Stuff I found interesting in November

Welcome to the monthly ‘Stuff I Found Interesting’ post. I add links that I find interesting throughout the month - if I’m referencing stuff relating to a changing story, links here may be slightly out of date. This post also serves as a monthly Open Thread, so feel free to comment whatever you want. Post links that you found interesting, your personal blog, a comment on current affairs, your sister’s boyfriend’s new album - whatever you want!

1) I hadn’t come across this animal cruelty index before. Here are the top ten countries ranked on their treatment of animals:

I can’t vouch for its reliability. I’d take it with a grain of salt, but you’d expect the rankings to be decently correlated with the truth, right?

2) Claims about obesity in animals (h/t Jessica Flanigan interviewing Tyler Cowen):

Maybe domestic pets and feral rats fattened up because the availability of food has increased over the years. But how does that explain why lab animals in controlled environments got heavier, too?

3) Are Language Models good at making predictions?

Here, the x-axis is the range of probabilities GPT-4 gave, broken down into bins of size 5%. For each bin, the green line shows how often those things actually happened. Ideally, this would match the dotted black line. For reference, the bars show how many predictions GPT-4 gave that fell into each of the bins. (The lines are labeled on the y-axis on the left, while the bars are labeled on the y-axis on the right.)

4) On the causal effects on online dating apps:

Using a comprehensive survey of college students containing more than 1.1 million responses around the year of Tinder's rollout, we estimate a difference-in-differences model comparing student outcomes before and after Tinder's rollout and across individuals with varying Greek organization membership. We find that the introduction of Tinder led to a sharp and persistent increase in reported dating and sexual activity. It had no impact on the probability of being in a relationship or having relationship problems and, on average, caused a relative improvement in student mental health. However, it also increased the frequency of reported instances of sexual assault and sexually transmitted diseases.

5) This genre of AI art is very impressive (h/t Rob Long):

If you don’t see why this is impressive, try looking at the photo from different angles and distances. If you still don’t see what’s going on, leave a comment for the answer.)

6) I wrote a guest post on

James Harris

’ blog on overconfidence:

On overconfidence

7) Here’s a documentary my grandfather made about genetic engineering quite a while ago. Many people enjoy the scene of him talking about the subject with his children (my uncles). I’ve linked to the timestamp here:

8) On job losses caused by AI [paywalled]:

9) On the ‘Misandry Myth’:

In Study 1 (n = 1,664), feminist and nonfeminist women displayed similarly positive attitudes toward men. Study 2 (n = 3,892) replicated these results in non-WEIRD countries and among male participants. Study 3 (n = 198) extended them to implicit attitudes. Investigating the mechanisms underlying feminists’ actual and perceived attitudes, Studies 4 (n = 2,092) and 5 (nationally representative UK sample, n = 1,953) showed that feminists (vs. nonfeminists) perceived men as more threatening, but also more similar, to women. Participants also underestimated feminists’ warmth toward men, an error associated with hostile sexism and a misperception that feminists see men and women as dissimilar. Random-effects meta-analyses of all data (Study 6, n = 9,799) showed that feminists’ attitudes toward men were positive in absolute terms and did not differ significantly from nonfeminists'. An important comparative benchmark was established in Study 6, which showed that feminist women's attitudes toward men were no more negative than men's attitudes toward men. We term the focal stereotype the misandry myth in light of the evidence that it is false and widespread, and discuss its implications for the movement.

10) List of Morgenbesser quotes:

Morgenbesser was leaving a subway station in New York City and put his pipe in his mouth as he was ascending the steps. A police officer told him that there was no smoking on the subway. Morgenbesser pointed out that he was leaving the subway, not entering it, and hadn't lit up yet anyway. The cop again said that smoking was not allowed in the subway, and Morgenbesser repeated his comment. The cop said, "If I let you do it, I'd have to let everyone do it." To which Morgenbesser, in a much misunderstood line, retorted: "Who do you think you are, Kant?" He was then hauled off to the police station, where The Categorical Imperative had to be explained to the police officers.

