bigamy that they had offered Henry VIII a few years before. Nevertheless the evangelical theologians’ belief that both their action and the bigamous marriage could be kept secret was a colossally naive political misjudgement, bringing bringing predictable and gleeful expressions of moral outrage from their Catholic opponents. Charles graciously granted an imperial pardon to his prince, who could in theory have suffered the death penalty for bigamy as newly laid down in the 1532. imperial law code (Lex Carolina). Melanchthon was so shattered by the self-inflicted catastrophe that he became gravely ill."
Puts a new spin on the oft-cited cry from the Protestant side about Catholic dogma and practices: "Where's that in the Bible?" I had heard it as Luther and others were consulted about Philip's marital problems, couldn't find a prescript for divorce - or rather couldn't settle on recommending it as a solution - but there was the Scriptural precedent of the Patriarchs having two (or more) wives.
Philip seems to have been really keen on this as a solution, for what it's worth, because he intensely disliked his first wife, was constantly committing adultery, wanted to regularise the whole position so he could eat his cake and have it, and even had the second wife picked out ready to go before the theologians got a start on the question: https://en.wikipedia.org/wiki/Philip_I,_Landgrave_of_Hesse#Bigamous_marriage
On the one hand, it's good that Substack is willing to put in a lot of work to change to UI to fit what Scott wants. On the other hand, it seems concerning that they weren't already doing all these things. My impression is that most commenters' complaints about the UI are glaringly obvious, and any competent designer could have fixed them without needing to be told. Or, perhaps more accurately, they could have fixed them if they had been prioritizing user-friendliness rather than engagement / revenue. For example, it seems clear to me that having big "Subscribe" buttons everywhere is bad for users (but good for revenue), so a designer who prioritized user-friendliness wouldn't do it, and wouldn't need to be told to do it.
My impression is that, of the sites that (still) have commenting, very few of them host the types of conversation that were typical at SSC. Mostly I see thanks and feedback, with occasional author responses, and a chronic problem with people getting into toxic discussions with each other. Given that, I'm not surprised that a site like Substack opted to go for a simple but functional commenting interface that would work well for the vast majority of users. We're weird.
As for the subscribe buttons, if they help generate revenue to keep Substack viable, I don't mind. Of all the causes of failure of sites like Substack, I'd suspect "lack of revenue" is one of the biggest, and "too many Subscribe buttons" is one of the smallest.
Well I think substack was attempting to be more of an email newsletter platform than a typical blogging platform, but it seems they've more and more had to face up with the fact that a lot of the authors want features that are more analogous to a blog than a newsletter. In the end they're probably going to mostly become a WordPress clone with very basic features for most authors, but custom features for the authors that bring them a lot of revenue.
I'm not surprised to see what you said about eugenics here, as I've gleaned over the years that your stance is roughly that. But I hadn't seen you crystallize your position like that before. Do you expect to ever devote a whole post to eugenics?
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.
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.
Has anyone tried deradicalizing conspiracy theorists through betting? I'm thinking specifically of political conspiracy theorists, as they seem to have more testable predictions - eg X country will do Y action by Z date, X person will win Y election, X person will be in jail by Y date, etc, but anything conspiracy theory will do. It seems like losing money, especially repeatedly, would make the person have much stronger incentives to update their beliefs than they otherwise would have. The other upside is instead of being annoyed with your conspiracy theorist friends/families, you'd have an incentive to communicate with to them, so you can make money. Maintaining ties to people with sane worldviews is important for this sort of thing.
What do people think of the pros and cons of an easily recognizable ACX symbol? Something that readers of the blog could use to identify each other in public, that would be commercialized either by Scott, or open for anyone to do so. I ask because a means to easily identify fellow readers in public would make it easier to find like-minded people, but I can of course see that it might bring risks given the ambiguous reputation we hold in some circles. It's worth it to me, but I'm curious if the cost/benefit analysis is equally favorable in the eyes of others.
Regarding sign-ups for Scott's new psych practice: I suspect what many of those readers would like (as a second-best solution anyway) is a referral to a different physchiatrist who is taking new patients and who has similar attitudes and views about the nature of the doctor-patient relationship, meds, research, etc. to Scott's. If Scott were feeling quite charitable, he might create a separate form where people can request a referral. I suspect that would siphon off most of the people currently ignoring the instructions not to try to sign up as a new patient of his. (Sorry Scott: I have never commented here before and don't know whether to address this to you or to the group!)
My blog, Naval Gazing, originated in the SSC OTs, and before the shutdown I usually put up links to new posts in the whole-number OTs. There appears to be at least some interest in me continuing to do so, so I should probably start with an overview of what I've been up to.
First, I finished the long-running Falklands series, which ended up at 24 parts:
@scott, >I support it in the sense of improving people's genetics and genetic outcomes - long-term through genetic engineering, medium-term through having things like the Nobel sperm bank (but less badly done) available for people who want them, and short-term through voluntary community-based efforts like Dor Yeshorim.
> dictionary, definition. eugenics. the study of how to arrange reproduction within a human population to increase the occurrence of heritable characteristics regarded as desirable. Developed largely by Sir Francis Galton as a method of improving the human race, eugenics was increasingly discredited as unscientific and racially biased during the 20th century, especially after the adoption of its doctrines by the Nazis in order to justify their treatment of Jews, disabled people, and other minority groups.
I'd be keen to know, what exactly sets you apart from the above definition, I do get it, you are not Nazi, and you are not a racist either, but how exactly your opinion disagrees with the rest of the dictionary definition of the eugenics.
The word eugenics, at least the way it is commonly used, implies a level of coercion/lack of consent that isn't present in the dictionary definition of the word, but is nevertheless what comes to mind when most people use it. The word also carries a baggage of being used by horrible people to justify some really miserable actions (and being used by other people to put those people in a bad light).
It's entirely possible to have a preference for improving the human gene pool without endorsing anything remotely resembling what people normally think of when they use the word eugenics.
Interesting question. However, I sense the topic veers towards CW territory (at least for me..) so I'd suggest saving further discussion for a more appropriate thread.
This is probably far too late to get any significant response, but I'll write out the post anyway to make rewriting it next OT easier.
I've been studying at a German technical university for a few years, and have been steadily confronting myself with the question: why wouldn't more American students follow a similar path?
I have to say, my confusion here has been pretty large, and I can't tell if it's due to:
-A unique situation that made it easy for me compared to the average American student.
-An American education (recommendation) system with a massive blind spot.
-A German visa system with a lack of interest in Americans.
The long and short of the steps that I took to get into this position is:
1. Learned German during an exchange year (not necessary, my university as well as perhaps 200 more German universities have Bachelors as well as Masters in English)
2. Applied for German universities (with a decent GPA/ACT/SAT [specific numbers on request], not much work is required. In fact, I found the 4 American universities I applied to much more difficult than the 10 or so German unis)
3. Was accepted to German universities.
4. Demonstrated a relatively reasonable level of financial stability (8,000€ in a bank account or a German family willing to vouch for you.)
5. Began attending the German university of my choice.
Current monthly costs, living in one of the 5 largest German cities, include:
*300-500€ (for me, just over 300€) rent in a shared flat, also studio apts (WiFi, electricity, heating, and water included).
*100€ food, relatively decadent with fresh vegetables, fruits, cheeses, and occasional meat.
*100€ stunningly inclusive health insurance.
*(You may be wondering, where's the university costs?)
*50€ University costs (this includes texts, computer labs, multiple student workshops, legal insurance, email, cloud service, and most of all, public transportation in the entire state + parts of other states)
*Around 10-50€ for a cell-phone plan. Figure this isn't too different from the states.
So if we're looking for the high-end of what I could pay, I come out with 800€ and 9600€ per month and per year, respectively. Throw one or two international flights home in there, for an extra 500-1500€, and I bump into 11k€. For my entire costs, from year one to year eight if I wanted to spend more time here.
Now, my comparisons to the US are based on anecdotes from friends, as well as a few websites with average cost estimations. Take this with a few grains of salt, but I seem to find $20k a year as a pretty common amount for in-state tuition and associated living costs, and $40k a year with a medium out-of-state tuition (plus living). This is either 100% more than what I'm paying, or 300% more. Not only that, the bachelor at my university (as well as all German universities) is a 3 year program, since GERs (general education requirements? Don't know my American university terms very well) aren't a part of the curriculum. So to get my bachelor, I'm spending 33k€ (to be fair, this is about $40k right now), whereas in-state public is spending $80k and out of state is spending $160k plus an extra year?
I'll assume most of you know someone, either around college age, raising children who will reach college age, or you're planning on having college graduates of your own some day. Reach out to them, let them know it's an option, and maybe save a year of their life and $100,000.
Or critique my comment and let me know exactly how I'm wrong. I welcome both.
Ningen Dock is something I'd never heard of, thanks for mentioning it. I've heard Mexico/Central America/Southeast Asia as alternatives for large surgeries in patients, even amongst some friends of the family.
I think it may fall prey to a similar problem as with universities, those patients/students who would most be in need of these alternatives (no insurance, no scholarships) probably tend to be those least aware of the options.
Definitely true that a lot of the value of education is social class and networking, but seeing as our society seems to be becoming more global (2020 maybe excepted), I see networking in the European Union as still beneficial for later opportunities. The social class seems to be similar in both places, whether I'm European college-educated or American college-educated, I'm still now a part of the college-educated class.
I appreciate the take, which dying niche would you say that Musk explored? The 'creating change in physical space' niche? Also curious as to which AI talk came to a point where 'sheds are for losers' was an opinion.
I find it hard to disagree on the point that the middle class norms hard. At the same time, I feel that 'study in another country for free*' could very well be an adoptable norm. Especially because it seems to be a norm for the middle/upper class in Asian/African/South American countries. More than 30% of my university are foreign students, almost entirely from those 3 continents (and Europe, but that one seems more of a gimme). North America doesn't even begin to appear at my university, I'm sure out of the 3000 foreign students I could count the Americans on a single hand.
I can recommend it! The higher you go up the education ladder, the easier it is as an English speaker. My University currently has more English Masters than German Masters, compared to a single English Bachelors in a field of about 20.
On the occasional meat, Grillkäse (grillable cheese) is a common BBQ replacement which is pretty common. Most supermarkets have anywhere between 5 and 20 different meat substitutes, using soy, gluten, milk proteins, etc. that are relatively close to meat in cost.
Did you go to Germany from a country already fairly near to Germany? How'd you find out about this possibility in the first place. I'm well past college age at this point, but when I was, I didn't know such an arrangement was even possible. American students aren't going to choose a path they don't even know is available to them. We have guidance counselors in high schools that steer you toward specific paths, and universities also directly advertise and recruit people.
Heck, when I was a teenager, I actually tried to conduct a fairly exhaustive search compared to just choosing the best known local schooll. Princeton published a guide to the best 308 universities or some number close to that, and I sought out additional information on a list of 20 selected from that book, and ranked them in a weighted basis along all of the dimensions I thought I cared about. Even going into that much detail, the possibility of a foreign university never even occurred to me. I also made a terrible initial decision in spite of trying that hard and transferred out after one semester.
My home country, and where I still have citizenship, is the USA. As to what I meant by relatively unique circumstances, I was the recipient of a US-Germany exchange program scholarship in high school. This was what originally made me aware of the possibility since our guidance counselor never had 'foreign university' as an option.
The best-laid plans can definitely go awry, I hope that the next university fit better!
Just a quick vote for "please no favourites". It'll just spoil the feeling of balance and make more standardised beliefs or ideas more visible while hiding the more interesting ones as people skip to the "Top Comments!!". It will create an echo-chamber-like atmosphere and kill discussion (like it did to metafilter.com) where those with differing opinions are accused of "not reading the room" - and thereafter silenced/shouted down by the majority.
It would be cool to be able to link posts without the comments, like was possible on the old blog. This is useful to send posts to people and have them be less scared about the size of the scroll bar.
I'm curious why you care about the read:unread ratio. Is it just to avoid having shelves and shelves of unread books, or do you actually want to conserve a certain number of unread books on your shelves?
Ways that I keep my library in check:
1. Avoiding charity shops unless I have a non-book reason to go in, because I almost always leave with one or more books since they're so cheap second-hand - Easier now they're all closed due to Covid!
2. Donating books I don't plan on rereading to charity shops/family/friends - harder now due to Covid!
3. Buying most books I want new on Kindle, which doesn't take up physical space - of course, this means my digital library has... 175 titles? Most of those are free Gutenberg editions of classic texts though, and a lot of those are unread.
There is something wonderfully decadent about the personal library. It calls to mind A Series o Unfortunate Events, where, each time the siblings moved, the personal library of their new caretaker was the symbol that they had arrived home. The word alone is sufficient to conjure up images of homeliness, coziness, comfort, and knowledge. It's old-fashioned, but if I had a permanent residence there would definitely be a reading room with some comfortable chairs, a lamp with a green glass lampshade, and wall-to-wall bookshelves.
Thinking about the fantasy of the personal library as safe space, it comes to mind that although I assume a certain feminine nature to the fetishization of libraries, Beauty and the Beast style, libraries are pretty agender. They combine both domesticity and academia, a kind of nurturing growth and the professionalism of a lawyer or doctor. Class-wise, the idea of a personal library strikes me as an 'old money' kind of thing, but perhaps it universalizes to educated people in general. I bet someone has made a decent term paper about the sociology or psychology of libraries - and if they haven't, someone sure could.
I know what you mean about the difficulty of keeping too many books, I left most of my childhood collection to my younger brothers and have only recently started filling a bookshelf of my own. I'm not sure if I'd ever dedicate an entire room to a library, I'd be much more likely to use my entire house as a library, assigning different books to different rooms for decoration and categorisation.
I currently have a rather eclectic collection, ranging from Dante's Inferno to the 99% Invisible City, but since most of it was bought from charity shops there's a lot of thriller/mysteries that I'll realistically only read once.
It's always worth having unread books though, my internet went down at the start of lockdown and I'd never have made it through the next few days without plenty of books to keep me company!
My strategy is to buy special editions where possible, it turns book shopping into a fun adventure, makes sure I don't buy too frequently, and also, I have really cool copies of a lot of books! I mean signed copies, really old copies, or any collections or anything that seem fun.
If anyone has recommendations on how I would go about getting Discworld in _any_ form, I'd love to hear it.
There is an illustrated version of the last hero which is absolutely beautiful
I've managed to read most of the Discworld series in a random order by borrowing it from libraries. Not the most efficient way to do it, but definitely the cheapest and most interesting.
A quick Googling suggests that eBay would be a good way to buy most of the books secondhand, which would be cheaper than buying them new:
https://www.discworldemporium.com/the-discworld-collector-s-library/441-discworld-collector-s-library-the-complete-collection
Obviously there's postage fees, but for that price I'd hope they'd send it to me in a replica of the Luggage, legs and all.
It's kind of sad how that site still describes their offering as "The Complete Collection... So Far!" but I guess the original author being dead doesn't always stop them from releasing additional work.
Jesus, that's a lot! I might save up to buy these all one day, but I'll more likely try to get used copies from little bookshops so that they all have more individual value to me.
Just finished watching the Amazon Prime adaptation of "Good Omens" and I was very sad that Pratchett was missing, it really felt like his touch was missing from a lot of things.
I believe that the "... So Far" refers to the fact that not all the Discworld titles have been released in collector's editions.
My strategy is to download the Kindle preview as a reminder that I wanted to read this book, and not to buy the book until I get to the end of the preview. But it doesn't work for books that haven't been released yet.
I do something similar with video games, but follow the inverse ratio. (I can "earn" a new game by getting 100% completion in 2 old games.) Your strategy makes sense for books, since you're increasing the visual impressiveness of your library while maintaining the same risk of embarrassment if someone grills you on a book you haven't read. Obviously, that doesn't apply to games at all.
Oooh, I like this! Reminds me of https://twitter.com/CaffeinatedCov/status/1278358375487680513?s=20 😂
I have a similar problem but a different approach.
Any 'shiny object' book I might want to read at some point goes on an Amazon wish list, and I'll occasionally check which ones are deeply discounted. I managed to accumulate ~200 unread Kindle books, many of which are anthologies or collections, and now I'm trying to bring that down to something more reasonable. I'll still buy books, but only if I plan to read them in the next month or if they're free, with a monthly cap of around $10 to spend.
This way I can still finish long series (Dune, most recently) where the later books are full price, but I also have incentive to finish books I already own and read more classics that are out of copyright.
Current known concerns about Substack comments (if you have more concerns, please post them as replies to this comment, one concern per reply; if you agree with someone else's concern, heart the comment):
1. Need to press "load more comments" to load more comments and "1 more reply" to see reply
2. Hearts need to be removed. SSC commentariat previously voted that likes/upvotes distort commenting behavior too much and they don't want them.
2a. People are concerned that you can't subscribe to comment replies without getting an email every time someone hearts it, but removing hearts should solve this.
3. Cannot currently edit comments.
4. Cannot currently report comments.
5. Cannot currently see which comments are new since last visit (as in old SSC where new comments would have a green square around them).
6. Cannot currently comment without it reloading the page, no easy way to get back to your old location.
7. Cannot currently without getting popup telling you to subscribe
8. Cannot currently link to some comments (link just goes to top of page)
Would be great in substack if there was a way to surface only comments that either the author made, or that the author replied to.
> 2a. People are concerned that you can't subscribe to comment replies without getting an email every time someone hearts it, but removing hearts should solve this.
