How governments spend far more on old people than on children, the most central chart in AI, the educational splits to the left and the right in British politics, a crucial obstacle to job automation, and much else. theupdatebrief.substack.com/p/government...

I've started a newsletter, The Update. Today I cover how journalists should use social science, EU-US unemployment convergence, the definition of AGI, and more. Please subscribe! theupdatebrief.substack.com/p/the-shrink...

Stefan Schubert (@stefanschubert.bsky.social) has a new blog! Instant subscribe. He's one of the little voices in my head by now when I evaluate stuff I read. theupdatebrief.substack.com

"It has gone largely unnoticed that time spent on social media peaked in 2022 and has since gone into steady decline." By @jburnmurdoch.ft.com www.ft.com/content/a072...

Zurich is the latest city to make parking more expensive for oversized cars like SUVs. Cities including Paris and Montreal have taken similar steps to counteract car bloat.

Most graphs of the fertility rate depict the 'period fertility rate', which is based on a single year's data and doesn't necessarily reflect how many children women actually have across their lifetimes. I've used data from the Human Fertility Database to show the cumulative number instead:

Twitter is a particularly stark example but in general the gap between the public and elites (left+right) on ID cards is huge

I much enjoyed David's new book. The first part of the book describes Singer's life and how the drowning child argument inspired the founding of effective altruism. The second part discusses objections to EA.

"French pensioners now have higher incomes than working-age adults" - extraordinary By @jburnmurdoch.ft.com www.ft.com/content/d419...

Financial Times analysis: US jobs at a high risk from generative AI have not been more likely to shed young workers since ChatGPT launched www.ft.com/content/99b6...

Study finds that Americans' policy views don't relate to basic economic literacy, and that such literacy doesn't make people think like economists. Also if Democrats and Republicans had had perfect basic economic literacy, their policy views would diverge. www.tandfonline.com/eprint/KRKRA...

Ask me anything! I'll be doing an AMA with EA forum today. You can ask me about anything, whether it's about topics I write about (life expectancy, fertility, mortality, global health, data, etc.), recommendations, writing or podcasting, or anything else. Feel free to reply with questions below.

❓Why do the Nordics & Dutch speak English so much better than the Germans, Italians & French? ➡️ New Working Paper: Out-of-School Learning: Subtitling vs. Dubbing and the Acquisition of Foreign-Language Skills w/ F. Baumeister & E. Hanushek www.nber.org/papers/w33984 A 🧵 1/12

Guess the nationality of the unnamed author of this @economist.com article on "the Big Three" 🙃

In the 18th century, there was a real chance of death at any point in life, and there wasn't a big peak in old age. It wasn't just higher infant mortality - the whole distribution was completely different. inquisitivebird.xyz/p/the-rise-o...

Be alert to uses of "by default" or "business-as-usual" in the context of projections of future trends. While they literally mean "in the absence of new policy", they often carry the association that they're likely (cf "usual"). But in fact, they're often highly unlikely. ->

I agree with this. Too often, people get away with this because of misguided norms of charity. And yes, we should focus more on authors (relative to the audience) than we do today when trying to improve discourse norms. www.lesswrong.com/posts/Zmfxgv...

In northern Europe, life satisfaction increases with age, whereas in southern Europe, it decreases. www.nber.org/papers/w33950

One thing outsiders often get wrong about academia is they say "Harvard published a study", "Oxford did such-and-such", etc - analogously to "Google did", etc. In most cases, that's wrong - it's groups of researchers (often at different unis) who do things, not universities.

Anthropic launches "initiative to support research and policy development focused on addressing AI’s economic impacts". www.anthropic.com/news/introdu...

"36% of all Chinese undergraduate entrants picked an engineering degree. In Britain and America the proportion hovers around 5%." Incredible numbers (assuming the definitions are consistent). www.economist.com/china/2025/0...

Do you believe in moral progress? In the latest episode of “Lives Well Lived”, Kasia and I speak with Josh Greene about why he believes that in the long run, the world is improving, despite recent setbacks. From fewer deaths in violent conflicts to more people living healthy, meaningful lives...

