Best of Twitter week of December 2-8, 2024

Fascinating paper on where 6000 global elites went to college. Billionaires, CEOs, heads of state, central bankers, etc.

In a word: Harvard.

Fully 10% of global elites went to Harvard. Elite US schools are over-represented (23% IvyPlus), but nobody comes close to Harvard.

🧵

— Paul Novosad (@paulnovosad)
7:06 PM • Dec 6, 2024

Bottom line is that the Oxford of 1920 "did things" that the current elite American universities evidently do not.  The English produced men we don't produce.

The USA shows a profound failure to develop, identify, and employ persons who are highly capable at dealing with the… x.com/i/web/status/1…

— Adam Van Buskirk (@AHVanBuskirk)
1:27 AM • Dec 5, 2024

The obvious reason for this is elite Oxbridgians could be trusted by the state because they were literally titled aristocrats or their chums with surnames that appear on military death rolls from a thousand years prior.

The USG does not have similar implicit trust of Ivy grads.

— Lyman Stone 石來民 🦬🦬🦬 (@lymanstoneky)
7:26 PM • Dec 5, 2024

x.com/i/article/1864…

— Paul Han (@pauljunsukHan)
9:09 AM • Dec 4, 2024

— digitronn2 (@digitronn2)
1:36 PM • Dec 1, 2024

this isn't hyperbole

— Aleph (@woke8yearold)
1:11 AM • Dec 7, 2024

Anyone who thinks AGI is imminent should be forced to fine tune an ML model on an advanced task for a week using a new training library.

This will disabuse them of that notion faster than trying to build a skyscraper with a hammer, boards and nails.

— Daniel Jeffries (@Dan_Jeffries1)
3:28 PM • Dec 6, 2024

In my opinion we have already achieved AGI and it’s even more clear with O1. We have not achieved “better than any human at any task” but what we have is “better than most humans at most tasks”. Some say LLMs only know how to follow a recipe. Firstly, no one can really explain… x.com/i/web/status/1…

— Vahid Kazemi (@VahidK)
9:04 PM • Dec 6, 2024

Things like this detract from the credibility of AI safety work, IMO -- it sounds spicy ("o1 tried to escape!!!") but when you dig into the details it's always "we told the robot to act like a sociopath and maximize power, and then it did exactly that".

— Nabeel S. Qureshi (@nabeelqu)
7:43 PM • Dec 5, 2024

A while back I worked w a PhD who had previously worked in a biochem lab. At the lab he had picked up a habit of never touching door handles if possible. I picked this up from him.

The rate at which I caught colds fell by 50%.

— ⓘ Dogs don't have thumbs (@MorlockP)
12:37 PM • Dec 5, 2024

Insurance companies are sin eaters for hospitals. They let hospitals maintain the image that they're righteous non profits and should also be allowed to charge whatever they like.

— Ryan Moulton (@moultano)
3:59 PM • Dec 5, 2024

breakthroughs of 2024

— gavin leech‎ ‎ ‎ (@g_leech_)
4:41 PM • Dec 4, 2024

The cinematic quality in the paintings of Henri-Paul Motte (1846–1922)

— Federico Italiano (@FedeItaliano76)
3:47 PM • Dec 4, 2024