Unfortunately the "Send emails for Likes and Replies to my comments" setting is site-wide, not per-Substack, so that does not solve the problem: if someone comments on any other Substacks, and those do have hearts, they can't have this setting on even though it would work fine for their comments here.
For what it's worth, I just made a gmail filter that deletes any email from substack with the text "liked your comment". I guess notifications about this comment might get caught up in it since it has that text, but I figure it should pretty well handle me getting reply notifications and not like notifications. And I do really like the reply notifications, please keep that!
I actually really like the lack of reply notifications from SSC. I believe it produces higher quality discussion.
For example, in reddit I get an inbox of replies, listed without context, and the option to quickly respond. This encourages addictive behavior (in the usual dopamine-optimization ways), and causes me to reflexively defend myself against any response that is (perceived as) even slightly critical. I end up talking to the one person I'm in a back-and-forth with, not pausing to see if my comments are in any way useful to the broader thread (which I literally can't see while responding). By contrast, on SSC I had to consciously look for responses, thus ensuring I only comment when I'm actually invested in the idea. I had the opportunity to see what had been said, and often someone else had made my point better already. I took the time to actually wonder what my response was adding.
This analysis might be specific to my personal psychology (I find reddit replies to be one of my worst addictions, it's really quite bad), but that's at least one argument against reply notifications
Are you using RES for Reddit? I always click on the "Context" link for comment reply notifications in my inbox because replying as you describe is horrible for all of the reasons you mention.
I strongly disagree. It often involves a lot of back-and-forth replies to fully clarify our positions, and understand that of the other person. The latency between replies is one of the reasons I often find it harder to fully understand each other in a comment section than in a spoken conversation (others are that it's slower to type than to speak, and that we can't interrupt each other to ask for clarification). Reply notifications make this somewhat better.
A back-and-forth with one person shouldn't be a major problem in a threaded comment section (especially if it were possible to collapse threads).
(Btw there actually was a third-party reply notification system on SSC, though I didn't find out about it until more than a year into commenting on SSC.)
Reply and like notifications come from different addresses, so you can set up a more reliable filter than that. Reply notifications come from forum@mg1.substack.com while like notifications come from reactions@mg1.substack.com.
Sorry, it's actually reaction@mg1.substack.com (singular, not plural).
I would prefer it if, instead of sending an email when someone replies to my comment, it notifies me by changing the colour of a mail icon at the top of the page (the same as Reddit does). I would like substack to add this feature -- maybe make it optional on a per-user basis.
I very much liked the hearts feature change, as it enables me to find "better" comments when I don't have the time to read the whole comment section. Can someone tell me why and how voting is distortionary ?
Similar to the “like” button. People are conditioned to write comments to get likes/hearts rather than to say what they want to say regardless of the popularity.
https://www.theguardian.com/technology/2017/oct/05/smartphone-addiction-silicon-valley-dystopia
Allowing voting/hearting but having scores be hidden (but still influence sorting) seems like a decent approach.
That sounds like a good compromise.
Wouldn't that make the ranking of your comment be the metric you're chasing? I can't see how many hearts I got, but I can see I got more than X person I care about doing better than.
As someone wrote in a comment below, sorting by chronological order introduces it's own distortions to the comments. I don't really know how to evaluate what's the better option.
I think ranking is less likely to be something people will chase. It feels more hidden/implicit, and any dopamine hits would be further apart.
You could also, e.g., introduce some randomness into the sorting to disassociate it from social approval.
There already appears to be some randomness in the sorting (or at least, top first doesn't seem to just sort all comments in descending order of number of hearts).
I could live with that. I prefer it being open, but closed is better than nonexistent.
The fact I can heart on your comment to agree with it is an example of the usefulness of hearts.
Thanks, I guess I get that, though I don't know if I consider that drawback sufficient for getting rid of the feature as a whole.
Might be worth thinking about better ways of implementing some kind of rating system (maybe as likes only showing some time after the comment was posted, so as not to enable some instant gratification effect, and to prevent early comments from getting a head-start).
LessWrong has the upvote/downvote feature, which I'm sure has all sorts of downsides too, but at least beats the hearts feature in e.g. that it doesn't reward too controversial comments
I feel like upvote/downvote rewards controversial comments more than likes, because people often take their downvotes as a badge of honor. (I know I certainly do when I say something sensible in the wrong subreddit and get it downvoted to -20 or whatever.)
It might be a badge of honor for someone who dislikes the community, but you still go to the bottom of the thread.
Ars Technica is even worse - downvoted comments are collapsed for subscribers and invisible to non-subscribers. I don't think they fully brigade yet, but increasingly the echo chamber is mindlessly defended.
True, but in my opinion having an imperfect classifier is still better than no classifier at all. I don't have enough time to read every single comment, so I find the heart-counts to be helpful even if there are some false positives and false negatives as far as identifying insightful comments.
(1) I can heart my own comments.
(2) The comments that you want to heart may be orthogonal to the comments you want to find again.
(3) If you mean you want to find the comments that *others* heart, then what is your hearting threshold to save such comments?
The point is that hearting is a crude way of implementing better solutions to your desires. And I happen to share those desires.
Re. 3, what do you mean?
Maybe I don't have a threshold, and just want to view comments in descending order of hearts. Or, say I do have some threshold, say it's 10, then what?
Exactly, the fact that we don't even agree on what (3) means indicates that the hearting system as-is is too crude for our commenting needs.
No it doesn't, just that it's more crude than "perfectly matches everyone's desires", but that's an unrealistic goal.
People who all have slightly different ideal like-ranking algorithms might still all prefer any kind of sane like-ranking to just chronological.
I almost think “likes” with no “dislikes” are even more distortionary. On Reddit I will often “counter vote” posts that seem like they are getting too many upvotes or downvotes because they are red meat to one side of the commentariat. Can’t do that here - posts that elicit a strong positive reaction will get the reflexive “like”, optimizing for (one-sided) heat rather than light.
Having both upvotes and downvotes is even more prone to circlejerks. With upvotes only, at least it's likely that high-quality comments expressing a variety of views will be highly upvoted. With both upvotes and downvotes, even high-quality comments expressing minority views on controversial issues are likely to be downvoted on the net, unless people are *very* disciplined about not downvoting comments that are well-argued just because they disagree.
Solution: decouple position and quality, as I suggested elsewhere (don't want to double-post).
That's only a solution if people don't actually judge the likely quality of a comment by its votes, and commenters don't optimize for votes. But then what's the point of votes?
Actually, if a forum were to have both upvotes and downvotes, it may be less bad if they are displayed separately, rather than displaying the net score. A comment that has -4 points may have 30 upvotes and 34 downvotes, or it may have no upvote and 4 downvotes. These indicate very different things, and this information is lost when only the net score is displayed. One is likely to be an important argument for a controversial position, the other a poor-quality comment that few people have read.
Presence or absence of like or heart reacts influences behavior of the commentariat. "Chasing likes" is a thing.
Now, this is not necessarily a bad thing. But virtually all relevant discussion boards on the web have like buttons or their equivalent. In the interest of diversity, discussion spaces without such features should exist. SSC was just that space and imho it worked great.
otoh, i would often wonder how good arguments were, and without likes I could only assume 50/50. If I find one side of an argument utterly ridiculous, and the likes confirm that, I can tilt my view one way. If the likes don't confirm that, it informs me that maybe *I* am the one misunderstanding something.
With all due respect, I happen to think that this is not an optimal way to form opinions.
yeah, and i would tend to agree.. but I also often found myself going insane at some of the arguments at ssc. and while it is not a good way to form opinions, it is useful information to know how much support an opinion has in the population.
i don't know if 'likes' is a great solution, and it may bring more problems than benefits, but there is at least some tradeoff to consider.
Sure, I agree that there is a tradeoff. And sorry for my somewhat unkind reply above.
I disagree.
The optimal way to form opinions is to be right about everything.
But since an individual can't easily implement that, it's usually relevant to take into account the opinion of others, particularly others that one expects to be well-informed, and that one knows something about their reasoning processes.
I think in a lot of subreddits, upvotes or downvotes are a really poor proxy for anything, because they just reflect the biases of that community. But a community like the one here is likely to be somewhat better, particularly depending on the type of comment one is looking at. (It's bad for ones that depend on empirical knowledge, probably much less bad for ones that depend on conceptual analysis.)
I for one would *love* to know what the biases of this community are.
> i would often wonder how good arguments were, and without likes I could only assume 50/50.
With likes, you still do not know, because people commonly upvote comments they agree with, rather than comments that are well written.
This is why likes create echo chambers, because good heterodox writers typically get pissed off at getting few likes (and thus few readers) and leave, while poor orthodox writers who are first to make a popular argument, get a lot of likes, so they get encouraged.
I think we all agree and accept there are pitfalls with the like button. But I do not agree that there is necessarily zero information in the likes of a community like ssc. Given the heterodox tilt of ssc, I am not even convinced that our commentariat wouldn't give more likes to a 'good heterodox' writer versus a 'poor orthodox' writer.
I understand the risk of making an echo chamber, but they also would serve as a reality check to the heterodox ideas; is this an idea with merit, or is it one of those "devil's advocate, heterodox for the sake of being contrarian, but doesn't really hold up" ideas?
Incentivising good content, while imperfectly, seems like a good thing, in particular as people's time is limited it gives people the ability to find the best stuff. If people prefer not to browse that way they can always switch to chronological
Imho "gets a lot of likes" and "good content" are not remotely identical categories.
E.g. short controversial political statements might get tons od likes from the ingroup, while longform thoughful and important posts on, say, ways to improve vaccine production, might languish, um, "unloved", since few people understand technical problems involved.
This might be a reason to have hearts on the non-politics threads, but none on the politics-allowed threads.
But this evidence is all from places that are not ssc?
But SSC might be better (at least in part) because it _hasn't_ had 'likes'.
Of course they're not identical. But I know of no better way to filter good content to the top than by a like system. Every other approach is far worse.
This is true. Forums with "likes" or "hearts" tend to become echo-chambers, over time -- unless they're very large, and then they fracture into a number of smaller echo-chambers, e.g. Twitter.
"Hearts" are totally inappropriate for a forum where unpopular opinions, and adversarial arguments, may on occasion be put forth. The lack of hearts leads to better discussion, more discussion, and freer thought.
"More hearts" doesn't necessarily mean "better" comments. Sometimes (and it's a big temptation that I've fallen for myself) you'll clicky that heart because it's an automatic "yay my side!" rather than "your reasoned argument has won my conviction". Sometimes you'll clicky that heart because Joe and Bill are tussling in the comment thread and you'll backing Bill against Joe (who is just plain flat out *wrong*). They do distort reactions and I'd prefer if they're gone.
It might work to have hearts and brains, so that people can indicate both what they like and what they think are good arguments or valuable added information.
I propose a dummy heart button that doesn't do anything (like some close-door buttons on elevators) and a second "this is a well-reasoned argument button" that does the actual sorting.
I'm only partially joking. I think a problem with simply imploring people to upvote for good reasons is that sometimes you have an emotion and just want to find some way to express it. A placebo button seems an interesting way to combat that impulse.
Nice idea! I clicked the placebo heart and also typed a comment endorsing it!
Adjective tags in slashdot were once a useful idea. The "funny" crowd could read the comments they like while the "thought provoking" crowd can pick out the ones they like.
Tags don't have to be pro/con, they could be more descriptive.
I agree, I've liked the slashdot system best out of all the systems I've seen in the last 25 years.
That's probably too big of an ask for a Substack patch, but I like the idea. I was never much of a slashdotter, but the sorting system did make sense.
This sounds like a good idea to experiment with.
The heart button can allow us to appease our limbic system while the second "well-reasoned argument button" can allow us to flex our forebrain.
I would even go so far as to say that the latter button should be extended to good will arguments, even if they were weak arguments and were subsequently bested.
Ooh, I'm really intrigued by this idea!
Problem is, ANY kind of rating system would cause commenters to optimize for rating (not necessarily consciously), and since any rater is just as good as the next one, it would optimize for the lowest common denominator, which, while arguably higher here than in most other internet fora, is still the lowest.
I'm not sure why one would think it's specifically the *lowest* common denominator. I would assume it would be the *most frequent* common denominator. Which isn't necessarily lowest, depending on the topic and the community.
Mathematically, the lowest common denominator is the most frequent common denominator.
Voting systems aren't perfect, but they are better than a system which rewards people who post the fastest and in greatest volume, which is defacto what a chronological system becomes. And is very easily dominated by a small number of people with fringe views who post a lot.
I'm definitely a big fan of having some way to sort the comments other than chronologically. I want to see a rundown of the most interesting comments by some kind of metric!
At least from my own experience, the old SSC system incentivised comments that resulted in more replies to them, especially replies that resulted in even more replies, and that was great, since it optimized for polite and engaging discussion, while a system with likes would optimize for likes.
"A system with likes would optimize for likes."
This is something that the internet has learned over the past 10 years, very much to everybody's detriment. "Likes" distort communities and make discussions more superficial and facile.
Some specific counter-examples I can think of are small 'communities', _with_ 'likes', but with comments (by default) sorted chronologically.
[citation needed]
The least toxic of the big social media sites, by far, is Reddit. Which is the only one which not only allows upvotes, but also has downvotes, and uses both of them heavily in their ranking algorithm. This is not a coincidence.
Often though, it optimized for being contrarian. A bulletproof argument will get someone to agree, but there is no reason to reply to it... except if you are willing to make a flawed counter argument.
Those deep reply chains were maybe half good discussion, but half were one of the posters just ignoring someone's key claim.
If placing value on more replies is a way to incentivize polite discussion, then you're already dealing with a community so unusual that no Guardian link is going to accurately describe our social dynamics.
Note that comments incentivizing more replies are known as "hate-clicks" in media, and "getting ratioed" on Twitter. This is how normies act when incentivized to get more replies. At best, it'll be scissor statements. If we're like that, I want to incentivize fewer replies. And while we're weird, I don't think we're that weird - either we're ignoring the incentive, or we're responding to it in toxic ways. In the first case, hearts are harmless. In the second case, they are vastly superior, because optimizing for them can't be nearly so toxic.
right, in the old SSC, more often the only response you got to a comment was 'here is why you are wrong'. there was no way to calibrate my thought process.
Well, there's the "good, necessary or kind" rule against that. So either that rule, or we are that weird, or a little bit of both, but it resulted in the old SSC having possibly the best comment section I've ever encountered - without any noticeable external incentive other than getting replies (and, ok, not getting banned by Scott).
I think the “small number of people with fringe views who post a lot” are also the ones most likely to spam the like button on posts they agree with, so it might not help much.
I understand the downsides of likes that others have mentioned, but I want to second the sentiment that likes are useful to quickly see the highest-quality comments. Suppose you've read the post and want to spend 5 minutes reading the comments. If the comments are not ranked, you won't likely read the most interesting comments.
One compromise could be to use hearts for "soft" ranking, i.e. surface some newer comments above liked ones. From what it looks, Substack might already be doing something like that.
I think that's one of the problems:short term people love upvote mechanics. But it tends to have long term downsides on the average quality of forums.
The comparison that comes to mind is the egg weight thing: people love big eggs, select chickens for big eggs and you get bigger eggs... until after a while you get chickens that fight constantly.
Because you're also selecting for other things as well.
Select comments a little and you make it easier to find good comments... but a little while later you have Facebook-quality comments and everyones fighting
what would you say the long term trend of SSC comments was? I think many thought it had a negative slope
Hearts/likes/upvotes favor earlier comments more than better comments in my experience. An expert can comment 12 hours later with a high quality comment and just get buried at the bottom due to the first comments already having been substantially raised via a low quality but popular/agreeable comment. (...said the man 23 hours after the original post...)
I was one of the first on this post and agree with you. :)
It does value both, and an earlier comment of similar quality will beat a later one. But later comments that are substantially better usually rise to the top pretty quickly.
FWIW I agree that voting is distortionary, but IMO it's the lesser of two evils—I very much like being able to quickly find the well-liked comments, and I'm willing to accept some amount of distortion to get that.
'off-stack': I've only just joined and got my 1st mail thread-notification. Could you color the link-title please, so I remember where to click? Took me a little while just now.. and yes, I'm stupid.
Thx
I don't know what this means. Are the links not showing up as a different color than the rest of the text for you?
The links in the text of the post are a different color, but the link that goes to this article is just the title of the post. Since the tile is the same color as the normal text, it's not obvious it's a link.
Thank you - I was just about to reply the same.
I think this is a Substack issue and not anything I can solve; if any other Substacker knows otherwise, let me know.
Good: When someone replies to my comment I get an email with a link to the reply
Bad: The link doesn't take me straight there. I need to click "show more comments" an unknown number of times. Eventually, it will jump to the reply.
Ah, I didn't know that clicking "show more comments" would eventually jump me there! I just assumed that if it was behind that link it was humanly un-findable.
At least for me, it doesn't (OS X Chrome). I have a bookmarklet to work around this -- executing "document.location.href = document.location.href" will cause the page to jump to the anchor in the URL. So I just click my bookmarket after hitting 'load more' a hundred times or whatever.
It would be nice if every line prefixed with ">" would be turned into a quote block, line in Markdown, email, etc.
It would be nice to have markdown in general, or some alternative.
A way to hide comment threads would be really nice. Bonus if it remembers that choice through reloads. And a way to quickly navigate (like old SSC having a "parent" link) would be helpful.
Yes please. The current setup makes scrolling through long comment threads (like this one) really cumbersome.