Surely that depends on what you love, and how good you are at it

New short post, you shouldn't make claims that assume the average person is really stupid andymasley.substack.com/p/my-are-you...

FDA announces a pilot program to slash drug review times www.washingtonpost.com/opinions/202...

I don't think these kinds of outrage posts over random members of the public are especially helpful www.bbc.com/news/article...

My second @sentientism.bsky.social interview is out! Thanks to @jamiewoodhouse.bsky.social for having me back to discuss The Moral Circle - as well as for giving me this awesome mug in London 🙂 You can watch the interview below, or listen elsewhere. Hope you enjoy! www.youtube.com/watch?v=e9ti...

Review of Seeing Like a State www.goodreads.com/review/show/...

Average bonnet height in new cars in EU+ has increased from 77 cm in 2010 to 84 cm in 2024. It poses a risk especially to children who can be hard to spot from a tall car. www.transportenvironment.org/articles/eve...

It's notable how much better China has done relative to neighbouring countries which also were relatively poor in 1990

Many are asking "if AI will destroy jobs, what new kinds of jobs will it create?" I think this question is often based on the naive assumption that the number of jobs in existing professions is fixed. Restaurant jobs aren't "new" yet have seen a lot of growth in the Internet era.

British universities got by far the most European Research Council grants in data published today. They were especially dominant in the social sciences and humanities, with >30% of grants. Universities in the former communist countries continue to lag behind. erc.europa.eu/news-events/...

Fin Moorhouse interviews @tobyord.bsky.social about the future of AI and the risks it involves. Clear and illuminating. pnc.st/s/forecast/5...

Yesterday I spoke to a Catholic priest who had been rejected from a Swedish university because his grades were in Latin (!) and there was no authorised translator to be found in Sweden. Example of the overly rigid Swedish admission system. (He has a PhD so is clearly qualified.)

New funding for UK progress/pro-science projects

Just from opening the PDF while I wait for my daughter’s swimming class to end, I feel like my favorite argument is missing: sometimes a paper is so bad one review is entirely sufficient for the world to see it’s junk. Right now, these papers are passed on from journal to journal>

guy who's on social media to save the climate

"(i) thousands of new constructs and measures are published each year (ii) most measures are used very few times, and (iii) there is no trend towards consensus or standardization in the use of constructs and measures"

The fall and rise of inheritance flows

I like Scott Alexander's series of posts on the norms of argumentation. I very much agree with this. I find it really annoying when people refuse to concede in this way. www.astralcodexten.com/p/but-vs-yes...

Many people complain about management consultants as shallow sellers of platitudes, but that's not my impression. I know several people who went from philosophy to top consultancies, and they are without exception really impressive. (Also, they liked it.)

This o3 analysis cited by Marginal Revolution seems like a big overestimate. In particular, I think it hugely underrates the "deployment friction" (which is part of why o3 has in fact not replaced all these tasks). I'm puzzled why this type of analysis is seen as so impressive..

A colleague just sent me this: an upscale flower shop in Østerbro, Copenhagen—no staff, just grab what you want and pay on your own. That's the definition of a high-trust society right there!

Top-tier sophistry

Really interesting, very in-depth on how open science and replication efforts were funded and gradually grew more influential. Great example of what thoughtful philanthropy can accomplish. goodscience.substack.com/p/metascienc...

There's an important difference between students using AI and people using AI as part of their regular jobs. Students need to demonstrate they personally know the material, whereas workers just need to get the job done in some way or another. And marking is a (part of a) regular job.

It's a myth that only pensioners want pensioner-friendly policies. It's tied to other myths, such as that we live in a gerontocracy and that voters only vote for their narrow self-interest. I think it's bad policy, but the conventional analysis of it is no good. yougov.co.uk/topics/polit...

We have a lovely new paper out, showing that even ignorant people can be very good at recognising who's knowledgeable based on minimal cues

Reading old history (e.g. 16-17th Europe), it's notable that people often talked about life (not just their life, but life in general) being miserable; and this was a motivation for religious feelings. That seems much rarer now. Possibly speaks against full hedonic adaptation.

"Beyond benchmark scores: Analyzing o3-mini’s mathematical reasoning" This more qualitative research seems informative and useful epoch.ai/gradient-upd...