Ask HN: Using AI to solve unsolved math problems to prove AGI | Hacker News Reader

Ask HN: Using AI to solve unsolved math problems to prove AGI

1 pointnutanc6 comments

Why are we wasting compute in trying to beat some artificially created benchmarks. I think ~~everybody~~ most people would agree we have reached AGI if an AI could solve some of the unsolved math problems. We have a wikipedia list too [1]

1. https://en.m.wikipedia.org/wiki/List_of_unsolved_problems_in_mathematics

View on HN

> I think everybody would agree we have reached AGI if an AI could solve some of the unsolved math problems.

I actually don't think everyone would agree. I certainly wouldn't. So, to be absolutely pedantic about this, you're wrong :)

Got it. Will update the OP so that we are right pedantically :)

Yeah, not even being pedantic, I still think you're wrong. AI may solve some problems by linking together various bits from its training, and put together very coherent responses, but at least currently, we're not headed in a direction that will produce the kind of leaps of imagination and intuition that humans are capable of.

Heck, nobody can even agree on what consciousness is, let alone intelligence. AI can certainly perform various kinds of rote "thinking," and put together vast amounts of resources, but in my opinion, it's gonna be a while before it will outdo the human capacity for creativity.

The Taylor Swift billboard 100 debut at 1 test.

A benchmark that's hopelessly beyond the current capabilities of the thing being tested is not very useful. You can't make a better car by checking how quickly and safely it gets to the moon.

Right. Now the problem is that certain activities need inherent creativity, and also need the (self-)development/evolving of this creativity. While creativity is usually associated with human arts, actually mathematics, especially theoretical mathematics needs a lot of it, to figure out proofs, to come up with new solutions and of course intuition, to at least have some slight idea where to start looking. (Certainly some parts of maths can be figured out with brute force also. But usually these hard nuts are a different category)

Our current neural networks are quite "static" compared to what's needed for this self-developing creativity. But looking at your profile, probably you could lecture me about this, so I wouldn't waste a lot of bits here.

Reply on news.ycombinator.com