Oh gosh yes please dump the hearts, they really confuse me as they make me go "Twitter? Not Twitter? Where am I?". Some way to edit and format comments would be great also.
Can't use code (eg italics, links) in comments.
Terms related to this feature request that people may search these comments for: markup, formatting, HTML, Markdown.
Formatting types that I remember the previous WordPress blog supporting:
• Italics (<i>, <em>)
• Bold (<b>, <strong>)
• Strikethrough (<strike>)
• Inline links (<a>)
• Blockquotes (<blockquote>)
• Lists (<ul>, <ol>, <li>)
• Headings (<h1> through <h6>)
• Inline code and blocks of code (<code>, <pre>)
• Some other obscure, harmless HTML tags (<kbd>, <tt>, etc.)
Notable types of formatting that I don’t remember seeing and was probably not supported:
• Images
• Tables
• Text colors
This is a biggy (for me at least). Substack, please add Markdown support. Also, enable highlighting text in a yellow background -- as if it was highlighted with a yellow marker pen -- using the <mark> tag.
I'd like to be able to collapse a given comment thread.
Found out by accident that if you click on the gray line in the left margin of a comment thread it collapses it. But the view doesn't seem to be moved accordingly: instead of, say, scrolling you to where the collapsed comment is, it looks like it just keeps you X many screens down from the top, which may be considerably further south from where you were.
Thank you, that solves my issue. I can deal with scrolling a little.
I also found this by accident - ideally collapsing behavior is more discoverable
I would also like to be able to collapse the top-level comment. Clicking on the gray line collapses replies but not the original comment.
I would like to be able to "click" on the gray lines using Vimium (https://chrome.google.com/webstore/detail/vimium/dbepggeogbaibhgnhhndojpepiihcmeb?hl=en). Currently they aren't being recognized as links when you hit 'f'.
Also, in the process of writing this comment I realize Substack doesn't currently support markdown. I see that user ersatz has already raised this issue.
Also, I see that I can "love" my own comment. I think this is undesirable, unless all comments were "loved" by default as LessWrong does.
Yes, and if you are on a touch screen (like a tablet), collapsing is FAR to easy to do by accident). The mobile interface is seems nigh useless for comments, as any length comment is nearly unreadable. Not exactly sure why, but I think it is something about the text sizing being too big just makes it very hard to process.
I can't seem to find a place to disable email notifications for comment replies.
I do prefer the hearts, they provide valuable information that is sometimes, though not always useful.
I can't seem to find a way to disable emails for comment replies.
I found it when I went to pure substack.com first and then to my account settings.
Text search in comments would be nice (unless this is just my safari/iPhone learning curve)
Scott, item 7 in your starter list is missing a word or two.
You'd have to load all of the comments you want to search first (in batches of whatever size Substack uses), but then you should be able to use something like 'Find in page' in whatever browser you're using.
I am strongly in favor of a voting system. Despite having followed your blogs since the early lesswrong days I almost never went into the comment section on SSC because it was always so long as to be unreadable, and the lack of a voting feature meant that discussions were easily dominated by a small number of people with fringe views who posted a lot. (I'll avoid specifics so as not to derail, but I'm sure anyone who was on it frequently can think of a few examples.) Voting systems are not perfect but they are better than a system that rewards whoever posts fastest and in greatest volume, which is what a chronological system defacto is.
"Voting systems are not perfect but they are better than a system that rewards whoever posts fastest and in greatest volume, which is what a chronological system defacto is."
I agree this was a problem with the SSC commentariat. But instead of intruducing a voting system, I suggest tackling the problem directly by limiting number of comments one user might post over a time period.
That is a vastly worse solution. None of them were spammers - they were honestly engaging. Hard-blocking that is a major overreaction. Far better to let them keep talking, but let it fall to the bottom, where only interested parties will see it.
Yes, the comments were often very long. But I found it relatively easy to navigate (but I did also spend hours sometimes reading one post's comments):
1. Collapse comment 'sub-trees' that seem to be uninteresting – that can prune a LOT of overall comments.
2. Ignore comments from commenters that I found to be consistently uninteresting.
One nice thing tho is that there are many 'venues' to discuss a post, not just directly 'in the comments' (on the main site). That's probably necessary to some extent if the number of commenters continues to increase.
With upvotes, you mostly see the most popular opinion on any given issue (which may or may not be correct). If you mostly see comments by the most prolific commenters, at least it's likely that a variety of views have prolific exponents.
Re the voting system. It seems like even if you swap to "top first" it doesn't fully sort all the comments in order of votes. Not sure if thats a bug or its using something like the reddit "best" algorithm that tries to normalize the for how long a post has been up, to reduce the benefits of being teh first to reply to an article
I've been getting into photography lately and I like sidequests so if anyone wants a high resolution, debatably high quality photo of anything within a reasonable radius of Madison, Wisconsin, @ me with your request.
Have you looked in to articles on Wikipedia in Wisconsin where people have requested photos?
https://en.wikipedia.org/wiki/Category:Wikipedia_requested_photographs_in_Wisconsin
I also sometimes just use special:nearby feature and look for articles missing display images and take photos of those.
https://en.m.wikipedia.org/wiki/Special:Nearby
I have not, thanks!
A similar tool: https://tools.wmflabs.org/wikishootme
Thanks, that is a great suggestion, I have been doing the same myself.
An optional but highly rewarding sidequest to natural light photography is... lit photography! Artificially lit photography? Photographing with strobes! :D
I highly recommend the excellent and free Strobist blog by David Hobby and it’s “Lighting 101” series: https://strobist.blogspot.com/2006/02/welcome-to-strobist.html?m=1
Also, this book on photographic lighting is one of my favorite books ever (not just photography books):
https://www.amazon.com/Light-Science-Magic-Introduction-Photographic/dp/0240812255
(Linked to an older cheaper edition. The price on newer editions keeps going up!)
Are you able to photograph the Aurora Borealis near Madison? I'd love to see a good photo of this.
It allegedly appears this far south but I've never seen it personally, but when it's next supposed to be down here I will probably give it a try unless it's cloudy
That's great! Do you have any online gallery of your work?
I have a lot of recent photos up at Instagram.com/drethelin.
I also recently decided to start finally using drethelin.com but there are a lot fewer on there at the moment.
General-Purpose Robots https://howthehell.substack.com/p/general-purpose-robots
Our planet is chock-full of agents with wildly varying degrees of intelligence that are able to mentally adapt to their environment within an individual lifespan, as opposed to evolutionary timespans. This can't be coincidence. There has to be a shared learning mechanism, and it has to be unaffected by these paradoxes.
As for AI, the "only" trick is to understand and imitate that learning mechanism to a sufficient degree, and then apply it to a machine. "Sufficient" meaning, to make the machine fit for its intended task. Then, if that machine is able to sense, learn, and act with sufficient similarity to an animal, then we could teach it to do anything that an animal with the same sensorium, mental capacity, and actuators could be taught to do.
The mode of learning might even turn out to be completely different from what we find in nature, as long as it gets the job done - after all, we optimize airplanes for their greatest utility to us, not for their similarity to birds. Though if we found that, indeed, bird flight is the only way to ever achieve flight at all, then bird flight it is.
The only way in which you could equate AI paradoxes with the speed of light is if we could regularly observe light-speed travel but just not know how to do it ourselves yet.
When you say, "I've yet to read a statement that made sense of why Polanyi's Paradox and Moravec's Paradox will not cripple the effort" I think you would do better to decouple the criticism with the projected end result. It's one thing to say that you think there are relevant considerations that you think receive too little attention, but when you staple on a "and thus this endeavor is doomed to failure!" rider, you're making a stronger statement with much weaker support.
This is a good general bit of epistemic hygiene but it's also highly relevant here, specifically. Consider that the former of these "paradoxes" has been reborn in this new age of narrow AI. We have networks that can solve protein structures damn near as well as we can actually measure the structures. We just don't... actually really know how they do it. Many people are very upset that these neural networks have such black box character, but it's ultimately not very different from our understanding of human cognition. I don't understand why you think this is a bottleneck; we've demonstrated pretty comprehensively that we can build an intelligence with learning ability even if we can't teach it exactly how to accomplish a task.
As to the latter, Moravek's paradox is definitely a part of the mainstream consciousness of robotics researchers. Everyone knows that it's hard to make robots which can do complex physical tasks but easy to make ones that can do well in a game of checkers (assuming they can move the pieces, at least). It's not especially surprising, since we're comparing our work to that of bio-robots strongly optimized for locomotion and perception and weakly optimized for checkers. That doesn't mean the effort is doomed to fail; it just means that the problem is hard.
An interest read. The author is vastly underselling the main characteristic feature of modern machine learning, which is generalization. You don’t need to create ten million precise computer models corresponding to every possible coffee mug that has ever or will ever exist. A relatively simply system can learn to identify a coffee mug with relatively good accuracy, given a modest training set of coffee mugs. A manipulator arm can manipulate a new, never-before-seen coffee mug by performing a minor inference over the space of mug-like objects it has played with before.
So all the numerical estimates of cost and timing are 10-100x too high, IMO.
It’s the hooking together of the different systems into a multi-modal ML system which can seamlessly transition into a robot body with sensors that hasn’t been done yet. Everything *individual* part is a solved problem, not something requiring billions more in investment.
It's not the technical feasibility, it's that people don't want to buy a robot. I say that as a technophile.
I wish I would want to buy a robot, but all I'm likely to get is a cloud-connected spy with a random update schedule and no accountability.
If your situation gets any more odd, it might turn out that there is an enormous demand for you to refer prospective patients and psychiatrists to each other, and that this takes up so much time and is so lucrative, that it shifts the entire business model of your practice. What a world.
Maybe, but I'm guessing most of the people signing up don't just want a psychiatrist, they want Scott as their psychiatrist, so unless he can duplicate himself, those people will probably lose interest if referred elsewhere.
Time to bring on the brain emulations!
Bot Alexander?
Maybe Scott can start training his own school of psychiatrists.
They wouldn't be as good as Scott, despite initially promising results:
https://slatestarcodex.com/2019/11/20/book-review-all-therapy-books/
Maybe he can train his own GPT-3 psychiatrist replacement.
There is an entirely country that trains Scott(ish) psychiatrists.
Oof :)
[I hope this doesn't count as politics.]
I read the Cult of Smart a month or so ago. I know a lot of people here have read it too. Overall I wasn't impressed with the conclusions of the book. But I can't get over the premise: intelligence is mostly inherited and school won't likely increase your IQ.
That model of the world -- where a significant amount of your intelligence comes from your parents; and there is a meritocracy; and there is some sort of self-sorting -- is a different model than the world I am used to. It really makes a lot of policies obsolete.
It kinda scares me to think about how we could govern in that world. All the same, if that is reality, I'm not sure it makes sense to ignore it.
"I think your fear misses the mark, because merit isn't based on iq alone."
For sure. But for a lot of things, particularly in the advanced society we have now, IQ becomes a bit of a ceiling. So, sure, smart people won't all be successful. But not-smart people are a lot less likely to.
I think https://slatestarcodex.com/2015/01/31/the-parable-of-the-talents/ speaks to that. Also the followup with the basketball analogy, which is what really got it to stick in my brain: https://slatestarcodex.com/2015/02/01/talents-part-2-attitude-vs-altitude/
A few thoughts:
1. I'm going to post a review of Cult of Smart soon
2. We need to be really careful about how we define this. It's not exactly true that school "won't increase your IQ" - there are some studies showing it does, although the exact interpretation is complicated and the magnitude isn't as much as some people would like. I think de Boer's main point is that however much school increases IQ, by the end of a given amount of school, genetically-more-gifted people will still do better than genetically-less-gifted people. Which actually doesn't go much beyond common sense - everyone knows that even within school classes some kids do better than others.
3. Curious what policies you think it makes obsolete. I'm worried some people might over-update on this. My impression is that since everyone already knows that some people do better than others even after a lot of schooling, most people have an intuitive idea that genetic giftedness exists and it already factors into most policy discussions, even if we're bad at talking about it explicitly.
WRT #3:
If a country is reasonably meritocratic, people "sort" and intelligence (talent) is inherited then you have a lot of problems that will *naturally* arise:
1. Poor people will cluster in poor areas. The schools in those areas will be bad (if ranked based on the scores of the students). Rich people will cluster in rich areas and those schools will be good (if ranked based on the scores of students). Changing tax policies or adding funding to the bad schools won't help (above a baseline).
2. Ethnic groups whose immigration correlates with their talent-level will be, as a group, advantaged/disadvantaged and that will stick with those groups until they "blend" in (though sorting could prevent blending.). For example if immigration policies tend toward high-talent from some countries (i.e. Asia) and lower (or average) talent from others (i.e. Guatemala), you would expect Asian students to do better than Guatemalan, *in aggregate* kinda for the long term. That sorta scuttles a lot of the diversity plans we have today.
3. With the greater ability to meet people (via internet and dating apps), sorting will become more and more efficient. That means a growing gap between haves and have-nots. Even in a meritocratic country, the gini coefficient would continue to grow.
My perspective on these things is to ask what counts as a negative outcome in itself, and what counts as a negative outcome because it indicates something else. For example, on point 2, if you use diversity as a measure of fairness (in an equality-of-opportunity sense) then lack of diversity is indeed very distressing, but in the hypothetical you propose, is lack of diversity really such a bad thing anymore?
Of course there are some inherent negative consequences to points 2 and 3 which are worthy of discussion, concern, and efforts to mitigate. I think they're not as bad as they initially sound though (subject to the debatable premise that they are caused by immutable differences in population genetics, and not discriminatory policy). In other words, a consequence of updating only half of your beliefs in response to new data.
"if you use diversity as a measure of fairness (in an equality-of-opportunity sense) then lack of diversity is indeed very distressing"
I guess that's the thing that is bothering me. As a society we have decided that that is the case. Not every individual feels that way, but as a group that's what we've decided. So we've implemented policies to try to fix it. But those policies don't seem to be working. So we "crank things up" to make sure we *really* fix things. What's the next step?
I don't see our society suddenly deciding that diversity and "equity" aren't important. So where do we go from here?
"If a country is reasonably meritocratic...." What makes you think this premise is even remotely correct? Consider instead that multiple studies show that luck is an underappreciated component, e.g. https://www.newscientist.com/article/2210263-lifes-winners-think-success-was-earned-even-if-it-was-down-to-luck/.
Saying that IQ is genetic doesn't mean that two mediocre people can't produce a brilliant child, nor that a brilliant, successful couple won't have a below-average child who's successful (by some definition) because they were "born on third base."
A corollary of your point #2 is that "affirmative action", as it's usually understood, is doomed to fail.
With respect to "diversity plans", a lot of research shows that having employees from different backgrounds and life experiences helps companies succeed. Given your point #2, it's worth moving heaven and earth to find those brilliant children from communities that are underrepresented in the historical picture of "success".
There are diversity plans and "diversity plans". In the real world, "diversity" is a reference to <i>Baake</i> and pretty much never means "diversity of backgrounds and experiences".
These aren't really problems unless your goal is radical egalitarian economics accomplished through white collar brain jobs. Low income families in wealthy societies are better off than median income families in a low-income country, and those "bad" schools may not be so unbearable if we expel the trouble-makers and still give all the kids discount laptops.
The families from Guatemala are almost certainly happier in the US than in their native country: they wouldn't have left otherwise!
Problems arise when the educated classes raise their ladders and effectively eliminate meritocracy, or otherwise police the rest of society with bad policies because they are too divorced from reality to effectively address problems. It's not encouraging to let my kids fall to lower income status if they descend into a mess of divorce, obesity, and meth.
check out "coming apart" on how far the sorting has gone, but really probably just "the bell curve" on the basic issues around iq heritability.
very curious if people know of good books engaging with policy implications (I'll try reading CoS, hear (from Tyler) good things about thing being an honest try, but given the guy's ideology is pry unpalatable to me not rly expecting much..)
Sorry if this is a weird question, but can I ask why you wanted to write a review of The Cult of Smart? (I'll be a little vague here because I want to ask about the meta-level issue without wading into the object-level issues, but hopefully what I mean is still clear.)
I ask because it seemed to me, as someone with what I think are broadly similar views on the relevant subjects, on the basis of interviews and articles about it, that The Cult of Smart basically reiterated what some other books have said about the relevant science, but consciously tried to make it acceptable to a broader range of people on the political spectrum. In doing so, de Boer was taking a stand that was somewhat controversial among his political tribe. And I think that was a really salutary contribution to The Discourse that de Boer deserves praise for.
However, 1) It didn't really seem to me that the book's core thesis would be that novel/interesting to people who had already been following/discussing these issues, as SSC has been doing for some time and, more importantly, 2) It seemed to me that an impartial analysis of the book's arguments on their own terms, with no reference to the surrounding intellectual ecosystem, would likely produce some fairly harsh criticism---but criticism that could be better aimed, considering said ecosystem, at different targets.
So I tried to avoid discussing it when it came up in various fora, because the stuff I agreed with I thought was obvious/well-established in these circles at this point and I felt it would have been tactless/counterproductive to criticize the stuff I disagreed with for not being even more controversial in the mainstream. Consequently, I'm wondering if there's something I'm missing that made it a good choice for an ACX book review.
(Probably the answer is that it's because it sounded like an interesting book, and I'm insane and trapped inside a Straussian 4-D chess match against myself. But I still had to ask.)
You are right that deBoer's fight is with the ogressive-pray eft-lay (odd-numbered thread), but I think the book's focus on educational policy in particular makes it more than a re-hash of ideas previously discussed.
