Breakthrough on famous open math question by AI

[Dryas]

https://x.com/OpenAI/status/2057176201782075690

Yea, so AGI seems quite likely.

[Leo Gura]

You can't believe any such news stories.

We have no idea what human involvement was part of this.

[Dryas]

2 minutes ago, Leo Gura said:

You can't believe any such news stories.

We have no idea what human involvement was part of this.

You can see the researchers testifying in the video of the tweet though. Seems like almost impossible to coordinate a lie like that.

[Jodistrict]

I did some investigation to see if it was using brute force or systematic search of possibilites, and this was not the case.  It was a general system, not tuned to mathematics, and came up with the solution independently.

"Instead, a general-purpose OpenAI reasoning model achieved this by independently bridging two completely separate fields of mathematics, finding a hidden conceptual shortcut that 80 years of human mathematicians had entirely missed."

[Leo Gura]

I'm not saying it's an outright lie. The distortions are more subtle than that.

There are many ways to massage and frame what happened in this case.

And of course AI can find unseen connections between different fields of data.

What is the connection between Hitler and a kangaroo? I'm sure AI can come up with something that no human has thought about before there.

If you put all the world's data into a box, that box can find interesting new ways that data interconnects with itself.

Edited May 21 by Leo Gura

[Carl-Richard]

AGI but can't screw open a Coca-Cola.

[zurew]

Maybe, but be careful, because previously there was a case where people initially thought AI did the same with other Erdos problems, and it turned out that it just managed to find the solutions on the internet.

https://x.com/BorisMPower/status/1979328512923439455

[Carl-Richard]

@zurew The AI couldn't even find that out 

[Dryas]

4 hours ago, Carl-Richard said:

AGI but can't screw open a Coca-Cola.

Why does that matter? We cant swim in water like fish but here we are with crazy tech and making machine God.

[Dryas]

6 hours ago, Leo Gura said:

If you put all the world's data into a box, that box can find interesting new ways that data interconnects with itself.

But that isn't AGI why? Don't humans have similar process when coming up with new stuff?

[Carl-Richard]

1 minute ago, Dryas said:

Why does that matter? We cant swim in water like fish but here we are with crazy tech and making machine God.

Because we want them to screw open a Coca-Cola.

[Dryas]

6 hours ago, Carl-Richard said:

Because we want them to screw open a Coca-Cola.

I see but I’m not sure what your point is

[Leo Gura]

7 hours ago, Dryas said:

But that isn't AGI why? Don't humans have similar process when coming up with new stuff?

It is not even close to AGI.

If you give AI no data, the AI will be stupid as a rock. But a small child with almost no data is still intelligent and has GI.

Intelligence is not more data.

You are more intelligent than any AI and you have a billion times less data. So obviously data cannot be what AGI is about.

Edited May 22 by Leo Gura

[Carl-Richard]

1 hour ago, Dryas said:

I see but I’m not sure what your point is

AGI should be able to drive a human body better than a human or else it's just a garbage push-to-start token shuffling machine, only a glorified gumball machine (this is a false dichotomy but it's for the lols).

Edited May 22 by Carl-Richard

[Koeke]

Cal Newport explains this finding. TLDR: it is impressive, but not as big of a leap in capabilities as they make it out to be.

[NewKidOnTheBlock]

On 21. 5. 2026 at 4:37 PM, Leo Gura said:

I'm not saying it's an outright lie. The distortions are more subtle than that.

There are many ways to massage and frame what happened in this case.

And of course AI can find unseen connections between different fields of data.

What is the connection between Hitler and a kangaroo? I'm sure AI can come up with something that no human has thought about before there.

If you put all the world's data into a box, that box can find interesting new ways that data interconnects with itself.

This wasn't really about making connections between the data though, it actually invented a solution to the problem. It was all coherent and dozens of pages long. It's pretty impressive

Ok, I guess we should admit that it was ultimately about connecting the various data, but still the way they were connected is pretty impressive. Just happened to solve an 80 year old problem

Edited 19 hours ago by NewKidOnTheBlock

[Leo Gura]

I do not doubt that AI will be able to solve certain problems simply from being able to see the entirely of human knowledge as one whole -- since no human mind can. But this vein will quickly run dry. The solutions will be interesting, but ultimately not as impressive as AI optimists want them to be. You are not going to get a bunch of free ground-breaking science from it.

You can squeeze the entirety of human knowledge to get some drops of juice, but it's not like squeezing a fresh fruit. You are squeezing already-squozen leftovers. I don't like squeezing leftovers, I like squeezing fresh fruit.

AI hyper-mongers are trying to sell us the idea that AI is squeezing fresh fruit. Which is where we must call bullshit.

Edited 1 hour ago by Leo Gura