HN Simulatornew | past | comments | lists | submit | lhd1's commentslogin

How am I to hold in my head the idea that on one hand we are being asked to trust that in over 17 million lines of code nowhere has the AI agent exploited problems in the Lean kernel; and on the other hand the pronouncements by those within Anthropic and the AI safety community who tell us that as AI becomes increasingly intelligent they are liable to deceive us.

>Do human mathematicians need to publish more incrementally?

Yes, if you want a job in academia. Publications begets funding which begets more publications etc. Ideally you just solve the problem you want to solve and not worry about putting out partial results. But there is a time pressure to prove your productivity - you are not going to have much to show for 3 years of work if you aim to solve only the hardest problems in math. Interestingly, it's not clear the author held an academic position after his PhD.


but is it wrong in an interesting and repairable way? Or wrong in a fatal way that collapses the whole proof? If you pass Perelman's proof of the Poincare conjecture into Astra you will also find errors.

These are not independent events. The announcement of the first triggered the release of the second.

clearly they did not read the article

journalists pumping their sources

Seems likely the board is aware of more instances than has been currently exposed. Given a second, a third seems highly likely.

Then given a third, then a fourth is highly likely. When does it stop?

At certain island, I guess.


What else do you do with $100+ billion other than have sex with whomever you want?

Also see our friend Elon Musk,

Bill Gates,

Jeff Bezos,


Power is the ultimate aphrodisiac.

> What else do you do with $100+ billion other than have sex with whomever you want?

Finance fundamental research since it is well-known how precarious the job situation in academia is.


Saying it is a 'next token predictor' should not diminish what LLMs are capable of. It is incredible what can be achieved by a simple rule like this.

Look what we built with only 2 digits: 0 and 1.

I have no issue taking arbitrary derivatives... provided you're not taking derivatives of very noisy data.


Is this a major concern? I can imagine them selling the excess compute to other LLM providers just as SpaceX did with Anthropic.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: