HackerLangs
TopNewTrendsCommentsPastAskShowJobs

andriy_koval

180 karmajoined 4 年前

comments

andriy_koval
·5 小時前·discuss
My bet they run gpt over dataset of 10k unsolved conjectures, it happened this one was solvable.
andriy_koval
·10 小時前·discuss
> to me and to many friends in their 30-40s, using AI models to achieve something we used our brains to achieve feels... empty?

you have to start using AI to achieve something which was way more challenging before, then you will feel lots of fulfillment and inspiration.
andriy_koval
·10 小時前·discuss
> Can someone explain how it in any way indicates how close we are to "AGI"?

I think it is historical name. At some point when benchmarking was very undeveloped, this was targeting abstract reasoning and generalization, hence AGI.
andriy_koval
·前天·discuss
sure, million dollar company can be automated using AI.
andriy_koval
·前天·discuss
[dead]
andriy_koval
·前天·discuss
you can spend some time researching the process, so you can come to better quality judgement.
andriy_koval
·前天·discuss
Big part of problem is that many companies hire foreigners not because it is "easier", it is not, but because they pay them little while holding in visa slavery.

On top of that, the way they hire foreigners and provide permanent residence path to them is straight fraud: they put job ads into some local newspaper no one reads to prove there is no local competition for the opening.
andriy_koval
·前天·discuss
Depending on scale of your products you may find that single person won't be able to cover all aspects. "Prompting" requires some domain knowledge and also quality control and guidance.
andriy_koval
·前天·discuss
My speculation is that frontier models are MoE, and they just have some number of experts for coding.
andriy_koval
·前天·discuss
> Today my "coding" sessions often enough begin with real life problems

intuition is that your sessions consists of 10% of domain related reasoning, and 90% of code plumbing. Those 90% could be moved to cheap and efficient specialized and focused model.
andriy_koval
·前天·discuss
I agree is that money is an issue, but not geo-boundaries. If EU decides to commit XXXB they may have a chance to catch up eventually.
andriy_koval
·前天·discuss
> The H-1B serves mainly to retain workers already here, not to recruit them from outside

that article looks like full of bs
andriy_koval
·3 天前·discuss
My bet is that in some industries like banking, PG share is minuscule.
andriy_koval
·3 天前·discuss
meta researchers from eu will create new company and build model using knowledge they have.
andriy_koval
·3 天前·discuss
> Not simply "don't want to live here", usually also "can't, there are no opportunities for income here".

also likely very underdeveloped infra
andriy_koval
·4 天前·discuss
it may not worth to pay you 200k if someone can also do that prompting..
andriy_koval
·4 天前·discuss
"prompting" is a skill, you pay for that.
andriy_koval
·9 天前·discuss
> Is this article trying to make me feel bad for these people?

Its so loaded question.

You picked two cases, while ignoring huge flaws of the system, because of which there is a chance innocent people get executed. Frank Athen Walls is likely very different person today compared to what he was 40 years ago, what is the rational to execute him now?
andriy_koval
·10 天前·discuss
I think runner knee was a bad example.

Sprint running is more sensitive to form in order to avoid injuries, especially in older adults.
andriy_koval
·12 天前·discuss
many people think opus 4.6 was the best