Hacker News new | ask | show | jobs
by azan_ 19 days ago
AI agents that can solve frontier math problems, something that few years ago was decades away.
3 comments

> something that few years ago was decades away.

There's no way of knowing this - I see articles fairly often on HN of mathematicians (sometimes grad students or younger) solving problems where progress previously had stalled.

What I meant was not that these problems wouldn't get solved for decades, but that few years ago (before advent of LLMs) if you've asked average researcher how far away are we from AI solving unsolved math problems, the median answer would be that we are far, far away from that.
If that's how your comment was meant, it seems an odd reply to the parent comment, which is firmly critical of AI's impact on society, not really debating about the progress of the technology.
I suggest you read the parent comment again. The poster clearly was laughing at the capabilities of AI and its utility:

> the value in having a personal assistant AI agent that can lie about the time for your appointment and autonomously delete your entire calendar

Sure, but 1) they're clearly being quite facetious with the above, and 2) it is also pretty clearly contextualised by the rest of what they said, which is making an equally sarcastic comment about how much AI is costing society.

The overall template is "look at how much we're giving away, but hey guess it's worth it for <insert stupid reasons here>".

Instead of immediately listing off whatever instance AI has shown some value in isolation, it might be worth considering its net effect on society as a whole. IMO that picture is not so rosy.

Thank god they can do it now! I'm willing to add thousands more to my bills, I'm sure AGI is around the corner and will make life so much easier.
You don't need AGI. If AI progress stopped right now, LLMs would still be amazing and extremely useful technology. It already makes life for many much easier. But it's easy to miss it when you are entombed in anti-AI bubble. But I've got something that may placate your fears - remember that horses did not vote.
It’s certainly easier to have a computer do all the thinking for you.
it definitely is if you were stupid to begin with

LLMs are a weird phenomenon. as always as of late, the sci-fi predictions are outdone by reality. Idiocracy seems cute.

If current LLM can do all the thinking for you then I'm terribly sorry. I use it mostly to do the boring and tedious tasks.
> You don't need AGI. If AI progress stopped right now, LLMs would still be amazing and extremely useful technology. It already makes life for many much easier.

No? I don’t? Because that’s the bright future they’re trying to sell, after they kick you out on the street.

> But I've got something that may placate your fears - remember that horses did not vote.

Neither did serfs. And that’s what they’re trying to establish. Also, very generous of you to believe that anyone outside of the tech bubble, or hell that majority INSIDE of the tech bubble, will lift a finger to defend the tech workers. Just look at how controversial a simple topic like work union is. It’s dog eat dog out there, and tech feudal lords know that they can just buy the most vocal ones.

Well, that's surely worth sacrificing people's livelihood for.
Yes, same as industrial revolution was worth sacrificing people's livelihood, because in the end we are much better off.
Should I take this as you volunteering your livelihood? Otherwise this comment rings incredibly hollow. It's very easy to say others should sacrifice what they've worked their whole life for, but it's not so easy to give it up yourself. If you truly believe it's worth it, though, you should be eager to do so.
> Should I take this as you volunteering your livelihood? Otherwise this comment rings incredibly hollow. It's very easy to say others should sacrifice what they've worked their whole life for, but it's not so easy to give it up yourself. If you truly believe it's worth it, though, you should be eager to do so.

Well yes, I'm not immune to potential displacement, I don't know where did you get it that I'm somehow in safe spot that will never get replaced with AI. And I'm ok with this risk.

Go ahead and quit your job then, if it's worth it. It's for the greater good, after all, and you're more than ok with it.
People losing jobs is price we pay for progress, not goal in itself. I'm really surprised I have to point it out. Hence me quitting my job will have zero positive impact. Programmers for years have automated other jobs and suddenly when their work is in danger somehow automation becomes bad.