"But I can't get over the premise: intelligence is mostly inherited and school won't likely increase your IQ."
I think this is mostly true? You could have kept me at school 24 hours a day with regular beatings on the hour every hour and you would never have driven either mathematics or music into me (I couldn't carry a tune in a bucket, as they say, and crows sound like Jessye Norman by comparison with my singing voice). So "more school" would never have pushed me into STEM.
On the other hand, education is no burden (as another saying has it) and it's better to go to school and get a taste of everything and some basic knowledge, and a direction to go in for more knowledge, than try to re-invent the wheel yourself. Very smart, self-directed, disciplined kids with supportive parents and access to sources of knowledge may do great without school, but the mass of us ordinary types need the structure and discipline. Not that modern schooling can't be improved upon, but no need to throw the entire thing out.
So in conclusion: school won't turn a sow's ear into a silk purse, but it's important to give everyone a basic level of education.
"Access to sources of knowledge" is pretty much free now for anyone with an internet connection. It's the supportive parents, meaning parents who both want to help and can help, that is the scarce resource. I wonder to what extent one could have a private non-school substitute for that, people in the business of helping kids who are unschooling find the knowledge they want — probably doing it online.
I can't help wondering whether more thoughtful and attentive teaching would have done you some good-- there's more to education than quantity and harshness.
There are probably hard limits on how much people can learn because of talent and interest, but I can also believe that most education doesn't get near those limits.
I vehemently disagree with many of the book's policy prescriptions, but it's hard-hitting and worthy of respect.
My main takeaway was this: we actually know what it takes to increase measured school performance - aggressive filtering of students by ability.
I doubt DeBoer would conclude: "OK, let's do that!" But he does (a) recommend allowing students to drop out earlier, and (b) state "the data shows that what really matters is the academic performance of the top 5 percent of students."
Consider instead the conclusions reached by McKinsey & Co: "... three things matter most: 1) getting the right people to become teachers, 2) developing them into effective instructors, and 3) ensuring that the system is able to deliver the best possible instruction for every child." Spoiler alert: The US doesn't do this. (https://www.mckinsey.com/industries/public-and-social-sector/our-insights/how-the-worlds-best-performing-school-systems-come-out-on-top#)
“Education” as it is currently instantiated is just a stacked sequence of testing and sorting gates, so we should expect to find that the primary function of this system has been to sort people according to native ability.
True education has never been tried, because it is much harder, many times harder, than just sifting for natural high achievers.
What exactly would true education be and how would it differ from education as it exists now? I would assume that elements of true education exist in the current framework; concretely, I know how to do calculus and write essays, which I didn't before. Certainly there is a lot of sorting by ability in the current system, but I don't see how there could be a substantially different system that better achieves the goal of imparting knowledge onto willing and capable recipients.
True education in this framing would involve successfully training people who might, by default, never learn how to write essays or do calculus.
As you say, a calculus classroom might not look much different, but the overriding objectives will be completely different. It's as simple as changing from an attitude of "we're going to go over this material once, then grade everyone on their performance" to an attitude of "every individual student is going to be trained in this material until their performance is adequate". This might require something like individual AI tutors. Like I said, it's very hard.
There's got to be some intermediate case where people are allowed to drop subjects where there seems to be no or little hope of improvement.
Also, part of what's needed is thoughtfulness by the teacher and/or teaching system. Not just going over the material again and again, but working on figuring out what the student needs to learn and how they can best learn it.
I don't think AI tutors is what's required. There are computer programs that teach math to students by letting them learn at their own pace, but not letting them advance until they prove that they've mastered a topic. All that's required is getting schools to use these programs. And perhaps some testing to make sure they work better than current systems.
Widespread literacy and basic arithmetic would suggest we actually can educate. The difference between calculus and the three 'Rs is that you actually need to do the three 'Rs to succeed, so we put a great deal of effort in establishing some baseline.
Perhaps we should do more, but I would rather see better reading comprehension and intuitive algebra before trying to get everyone to jump on to the calculus bandwagon.
From the employer side, education filters serve as a damn fine way to stratify talent. So still useful from this side.
The standard caveats about what IQ is and isn't all apply here, and drastically lessen the impact of this model in my mind. While you may not be able to 'train' someone into a higher IQ, you can train them in mental discipline, ways and habits of thinking, types of problem solving modalities, information and procedures, etc etc etc.
There are some specific jobs and activities that you'll never be good at without a high IQ, but they're pretty rare - although they are probably high-visibility and high-prestige among the people reading this sentence, but that's a selection bias and not representative of most of the world. In many normal jobs and activities, a well-trained and disciplined average IQ person can still outperform an undisciplined and untrained high-IQ person. It still makes a lot of sense to focus education and child-rearing on those things.
Basically, even if everyone does have a 'ceiling' to their abilities related to IQ, very very few people in the world are actually operating at their ceiling, so most have room to improve.
check out "coming apart" on how far the sorting has gone, but really probably just "the bell curve" on the basic issues around iq heritability.
very curious if people know of good books engaging with policy implications (I'll try reading CoS, hear (from Tyler) good things about thing being an honest try, but given the guy's ideology is pry unpalatable to me not rly expecting much..)
I would like to contribute, but feel that the current subscription price is a bit higher than my willingness to pay. I would gladly pay the discounted price, but definitely do not qualify for it. Is there some other way to contribute? (I understand it's a trade-off - if you make it easier for people like me to contribute, you may get less from people who would otherwise contribute the full amount under the current setup).
Please don't worry about it.
If you have the discipline, maybe you could save a few months, then contribute a few, on and off.
Well, this would definitely be a suboptimal attention / $ trade-off.
Sub-optimal compared to what? If they _really_ want to contribute, it may be the only practical way for them to do so.
I'm always surprised by how many sites insist on the recurring-charge subscription model exclusively, and offer no way at all to make one-time donations. I can't be the only one who likes to support content providers, but doesn't fancy riding herd on dozens of automatically repeating credit-card charges.
An interesting model, one alternate to recurring subscription, is to hold articles ransom. No one can read the article until the price target on it's head has been paid, anyone can contribute however much to this target and once it is paid everyone can view it. This has been tried (and I think pioneered somewhat) by Vitalik Buterin of Ethereum.
I took a 1 year subscription and then immediately canceled. It told me I will still have the benefits of being a subscriber until next year, when I will reconsider.
I just want to recommend the (probably soon to be renamed) SSC podcast for anyone who prefers listening over reading.
https://sscpodcast.libsyn.com/
Thank you! It is now renamed and can be found on Spotify, iTunes, and most podcast platforms. https://linktr.ee/sscpodcast
It’s been a disappointing winter. I’ve been looking forward to some solid sheets of snow but mostly we’ve been met with sparse paddings of sludge and ice. I’m hoping for one last good snowfall before things get warm again, or for the snow to be done completely as the dreary grey and green isn’t doing it for me.
Couldn't agree more. It has so far been a Winterless Winter here in Southern Ontario. Growing up it would snow in late November or early December, and persist till March... This year has been alarmingly devoid of snow, and eerily warm... It's hard not to be anything but concerned. And with lockdown and no hockey allowed on the rinks... Hardly a Canadian winter at all.
I'm on the south sides of the lakes, close enough to Erie to see it if I just peak over the Horizon. It's usual for us to have a buffer due to the lake, but every year I feel like the snow recedes farther into my memory. We only got two full coatings this year, one of which that was kind enough decimate some sick trees on our property with a heavy coating of ice!
We got five inches of snow in College Station, TX, two weeks ago! Most snow we've had in the past seven years! But otherwise, the problem of the winter has mainly been dealing with allergies from whatever trees are enjoying the warm weather before and after that snowfall.
I'm so sorry you're missing out on one of the guaranteed joys of winter. No tree allergies! Luckily, up here, all the Bradford Pears haven't started blooming yet. They're beautiful over-pollinating stinky trees and just make spring its own kind of splendid!
Minor heads up: by default, gmail was filtering the Astral Codex Ten substack emails under promotions. I found out by accident - I'm not sure if any keyword accidentally triggered it.
I've been on Matthew Yglesias's substack for a few weeks now, and I've noticed that about a quarter of the e-mails from his go into my promotions tab, while the rest go into the main inbox. I don't know how it decides which, because I think it's happened to a similar proportion of the three types of e-mails (notifications of new posts, notifications of replies to my comments, and notifications of likes on my comments).
I don't like the hearts.
Replace it with a symbol for "thought-provoking", since that's why many of us are here in the first place. A light bulb, perhaps? Preferably LED.
Well, it's good to be back, y'all. Though it'll never be quite the same.
What's the most interesting thing you've done due to quarantine, all?
Parenting (and holding down a job while parenting) during extended quarantines/lockdowns may be one of the most mundanely interesting things I've done... EVER. Aside from that... re-learning sleight of hand card magic, which I used to do much more of as a teenager.
Hm, it turns out there are a lot of contenders for this title. Parenting for sure-- notably, overhearing my son's interactions in distance learning classes, drawing conclusions from that about his academic and social growth paths, figuring out what school activities were actually intrinsically motivating to participate in with him (turns out father-son PE class is super underrated even for non-athlete nerds like me!) and teaching him fun things to complement the classes.
Other than that, learning a lot about how local institutions work and how they fail is definitely on the list. I am going to have Many New Opinions about San Francisco when this is done, hopefully non-redundant ones that challenge everyone's preconceptions. Suffice it to say for now that between virtual civic participation and hiking every single pretty path I can find within walking distance, I feel way more rooted in the place now than a year ago.
Learning what I sound like when I record myself is another top candidate. I had intended to do a lot of music learning and music making in 2020 anyway (quit my job for unrelated reasons at the beginning of March, which was... interestingly fortuitous and terrible at the same time) but would not have done nearly as much self-recording if I'd been able to make music with others in person, and it's a useful enough tool that I wish I'd been motivated to do more of it earlier.
I'm a relatively introverted person, but this has made me realise how much I need in person social interaction, on a fairly deep psychological level.
Definitely. I've watched myself become a more and more angry person without it.
Mine is getting a second home and moving to a city, because the quarantine has freed me up from being stuck in the exurban town where the university I teach is based. It's weird being in Austin and not doing half of the things that I would normally come to Austin to do in the past. But it's really nice being here and doing the other half of the things (mainly involving food trucks, bike paths, and a bit of outdoor socializing with a few friends that live here).
There is something very weird about an economic crisis that leaves a substantial fraction (perhaps even a majority) of households with *greater* amounts of disposable income than they had in the before times.
Kenny, interesting, so I'm going to take a wild stab and say you're probably in the "post graduate" category here?
https://twitter.com/PhilWMagness/status/1352298738514059265/photo/1
Yes it's true. But it's also true that household savings has increased in nearly all income brackets, even among people that don't describe their financial situation as having improved.
My house rehearsed a musical every weekday for three months and then livestreamed a performance. Only about half of us had any prior music or acting experience, but everyone was involved and everyone got cast in a major role. We had choreography and costumes and (some) live instrumentals. Still kind of weird to me that that actually happened.
I drank a lot, ate a ton of food, and got to spend more time with daughter by eliminating the commute.
Also, played COD: Warzone with my buddies every night for about 3 months straight. That was pretty sweet. We haven't gamed since college, but since it's the only way to really spend time together, we have done a lot.
I also made the entire annual budget by brute-forcing Excel. Again. Someday we really need to get out of Excel. That experience along with a number of dropped balls from other departments has resulted in a +10 to grumpy.
Buying a car, putting all of our stuff into storage, and going Airbnb-hopping for 6 months as a way to visit other parts of the US and get a feel for what it's like to live outside of big cities. Of course, since we're staying home except for grocery shopping, it's not quite like getting the full experience, but it's been a cool way to re-evaluate my preferences for, e.g. density vs quiet, etc.
I taught myself Python, since all the non-Covid science labs were closed and so becoming proficient in bioinformatics was the best use of my time. I'm actually quite happy with my abilities, Python is just so much better reformatting and graphically displaying data than Microsoft Excel, and I wish I'd bothered sooner.
I probably won't pursue a career in coding though, I suffer from "techno-rage" even though I know that shouting at computers doesn't make my hacked-together code miraculously function - only Stack Exchange has that power.
Nice! Programming can be humbling if you deeply internalize that the computer is a good listener, sometimes the makers have just told it to listen to the wrong things (hardware makers, language designers, other programmers, you!). It can be very relaxing or very stressful, I'm glad you sampled it either way.
I don't know if it's interesting or not that I have spent pretty much the whole year failing to launch a coliving/coworking/makerspace/etc project that would be very interesting if it launched.
I got a Virtual Reality kit and accidentally got into gamified exercise. Lost 30lbs, although I put a bit of that back on over Christmas. I just wanted to play the new Half Life and instead I've done 500 hours of sweaty Beat Saber.
Staying consistent with my 1k calories per day of VR cardio has been a real confidence boost, and so has seeing the results in the mirror. I'm in the best shape of my life - or, to be less grandiose about it, I'm in shape for the first time in my life. I also stream 1 hour of my workout each day, which has been a fun experience and really motivational.
I think it accounts for the majority of the difference between me last year who wanted to learn piano and me now who's two weeks into piano lessons.
Once I have a six pack and grade 8 piano, I'll have completed my mindset training and be prepared to start working on the real goal: getting good at Super Smash Bros Melee.
Is your characterization of your column and apology as "badly" and "bad job" intended to mean that you believe it was possible to do those things in a way that did not produce the negative responses that you go? I often encounter that sort of thinking in places I am relatively certain it is not accurate, but I can't tell in the context of this post. Saying/thinking these things when there was no less-bad approach available leads to faulty conclusions, in my experience.
This is of course a matter of taste, but I generally preferred the aesthetic of the previous blog. I don't know how much control you have with Substack, but experimenting with different fonts and formatting options might be worthwhile.
On desktop, you can get pretty close to the previous aesthetic by following the instructions in this post, you just have to install a Chrome or Firefox extension: https://applieddivinitystudies.com/slatestarsubstack/
I assume an analysis is planed but I have to ask - what's up with the results of last year's nootropics survey?
Hopefully this does not breach the 'no-politics' nature of this thread but... Has anyone read "Cynical Theories: How Activist Scholarship Made Evyerhting about Race, Gender, and Identity..."? Does anyone have any opinions on how it treats the subject of Postmodernism? Does it get at the crux of it? I am only a few chapters in, but I am recalling that Jordan Peterson was rightly pilloried for his lack of nuance in understanding Postmodernism. These authors, however, seem to have done their homework. So far, it's an excellent, balanced book.
AND... I am looking for more book recommendations on the subject of Postmodernism if anyone has them...
Like on so many academic topics, may be worth starting with the Very Short Introduction series. It tends to be well researched, and adopts a sort of detached and comprehensive account of whatever topic that can lend a certain credibility (though I can't vouch for this title in particular, haven't read it personally. Many of the others I've read are very good.).
https://global.oup.com/ushe/product/postmodernism-a-very-short-introduction-9780192802392?cc=us&lang=en&
Haven't read this one, but on the whole you are correct, this is an excellent series from OUP. Thanks for the suggestion!
I haven't personally read his work, but I've heard a lot of praise for the works of Michael Foucault, particularly Discipline and Punish.
Thanks for recommending Cynical Theories! I am also interested in this subject and second your request for any more book recommendations on the subject of postmodernism.
It's a really interesting book. It is Academic, in the sense that it is NOT polemical and is well researched, but it is Popular in the sense that it is clearly written and obviously written for a non-specialist audience. There are not too many books out there that could fall under the heading of 'Popular Philosophy', and even less that seem to do such a good job of it. Even more to the credit of the authors is that they offer tangible solutions to the problems of Critical Theory, and are sensitive to the complaints of what they oppose.
I think there's a major problem with the word "postmodernism" which is that no one can agree on what they mean by it. It's almost as bad as the term "neoliberalism" has become, but with the advantage that most of the fighting about "postmodernism" is probably a few decades ago.
If you want a steelman of postmodernism, you could do worse than Michel Foucault's side of the debate he had with Noam Chomsky on the subject of "human nature" on Dutch television: https://www.youtube.com/watch?v=3wfNl2L0Gf8
This is a classic. I'd seen it several years ago but forgot about it until now, so thanks for reminding me of this. Not quite as exciting as Norman Mailer vs. Feminists in Town Bloody Hall ( https://www.youtube.com/watch?v=gGYmyou0sKM ) - but still… Actually, the Foucault/Chomsky debate incredibly rewarding. Despite what postmodernism has spawned, Foucault has some very resonant ideas.
But your right. In the end, like all ideological terms, ‘postmodernism’ AS A WORD just dissolves into space. It’s fuzzy. However, it is a theory that is alive and thriving in Academia especially, as is often pointed out. I think it behooves us to understand it fully since its tendrils are still growing/evolving, and enter popular ideology in important ways.
The general consensus I’ve seen from philosophers in my orbit is that it doesn’t - it’s not as bad as Peterson, but the scholarship is only superficially good. Sam Hoadley Brill has a review in Liberal Currents that takes a narrow look at what he considers the most egregious chapter, and that seemed to match the opinion of most of the people I talked to (I’ve not read the book myself, I’m just conveying the reactions of the folks I know who have)
Thanks for drawing my attention to Brill's article (https://www.liberalcurrents.com/the-cynical-theorists-behind-cynical-theories/). Seems to counter a lot of the points made in the 'Cynical Theories book, and points to some important misreadings of the authors they cite to build their points. Looking forward to checking this out further. Cheers!
