Can AI do math, or does it just act like a calculator?

Science Quickly

Scientific American

Science

4.4 • 1.4K Ratings

🗓️ 25 March 2026

⏱️ 15 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

In this episode of Science Quickly, Kendra Pierre-Louis and SciAm reporter Joe Howlett explore a new math challenge designed to test whether today’s AI models can truly tackle cutting‑edge proofs. They break down how the experiment works, what early results say about AI’s mathematical abilities and why researchers are still debating whether these tools are revolutionizing math or just offering a high‑powered assist. Recommended Reading: As AI keeps improving, mathematicians struggle to foretell their own future AI just got its toughest math test yet. The results are mixed Is AI on the precipice of revolutionizing math? It depends E-mail us at sciencequickly@sciam.com if you have any questions, comments or ideas for stories we should cover! Discover something new every day: subscribe to Scientific American and sign up for Today in Science, our daily newsletter. Science Quickly is produced by Kendra Pierre-Louis, Fonda Mwangi, Sushmita Pathak and Jeff DelViscio. This episode was edited by Alex Sugiura, with fact-checking by Shayna Posses and Aaron Shattuck. Our theme music was composed by Dominic Smith. Learn more about your ad choices. Visit megaphone.fm/adchoices

Transcript

Click on a timestamp to play from that location

0:00.0	Now, are all the traitors present?
0:02.0	Let's get started, shall we?
0:04.0	From rags to riches.
0:06.0	I'm so sick of this.
0:07.0	Working like a dog and being treated worse.
0:09.0	Yorkshire to New York.
0:10.0	Poor climbers, you and me.
0:12.0	A life dedicated to revenge.
0:14.0	Let's make this an occasion to remember.
0:17.0	A Woman of Substance on Channel 4.
0:19.0	Stream now.
0:20.0	... A Woman of Substance on Channel 4, stream now.
0:36.9	For Scientific American Science Quickly, I'm Kendra P. Lewis, in for Rachel Feldman.
0:49.2	In 1997, Deep Blue, a supercomputer built by IBM, did the unexpected. It defeated chess giant Gary Kasparov at his own game, leading to a flurry of headlines about whether Deep Blue
0:53.8	was truly intelligent
0:55.4	and if computers could now outthink humans. The answer, at least then, was mostly no. But it's now
1:02.9	26 and we have a growing number of generative AI models that are once again making us wonder. Can
1:09.4	machines outthink us? To dig into this
1:12.1	question, a group of researchers aren't turning to chess this time. They're looking to math.
1:17.4	To learn more about that, I talked to Joe Howlett, a staff reporter here at Siam, covering math.
1:23.3	Thanks for joining us today, Joe.
1:25.0	Thank you for having me.
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from Scientific American, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Scientific American and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.