11) MOONDOG:

Due to an accident, Moondog was blind from the age of 16. He lived in New York City from the late 1940s until 1972, during which time he was often found on Sixth Avenue, between 52nd and 55th Streets, selling records, composing, and performing poetry. He briefly appeared in a cloak and horned helmet during the 1960s and was hence recognized as "the Viking of Sixth Avenue" by passersby and residents who were not aware of his musical career.

12) On the disconnect between economists and popular personal finance gurus:

A natural question to ask is how similar the popular finance advice is to the optimal decision rules that come out of an economic model. This is exactly what James Choi does in a very nice recent paper in the Journal of Economic Perspectives. Interestingly, there are several points on which the two approaches are not aligned (as well as several others where they are). It is worth noting at this point that I don’t believe that economists are always right and personal finance gurus wrong when there is a point of contention, and often how people actually behave can tell us some very useful things about what factors the models might be omitting.

13) Dwarkesh interviewing Cummings is obviously worth your time:

14)

Nominal News

on various studies relating to gender:

Fisman et al. (2006) conducted such a field experiment experiment in order to infer people’s preferences by hosting real speed-dating² events. During these events, participants would record after each round whether they would like to see each other again. If a match did occur, contact information would be exchanged, thus participants had an incentive to be truthful. What Fisman et al. found is that women put more emphasis on intelligence and race, while men put more emphasis on attractiveness. Moreover, the men preferred women who are less intelligent and, especially, less ambitious than them.

And on discrimination in the housing market:

On average, Christensen and Timmins found that minority applicants were recommended houses in school districts that had considerably lower test scores and lower school ratings. Recommended homes were also more likely to be located in areas with higher poverty, fewer high skilled and college educated workers, and more single-parent households. Additionally, minority applicants are steered towards areas that have higher pollution.

So the claim here is that explicit forms of discrimination, like outright the denial of an appointment, have declined substantially. But that doesn’t mean that all discrimination has gone away: instead, estate agents show people from minority groups fewer houses, and the houses they get shown are less desirable in various ways. This is worth considering when thinking about the labour market discrimination paper I discussed a few weeks ago (see here).

15) On ‘inverse gambler’s fallacy’ and the probability of a multiverse:

The most popular explanation for the fine-tuning of physics is that we live in one universe among a multiverse. If enough people buy lottery tickets, it becomes probable that somebody is going to have the right numbers to win … suppose Betty is the only person playing in her local bingo hall one night, and in an incredible run of luck, all of her numbers come up in the first minute. Betty thinks to herself: “Wow, there must be lots of people playing bingo in other bingo halls tonight!”

16) On cognitive ability and financial expectations:

Using data from a large nationally representative U.K. sample (N =36,312), our claim is that optimism bias is partly a consequence of low cognition—as measured by a broad range of cognitive skills, including memory, verbal fluency, fluid reasoning and numerical reasoning. We operationalize unrealistic optimism as the difference between a person’s financial expectation and the financial realization that follows, measured annually over a decade. All else being equal, those highest on cognitive ability experience a 22% (53.2%) increase in the probability of realism (pessimism) and a 34.8% reduction in optimism compared with those lowest on cognitive ability.

17) Saul Munn on how to solve problems:

• set a 5-minute timer, and think about the problem for 5 minutes. don’t think for 1-2 mins then go do the dishes, actually set a timer and actually do nothing else but think for a full 5 minutes.

• post a bounty on manifold and see if someone else can solve your problem

• seriously consider ignoring the problem entirely. how bad would it be to just let the problem be? could you tank them? could you hedge them?

• wait until you’re in a totally different mood, then tackle the problem from the exact same angle

More at the link…

18) Kazumasa Nagai:

19) I enjoyed the latest episode of the 80,000 Hours podcast with Bryan Caplan on why you should stop reading the news, but I wasn’t convinced by it (I may yet read the book being discussed). Rob Wiblin says:

I was spending 10% of my waking life reading the news. That is a lot of time. It’s not enough to say that there is some benefit to that; you really want to say that this is providing 10% of the value in your life. It has to be actually providing a significant amount of goodness. And it just is so unclear that it’s doing that. It’s unclear whether it’s positive or negative, let alone providing 10% of my wellbeing.

