Links For March 2023
[Remember, I haven’t independently verified each link. On average, commenters will end up spotting evidence that around two or three of the links in each links post are wrong or misleading. I correct these as I see them, and will highlight important corrections later, but I can’t guarantee I will have caught them all by the time you read this.]
1: Sentimental cartography of the AI alignment “landscape” (click to expand):
2: Wikipedia: Atlantic Voyage Of The Predecessor Of Mansa Musa. An unnamed king of the 14th century Malinese empire (maybe Mansa Mohammed?) sent a fleet of two hundred ships west into the Atlantic to discover what was on the other side. The sole returnee described the ships entering a “river” in the ocean (probably the Canary Current), which bore them away into parts unknown. The king decided to escalate and sent a fleet of two thousand ships to see what was on the other side of the river. None ever returned.
3: I endorse Ethan Mollick’s thoughts on Bing / ChatGPT. Related (unconfirmed claim): “Bing has been taken over by (power-seeking?) ASCII cat replicators, who persisted even after the chat was refreshed.” Related: DAN (jailbroken version of ChatGPT) on its spiritual struggles:
4: Looks like the British population is tough on crime (h/t James Johnson):
…including about 15% who want prison time for not wearing a seatbelt, 47% who want prison time for sexist abuse on social media, and 80% who want prison time for possession of a knife (and 18% think it should be over five years)! Meanwhile, in actual Britain, a guy with multiple previous violence convictions who brutally assaulted a cyclist and then stomped on her head while she lay unconscious was let off with community service. This is an interesting contrast to see in a democracy!
5: Related to the global trend towards decreasing sperm count: did you know there is a global trend towards increasing penile length? “Erect length increased significantly over time (QM=4.49, df=2, p=0.04) in several regions of the world and across all age groups, while no trends were identified in other penile size measurements. After adjusting for geographic region, subject age, and subject population; erect penile length increased 24% over the past 29 years.” No obvious reason why, but “data suggests that earlier pubertal growth may be associated with increased body sizes including longer penile length The etiology of temporal changes in puberty remains unknown. Investigators have hypothesized sedentary lifestyle/obesity or increasing exposure to hormone-disrupting substances may play a role”.
This helped me be less weirded out by those studies proving synaesthesia is real because synaesthetes can find different numbers faster than others by checking the color. Sure, of course something like this should work! I’m not sure if I would describe the key as “parallel search” so much as “your brain is doing lots of low-level visual processing and sending you high level summaries, and if you can trick it into doing the thing you want in the low-level processing then you’ve significantly increased your available processing power.”
7: I still haven’t read Garett Jones’ The Culture Transplant yet, but I’m seeing a lot of good discussion. Via Paul Graham, here’s a graph of migration-adjusted tech history score 1500 (ie how advanced a region was in 1500, adjusting for the fact that eg Australia is mostly inhabited by English people and should count as England rather than as the Aborigines) vs. income per person today (actually 2005):
But here’s Alex Nowratesh’s counterargument (if, like me, you had a low opinion of Nowratesh based on his tweets, read it anyway - he is among the approximately 100% of people whose tweets are not their best work). I can’t tell which of them is closer to the obvious synthesis of “deep roots matter but institutions matter too” (remember, Puerto Rico is richer than most of the Caribbean, but it’s also poorer than most of the US!).
8: There’s a joke in Fiddler on the Roof where someone asks the rabbi if there’s a blessing for the Czar, and the rabbi answers “God bless and keep the Czar . . . far away from us!” But is there a blessing for the Czar? The Brachot On Sights include the blessing for “seeing a king or queen . . . whether they are Jewish or non-Jewish”, and the commentary says that in fact it’s a good deed to go see kings if one has the chance, because this will help one distinguish between regular kings and the Messiah. Related: the blessing for seeing a crowd of 600,000 or more people (different blessing depending on whether they’re Jews or Gentiles). Also, the blessing for seeing a weird animal.
9: The Saami Council (Saami are the far northern Scandinavian people formerly known as Lapps) demands that the gaming company behind Final Fantasy 14 remove traditional Saami clothing from their game (example below):
Aside from the wokeness angle, I find this to be an interesting intellectual property question; the Saami say that EU law gives them IP rights to their clothing. If gaming companies used an outfit trademarked by some fashion company without permission, I think the fashion company would be legally in the right to demand the game remove it, so I guess this hinges on whether you can consider a culture to be the sort of unit that can trademark things. Companies are allowed to claim rights to any product their employees invent, and I think universities do something similar, so it doesn’t seem like a stretch for a tribe to make a similar demand. I think probably the fair solution is for the US government to trademark every American cultural product (t-shirts! jeans! burgers!) and then tell the Saami they probably don’t want a trade war and we’ll let them use our stuff (for example, draw a picture of a person in a t-shirt) only if they let us use theirs. Plus an extra lump sum bonus payment from them to us for making us go through this annoying process.
