Original post unavailable.
Sentiment
Many users are optimistic that AGI will skyrocket employment by creating new problems to solve like space travel, while others worry about mass unemployment, LLM psychosis, and harsh economic effects on most countries.
Pos
55.6%
Neg
44.4%
9 comments with sentiment.
Cluster Engagement
Digg Deeper
No Digg Deeper questions have been answered for this story yet.
Related links
Posts from X
Most Activity
Most Activity
VIEWS44KBOOKMARKS142LIKES937RETWEETS54REPLIES124
I've always been skeptical of the rosy picture painted by some people of what post-AGI employment will look like. In fact, I've said before I think it's going to hit human employment like a meteor. However, if what's going on with math right now is any indication, we are going to need way, way more experts in every field to check the results produced by amateur+AI teams. Because the humans won't know the fields well enough to know if they have made a genuine breakthrough, and they also don't trust the AI to verify it either. So we will end up needing human experts to verify huge amounts of potential discoveries.
'Execution is becoming abundant, fast, and scalable. The complement to execution is verification: the capacity to know whether what was executed is what was intended.'
@AndrewCurran_ @carlfranzen Verification is the bottleneck https://arxiv.org/html/2602.20946v2
@DeepDishEnjoyer The OP was an extremely well qualified expert who works at Anthropic though, I'm not sure people would have trusted the models as much if it was an anon with 50 followers.
@AndrewCurran_ @carlfranzen Verification is the bottleneck https://arxiv.org/html/2602.20946v2
@AndrewCurran_ @DeepDishEnjoyer Yeah I find it interesting where even if I think I have multiple results and it’s been checked independently by multiple different models that I have nobody around me in real life is even good enough to check the math done by the models at this point…
@AndrewCurran_ @DeepDishEnjoyer I think im decently worried about developing some form of LLM psychosis since I’ve seen so many people get it. So I just don’t do anything with the results…
@AndrewCurran_ So this is the thing. Will we really need full human employment to ride herd on AI-powered production? Maybe not. Will it be very very easy to manufacture millions of dubiously necessary what-the-heck-can't-hurt make-work jobs like those just to give people a sinecure? Yep.
@AndrewCurran_ How long do you think this will last though?
@AladinKumar1 But there is going to be a very, very large amount of verification needed.
@AndrewCurran_ To me I think we are about a weird surge of hiring. Where companies basically conflate more people with more prompters and more verified output. But once models can discern good outputs and bad it's gonna hit like that meteor.
@ccatalini @AndrewCurran_ @carlfranzen What’s the solution then?!
@AndrewCurran_ But verification generally takes less time then solving the problem
@createtoremove @AndrewCurran_ @carlfranzen What @AndrewCurran_ wrote! More verifiers! As well as better ground truth that can feed better models needing less verification.
@FeltSteam Well, nothing lasts forever.
@AndrewCurran_ my guess is models will catch up quicker than human experts (though they will be necessary, just fewer of them)
@ccatalini @createtoremove @AndrewCurran_ @carlfranzen I bet AGI will solve @AndrewCurran_ bottleneck of more verifiers much more efficient as so to eliminate his “bottleneck”. China will not build platforms with safeguards that bottleneck their progress
@AndrewCurran_ Seem to be forgetting the role of the experimental physicist
@AndrewCurran_ Post AGI imo employment will skyrocket at first, there are so many problems that need solving. Space travel,Colonies on different planets etc etc. This happens only if we get true AGI tho. If we have AI models that are helpers/CoWorkers then yes employment sector is cooked.
@DeepDishEnjoyer @AndrewCurran_ were people plugging into llms? you could just use wolfram or literally napkin. it’s like a 2 minute calc 3 problem
Load more posts
@AndrewCurran_ that's not how people verified the jacobian conjecture disproof. literally everyone plugged it into their favorite llm lol