Profile avatar
stefanschubert.bsky.social
Effective Altruism and the Human Mind (with Lucius Caviola) is available for free at: https://academic.oup.com/book/56384 For physical and audiobook versions, see: https://stefanschubert.substack.com/p/physical-and-audiobook-versions-of
736 posts 3,189 followers 630 following
Regular Contributor
Active Commenter
Posts 22 Comments 28

Some history takes: You generally learn much more from the recent past than the distant past. Rome is not a good analogy to the US. Still, we underrate prehistory (relative to, e.g. Antiquity). It can teach us about our evolutionary roots. ->

submitted 19 hours ago • 1 comment

Thoughtful article by Ege Erdil of @epochai.bsky.social on AI progress in 2025 epoch.ai/gradient-upd...

submitted 1 day ago • 0 comments

There might be more context to these numbers, but they look striking www.bloomberg.com/news/article...

submitted 1 day ago • 3 comments

Claude is overall not terrible at writing, but it often does something reminiscent of hallucination in the way it structures text. It simply ignores or forgets that some things don't hang together at all, in a way a human writer that's otherwise ~Claude-level wouldn't do. ->

submitted 1 day ago • 1 comment

Study found that people think that deliberative thinking is superior to intuition, and that deliberative thinkers are smarter and more trustworthy than intuitive thinkers. So whatever else it may be, "Blink" is counterintuitive. osf.io/preprints/ps...

submitted 2 days ago • 0 comments

Yes, Democrats need to increase their sense of urgency; that they really need to win. "The path forward for Common Sense Democrats" by @mattyglesias.bsky.social www.slowboring.com/p/the-path-f...

submitted 2 days ago • 0 comments

China’s sex ratio of young adults is predicted to be ~115-120 men per 100 women for decades

submitted 2 days ago • 2 comments

As angry as I am at the Republicans who enabled Trump, I am trying to resist the temptation to attack them when they criticize Trump. We need them to do that.

submitted 3 days ago • 1 comment

Most books are not worth reading, most people are not worth hiring or following, most things are not worth doing. By historical standards, modern society is huge and offers tremendous choice. But an underrated implication of that is you need to up your rate of rejection.

submitted 3 days ago • 1 comment

Changing conspiracy theory beliefs is very hard, but a replicated finding shows a short chat with GPT-4 changes people’s belief in conspiracy theories for the long term. Why? It isn’t rhetorical tricks, it's that AI provides relevant facts and evidence tailored to each person's specific beliefs.

submitted 4 days ago • 8 comments

There's widespread nihilism about Trump's incessant lying. "His statements are not intended to be accurate." That makes it worse, not better.

submitted 4 days ago • 2 comments

Yes, we lose sight of the bigger picture. And we're too negative. www.thenewatlantis.com/publications...

submitted 5 days ago • 2 comments

People take many long-term projections much more seriously than they deserve. Presenting one scenario "with migration" like this is silly. Obviously you can have very different levels of migration, leading to very different population sizes. Ridiculous. www.theguardian.com/world/ng-int...

submitted 5 days ago • 1 comment

Age and class are stronger predictors of UK voting than gender. The EU referendum vote remains a strong predictor (unsurprisingly). (YouGov) ygo-assets-websites-editorial-emea.yougov.net/documents/Vo...

submitted 5 days ago • 0 comments

An extremely common mistake is viewing the EU as a unitary political actor, akin to the US. It's better seen as a collaboration between independent countries than as analogous to the US. And those countries often fail to coordinate (e.g. now re support to Ukraine/defence).

submitted 5 days ago • 0 comments

In light of recent events, some talk about more European integration. While some of that may make sense re defence, overall I think federalists should proceed with caution. The last decades have seen a lot of resistance towards more integration, and the UK even leaving.

submitted 5 days ago • 2 comments

I look at this, at Denmark and Sweden, and then at some bigger #EU states and despair.

submitted 6 days ago • 8 comments

Some argue that AI has already boosted productivity a lot, but that it doesn't show up in the statistics because staff using AI systems simply work less. I'm sceptical that this happens on a wide enough scale to have a big effect. Companies tend to be able to make use of productivity gains.

submitted 7 days ago • 0 comments

People tend to conflate pro tanto-arguments (everything else equal) with all things considered-claims. This is a super common cause of confusion in debates.

submitted 8 days ago • 0 comments

The fraction of US GDP spent on computers and related equipment peaked during the dot-com boom and hasn't recovered to that level epoch.ai/gradient-upd...

submitted 8 days ago • 0 comments

If you can't understand a book, is that because of shortcomings of yours, or because the author is unclear? Stanislav Andreski suggested a test in *Social Sciences as Sorcery*: read Carnap or Russell, or physics, and see if you get it - if you do, the fault is with them.

submitted 8 days ago • 3 comments

Stockholm has rent-regulations, and in some areas the rent is far below 50% of the market price. Flats are allocated based on time in the queue. In the most attractive areas in central Stockholm, the average queuing time for allocated flats is more than 20 years.

submitted 10 days ago • 1 comment