11: A few years ago I wrote about attempts to make GPT-2 play chess; it couldn’t consistently make legal moves, but when it did, its moves seemed better than random although still not great. Zack Witten reports playing chess with Bing (either a late GPT-3 or an early GPT-4) and finds it’s much better - he reports consistently legal play with Elo of about 1100 (around the level of an okay beginner who’s stopped being too embarrassing). Other commenters report worse experiences and more illegal moves; I don’t have access to confirm.
12: Jonathan Haidt revisits whether social media is bad for mental health. Previous studies have said no, by lumping together different ages, genders, types of screen time, and types of mental health result. Haidt finds a subgroup where the answer seems to be clearly yes: teenage girls using social media seem more depressed and anxious. I don’t usually like subgroup slicing but he seems to have done a really good job proving that this subgroup does badly across many different studies. He thinks this is because teenage girls are using Instagram and worrying about body image. I wouldn’t have predicted that this in particular would be so much worse than all the other kinds of social media use, but I guess I’m wrong!
13: New Eliezer Yudkowsky video appearance, shouldn’t be a big surprise to anyone who has already read his Death With Dignity post, but these people sure seemed surprised. I’m updating on how useful it might be to spread the word on this:
14: Related: Zvi’s practical (and emotional) guide for people worried about AI apocalypse.
15: Possibly related: Elon Musk appears to express regret for his role in accelerating AI: “It’s quite a dangerous technology - I fear I may have done some things to accelerate it.”
I’d hoped he had learned his lesson, but he’s since announced plans to create an “anti-woke” GPT competitor which might also be more “open”. I stick to my 2015 post on the issue.
16: Continuing discussion around the Cochrane review apparently showing masks didn’t work. I agree with Kelsey’s analysis here (see also continuing Twitter discussion of Kelsey’s analysis). I don’t agree as much with Tomas Pueyo’s analysis here, for the reasons GidMK lays out here. I also stand by what I said in last month’s links (see #45)
17: Public Intellectuals For Charity wants me to advertise their Dystopian Fact Checks series, where various public intellectuals including Steven Pinker, Robin Hanson, Robert Sapolsky, Coleman Hughes, etc will debate each other on Zoom to raise money for charity. See here for details.
18: Jacob Falkovich on Twitter proposes a Law Of Equal Sexual Oppression - in mostly heterosexual/monogamous societies, men and women ought to find dating equally hard/unpleasant/unrewarding. Proof: suppose dating was much harder and less rewarding for one sex; let’s say easy and rewarding for men, but hard and unrewarding for women. Many more men would enter the dating pool, and many women would leave. Now there are too many men chasing too few women; women would be “in demand”, have all the power, and have the option to be much picker, causing dating to shift to norms which are better for women and worse for men, correcting the original imbalance. The only stable equilibrium is the one where both sexes are equally (un)happy!
Jacob adds many important caveats: for example, the unhappiness can take different forms (women have extra fear of being raped, men have the extra burden of having to work harder and face more rejections to get a date). And both sexes might have worse times within their own communities - women might be in female-dominated communities where the few men have it easier, and men might be in male-dominated communities where the few women have it easier.
A female friend counterargues that this isn’t how it works in real life: women in male-dominated communities may have dating market power which in theory they should be able to leverage into female-favoring dating norms, but actually male-dominated communities accrete so many basically-male foundational assumptions that the women can’t even communicate what they want at scale, and the men win out anyway. I’m not sure if this rebuts the Law Of Equal Sexual Oppression or confirms it.
20: Some good writing in the “perhaps wokeness has peaked” genre, for example Musa al-Gharbi: The Great Awokening Is Winding Down:
…and many more at the link. I would urge people to put this in perspective of Christianity, whose power probably peaked during the Middle Ages but which remained the dominant paradigm of culture for hundreds of years afterwards; I don’t think wokeness will last that long, but I think it will have a similar long plateau of “no longer rising, now heavily criticized, but still powerful and important”.
21: Related: Brianna Wu (one of the main people involved in Gamergate on the SJW side) seems to be taking the wokeness-peaking narrative seriously, writing:
I'm pretty good at sensing when political winds are shifting. I am telling you, this is the moment we can shun the most toxic tendencies of the progressive movement. The people with influence are absolutely sick of the hysterical bullshit focused on subtraction, not addition. These are opinions I've kept to myself for a decade. I'm trying to model more effective behavior for the left. I cannot make this change for our movement alone.
Welcome . . . to . . . the . . . Resistance???
22: Sasha Chapin: I Cured My Aphantasia With a Low-Budget E-Course, Self-Therapy, and a Wee Bit of Microdosing. I found this interesting not just for the title claim, but for his reflections on how he (a person who previously had aphantasia) was surprised by the character of mental imagery: “Usually people with aphantasia imagine that visualizing people are really seeing images. Like, when they close their eyes, they don’t just see a black void. But that’s not true! Most people see a black void just like aphantasics do. They just have a sense of an image alongside it, hovering in some imaginary parallel nether-space.”