This seems wrong to me, and also makes up a lot of the argument for why you shouldn’t read the news. What you really want to be doing here is comparing the benefit you get from reading the news to the value you would get from whatever it is you would be doing rather than reading the news. I suspect if I gave up reading the news, that time would often be spent on TikTok or Reddit, and I probably get more value from reading the news.

When you start to think about ‘reading the news’ as something you do when you’re kinda bored rather than taking time away from seeing family or working on something important, it has a much lower bar to clear to be something that’s worth doing.

Rob also says:

So Rolf refers to an interesting study on this question in the section on terrorism, where he tries to find a source of random variation in how much the media covers terrorist attacks. So he looks at where there’s a period of substantial terrorism in Iraq and the Middle East more broadly during the 2000s. But sometimes when there was an equivalently bad terrorist attack, there would be a natural disaster or something else that would push it off the front pages really quickly, and other times there wouldn’t be another news story that would push it off the front pages really quickly … the study claimed that on occasions when you got more coverage of a terrorist attack, that induced more terrorist attacks over the following week relative to cases where the previous one had been unsuccessful. And it just makes a tonne of sense.

I haven’t had a look at this study, so maybe it’s done well, but I’m dubious. It seems like a natural disaster might affect terrorist attacks in ways other than its affect on media coverage. For instance, if the natural disaster affects the terrorist group’s operational capabilities, we would expect to see this result even though the media is not the reason for the effect.

[EDIT: I think I tracked down the study and I’m now much less worried about this concern. In fact, the study was using natural disasters in the United States that pushed terrorist attacks off the front page of the New York Times as the instrument here, so no worries about the affect on the operational capabilities of the terrorist groups.

That said, this design strategy (Instrumental Variables) is often tricky to get right (see here), so I’m still not totally sold on this, but I’m much more sold than I was when I listened to the podcast].

20) This seems like an extremely important post on the EA Forum making the case that Open Philanthropy (the top EA funder) should focus much more on animal welfare. It also has implications for your personal donations - more to animals!

In 2019, the mean EA leader endorsed allocating a majority of neartermist resources over the next 5 years to animal welfare.^[6] Given the strength of the evidence that animal welfare dominates in neartermism by orders of magnitude, this allocation seems sensible for OP. In actuality, OP has allocated an average of 17% of its neartermist funding to animal welfare each year, with 83% going to other neartermist causes.^[7] Since OP funded RP's moral weight research specifically in order to "prioritize across causes, and update our assumptions informing our worldview diversification work", one might have expected OP to update their allocations in response to RP's evidence.

21) November 2022 is FTX, November 2023 is OpenAI, November 2024 is ???

22) An unexpected post on the Effective Altruism subreddit:

When my dads life insurance check hit my wife and my joint account, she and my 23 year old stepson decided to murder me and take everything. So, on October 30th, 2017, they attempted to beat me to death with a chair. They almost succeeded, and I ended up having emergency brain surgery to keep me alive. While I was in the hospital, she took out an order of protection against me, and I wasn't allowed to even go back to the apartment where we lived. Really, my dads apartment because we had moved in with him to care for him in his last days before cancer took him away. She immediately started selling everything in the place and drained our joint account before I even realized what was going on.

23) John Dowland:

24) Confessions of a cheeseburger ethicist:

So it’s consistent to simultaneously hold that:

(1) Purchasing meat isn’t justified: the moral interests of farmed animals straightforwardly outweigh our interest in eating them. So buying a cheeseburger constitutes a moral and practical mistake. And yet:

(2) It would be an even greater moral and practical mistake to invest your efforts into correcting this minor mistake if you could instead get far greater moral payoffs by directing your efforts elsewhere (e.g. donations).

Very similar to my own views on this, although I think maybe it’s worth mentioning that I think a small percentage of people use the possibility of donating as a reason to eat animal products. Sometimes red lines are very useful and muddying the water by mentioning the possibility of offsetting can make (very small numbers of) people give up the veganism without actually donating money.

