Vinamr

This, exactly: "It doesn’t matter how accurate or useful the information [in AI-generated PR descriptions] is, I’m instantly put off by the thought that it will take me longer to read than it took the driving author to write... even when that driving author is me!"

Singapore's 1966 Land Acquisition Act ≈ land acquisition at *market-value* + a land-specific capital gains tax (since, in expectation, an x% capital gains tax = a 100% capital gains tax applied on x% transactions selected randomly). More efficient than what I thought it was!

Using two independent classifiers, accuracy can be estimated without knowing the truth! When applied to trials with verdicts from both juries & judges, it's found that judges may be slightly more accurate, while juries may give fewer wrongful convictions. ipr.northwestern.edu/our-wor…

One reason why some populations are more violent may be that they're/their ancestors were organized into segmentary lineages -- kin-based institutions that bind clans via honor & vengeance obligations Explains chronic conflict in Africa, recruitment into Islamist terrorist groups, etc. From @JoHenrich's The WEIRDest People in the World

"[T]wo facts about the world explain the anger we see [against Indians]. 1. People have a preference for light skin. 2. The preference for lighter skin can be compensated for with a kind of masculine charisma, which blacks have and Indians don’t."

mattbruenig.com/2023/02/16/t… Since TFR is the average number of births women of all ages would have if current rates held for a lifetime, delays in childbearing make TFR fall even if lifetime births don't. ourworldindata.org/grapher/t…

The "tall buildings & LED lights = higher GDP" fallacy is rampant amongst Indians who believe China is much further ahead of India than it actually is (especially, but not only, if you account for overstatement).

AI exposure not/negatively correlated with labor market distress. Most-exposed workers less likely switch to less-exposed jobs than they were before the gen AI boom. => The anticipated big AI job displacement hasn't happened (yet). xitter.gq/vinamrsachdeva/status/…

.@psychosort: "... when you're looking at … a wrestling match … and you see their bones stick out, you don't need to be a doctor to tell that they're injured …" @robinhanson: "There's this old principle in decision analysis that says if you've got two options and you're finding it really hard to pick between them, it probably doesn't matter so much. … Unfortunately, I think academia works as an institution for credentialing impressiveness and allowing people to associate with credentialing impressiveness. That means people look for the hardest things they can find to do, and do those, even if the hard things aren't much more useful, or even any more useful, than easy things." open.spotify.com/episode/3OP… (relevant discussion from ~15:55 to ~24:38)

Fully binding arbitration would actually nullify a lot of gov't regulation (esp the kind that's usually bad/not needed imho), so gov't keeps the power to overrule it in many cases (esp those that help nullifying regulations)

.@robinhanson's idea to increase fertility: allow parents to buy their child's debt/equity (if prediction market odds of child's suicide w that deal r <x%) to pay for costs & an alt that brings the gov't in the middle bc people find that less offensive! open.substack.com/pub/overco…

A declining population harms not only those who are never born, but also others, because those who would have been born could have produced new knowledge that increases overall productivity. web.stanford.edu/~chadj/empt…

Retrospective voting now: vote to re-elect the incumbent if you're doing better than expected during their term, else replace Alternative: candidate with the highest prediction market odds of being approved by voters at the end of their term serves! slatestarcodex.com/2020/06/1…