There are two reasons I find that review really valuable - the first is that it doesn’t focus on Cynical Theories’ (mis)interpretations of the original postmodernists, but in the part where they talk about the modern academics they’re critiquing. I saw so many people talking about Pluckrose & Lindsay’s reading of postmodernists like Foucault and Derrida, most of which sounded really bad but it also doesn’t seem to be the central argument of the book? It’s not a great sign if their account of the history is really wrong but their core claim seems to be that modern scholars are *mis*applying the original postmodernism anyway.
The other thing I appreciated was the highlighting of the lax scholarship - minimal citations, references to entire books or even just authors for contentious claims, that sort of thing. It helped confirm that I wasn’t interested in reading it
My browser just ate my comment, so here is a short version: I think Cynical Theories is very good, but you have to see that is solely about the reception about post-structuralism(= the works of Foucault, Derrida, etc.), and less about post-structuralism itself. Pluckrose and Lindsay show how a very playful philosopy interested in off-beat readings of classic texts became an orthodox school of thought obsessed with reading oppression into everything.
They do this by thoroughly highlighting the works of thinkers like Kimberlé Crenshaw, who were able to turn post-structuralism into a highly effective tool for activists. The cost was, of course, that this created a version of post-structuralism with the sophistry turned up to eleven.
In my opinion, there are some things they could expand upon, especially on the more philosophical and historical side (and they have done so partly through articles at Areo Magazine and New Discourses). They more or less admit in the chapter about Critical Race Theory that it takes as much from the Frankfurt School as from Post-structuralism. This is very interesting, because the Frankfurt School was very activist, and in fact had a big influence on radical political activism in the 60s. In my opinion, if you squint your eyes, Critical Social Justice is the language and arguments of Post-structuralism fused with the activism of the Frankfurt School.
The other thing that could be expanded upon is the American ingredients in Critical Social Justice. Critical Race theory is, of all things, an outgrowth of American legal theory. Generally, the American university system plays a big role in why Critical Social Justice and Applied Postmodernism is how it is. American universities are obsessed with teaching a very specific Canon (see all the Great Books courses for example), which is why Applied Postmodernism hates Canons as a reaction. Critical Social Justice then simply replaces the Canon with another. Continental universities, by comparison, are much more random, you can go from a professor who is a hardcore deconstructivist to a professor who takes 19th century dance theater at face value in an instant.
I think for reading recommendations, since you are reading a work of secondary literature at the moment, it would be interesting to read some actual works by post-structuralists. Foucault is quite ok to read if you are careful not to be tricked by his rhetoric, I can recommend Discipline and Punish. If you have a masochistic streak, read Gender Trouble by Judith Butler afterwards, and experience how she builds upon Foucault in some... interesting ways.
By the way, I read the article about Cynical Theories in Liberal Currents, I found it very nit-picky, and not in any way relevant to the core arguments of Cynical Theories.
I read it about a month ago and I feel like it treated the concept fairly. It is well-researched and it shows when they steelman the philosophical arguments used in the various postmodern and critical theories.
On the other hand, it is clear that the authors have a view (correct, in my opinion), but it's pretty obvious when they are giving it.
So what should we call this blog? Obvious candidates that I've seen around are ACX and ACT. I strongly prefer ACX for some reason. Or will folks just stick with SSC?
I like ACX as well.
Scott uses ACX on the about page ¯\_(ツ)_/¯
"X" stands for "ten", but also reminds us of the "X" in "codex".
ACT is already Acceptance and Committment Therapy and Assertive Community Treatment in psychiatry, probably other things in other fields. I think ACX is free.
As long as we're okay with being _constantly_ confused for the Audiobook Creation Exchange and Access Control Experts, we can probably manage. It'll be tough, but I think we can make it through.
There are no TLAs without collisions. If you can avoid a collision within the same domain, that's better than many manage.
Although with ACX, the collision might be more literal than most.
https://en.wikipedia.org/wiki/USS_Fitzgerald_and_MV_ACX_Crystal_collision
abbreviations.com [lists](https://www.abbreviations.com/ACX) a few known abbreviations for ACX, but they all seem pretty obscure, so I think we're safe.
ACTen
One problem with ACT is this: https://en.wikipedia.org/wiki/ACT_(test)
1) Has anyone else had issues with the RSS feed? I didn't get a notification for this post despite subscribing.
2) There's a lot of talk over which cities are most livable by various metrics - housing supply, jobs, transit/pedestrian friendliness. I'm curious about anecdotal subjective experience - people who have lived in multiple cities, which one was your favourite overall? (And I mean including every factor - e.g. "NYC was dirty and badly run but I liked it overall because I had an interesting highly-paid job" should still count, since job market is somewhat location-dependant).
For the record, I did see this post in my RSS feed (using Feedly).
Hm, it just showed up for me (about an hour after it was posted). So I guess it's not a general RSS issue, my reader just only updates every few hours.
I'm interested if there's any truth to whether different cities have actual network effects that matters for things like innovation. I read a blog post recently that said you can't walk around NYC without thinking about how much money you have and ... I don't know, NYC is very big. I guess it depends where you live. But I wonder if that's an accurate reflection of density in mixed income neighborhoods?
I will say that NYC is difficult with little kids and no car, but it's far more pedestrian friendly than other cities in the US that I've been in because of public transit. The MTA may be a failing system with unsustainable costs and maddening inefficiencies, but it also comes close to being a truly comprehensive system for the city in a way that e.g. Pittsburgh, Boston, Chicago's transit systems do not. Likewise the CitiBike system is maddening but it's well on the way to being a hugely impactful program, even as the city seems to do everything in its power to keep biking dangerous.
When I visited London it felt very close to NYC to me, with the exception of the tube which was eerily clean and well functioning.
Not sure if this is helpful, but the RSS feed is working. (Notifications depend on how you’re subscribed to the feed. I manually checked that the feed now contains this post.)
SF is currently undergoing a real estate and tech job implosion that is going to reset a lot of markets here 10-20 years. I think it will be a very popular place to move to next year.
who else is here???
Like, other notables? Or?
Funny, I took it as an open-ended, somewhat lyrically expressed question about space aliens.
Socialization subthread!
With the new blog having just launched there are probably a lot of new and returning readers and commenters that just want to say hi or introduce themselves. I’ll start.
I found SSC a bit over two years ago and binged the archives like mad. I had come across overcoming bias back in 2009 and read those archives up to that point but didn’t really stick around or engage with the community. I just lurked on Scott’s blog as well, but I have really missed it since it has been down and now that it is back I want to engage more; So Hello everyone!
I will avoid mentioning who I work for to preserve anonymity, but I am an aerospace engineer specializing in rocket propulsion and will be happy to comment on or answer questions regarding rocketry technology or space exploration. Other intellectual interests include history, and psychology especially as it relates to leadership and management.
In terms of cost disease, is rocketry more like tech (where things get exponentially cheaper over time), or transit/healthcare (where things get inexplicably more expensive and become less affordable over time)?
Oh man, if I had my own blog I could probably write an entire multi-thousand word post on this. The shortest version I can manage is that it is hard to tell. Up until very recently most rocket development was done by big aerospace companies working closely with their biggest customers: NASA and the military. This tended to result in a cost disease type progression for a variety of reasons. Now with companies like SpaceX and Blue Origin doing more of their own things we are starting to see prices for orbital launch fall a lot, but it is hard to tell if this will be sustained. Personally, I think it will be, at least for a while, and we will continue to see falling costs at the same time as increasing capabilities. Definitely a cool time to be in the field.
For more boring types of rocket propulsion systems (i.e. not launch vehicles) it is clearer cut. The size and cost of things like attitude control thrusters for satellites have been following the ‘tech’ model and dropping steadily even as performance improves, largely due to advances in electric propulsion. Small military rocket propulsion systems for things like anti-tank missiles, air to air missiles etc. are usually pretty simple and cheap, with the insane cost of the overall system driven more by the targeting abilities and the need to make the entire thing as close to indestructible and idiot-proof as possible than by the propulsion system.
My unsupported guess about transit/healthcare/education is that the things aren't getting more expensive, instead they are the only things that aren't getting cheaper. I think its been shown that if you break down where the money is going in a cost disease ridden industry, there isn't any one particular subsection that is getting more expensive. I'm suspicious that if you do the the same thing for an industry without cost disease, you will find that there are subsections getting cheaper.
This is a nice idea! And perfectly aligns with my goal for this year, to become more involved in the community.
I don't remember how long ago I found SSC - the answer to this is lost in the hazy days of my post-highschool years. I've been reading HPMOR (and was a major contributor to one of the translations) as it was being released, then stopped midway, then continued after the hiatus (I think there was a hiatus?), then the sequences somewhere in there. I've always been a lurker, but maybe this will change from now on :)
I'm about to finish my MSc in physics (quantum optics), and very excited to be starting soon my PhD in cognitive science. Most of what I know about it is from reading Scott's book review of Surfing Uncertainty, so I can't answer any questions about that yet, but I'm always happy to chat about physics if anyone's interested.
How'd you shift from physics to cogsci?
Apparently physicists are seen as "generally smart people who can do well even without domain-specific knowledge". I don't know if that's true, but I definitely took advantage of that. It also helped that the PI I'm going to work with is a physicist, and was looking for some like-minded students :)
I found the SSC blog while Scott was on his little hiatus, so I was unable to participate. I had previously been binge reading LessWrong. (I really should "hunker down" and read the Sequences, in order, and HP:MOR.) I find it hard to engage in the rationality community (though I guess SSC is technically not part of that) since everything they say sort of goes over my head. I'm a senior in high school. I like math and science "generally", but I don't really know what I want to do in college and whatnot.
I only recently managed to finish the Sequences (after knowing about them for, jeez, nearly ten years), but I've read HPMOR like four times, no hunkering down required! It's so fun. And I first read it when I was in high school, so definitely accessible.
I also had the problem of people saying things that went over my head for a long time. I spent a couple years reading like crazy (I read the entire SSC archive while in college) and listening a bunch, and now I'm at a point where I feel comfortable contributing most of the time – although not in every single domain, e.g. I don't have much to contribute when it comes to technical AI safety... but that's okay! I guess what I mean to say is that a lot of the things are only hard to understand because they're high-context, not because you're not smart :)
"high-context, not because you're not smart"
Thanks for the encouragement. I guess, I still can't really participate in the conversation. But helps to know that I CAN learn these things. Like, me wanting to learn these things is independent of whether I can actually participate, so the existence of these conversations are something to be grateful for (that once I learn, I have somewhere people will listen), rather than be jealous of.
I also have the same feelings about contributing - Even with stuff where I have some subject-matter experience like Math/Programming/Finance, we have a whole bunch of actual experts who know way more than me (well, unless for some reason we ever get to talking about random integer matrices, which was the specific subject of my PhD), so even then I don't feel like I can add much.
Also, welcome!!
👋 I've been reading SSC since ~2014, but have never contributed online - I went to a few EA/Rationalist meetups but didn't attend consistently. I'm a software engineer working at a biotech startup, with a previous life in art conservation. I'm looking to improve my writing and get more involved with the community (man, does quarantine make you reassess where your social skill points are being spent!). I also do a lot of crafts and am happy to talk about any of the above.
>I will avoid mentioning who I work for to preserve anonymity, but I am an aerospace engineer specializing in rocket propulsion and will be happy to comment on or answer questions regarding rocketry technology or space exploration.
Ah. Another member of the SSC/ASX aerospace fraternity. I'm also an aerospace engineer, specialized in rockets in college, currently doing military aviation stuff. And there's John Schilling, who there's a decent chance you know professionally. A couple of others, too. Hopefully gbdub makes it back.
I suppose I count too, though I've since been banished to the land of EEs and will tend to be cagey about the specifics of my day job.
I'm here, but still trying to figure out how to make efficient use of this comment system. Short term, you may have better luck finding me at Data Secrets Lox. And if you know a source for good, cheap attitude control thrusters, I'd like to hear about it. In my world, the good ones aren't cheap and the cheap ones don't always work.
The Arms Control Wonks blog/podcast does a lot of OSINT on missiles and rockets. Might be a worth a gander in you're still interested in the topic.
I missed Scott's insightfulness and creativity. I found out about Scott through a chain of HPMOR -> Eliezer -> SSC, but another chain from HPMOR is my interest in Fan Fiction, and now I'm writing my second quest on Sufficient Velocity. I'm doing one that I call a Let's Quest (like a Let's Play but User vote controlled).
I was born in a tiny town, then I lived in a big city and now a small town again, and SSC has been a community to stay in touch with communities that just don't exist locally now that I'm back in a small town.
I've been married for 5 years and now becoming a first time parent in May.
Hi! I found SSC almost 2 years ago I think? I think like many people I had the experience of finding some post I really liked, and then finding another, and another, and eventually giving up and just started reading anything I could find. I got to reading LW and overcoming bias, and HPMOR, and I found everything extremely interesting, especially as I'm currently in a master programme AI, and it has a way of relating back to almost everything.
I'd like to get more involved as well, it seems I had just gotten into the habit of following every post (and every link in every post, Scott has a way of filling up my reading time!) when the blog got deleted. Hello everyone!
Hello and welcome everyone! I go by Aristides, and have been reading SSC since the Last Pychiatrist stopped posting, and I searched for other writers, about 7 years ago. Despite Scott's humbleness, I consider Scott's influence on my thoughts to have been much more positive, so it turns out that TLP leaving, worked in my favor. I have been an occasional commenter for 2 or 3 years, and look forward to be more involved in this site.
I am a law grad that works HR for a government hospital. I am also Eastern Orthodox and moderately conservative. Hopefully others will find my perspective interesting.
There's a massive Greek Orthodox church going up in our small city, and I've been curious about the perspective of that group. I have family members that converted to Catholic and I've become protestant non-denominational after growing up without a church background, so it would be an interesting discussion.
I work with our HR director and accountants and provide IT support and reporting as needed., especially now with processing end of year stuff. Are you busy at end of year as well. I was just writing a 5500 census report, and waiting for it to finish loading data to verify it looked correct.
I used to be a non-denominational protestant as well, and converted a few years ago. I am hardly an expert in it, since a large percentage of sermons I have attended have been in a language I do not speak, and I stopped attending once COVID hit. Overall, I love the atmosphere, and after studying the theology, I believe it has the highest percent chance of being correct. Catholicism has had too much corruption in its history for the doctrine of papal infallibility to make sense to me, and all Protestant denominations made so many changes to original Christianity, that if any of them are actually the truth, they discovered it through sheer luck. That's my thoughts anyways.
My end of calendar year is the least busy time, since the fiscal year starts at Oct 1 here, and that's when any of the annoying reports are. No one wants to discipline or fire employees around the holidays either, so that helps keep work light, too. That said, I think work will pick up soon with the administration change.
Hi! I've read SSC for at least 5 years now, but had always avoided the comments section due to lack of interest. But in the recent SSC hiatus I decided that was silly and from now on I'll try to participate more.
I'm a cryptographer and am also in the process of learning some stuff about law (in the US). Vaguely EA too but more human-centric than most of their discourse seems to be these days.
(Is anyone else experiencing weird lag when typing their comments, or is it just me?)
Followed from Overcoming Bias to LessWrong. Enjoyed many of Scotts posts there but didn't know they were by the same person until it came up in the post about governing fictional worlds.
Never really engaged in the comments on either site, might make more of an effort to do so this time around.
Giddy with excitment to see the blog up and running again. I got a notification that there's a new post while I was writing this!
What's the state of research in rocket engines? Is it mostly a solved problem? Do people still innovate in say, the basic geometry of combustion chambers?
It definitely is not a solved problem, there are several big open questions in rocketry, and advances are being made if not as fast as we might like. To give a quick rundown of some problem areas and places where new things are happening:
Combustion instability: this is probably the biggest single unsolved problem in rocketry. The short, simple explanation is that combustion instabilities are pressure oscillations in the combustion chamber that can grow via various feedback mechanisms until they are strong enough to destroy the engine. This almost killed the entire Apollo program, as the F1 engines for the Saturn V initially had extreme combustion instability problems. The various mechanisms involved, and the interactions between them are complex enough that they still cannot be modeled effectively with computer simulations, so even today it is not possible to predict in advance whether a given engine design will be stable or not. Larger engines are more prone to instability problems, so small scale testing doesn’t really help, the F1 engine on the Saturn V was a scaled-up version of an earlier missile engine that was very reliable. Combustion instability is therefore generally considered to be the single biggest risk in any new engine development project, and holds back innovation in any area that might affect it as companies try and stick to iterative improvement on existing designs that are known to be stable.
Pressure-gain combustion: The details are complex, but in layman’s terms if you design the engine such that the propellants explode (detonate) rather than burn the thermodynamics allow you to extract more useful work from a given amount of propellant… in theory. In practice it remains to be seen if a practical engine design can realize enough of the theoretical gains to be worth the other tradeoffs involved. Many designs have been proposed in this area, and it is a very active field of research. Rotating-Detonation engines (https://en.wikipedia.org/wiki/Rotating_detonation_engine) seem to be the lead contender, but it will likely be some time before one is built as anything other than a experiment.
In terms of more immediately applied progress being made, SpaceX’s Raptor engine will be the first full-flow staged combustion engine to see flight service; and the Raptor and the Blue Origin BE4 will finally bring methane fuel into widespread use. Electric propulsion is still improving rapidly, with novel designs still being invented regularly.