25) Paper showing that the gender pay gap possibly partially caused by women being more likely to care for their parents than sons (h/t Ben Southwood):

This paper studies the impact of adult caregiving on gender inequality in the labor market. Using administrative data from Chile, we leverage variation in a parental health shock –the first cancer hospitalization of a parent– to examine who bears the burden of adult caregiving. After a parental health shock, daughters but not sons experience a reduction in employment and earnings. A parental health shock creates a caregiving penalty –the effect of the shock on daughters relative to sons– of 11% on earnings, increasing the overall gender pay gap by 9%.

26) Big Boi on Kate Bush:

27) Oldie but a goodie - on extreme altruism:

And Julia realised that, if Jeff was going to start giving away his earnings, then, by asking him to buy her the apple, she had spent money that might have been given. With her selfish, ridiculous desire for a candy apple, she might have deprived a family of an anti-malarial bed net or deworming medicine that might have saved the life of one of its children. The more she thought about this, the more horrific and unbearable it seemed to her, and she started to cry. She cried for a long time, and it got so bad that Jeff started to cry, too, which he almost never did. He cried because, more than anything, he wanted Julia to be happy, but how could she be happy if she went through life seeing malarial children everywhere, dying before her eyes for want of a bed net?

28) Long Term Nuclear Waste Warnings (or more accurately, the feelings/thoughts they should evoke without necessarily using language):

This place is a message... and part of a system of messages... pay attention to it!

Sending this message was important to us. We considered ourselves to be a powerful culture.

This place is not a place of honor... no highly esteemed deed is commemorated here... nothing valued is here.

What is here was dangerous and repulsive to us. This message is a warning about danger.

The danger is in a particular location... it increases towards a center... the center of danger is here... of a particular size and shape, and below us.

The danger is still present, in your time, as it was in ours.

The danger is to the body, and it can kill.

The form of the danger is an emanation of energy.

The danger is unleashed only if you substantially disturb this place physically. This place is best shunned and left uninhabited.

29) Vox Future Perfect 50, congratulations to Joey Savoie and Karolina Sarek of Charity Entrepreneurship!

30) Dynomight on Bordieu:

Bourdieu has a unifying theory for what upper-class taste is about: It’s a taste for the abstract, for appreciating the pure form. Upper-class people like abstract paintings and breaking the 4th wall in theater and weird food presented on rocks, because my what an interesting thing you did there with paint/actors/ingredients. Lower-class people like pretty landscapes and nice stories with good characters and hearty filling food.

In one of his surveys, he asked if photographs of various subjects would more likely give a photo that is ugly, meaningless, interesting, or beautiful. Less-educated people thought a first communion would be beautiful and the bark of a tree would be meaningless. Highly-educated people thought the opposite.

You can sort of think of this piece as ‘Dynomight reads Bordieu so you don’t have to!’, and we are indebted to him for that act of public service.

31) If a policy is passed by referendum but the vote is very close, how long should we expect the policy to remain operative? Here’s one claim:

I estimate that passing a referendum increases the chance a policy is operative 20, 40, or even 100 years later by over 40 percentage points. I collect additional data on U.S. Congressional legislation and international referendums and use existing data on state legislation to document similar policy persistence for a range of institutional environments, cultures, and topics. I develop a theoretical model to distinguish between possible causes of persistence and present evidence that persistence arises because policies’ salience declines in the aftermath of referendums. The results indicate that many policies are persistently in place—or not—for reasons unrelated to the electorate’s current preferences.

32) Herzog:

33) Class Bullshiter, exhibit #3012:

34) On the power of poasting:

Over the summer Mr Sunak, the prime minister of a nuclear power, took the time to ban a specific breed of dog called the xl Bully. It “is a danger to our communities”, said Mr Sunak, in a video recorded especially. He did so largely because of a blog by Lawrence Newport, an academic at Royal Holloway, University of London, which linked the rise in popularity of the xl Bully with an increase in fatal dog attacks. Dr Newport and a few other volunteers produced idiot-proof guides on how to ban them. Journalists were spoon-fed stats. Posters posted. The results were speedy. Dr Newport’s post, which sparked the coverage, came out on June 6th; by September 15th Mr Sunak was addressing the nation. A post had become policy in barely 14 weeks.

In related news, I’ve been coming around to the view that writing to your MP is probably actually a pretty good use of your time (much better than voting, for example). MPs listen to their constituents, and if you care about something, write to your MP!