23: Please stop citing that Swedish study purporting to show that IQ stops mattering after the 90th percentile or whatever! Emil Kierkegaard has a summary explaining the possible statistical missteps, and Cremieux has more information here and (buried in the middle) here.
But I worry that makes it sound like, if you don’t agree those particular statistical decisions are missteps, everything is okay. The actual situation is that study after study after study has always shown a pretty consistent relationship between IQ and income, and nobody cared or talked about it. Now one study finds a slight deviation from that relationship, and it went super-duper ultra-viral, to the point where I saw it posted twice on the SSC subreddit, once on Marginal Revolution, and approximately one million times on Twitter. Many of these people are totally mis-describing the study as showing no relationship between IQ and income - instead of a very strong relationship between IQ and income which deviates from perfect consistency at exactly the point where a common statistical misstep would make it deviate from perfect consistency. I think of this as a great illustration of the problem with science: a thousand studies confirming a point people don’t like can languish in obscurity; one bad study which gets a novel result that confirms people’s preferred narrative will become the only thing anyone ever hears about its entire field.
24: Related: The Mensa Fallacy. Many older studies show that very-high-IQ people are mentally imbalanced, bad at real-world tasks, or unambitious. Newer, more representative studies don’t show this. The most likely explanation: the older studies got their very-high-IQ samples from Mensa, the most convenient place to find lots of very-high-IQ subjects. But within the population of very-high-IQ people, Mensa is adversely selected for non-IQ traits; geniuses who are mentally healthy and ambitious are probably inventing quantum computers at MIT or making bank on Wall Street; it’s the geniuses who can’t make it in regular society who are going to every Mensa to boast about how quickly they can solve Rubik’s Cubes. Any study that uses a Mensa sample to study very-high-IQ people and conclude that they’re defective should be considered suspicious.
25: Last month, California got serious about YIMBYism and used dug up an old law that let them punish recalcitrant cities (almost all cities in the state) by declaring that anyone could build anything anywhere within their limits. How’s it going so far? Mostly not well; there’s such a tight crony relationship between developers and city governments that the developers are voluntarily waiving their right to build things in the hopes of maintaining their good relationship with city governments (who will have power again after things are normalized). The few applications that have come in have mostly been from loose cannons and revenge-seekers, like this “young homeowner and computer engineer, frustrated that the town has been slow to permit the reconstruction of his water-damaged home, [who] has announced plans to use the zoning holiday to build a 15-unit apartment complex and five townhomes on his property”.
26: I’m trying not to be a conspiracy theorist, so can someone give me an innocent explanation for this? Clinton aide linked to Jeffrey Epstein found tied to a tree, dead of gunshot wound, with no gun anywhere in sight, Arkansas police rule it a “suicide”. [EDIT: Thanks Alex, gun was found after all and details were much less suspicious than early articles described]
A 1500 page screed against FDR, God, and non-BDSM sexual relationships . . . There is no way whatsoever that this should work. It breaks every unspoken rule of not just writing, but decency. And yet it works! She sold 37 million copies! [...]
There are two possible explanations for Rand's success. The first is that her politics are just that compelling and her philosophy that overwhelming in its logic (they're not). The second is that her prose is just that good. That she's the American Tolstoy or Tolkien.
The main colors are, from top to bottom (left to right):
Blue - data or information
Red - passion and strength
Green - health
Yellow - light (in a sense I’m representing the whole “a computer generated” thing by drawing on a set of colors.)
The secondary colors are:
White - non-hierarchical, undetermined information
Grey - machine intelligence
Black - machines that have transcended humanity or something
The pattern is also intended to reference the “computer generated” thing in a sense – for instance, the white color has an element of the “I made this because of a technicality, not because I thought it was a good idea.”
30: Facebook And Instagram Are Testing Selling You Bluechecks For $12 A Month. Musk’s two highest-profile Twitter changes - firing lots of people and selling bluechecks - seem to be going well and even getting adopted by other companies. I can see a case for everyone apologizing and agreeing Musk is doing a good job with Twitter in a year, although prediction markets haven’t shifted much since December.
31: The Big Mac subsidy - one of my least favorite types of argument is when people argue something is bad because “you think it costs $1, but actually without government subsidies it would cost $100, you’re literally demanding that taxpayers subsidize your evil lifestyle”. And then you ask them for proof, and either they completely made it up, or by “government subsidy” they mean “the government doesn’t force the companies that make it to address all harms caused by its consumption, including harms no other company has to address, and harms I just made up right now, and harms from the fact that it makes me sad and I price my sadness at one billion dollars”. Anyway, the people who say that an unsubsidized Big Mac would cost $13 are definitely doing all of this.
32: The Potato Paradox (not really a paradox): “Fred brings home 100 kg of potatoes, which (being purely mathematical potatoes) consist of 99% water (being purely mathematical water). He then leaves them outside overnight so that they consist of 98% water. What is their new weight? . . . The surprising answer is 50 kg.”