Menu
Sign In Search Podcasts Libraries Charts People & Topics Add Podcast API Blog Pricing
Podcast Image

Deep Questions with Cal Newport

Did AI Just “Solve” Math? (Let’s Take a Closer Look) | AI Reality Check

28 May 2026

Transcription

Transcript generated automatically by AI and may contain errors.

Chapter 1: Did AI just solve a significant math problem?

0.031 - 17.394 Cal Newport

Last week, OpenAI published a press release titled, An OpenAI Model Has Disproved a Central Conjecture in Discrete Geometry. They were talking specifically about the planar unit distance problem, which was first posed by Paul Erdos in 1946.

0

Chapter 2: What did OpenAI achieve with their recent announcement?

17.414 - 34.064 Cal Newport

Now this is actually a pretty simple problem to state. It basically says, What is the maximum number of pairs of points in a set of endpoints in a flat plane that can be exactly one unit of distance apart? Now, back in the 1940s, Erdos proposed an answer to this question.

0

Chapter 3: Is the result of AI's math breakthrough truly important?

34.124 - 36.747 Cal Newport

He couldn't prove it, but he thought he knew what the answer was.

0

Chapter 4: Are LLMs now smarter than human mathematicians?

37.428 - 60.582 Cal Newport

Last week, OpenAI essentially announced that they had used an LLM to prove that Erdos' proposed answer was, in fact, incorrect. The OpenAI press release was accompanied by a video that featured dramatic music and a group of researchers writing earnestly on a comically small blackboard as they explained why this was a big deal. Here, let's play a clip of that video.

0

64.274 - 80.697 Unknown

This is the first mathematical breakthrough due to an AI. It's been described as the most well-known problem in combinatorial geometry. So for a whole subfield of mathematics, it's like maybe the best known problem there is.

0

80.677 - 93.691 Cal Newport

The mainstream press soon picked up on this story with enthusiasm. Here's the new scientist headline. Mathematicians stunned by AI's biggest breakthrough in mathematics. People on X predictably went even more wild.

0

Chapter 5: Will AI conquer all equally hard mathematical challenges?

94.212 - 116.738 Cal Newport

Peter Diamandis tweeted the following. An open AI model just proved an 80-year-old math conjecture from Paul Erdos, one of the most prolific mathematicians in history. We're going to solve everything. All right, so what's actually going on here? Did AI just reach genius level? Has math as a discipline just been automated?

0

Chapter 6: What is the future of mathematics in the age of AI?

118.119 - 132.739 Cal Newport

As a theoretical computer scientist myself who has published a lot of applied mathematics research in my days and someone who proudly boasts an Erdos number of three, which you can look up if you don't know what that means, I am, for obvious reasons, particularly interested in these questions.

0

132.854 - 151.521 Cal Newport

Well, it's Thursday, which means it's time for an AI Reality Check episode of this show, which is the perfect opportunity to seek some answers. So that's exactly what we're going to do. As always, I'm Cal Newport, and this is Deep Questions, the show for people seeking depth in a distracted world.

0

158.403 - 167.256 Cal Newport

All right, so we need to start by getting more specific about what exactly OpenAI actually did, and then we can get into the implications of what that means for the rest of us.

0

Chapter 7: How does this AI breakthrough affect mathematicians?

168.358 - 176.41 Cal Newport

All right, so we're looking at this unit distance, planar unit distance conjecture. Erdos was convinced that he had identified the answer to the question.

0

Chapter 8: What are the implications of AI on mathematical reasoning?

176.631 - 184.663 Cal Newport

I don't want to get too mathy here, but just to say it quickly, Erdos thought that if you were placing endpoints into the plane,

0

185.015 - 202.956 Cal Newport

the maximum number of points that you could get to be a unit distance apart would be upper bounded by n raised to the power of 1 plus some constant c divided by the double log of n. Now, as you're probably noticing as you listen to me, that second term in the sum is going to tend towards 0 as n increases, asymptotically speaking. So this result, the answer...

0

204.05 - 225.212 Cal Newport

asymptotically is going to approach plain linear as the point set increases. That's a really elegant answer. Erdos was convinced that was right. A lot of other mathematicians just assumed that that was right because Erdos is usually right. And so people tried for a long time to prove that was indeed the fundamental limit. Now, what OpenAI did was they released a paper that said, no, that's wrong.

0

226.173 - 249.134 Cal Newport

We actually have a counterexample. We have a way of placing points that That is going to have more points. We're going to feature more points at unit distance than that limit, even as n increases. I believe the actual bound is something like n plus 1 plus some small fixed constant epsilon that stays fixed as you increase n as opposed to approaching 0. So they had a counterexample.

0

250.456 - 271.233 Cal Newport

They didn't say... Here's the right answer. Here is what the limit is. Here's the best. Here's the most possible point. So they didn't replace Erdos' conjecture with a better conjecture. They didn't prove Erdos' conjecture, but they provided a counterexample construction that showed the thing he thought was the right answer couldn't possibly be right. Now, how did they do this?

271.273 - 295.06 Cal Newport

Well, they used a reasoning LLM. So that's an LLM that has been tuned to essentially talk out loud, to sort of think out loud and wander with its thoughts. We first saw the first reasoning models back in 2024 with O1 and the O models, a deep sequence reasoning model as well. Basically, reasoning models are a way of taking an LLM, which are static and have no memory,

295.58 - 316.004 Cal Newport

and having them approximate something like more dynamic computation with memory, because it can sort of, as it rambles, right? It's looking at everything it said so far when it produces the new token. So it can, if you're rambling, you're thinking out loud, you can use all of that thinking in producing the new token. So it's like you have some memory and this wandering can be somewhat dynamic.

316.044 - 332.73 Cal Newport

You can get some basic like iterative or looping type thinking in it, right? So they use the reasoning model. And what they did is, I don't know how many times they prompted it or on what questions they prompted it, but on one of the times they prompted it about this particular problem, the model spit out a very long transcript of an answer.

332.791 - 343.781 Cal Newport

And a team of expert mathematicians poured over this answer, and in this long chain of thought transcript, they identified in there... the core idea that became the counterexample.

Comments

There are no comments yet.

Please log in to write the first comment.