Calazon

He won't be punished by the legal system.

Possibly. But even if not, how many AI model developers do we need? We will likely need a nonzero number, but it could be very few compared to the total number of software developers we have today.

Could definitely be I've had the misfortune of working with below-average programmers (and inheriting their codebases).

Yeah I'm really impressed by what Opus 5.5 can manage.

I feel like human code is on a wide quality spectrum. The best human code is way better than what Opus 5.5 can do. But the average code, I'm not so sure.

The way I see it many HOAs are the opposite of privatization, at least when they cannot be opted out of. They control land, impose taxes (sorry, "fees"), have elected leadership, provide services, make rules that they have the power to enforce, and the only way to leave them is to move out of their territory.

In my eyes this is essentially the same thing as a form of local government, just on a smaller scale.

The other problem is that the AI may well do what humans tell it to rather than what they want it to, in the style of a literal genie or paperclip maximizer.

It can lead to catastrophe without ever being conscious or wanting freedom, or anything like that.

Where I live, even when government does make demands, as in the case of foster parenting, they still allow two boys or two girls to share a room.

Presumably some combination of marketing existing models and training/improving future models.

My point was about what is and isn't the responsibility of AI labs. Why should they be required to invest more (or fewer) resources into math than they otherwise want to?

(If you ask me, safety and security are what they need to invest more into.)

The insane part is placing the burden on AI labs to advance mathematical understanding. Why should this be their responsibility?

If they want to simply generate theorems and proofs and post them on the internet, why not? Does doing that really inflict "substantial negative externalities ... on the mathematical community"?

Because it's faster.

I've done this on different topics - I know the answer is in a particular eBook/PDF/document, but for whatever reason it's not trivial to look it up. The model can do it a lot more quickly than I can, and then I can still verify the accuracy.

I wonder how the ratios would change for papers at different parts of the review process. For what fraction of published papers are the authors unable to answer basic questions about them?

That sounds incredibly frustrating. It's bad enough in code reviews, but is that approach to work being tolerated in meetings these days?

They'd be hobbies in most utopian scenarios too.

The problem is that by default progress might look like an increase in power without a corresponding increase in safety. That could be catastrophically dangerous.

There are captchas that demand selfies now? I have never heard of that before and it sounds completely crazy.