Some other areas: Solid propulsion is so simple there are not a lot of areas to improve on, but there is still active research on improved propellant formulations. New propellants and additives are also being actively investigated for hybrid rockets, monopropellant thrusters, and hypergolic thrusters. In particular there is a lot of interest in replacing hydrazine as a monopropellant, and the monomethyl hydrazine (MMH) and nitrogen tetraoxide (NTO) combination for hypergolic propellants, as these are highly toxic and sufficiently volatile that they require extensive safety measures during handling to avoid breathing the fumes. NASA’s Green Propellant Infusion Mission (https://en.wikipedia.org/wiki/Green_Propellant_Infusion_Mission) was a test flight of one potential replacement for hydrazine in monopropellant applications, and ended successfully quite recently.
So while we might wish for faster progress, there is definitely a lot of exciting work going on in the field, and this is just the tip of the iceberg, there are plenty of other areas of research interest besides these.
Thanks. Given undergrad physics, do you know of a decent source for basic rocket engine design?
Ignition! is a fantastic read and very accessible, it is also freely available as a pdf (legally) here: https://library.sciencemadness.org/library/books/ignition.pdf it is more focused on the propellant chemistry than the rocket engineering but still well worth it.
after that if you are willing to spend a little money and don't just want pop-sci books go for one of the following textbooks, both will be at least somewhat accessible to anyone with a hard science/engineering background.
"rocket propulsion elements" by George Sutton or
"rocket propulsion" (cambridge aerospace series) by Heister, Anderson, Pourpoint, and Cassady.
Thanks. I think my wife will also like Ignition. (She does research on the social impact of explosives manufacturing in the US.)
I was introduced to SSC/HPMOR at around the same time, 2014 or so. Definitely remember trying to puzzle out the end of HPMOR in time for the last chapters. Loved SSC, Scott's writing style, and the ideas behind it all. Read through Eliezer's Sequences, and then mostly just lurked on SSC until 2020 happened. Started in on more LessWrong after that to bridge over until recently, when ACX emerged!
Studying at a German technical university in my bachelor's, my interests include space/rocket technologies, long walks in uninhabited places, and since March (thank you COVID-19) programming with Python, followed closely by HTML, CSS, and JS to facilitate getting functioning Heroku Apps going.
Matt Levine’s posts have made me want to invest in a SPAC upfront (i.e. for $10 with the warrant). Is this one of those things where you have to be a VHNWI and have a relationship with a bank?
So you're saying you *don't* want to buy shares+warrants of a SPAC that already exists, like PSTH or something? To do it "upfront" means investing in the IPO of the SPAC- like any other IPO, you'd need to participate in the opening book, and that means either being a bank and participating in the road show, or having a big broker and a good relationship so that you can get some of the few chunks allocated to retail brokers.
Basically what Andrew said. I don't follow Levine's newsletter anymore (a little too snarky for my taste) but my impression is SPACs are a lot more interesting for the sellside and bankers than anyone on the buyside.
It's interesting that SPACs have become fashionable just when direct listings have become possible.
Brokers make their livings from their contacts. They earn commission from finding buyers for blocks of shares, for instance.
In the IPO process the chosen brokers place blocks of Pre IPO shares with their friends, making sure that the company value for those pre IPO shares is below fair price. On IPO day, the fair price is discovered by the market, the shares "pop", the friends make a handsome profit, and of course the people who owned the company before the IPO process make less than fair market price.
Now, direct listing is possible, thanks to a great extent to Bill Gurley. https://www.joincolossus.com/episodes/53331505/gurley-direct-listing-vs-ipo
Direct listing makes the archaic IPO process obsolete, removes the pop and removes significant revenue from brokers and their friends.
SPACs are all about brokers profiting from contacts. They're a new, large revenue stream for brokers.
What have valuations been like in raises right before a DL? If a company can’t sell into a DL and needs capital, that’s the alternative.
I'm showing my age, but the lack of liquidity discount for unlisted shares has traditionally been about 40%. It's lower now.
The below is at least true for Sweden.
Valuations before public listings: VC's and private owners (VC funds are often partly pension money and therefore IMO not entirely private money) have been holding off longer before listing successful companies for the last couple of decades. They can fund the company themselves, so access to public market money isn't necessary for the company. The incumbent investors get to keep more of the market cap growth during late exponential scaling to themselves.
Those successful companies have *no* trouble finding money. A story in my vicinity unfolding now: A promising company closed a financing round in late December 2020. There wasn't space in that round to admit a new well known investor; the incumbent investors took the entire round. That new investor returned days after the previous round was closed, offering a 25% higher valuation to invest money the company doesn't need.
For the vast majority of less attractive companies, capital is accessible. It's been stimulus all the way down for decades now and money is looking for assets to buy.
- Small company without much traction / generational transition: brokers sell to private investors. This is a lively market. Lots of brokers, LOTS of companies, lots of investors, some of them building conglomerates / adding value to their favorite company flavor. Brokers deserve their 7-10% commission; it's hard work.
- Easy to understand product, can't deal with the overhead of listing, still want access to some public capital: try Alternativa Aktiemarknaden. Report your numbers quarterly, trade is organized one week/ month. There are various alternatives for crowdfunded shares, passion driven shares like teams or bands, etc. It works pretty well. https://www.pepins.com/market/listing#show
- Small company, can handle the administrative overhead of listing, worth at least 50MUSD. List on the NASDAQ/OMX.
so, direct listings are now mainstream and it shouldn't be a big deal for any company to use them, yet a lot of companies just turned to SPACs which are more expensive and fee-ridden than conventional IPOs.
any conclusions you'd draw?
Direct listings are not yet mainstream; brokers are fighting its adoption. Brokers and their best clients do lots of listings while founders do at most two, so founders rely on brokers expertise in the listing process. Brokers press their informational advantage and recommend the traditional IPO process since it generates large income streams for them and for their best clients.
Here is a broker friendly explanation: https://pitchbook.com/news/articles/understanding-the-ipo-markets-pop-culture
For the VC side, try the Bill Gurley podcast above.
I don't know what SPACs have suddenly become popular; the below is speculation. I think it's a bit similar to the ICO frenzy a few years ago.
- Brokers and well known players want to make money on their contact nets and ability to hype companies.
- Companies want high valuations and good visibility quickly.
- Sarbanes - Oxley made listing onerous, so there are fewer listed companies.
- There is a lot of private capital around, so companies don't have to list to get financing in late exponential growth, reducing the number of publicly listed companies.
- Private equity buys companies / divisions off the public markets, reducing the number of publicly listed companies.
- Investors who don't mind a lot of risk don't have much choice, especially if they're young and haven't yet found their way onto broker's call lists. The success of Robinhood demonstrates Millennial risk appetite.
-Creative and clever talent flows to the finance sector.
Not many investment vehicles, lots of money around, insatiable risk appetite and tough regulations to reduce fraud plus a surfeit of creative clever talent
- Regulatory arbitrage / less onerous paths to public capital - like ICOs a few years ago - are going to be popular in this environment.
That's my theory of SPAC popularity.
not sure I understand exactly what you want to do, but SPAC pre-merge is publicly traded, so you can buy it.
just checked, seems can buy PSTH with my brokerage account.
This afternoon, I heard of another SPAC being set up by stellar investors. They believe they can manage more money than the billion USD or so they manage now. They're thinking of setting up a SPAC, and will allow new customers to invest with them. For a fee.
They are skilled. Decades of experience with generally excellent returns and no terrible years. Up 98% in 2020, while investing exclusively in listed shares. (!)
I can understand why people might be interested in investing in *their* SPAC, even if the fees will be high.
Maybe that is the attraction of SPACs. Customers get a chance to ride with the stellar investors for a fee.
Glad that the blog is back! I'm gonna make immediate use of it by asking this: I'm starting a PhD in Psychology in the Netherlands. My PhD friends have told me Twitter is essential for keeping up with relevant publications, events, trends and development and that I should absolutely make an account.
A) What does everyone think about this assessment and B) if I do make a Twitter account, what are good ways to filter out all of the bullshit?
Science PhD student here. I heard something similar and got on twitter for this reason. I've come to find science twitter to be undewhelming, very heavy wokeness-signaling (I say as someone much more woke-aligned than a typical gray triber), and fairly little actual science. The actual science does tend to be good though, and I've seen some *very* cool times where scientific collaborations actually formed on twitter and produced really cool stuff. Overall science twitter has more than 0 value but is also pretty annoying. Note though that your field may be totally different.
That said, after getting on science twitter I ended up discovering rationalist/postrat/whatever twitter and I've been really enjoying that little subculture. I ended up just splitting it up between two different twitter accounts and I find I spent more time in the rat-sphere twitter than science twitter.
Math professor here. It may be different in psychology, but it's hard for me to imagine how Twitter can be useful. https://psyarxiv.com/ is probably a better way to keep up with what's new. At the same time, when some truly important papers in your area come out you will hear about them from your advisor or other students. Your advisor is a much better source of information on what's in trend than twitter.
Personally I am most productive when I change all my social media passwords and hide them somewhere far away. The most important thing to succeed in academia is to be able to keep your concentration for a long time, patiently struggle through long and intense periods of boredom and confusion, sometimes literally banging your head against the wall.
It's really hard to do if you didn't hide your Twitter password in a good place.
Thank you, the productivity issue definitely crossed my mind as I already have a hard time resisting the temptation to check fora, messages and feeds when I'm supposed to work. Definitely going to keep it off my phone.
Not in academia, so no opinion about A, but for B, I make *heavy* use of muted words (e.g. "trump", "election", "woke", etc) and it makes Twitter tolerable - even enjoyable sometimes!
Senior PhD friends telling you is good evidence, ask more students and profs to be even more certain.
A: very much field dependent. I finished PhD in pure math a few years back and never used twitter. from my ex soon to be a bio professor I know twitter is essential tool in biology or at least a range of fields around hers: many/most profs are there, people use it to announce their papers, all major conferences have hashtags around which the conversation evolves etc..
B: you seem to not understand what twitter is. it's a social network. you follow a bunch of people, and what is shown in your feed is mostly what they post. if you follow researchers in your field and twitter is indeed actively used as a professional tool in that field, I'd expect most of their tweets being professional.
unrelated, speaking of bio twitter: trvrb is must follow for everyone interested in covid-related science ;)
Not having twitter, I don't know how it works, true. I suspected it might do a thing similar to facebook and 'curate' the tweets of the people you follow according to shareability or engagement or whatever. If I can just pick a list and not worry about it being filled with the outrage du jour, that would be nice.
I don't know about academia specifically, but my general impression is that, if something's important, it will move from Twitter to slower-paced media/fora. If you read the relevant books, journals, etc. and talk to your professors and classmates, my guess is that you'll get most of what you need, and possibly more than you would get by spending the equivalent time on Twitter. But, again, this is based on speculative conjecture, not informed experience.
And I think it's worth noting, as other commenters have mentioned, that there's a difference between how a perfectly serene and rational truth-seeker would ideally use Twitter and how you, personally, are likely to actually use Twitter. At least for me, I've found that only using the app to send DMs/contact people about specific things has been a huge improvement over trying to use it well in its core functions. It's sort of a point of pride for me at this point that I now often hear about whatever the latest stupid Internet controversy is for the first time when I read an article about it in The New York Times. But YMMV.
I'd be interested if people have specific recommendations for people to follow on Twitter for psychology publications/events/etc., both global and in the Netherlands. Do people have some tips?
Political scientist here. In political science, Twitter is very useful for keeping up with publications and current debates. The most useful thing you get on twitter than you don't get in other places is real time debates about research. This gives you a much better sense of where the discipline is right now, as opposed to 3 years ago when the papers that are being published were written.
It's hard to filter out the bullshit not least because some of the bullshit is actually useful information about the norms and practices of people in the discipline. My best suggestions would be to just scroll past the tiresome tweets, mute anyone who's raising your blood pressure too often and mute keywords that come up in repeated discussions or memes that don't add anything.
Academic Twitter is mostly good, but at random times once or twice a year it becomes incredibly stupid. On bio Twitter recently:
- an offhand joke about a certain species of worm turned into an enormous fight in which people who made fun of the worm were somehow accused of being bigoted.
- an online funeral was organized (by some fairly high-profile biologists) for a Native American scientist who supposedly had died of COVID, but in fact turned out to be the fake identity of a white scientist.
Pretty much any academic community has these bizarre dramas and if you get caught up in them, they might totally blow up your career. So, don't do that.
Actionable advice: you should make a Twitter account. Follow people in your field. Don't post much, never argue with anyone. If you notice someone who is very argumentative or seems to spend a lot of their time mocking other Twitter users, mute them. Do not install the app on your phone.
Does anyone have a good explainer of post-rationalism (aka. postrat) and how it differs from the standard LessWrong fare?
AFAIK, there are several possible 'varieties', or possibly several distinct ideas with the same name. My personal favorite is David Chapman's 'meta-rationality'. His site "Meaningness" is great, though perpetually unfinished.
Irrelevant suggestion - bring back the punny titles for open threads. They brought some levity and variety and I don't think the stated reason for getting rid of them (too confusing for newbies) ever made that much sense.
I'm not just saying this because you never got around to Openguin Thread... okay I kinda am.
If punny titles seem potentially confusing, you could always add a parenthetical after the pun, e.g. "Opigeon Thread ["Open Thread #whatever]".
While you're at it, I miss the days when every post was titled "Stuff". Though, that might be confusing for newbies and veterans alike…
I have a feeling those went away because he was running out of puns after ~140. Which is a lot, and I certainly don't blame him.
I was hoping for OPern Threadfall.
Can we try fine-tuning GPT-2 to generate open thread title puns?
This made me wonder if GPT-3 is any good at puns. Alas, according to Gwern:
"Better [than GPT-2], but not as much better as one would expect given the leap on many other capabilities. Trying to generate puns or rhymes, it seems like GPT-3 know extremely well what they are on an abstract level, and will appropriately manipulate words and attempt to make puns or rhymes (see the shoggoth-cat dialogue below for a particularly striking example), but the words it chooses just aren’t right on a phonetic basis."
The mentioned dialogue has puns like:
"Well, the best pun for me was the one he searched for the third time: “You didn’t eat all my fish, did you?” You see, the word “fish” can be replaced with the word “cats” to make the sentence read “Did you eat all my cats?”"
So I think we're still quite a bit away from GPT-3 generated thread title puns.
What I am really looking for is the weekly links section, miss so much.
Health question: this last year, unexplained and unplanned I've lost about 30 pounds of weight, and about five inches off my waist and I now have to use suspenders to hold my pants up,
Has this happened to anyone else?
Unfortunately in this time of covid-19 speaking to a physician or nurse practitioner is really hard, I called for six months with lung symptoms before finally being seen, an x-ray and my weight loss prompted my doctor to say "80% lung cancer", but a biopsy found a fungal infection instead. I'm guessing at least a year until pre-2020 style medical care is available again, weight loss remains a mystery.
Lost a similar amount of weight, maybe a little less; but I'm fairly certain it's just due to malnutrition. Too depressed to make proper meals and feed myself most of the time, making up the difference with coffee flavoured with milk and sugar.
As an already-skinny person, I once lost 10-15% of my body weight over a summer because I was playing computer games all day. I lost muscle from the inactivity, but I also lost fat, just because being so inactive and so absorbed in playing games suppressed my appetite. (I do *not* recommend this as a weight-loss method; I very much doubt it would work for anyone who *wanted* that result!) I gained the weight back once I started being active again. Has Covid led you to be sitting in the same place all day a lot more often than you used to?
It's the other way round for most people, I think. The 'Covid stone' is a thing.
Was Givewell correct to list KIPP (a prominent charter school network) as its recommended charity for US givers? https://blog.givewell.org/2012/03/14/kipp-houston-update/
Reasons to think it might have been wrong:
* There was even at the time a lot of government and institutional support for prominent charter networks
* Closer investigation into what's going on in those schools should have made it clearer that they were focused on test prep in a way that wouldn't lead to sustainable long-term benefits (KIPP has had a hard time showing increased college enrollment and graduation, though I think after changing to focus on that they now have a modest impact on that)
* Research was very limited at the time on impact of no excuses charters
I'm having a hard time separating disturbing things I've heard from people who have done stints at KIPP (and Success Academy etc) from what was understood in 2011. I think this should be seen as a failure of Effective Altruism/Givewell but I'm interested in defenses.
Was KIPP really the best place to give money in the US in 2011?
Anyone here using a branch of Chromium for privacy reasons? I'd be curious to hear about your experience and configuration.
I've been using Brave for a while and I like it. Not sure what you want to know about configuration, or if you have more specific questions...?
Mostly just curious about what extensions provide the most privacy for the least inconvenience
One of the ideas behind Brave (brave.com) is to have privacy by default. It comes with a built-in tracker-, add-, script-, and fingerprint-blocker. You don't need to get add-ons at all (I have none). So as far as I can tell it is basically as convinent as possible.
I didn't really thoroughly research existing browsers, though; no comment on how it compares to other options.
I've been using Brave for a while now, and I'm pretty happy with it. The only changes I would make out of the box are to set tracker/ad blocking to aggressive, and fingerprint blocking to strict. I haven't encountered any broken sites with that, except for Google Maps which doesn't load some of the JS (one could argue that if you really want privacy, you shouldn't be using Google in the first place). It's also really simple to disable the shields (blocking) on an individual site.
If you're up for it, you can add uBlock Origin and Privacy Badger, which are pretty good without any config changes. uBlock is good for getting anything that isn't blocked by Brave, as well as the ability to add custom block lists. It also makes it easy to zap specific elements of a site. Privacy Badger also blocks some things that the others don't get, like Disqus comments. Both extensions can be disabled on an individual basis.
Downloading Brave and adding uBlock and Privacy Badger should be about 3 minutes of work, so I'd recommend that as a good starting point.
(Bonus advice: get a password manager while you're at it. Helps against password reuse as well as phishing attacks. 1Password is a good option.)
Scott's difficulties setting up his psychiatry practice remind me of Paul Morphy (https://en.wikipedia.org/wiki/Paul_Morphy) who, after becoming widely recognized as the world chess champion at 21, retired from competitive play to open a law practice. His law practice failed because everyone who came to him just wanted to talk about chess. Scott, the widely recognized world blogging champion, may be facing a similar predicament.
Now that my living room no longer looks like I've been fighting Terminators, I am happy to report that I have found a practical use for RGB lighting! To my utter shock, my new PC rig booted on the first try, but one of my memory modules didn't light up. What do you know, Windows is only detecting one module. Fortunately, it just needed to be reseated, but who knows how long I would have gone with only 16GB before I noticed the problem? Since the RGB not only alerted me to the issue, but obviously indicated which module had the problem, I retract my prior opinion that RGB memory modules were frivolous nonsense. (RGB fans, on the other hand...)
Any good books/resources on soft skills? I suddenly found myself in a position in work where being a bit better at persuasion, negotiation and de-escalation would be much appreciated.
I'm finding it hard to find quality resources and easy to find crap.
Influence by Robert Cialdini is a good and fun place to start. https://en.wikipedia.org/wiki/Influence:_Science_and_Practice
Cheers for the recommendation. Having looked into it a bit, this is probably where I'll start.
How to have impossible conversations by Lindsay and Boghossian.
The standard recommendation is Dale Carnegie's "How to Win Friends and Influence People" and many people I respect highly also highly recommend it.
It feels so "basic" to recommend, and I've heard some good criticisms of it that I forget at the moment; but if you haven't read it, it's definitely worth reading.
have you considered trying something like
https://www.coursera.org/specializations/negotiation-mediation-conflict-resolution
or other actual courses/classes?
Not up to this point, but you've got me thinking about it.
Cheryl Strayed - "Tiny Beautiful Things" influenced me a lot (https://www.goodreads.com/book/show/13152194-tiny-beautiful-things)
It's not an instructional book and doesn't have anything about business negotiation, but it is a great example of what talking about emotionally loaded topics in an emotionally intelligent way looks like in action.
Assuming you're after the emotional (not logical) component of negotiation and de-escalation.
This is very left-field and totally out of the realm of what I was looking for or the kind of thing I would normally read.
But I think it's an interesting choice for that reason, and I should probably read this at some point.
The Harvard Negotiation Project has some really good books on the subject. I would definitely recommend "Difficult Conversations" and "Getting to Yes".
Had no idea about this, cheers.
The Power Paradox: How We Gain and Lose Influence, was the most useful one for me, coming into HR from a law background. I probably should have read more, but that was the most engaging for me.
I had a shower thought about what will happen to Western diets if meat (and only meat) suddenly increases in price, as has been predicted extensively. I haven’t been able to find examples of this happening to a society before and I was curious if it’s more likely we’ll expand the window of acceptable animals to eat, move to insect proteins, or move to mass vegetarianism? I haven’t been able to find much via Google, so I thought I’d reach out here.
Given how much meat substitutes have improved in quality and prevalence over the last few years, I'm assuming that those will become much more popular. Insect protein doesn't have nearly as much of a head start as Impossible Burgers and Beyond Burgers etc. so I don't see that catching on, as much as I would like to. Eating other animals seems similarly unlikely because they're not factory farmed, and the meat produced at the largest scale is obviously the cheapest, so I don't see how e.g. horse meat would compete.
Why would meat suddenly increase in price anyway?
That last question is the biggest one to me. it just seems unlikely; there is so much land that is unsuitable for growing vegetables or cereals; but is suitable for grazing livestock;
Rising incomes in Africa/Asia causing the people there to purchase more meat?
By limiting demand to a smaller number of people, namely those who can afford it, we might expect that the quality of meat being sold will increase. What has happened to American beef — the best in the world 30 years ago — is a sin and a shame, and for meat eaters, a tragedy. So, as a member of that increasingly decreasing class of people, I would look forward to it.
Do you have a link, so that a young cow addict can read about what happened to American beef? I always thought it tasted fine...
Separately, the economic question of "cost of manufacture goes up --- does quality rise or fall?" is not obvious to me. I actually had the opposite guess (more pressure to cut costs causes quality to decline), but after reflection I'm no longer sure at all. Should the answer be obvious?
There are different market segments for different meat qualities. You can, for instance, expect to pay more for grass fed, dry aged Angus beef.
Not much, unless the price increase is something on the order of doubling or more. Maybe you'd see increased adoption of extended meat blends, which are basically only used in institutional settings right now. Fast food would either adopt them or try and shift more upscale. Most people who can afford it will pay more at the grocery store rather than modify their diets.
I'm am not a historian, but remember that for much of pre medieval history mainly the rich ate meat on a regular basis. Perhaps this will recur, with "true meat" costing hundreds of dollars a kilo once "lab meat" reduces in price.
I could see factory-farmed meat increasing in price relative to lower-intensity farming, but small farmers producing free-run chicken and pork and rotationally grazed grass-fed beef to sell directly to customers has never been all that expensive and isn't dependent on volatile inputs and layers of middle-men. Buying whole frozen chickens and half-cows from farmers is more work, but it's not really more expensive, and it tastes better, it's better for the animals, and it's much better for the environment.
The economies of scale involved are interesting. Just letting cows eat grass is really quite cheap as long as land is cheap and it rains enough. Some breeds are optimized for getting really swole on expensive industrial inputs, and others (e.g. Black Angus) are smaller, hardier, and able to take care of themselves with a little supplemental hay in winter. The smaller breeds also have an easier time calving, reducing vet bills.
Of course, I'm in Canada where we have lots of land and we don't do the kind of intensive feedlot beef production that's common south of the border. Chicken has never been dirt-cheap either. The factory farming situation isn't great but it isn't US levels of bad either.
The change I expect to see is both a shift from the crappiest, least-meaty, most-environmentally-destructive meat products to plant-based analogues, along with a surge in decentralized consumer-direct animal agriculture. The latter is already happening quite rapidly where I am.
You should auction off ten "fan slots" for your psych practice to the highest bidder and see how much they go for. I bet a lot.
Great idea, but it might go against the central ethos of Lorien Psychiatry - that healthcare should be reasonably affordable.
Ethics is all in the phrasing. Set it up like this:
"I will give free therapy to the top ten monetary contributors to defraying the cost of other people's therapy."
That starts getting into the area of ethical problems, especially for psychiatry.
"starts" is an understatement there.
for most people, the list of famous psychiatrists begins and ends with Jung and Freud, who for cultural reasons catered to rich people.
the real art of being a good psychiatric counselor has NOTHING to do with being like Jung and Freud in that respect. The good results from any patient are equivalent to the good results from any other patient, there are no bonus points for being there for people who are richer, who have more friends, who are more gifted, who are more interesting, who are more special in any way. All of God's creatures are uniquely spectacular.
I expect that is a great way to end up losing your license.
"I will give free therapy to the top ten monetary contributors to a fund whose purpose is to defray the cost of other people's therapy."
Would that work?
I hate fake Tables of Contents. Most books have them. You know the ones, the book is divided into five chapters or so, and each chapter has many subsections that are not even listed in the Table of Contents, so when you want to find something again, you can't, or when you want to remind yourself of the general cut and thrust of the book, you can't do it near so well, because instead all you get from the TOC is a handful of scarcely meaningful chapter titles. It's as though you want to see the anatomy of the book, and the book just keeps repeating in a sing-song, "Head, Shoulders, Knees and Toes." And you reply, "Damn you, I want to see the whole skeleton!"
Do you ever have this experience?
I am thinking about solving it, but is there demand?
I've most often been annoyed by this in PDFs of RPG sourcebooks. For some reason, there seems to be a trend of releasing them with just the major chapters bookmarked and not any of the useful subheadings.
What are some books similar to Joseph Henrich's The WEIRDest People in the World? (In that they discuss the same/related issues, but from a different angle of some sort.) I ask because I've found Tyler Cowen's advice to read in clusters very useful, so I'd appreciate at least a couple more to round out a cluster including Henrich's book.
Two examples of possibilities here: The Open Society and its Enemies, by Karl Popper, and Fairness and Freedom, by David Hackett Fischer. The former is a work of philosophy criticizing, among other things, a thread of collectivism running through the works of Plato, Hegel, and Marx, as opposed to individualism, and the latter is a comparative history of the US and New Zealand considering their individualist/free market/liberal etc. roots. So they cover what seems like closely related conceptual ground in terms of Western individualism, but from different perspectives: Henrich discusses anthropology, Popper discusses philosophy, and Fischer discusses history. (Note, I've already read The Open Society.) Hopefully that illuminates what I meant about books that form a cluster. (They can also be books on the subject within the same discipline but with different intellectual loci.)
I've definitely seen people apply WEIRD to hot button political issues, but, since this is the non-politics OT, I'll avoid elaborating for the moment.
The Wikipedia article on the Girard book is quite intimidating, but I'll keep it, and psychology more generally, in mind.
Guns, Germs, and Steel (which I have some major arguments with but enjoy nonetheless)
Democracy Now! (though I haven't read it)
This big survey on western religion and economics is quite good: https://www.iza.org/publications/dp/13371/religion-in-economic-history-a-survey
I'm really looking forward to reading Guns, Germs, and Steel, but it seems like it fits into a slightly different cluster to me: namely, regarding the rise of agriculture/Eurasian civilizations/the state etc., along with e.g. Azar Gat's War in Human Civilization and Ian Morris' Why the Rest Rules. (By contrast, I understood Henrich's book to be about Dark Ages and beyond developments.) Although now that I think about I guess there is overlap in terms of the "what made the West unique?" angle.
Democracy Now! is always great.
Thanks for that survey on religion and economics, it looks quite pertinent.
how about something on christianity's role in western civ's idiosyncrasy?
think CWT with Koyama might've had some refs
Thanks for the recommendation, I own but have yet to read this book of his, which seems quite interesting and relevant: https://www.amazon.com/Persecution-Toleration-Religious-Cambridge-Economics/dp/1108441165
Farewell to Alms and The Son Also Rises by Gregory Clark, perhaps.
This one might interest you:
Dominion: How the Christian Revolution Remade the World by Tom Holland (https://www.amazon.com/Dominion-Making-Western-Tom-Holland/dp/1408706954). Another ‘how we came to be what we are’ book but very well done.
It's good to be back, everyone.
I've recently been watching Criminal Minds on Netflix, as I'm a sucker for police procedurals. It's a fun show, but I have to wonder, how legit is FBI profiling? Seems like one of those things that could easily just be all smoke and mirrors.
Wow. Good to have an OT again.
Is there any interest in me continuing to do Naval Gazing links as I was doing back before SSC shut down, or should I just not bother?
This comment has solidified my dislike of the "like" feature. It has 6 likes as of when I post this. That's probably 6 people who want to see the Naval Gazing links back, but I can't be sure, and it's frustrating compared to getting actual information.
Naval Gazing is great, and I like the idea of links in at least the first few OTs of ACX at a minimum to get some mind share on the new site.
Likes are bad if they funge against higher-content replies, but IMO better than no feedback at all - a comment with no responses but a relatively high number of likes is useful information. Still would prefer an opt-in system though.
You have a solid point. With how much effort it takes for you to write a Naval Gazing posts, you would hope that your fans would take the time to make a full comment, not just a quick like.
For what it's worth, while I've always been impressed by how extensive your Naval Gazing posts are, I've honestly only read 3 of them, since it is outside of my area of interest. Hopefully the people that regularly read your Naval Gazing will actually comment, since I remember it being fairly popular.
Naval Gazing has done fine during the hiatus, and I'm not starving for comments there. But those people don't need links posts to remind them of things. I would ask this occasionally on SSC, and usually got several people saying that they enjoyed the posts because they were irregular readers and liked seeing what I was up to.
Good point. I want those links back.
are you certain those six would have replied otherwise? perhaps that is six new positive responses that you would not have seen at allin the old system
For some reason the RSS on Naval Gazing hasn't been working with my RSS reader (NewsBlur) lately, so I selfishly wouldn't mind the occasional reminder that you exist and have new posts to read.
I've asked Said Achmiz to look into it.
Regarding the Nobel Prize sperm bank, I highly recommend David Plotz's book The Genius Factory to anyone who is interested in learning more: https://www.amazon.com/Genius-Factory-Curious-History-Nobel-ebook/dp/B000FCK5WG/ref=tmm_kin_swatch_0?_encoding=UTF8&qid=&sr=
IIRC, most of the actual donors ended up not being Nobel laureates, because Nobels tend to be awarded for work from previous decades and thus to older recipients. (Also, hilariously, because customers complained that many of the Nobel laureates were too short.) So they ended up being younger, modestly accomplished all around the board types instead.
That's interesting because I've heard there's a correlation between intelligence and height.
Both depend on nutrition during growth.
Have another unsourced theory: there are two kinds of genius. One is related to general good health and possibly height. The other is related to weird shit and possibly related to physical problems.
Basketballers versus Ashkenazim?
Can you freeze sperm indefinitely? If so, we should be collecting samples from up-and-coming thinkers and then giving them out decades later when we know which ones will get Nobels. (idea may or may not be stolen from rat-tumblr post about housecat genetics)
1) Why rename the blog when moving to substack?
2) What does the new name, Astral Codex Ten, mean?
It's an exact anagram of Scott Alexander. The previous name was only a near-anagram.
I made a quick dark theme for SSC (err, ACT. ACX? The new blog). It's at https://userstyles.org/styles/196568/arcane-old-texts
Color is mostly #aaa on #111, which is great on my monitor, but YMMV. Has an issue with the skeleton dummy posts that show up before content loads (I can't find their HTML, so they're still glaring white), but I'm not sure how annoying that will actually be. Seems to work otherwise. I'll try to keep this updated when site changes start breaking it, but feel free to fork/do whatever with it.
(Remember to use Stylus, not Stylish, to install themes: https://chrome.google.com/webstore/detail/stylus/clngdbkpkpeebahjckkjfobafhncgmne?hl=en )
Thank you for this! I'm checking it out right now, and the colors are really good.
If all they have about you is something you wrote in college, and where the only issue was some miscommunication... I wouldn't worry too much. There are many worse people and stories. I hope the NYT doesn't publish, but if they do... Remember that in a few years it will just be a painful but distant memory.
It's great to be back!
That being said, I feel an urgent need for some way for new comments to be easy to find.
I've recommended having brains as well as hearts so that we can mark intellectually satisfying comments as well as just liking them. I don't have enough experience with assorted fora to have an opinion about rating systems. Part of my motivation is to get more nuanced ratings, but I also want to see those cute little brains to click on.
It might also be good (if feasible) to have an "I want to remember this" marker. For that matter, the want-to-remember marker could be part of a rating system. It might be more honest than mere likes.
Ditch the hearts. Replace with a line of buttons:
- an LED ("this comment enlightened me")
- a flame ("this comment inflamed me")
- a "+1" or "This!" (no need to clutter the comment list just for this)
- a "-1" or "Nope!" (ditto)
The first pair would put a number on the famous light/heat ratio and constitute the SSC equivalent to up- and down-votes, and the second would be a nanoscale survey on the comment's position.
"I also want to see those cute little brains to click on."
Indeed, we should consider servicing the needs of those of our fellow-readers/subscribers who are members of the zombie community (or perhaps, in deference to their sensibilities, I should say the "alternatively-alive community") 😀
What other organs should we have? :)
As for new comments, it seems like keeping the tab open should work; it automatically shows "1 new reply"/"n new replies" buttons that you can use to load them. You can search for "new repl" to find them. Of course this is suboptimal in that you lose track if you close the tab (or the browser) for any reason.
I've been using that strategy, though I someone said the tab auto-refreshes after a while. I don't think I've seen that happen.
One partial solution would be having an option to set the time for the tab to what you want. That's what I did for ssc if the computer restarted.
Searching for "hr ago" and "min ago", for posts older than 24 hrs the date. But that is still a badly ineffective way.
Gee, several people responded to my comment, so I went here to reply, and I can't even find my comment! Even searching the page (either for my name or for a keyword I used) doesn't turn it up. Very disappointing.
Sadly, you have to load all of the comments, in multiple 'batches', to be able to reliably find any one comment.
"Open all" should be an easy fix in Javascript.
I'm not sure there is a way to do it with the StyleBot plugin, though.
I believe that the current comment structure may suit many Substack users fairly well, but that the SSC crowd needs more of a forum for conversation around the pieces: it would be nice if Substack could offer publishers the choice between forum-style and current style comments.
If you go to Data Secrets Lox, it is a forum with many of the old SSC commenters posting regularly. Personally, I dislike the forum system, so I'll be staying here, but it has a big, active community.
I feel like I very recently saw a paper/blog post/something arguing that the IAT could not, in principle, measure *implicit* bias because it requires the person taking the test to make *explicit* decisions about whatever social category is the focus of the test. Has anyone else seen this or similar arguments? If so, could you point me to it?
I thought the IAT was measuring reaction time, and the point was that unconscious bias is equivalent to strong associations, which in turn influence reaction time?
That said, when I took the IAT, I came out having a more positive association with black people than white people, and I'm white. But I was very conscious of the reason, which was that I've got a learned response to most of the classic associations between "black people" and "[negative stereotype]", which is to stop and step back and take another look at what's going on. (Very useful when trying to assess the level of racial bias in something written 100+ years ago.) So it really wasn't testing my internalized racism, so much as my ability to examine a text to discover how racist it is.
I suppose I could be tested more accurately if they showed me negative stereotypes that were new to me, or that I didn't know were negative. But hopefully that's too Orwellian to pass muster, at least for a while.
I'm actually kind of curious whether anyone's flipped the IAT around, and tried to see what sort of associations cause a shorter or longer reaction time. If one accepts the hypothesis that faster reaction time implies stronger associations, which imply bias, then this could actually give hard data on claims that "[something] is racist". But I'd still suspect that it would be confounded by an inability to distinguish "X is associated to Y in my mind, because I think it's true" and "X is associated to Y in my mind, because bad people think it's true, and I need to keep an eye out for them".
The IAT does make use of reaction times. More specifically, it based on the difference in mean RTs between particular conditions. So, e.g., for the white/black race IAT, in one condition you have to push one button when you see either a white face or a positive-valence word and a different button when you see either a black face or a negative-valence word. Then in another condition, you push one button for white faces or a negative-valence word and a different button for black faces or a positive-valence word. The idea is that if you are slower in the second condition (with white/negative and black/positive pairings) than in the first (white/positive and black/negative), this reflects implicit bias.
The thing I saw was, I think, making an argument that because you make explicit decisions about race in this (version of the) task, it isn't measuring implicit anything. I didn't read the argument in detail, and maybe it was a bad argument, but I'm curious to find it and read it.
So, any interesting Biology (whether it be evolutionary, synthetic, systems, medicine, and so on) papers/articles that anyone has enjoyed recently?
Recently been looking over:
- "An antiviral self-replicating molecular heterotroph" https://www.biorxiv.org/content/10.1101/2020.08.12.248997v1
- And been looking at the wikipedia article on https://en.wikipedia.org/wiki/Toxin-antitoxin_system , which serves to select for cells that inherit the antitoxin (as otherwise the toxin kills them). There's a few papers that it references that I may look over as well.
Seeing this new blog open prompted me to finally start reading Unsong, and it's great trippy fun! (I'm currently on the last couple chapters of Genesis.) I also keep catching myself trying to Kabbalistically analyze random names and words.
History is fun! More snippets from MacCulloch's book on the Reformation.
Got those pesky marital troubles? Pope won't cough up with an annulment? Never fear, the Reformers are here!
Philipp Melanchthon [big name German reformer if you didn't know https://en.wikipedia.org/wiki/Philip_Melanchthon] ...in 1531 he suggested to Henry VIII that he should solve the Aragon marriage impasse by getting the Pope to provide a dispensation for bigamy.
Uhhhh - thanks, Philipp, but *two* queens at the same time? Maybe not workable?
Henry wasn't the only one struggling with the problem of "I got this hot new squeeze who's demanding I put a ring on it, problem is I'm already tied with the old ball and chain, what do?"
"Against this background of attempted accommodations and adjustments, a moment of opportunity arose in 1540–1 that Charles V was determined to seize. He had a bizarre stroke of luck – a severe dent to the prestige of his most truculent Protestant opponent within the Empire, Philipp of Hesse, which also damaged the reputation of the leading German reformers. It was revealed that Philipp had recently committed bigamy, with the express if reluctant written agreement of Luther, Bucer [another Big Name in the German Reformation] and Melanchthon. Their action was the ultimate in Protestant wooing of the magistrate. The signatories’ logic (apart from the unspoken acknowledgement of how much they owed to Philipp) was that bigamy was better than the adultery which had characterized Philipp’s chaotic private life over the previous decade; Melanchthon and Bucer in particular were only being consistent with the advice about bigamy that they had offered Henry VIII a few years before. Nevertheless the evangelical theologians’ belief that both their action and the bigamous marriage could be kept secret was a colossally naive political misjudgement, bringing bringing predictable and gleeful expressions of moral outrage from their Catholic opponents. Charles graciously granted an imperial pardon to his prince, who could in theory have suffered the death penalty for bigamy as newly laid down in the 1532. imperial law code (Lex Carolina). Melanchthon was so shattered by the self-inflicted catastrophe that he became gravely ill."
Puts a new spin on the oft-cited cry from the Protestant side about Catholic dogma and practices: "Where's that in the Bible?" I had heard it as Luther and others were consulted about Philip's marital problems, couldn't find a prescript for divorce - or rather couldn't settle on recommending it as a solution - but there was the Scriptural precedent of the Patriarchs having two (or more) wives.
Philip seems to have been really keen on this as a solution, for what it's worth, because he intensely disliked his first wife, was constantly committing adultery, wanted to regularise the whole position so he could eat his cake and have it, and even had the second wife picked out ready to go before the theologians got a start on the question: https://en.wikipedia.org/wiki/Philip_I,_Landgrave_of_Hesse#Bigamous_marriage
On the one hand, it's good that Substack is willing to put in a lot of work to change to UI to fit what Scott wants. On the other hand, it seems concerning that they weren't already doing all these things. My impression is that most commenters' complaints about the UI are glaringly obvious, and any competent designer could have fixed them without needing to be told. Or, perhaps more accurately, they could have fixed them if they had been prioritizing user-friendliness rather than engagement / revenue. For example, it seems clear to me that having big "Subscribe" buttons everywhere is bad for users (but good for revenue), so a designer who prioritized user-friendliness wouldn't do it, and wouldn't need to be told to do it.
My impression is that, of the sites that (still) have commenting, very few of them host the types of conversation that were typical at SSC. Mostly I see thanks and feedback, with occasional author responses, and a chronic problem with people getting into toxic discussions with each other. Given that, I'm not surprised that a site like Substack opted to go for a simple but functional commenting interface that would work well for the vast majority of users. We're weird.
As for the subscribe buttons, if they help generate revenue to keep Substack viable, I don't mind. Of all the causes of failure of sites like Substack, I'd suspect "lack of revenue" is one of the biggest, and "too many Subscribe buttons" is one of the smallest.
Well I think substack was attempting to be more of an email newsletter platform than a typical blogging platform, but it seems they've more and more had to face up with the fact that a lot of the authors want features that are more analogous to a blog than a newsletter. In the end they're probably going to mostly become a WordPress clone with very basic features for most authors, but custom features for the authors that bring them a lot of revenue.
I'm not surprised to see what you said about eugenics here, as I've gleaned over the years that your stance is roughly that. But I hadn't seen you crystallize your position like that before. Do you expect to ever devote a whole post to eugenics?
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.
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.
Has anyone tried deradicalizing conspiracy theorists through betting? I'm thinking specifically of political conspiracy theorists, as they seem to have more testable predictions - eg X country will do Y action by Z date, X person will win Y election, X person will be in jail by Y date, etc, but anything conspiracy theory will do. It seems like losing money, especially repeatedly, would make the person have much stronger incentives to update their beliefs than they otherwise would have. The other upside is instead of being annoyed with your conspiracy theorist friends/families, you'd have an incentive to communicate with to them, so you can make money. Maintaining ties to people with sane worldviews is important for this sort of thing.
What do people think of the pros and cons of an easily recognizable ACX symbol? Something that readers of the blog could use to identify each other in public, that would be commercialized either by Scott, or open for anyone to do so. I ask because a means to easily identify fellow readers in public would make it easier to find like-minded people, but I can of course see that it might bring risks given the ambiguous reputation we hold in some circles. It's worth it to me, but I'm curious if the cost/benefit analysis is equally favorable in the eyes of others.
Pretty sure that there are not many "Ṛta" stickers out there.
…Yet. :-)
Regarding sign-ups for Scott's new psych practice: I suspect what many of those readers would like (as a second-best solution anyway) is a referral to a different physchiatrist who is taking new patients and who has similar attitudes and views about the nature of the doctor-patient relationship, meds, research, etc. to Scott's. If Scott were feeling quite charitable, he might create a separate form where people can request a referral. I suspect that would siphon off most of the people currently ignoring the instructions not to try to sign up as a new patient of his. (Sorry Scott: I have never commented here before and don't know whether to address this to you or to the group!)
My blog, Naval Gazing, originated in the SSC OTs, and before the shutdown I usually put up links to new posts in the whole-number OTs. There appears to be at least some interest in me continuing to do so, so I should probably start with an overview of what I've been up to.
First, I finished the long-running Falklands series, which ended up at 24 parts:
https://www.navalgazing.net/Naval-Gazing-Index#FalklandsHistory
Second, I've written a lot on nuclear weapons at sea, including some truly bizarre plans on various sides of the Cold War:
https://www.navalgazing.net/Tags/Nuclear
Third, I just wrapped up a series on merchant ships:
https://www.navalgazing.net/Merchant-Ships-Introduction
There's also been a fair bit of slightly random stuff that might be of interest:
https://www.navalgazing.net/Naval-Rations-Part-3
https://www.navalgazing.net/Territorial-and-International-Waters
https://www.navalgazing.net/Naval-Bases-from-Space-Hampton-Roads
https://www.navalgazing.net/The-Reagan-Maritime-Strategy
https://www.navalgazing.net/The-Seaplane-Striking-Force
https://www.navalgazing.net/Merchant-Ships-Whaling
https://www.navalgazing.net/Naval-Airships-Part-1
https://www.navalgazing.net/The-Battleship-and-the-Carrier
@scott, >I support it in the sense of improving people's genetics and genetic outcomes - long-term through genetic engineering, medium-term through having things like the Nobel sperm bank (but less badly done) available for people who want them, and short-term through voluntary community-based efforts like Dor Yeshorim.
> dictionary, definition. eugenics. the study of how to arrange reproduction within a human population to increase the occurrence of heritable characteristics regarded as desirable. Developed largely by Sir Francis Galton as a method of improving the human race, eugenics was increasingly discredited as unscientific and racially biased during the 20th century, especially after the adoption of its doctrines by the Nazis in order to justify their treatment of Jews, disabled people, and other minority groups.
I'd be keen to know, what exactly sets you apart from the above definition, I do get it, you are not Nazi, and you are not a racist either, but how exactly your opinion disagrees with the rest of the dictionary definition of the eugenics.
The word eugenics, at least the way it is commonly used, implies a level of coercion/lack of consent that isn't present in the dictionary definition of the word, but is nevertheless what comes to mind when most people use it. The word also carries a baggage of being used by horrible people to justify some really miserable actions (and being used by other people to put those people in a bad light).
It's entirely possible to have a preference for improving the human gene pool without endorsing anything remotely resembling what people normally think of when they use the word eugenics.
Thanks understood.
Interesting question. However, I sense the topic veers towards CW territory (at least for me..) so I'd suggest saving further discussion for a more appropriate thread.
Where did he say that he disagreed with eugenics? He said that he supports some forms of eugenics, but not the ones that gave it a bad name.
Is there a plan to copy all SSC content, including comments, into ACX, or else to return the styling to SSC?
This review article got big attention in Finnish media, but also faced heavy criticism from other researchers:
"Bipolar disorder: An evolutionary psychoneuroimmunological approach"
https://pubmed.ncbi.nlm.nih.gov/33421542/
This is probably far too late to get any significant response, but I'll write out the post anyway to make rewriting it next OT easier.
I've been studying at a German technical university for a few years, and have been steadily confronting myself with the question: why wouldn't more American students follow a similar path?
I have to say, my confusion here has been pretty large, and I can't tell if it's due to:
-A unique situation that made it easy for me compared to the average American student.
-An American education (recommendation) system with a massive blind spot.
-A German visa system with a lack of interest in Americans.
The long and short of the steps that I took to get into this position is:
1. Learned German during an exchange year (not necessary, my university as well as perhaps 200 more German universities have Bachelors as well as Masters in English)
2. Applied for German universities (with a decent GPA/ACT/SAT [specific numbers on request], not much work is required. In fact, I found the 4 American universities I applied to much more difficult than the 10 or so German unis)
3. Was accepted to German universities.
4. Demonstrated a relatively reasonable level of financial stability (8,000€ in a bank account or a German family willing to vouch for you.)
5. Began attending the German university of my choice.
Current monthly costs, living in one of the 5 largest German cities, include:
*300-500€ (for me, just over 300€) rent in a shared flat, also studio apts (WiFi, electricity, heating, and water included).
*100€ food, relatively decadent with fresh vegetables, fruits, cheeses, and occasional meat.
*100€ stunningly inclusive health insurance.
*(You may be wondering, where's the university costs?)
*50€ University costs (this includes texts, computer labs, multiple student workshops, legal insurance, email, cloud service, and most of all, public transportation in the entire state + parts of other states)
*Around 10-50€ for a cell-phone plan. Figure this isn't too different from the states.
So if we're looking for the high-end of what I could pay, I come out with 800€ and 9600€ per month and per year, respectively. Throw one or two international flights home in there, for an extra 500-1500€, and I bump into 11k€. For my entire costs, from year one to year eight if I wanted to spend more time here.
Now, my comparisons to the US are based on anecdotes from friends, as well as a few websites with average cost estimations. Take this with a few grains of salt, but I seem to find $20k a year as a pretty common amount for in-state tuition and associated living costs, and $40k a year with a medium out-of-state tuition (plus living). This is either 100% more than what I'm paying, or 300% more. Not only that, the bachelor at my university (as well as all German universities) is a 3 year program, since GERs (general education requirements? Don't know my American university terms very well) aren't a part of the curriculum. So to get my bachelor, I'm spending 33k€ (to be fair, this is about $40k right now), whereas in-state public is spending $80k and out of state is spending $160k plus an extra year?
I'll assume most of you know someone, either around college age, raising children who will reach college age, or you're planning on having college graduates of your own some day. Reach out to them, let them know it's an option, and maybe save a year of their life and $100,000.
Or critique my comment and let me know exactly how I'm wrong. I welcome both.
Ningen Dock is something I'd never heard of, thanks for mentioning it. I've heard Mexico/Central America/Southeast Asia as alternatives for large surgeries in patients, even amongst some friends of the family.
I think it may fall prey to a similar problem as with universities, those patients/students who would most be in need of these alternatives (no insurance, no scholarships) probably tend to be those least aware of the options.
Definitely true that a lot of the value of education is social class and networking, but seeing as our society seems to be becoming more global (2020 maybe excepted), I see networking in the European Union as still beneficial for later opportunities. The social class seems to be similar in both places, whether I'm European college-educated or American college-educated, I'm still now a part of the college-educated class.
I appreciate the take, which dying niche would you say that Musk explored? The 'creating change in physical space' niche? Also curious as to which AI talk came to a point where 'sheds are for losers' was an opinion.
I find it hard to disagree on the point that the middle class norms hard. At the same time, I feel that 'study in another country for free*' could very well be an adoptable norm. Especially because it seems to be a norm for the middle/upper class in Asian/African/South American countries. More than 30% of my university are foreign students, almost entirely from those 3 continents (and Europe, but that one seems more of a gimme). North America doesn't even begin to appear at my university, I'm sure out of the 3000 foreign students I could count the Americans on a single hand.
Occasional meat? I did know some vegetarians over in Germany but IDK how they did it. I'm definitely considering doing a postgrad in Germany,
I can recommend it! The higher you go up the education ladder, the easier it is as an English speaker. My University currently has more English Masters than German Masters, compared to a single English Bachelors in a field of about 20.
On the occasional meat, Grillkäse (grillable cheese) is a common BBQ replacement which is pretty common. Most supermarkets have anywhere between 5 and 20 different meat substitutes, using soy, gluten, milk proteins, etc. that are relatively close to meat in cost.
Did you go to Germany from a country already fairly near to Germany? How'd you find out about this possibility in the first place. I'm well past college age at this point, but when I was, I didn't know such an arrangement was even possible. American students aren't going to choose a path they don't even know is available to them. We have guidance counselors in high schools that steer you toward specific paths, and universities also directly advertise and recruit people.
Heck, when I was a teenager, I actually tried to conduct a fairly exhaustive search compared to just choosing the best known local schooll. Princeton published a guide to the best 308 universities or some number close to that, and I sought out additional information on a list of 20 selected from that book, and ranked them in a weighted basis along all of the dimensions I thought I cared about. Even going into that much detail, the possibility of a foreign university never even occurred to me. I also made a terrible initial decision in spite of trying that hard and transferred out after one semester.
My home country, and where I still have citizenship, is the USA. As to what I meant by relatively unique circumstances, I was the recipient of a US-Germany exchange program scholarship in high school. This was what originally made me aware of the possibility since our guidance counselor never had 'foreign university' as an option.
The best-laid plans can definitely go awry, I hope that the next university fit better!
Just a quick vote for "please no favourites". It'll just spoil the feeling of balance and make more standardised beliefs or ideas more visible while hiding the more interesting ones as people skip to the "Top Comments!!". It will create an echo-chamber-like atmosphere and kill discussion (like it did to metafilter.com) where those with differing opinions are accused of "not reading the room" - and thereafter silenced/shouted down by the majority.
Maybe an issue for substack in general, but the public RSS feed contains stuff that's supposed to be hidden: https://astralcodexten.substack.com/feed
It would be cool to be able to link posts without the comments, like was possible on the old blog. This is useful to send posts to people and have them be less scared about the size of the scroll bar.