Navigate Select ESC Close

Grant Sanderson (@3blue1brown) – AI and the future of math

2026-06-30 Science & Technology
75.9k
2.6k
260
Dwarkesh Patel
Dwarkesh Patel
1.4m subscribers

Unlock all features

FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.

Description

Always so much fun to chat with @3blue1brown AI has been making much faster progress in math than in other fields. As a result, mathematics is showing us, very concretely, what AI progress in other fields will look like. Even within mathematics, there's a jagged landscape. What does it look like? What is the nature of the most important conceptual breakthroughs in the history of mathematics, and how different are they from what AIs are currently able to do? Does AI (on net) increase or decrease human understanding of the field? How big is the overhang from having AIs systematically try to connect ideas already in the literature? And what advice does Grant have for aspiring mathematicians, coders, and other students who are passionate about fields that are being most transformed upon by AI? 𝐄𝐏𝐈𝐒𝐎𝐃𝐄 𝐋𝐈𝐍𝐊𝐒 * Transcript: https://www.dwarkesh.com/p/grant-sanderson-2 * Apple Podcasts: https://podcasts.apple.com/us/podcast/grant-sanderson-ai-and-the-future-of-math/id1516093381?i=1000774870615 * Spotify: https://open.spotify.com/episode/0X3t4uRlpVT4MXPYDIrNYX?si=HZf_0Ky2Q42tOWYZNvWi6w 𝐒𝐏𝐎𝐍𝐒𝐎𝐑𝐒 * Gemini 3.5 Live Translate is what I wished I'd had on my last trip to China. It detects more than 70 languages and translates them in near real-time… and it preserves your original pacing and intonation. If you're building an app that needs live translation, you should check out Gemini 3.5 Live Translate. Get started at https://ai.studio/live * Cursor’s harness lets me use models for a huge range of tasks at the podcast. For example, Cursor cuts out the ads from each episode I produce so I can post them on Bilibili. It also helps me prep for interviews — I have a repo full of books and papers that Cursor sorts through to find the exact right file for any given question. Try Cursor yourself at https://cursor.com/dwarkesh * Jane Street sponsors 3Blue1Brown, so Grant has gotten to spend a lot of time with various Jane Streeters. He actually just recorded an interview with a few of them, so when we sat down for this episode, he told me about some of the things he learned, like how Jane Street keeps their role definitions fuzzy to make sure their people keep learning and growing. Go check out Grant’s full interview at https://3b1b.co/janestreet To sponsor a future episode, visit https://dwarkesh.com/advertise. 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 00:00:00 – AI is discovering new proofs. Is that AGI? 00:11:32 – The verification loop on conceptual breakthroughs can be a century long 00:26:12 – Will we understand an AI proof of the Riemann hypothesis? 00:38:08 – Can AI find the hidden bridges between fields? 00:53:48 – Why real-world tasks don’t fit into RL environments 01:07:07 – Good writing requires theory of mind that AI still lacks 01:16:02 – Why learning will still depend on human curation

Top Comments (10)

@vineetgundecha7872 2026-06-30

It's weird to listen to 3blue1brown without the cute visuals

455 10 replies
@vivekviswanathan2283 2026-06-30

Dwarkesh, I'm kind of bummed that you forgot the most salient question for Grant, which is "How'd you get those triceps?"

339 20 replies
@keithwins 2026-06-30

Omg " build a man a fire and he's warm for one night, set a man on fire and he's warm for the rest of his life" Grant I never knew your humor got that gritty! Fantastic

170 1 replies
@elasticlogic 2026-06-30

All these hints of the Galois quintic 3b1b video make me salivate

46 1 replies
@WanderingJoy 2026-07-01

Grant’s thinking is amazingly clear

29 2 replies
@itssoaztek4592 2026-06-30

(First time ever for me to comment before actually watching the show.) Absolutely awesome to have Grant on your show. His work on YT meant more for my understanding of math than 10 years of public school/university. I am so looking forward to this :)))

5
@b1walker 2026-07-01

Very happy about this collab!

3
@bmenrigh 2026-07-02

One thing I'm very excited about is the formal verification efforts being accelerated by AI. I have a hope that as the library of formalized lemmas, theorems, and definitions grow, it's going to enable even more machine automation and verification. I don't think it's crazy to think that in perhaps 10-15 years almost all mathematical knowledge has been formalized in a way that enables models to add new knowledge and connections to math quite rapidly.

3
@matyourin 2026-07-02

about "learning with LLms": i have been a teacher for maths, physics and computer science for a few decades, taught "in presence" and during the pandemic online in zoom meetings and with video lessons and narrated screen captures with/without me visible in the corner and i have been using LLMs basically since they emerged for coding, learning, summarizing, mathematical "work" like turning PDFs into latex and stuff like that. my key learnings from my experience in the "using them for leaning/teaching" are: 1) the style/character/enthusiasm/experience/knowledge of the teacher plays a very large role in teaching/learning and it is more important than the subject, the method, the medium, the material used. 2) with a "good" teacher: students learn better in a setting with the teacher physically present than with the same teacher saying the excact same words in a video or in an online setting 3) a student already curious and used to learning on his/her own with books, online material, ... and trained in critical thinking and questioning sources and can profit a lot from using an LLM. i think the reason is that they already have an internal structure and mental model for what they know and what they dont know, so they are better asking relevant questions and seeing the gaps of their own understanding and the explanations they are getting. e.g. if you just ask an llm "teach me linear algebra" it is a way worse experience than asking "i am learning linear algebra and did not understand how inverting a matrix works and what it is used for. can you give me an explanation with numerical examples and usecases". 4) as a teacher using an LLM was really useful for: creating variations of existing teaching material, like more practice questions (i give it 10 integrals and ask for similar problems, especially gemini was good at this), formatting latex / typst documents or turning scans into latex/typst, creating software solution that create practice material for students, planning courses / breakdown topics into smaller lessons.

2
@new_memeplex 2026-07-02

Fascinating. One that will be very interesting to come back to in 10 years time. You are doing a great service Dwarkesh and deserve your success.

0

Unlock the Data Inside
Turn Videos into Knowledge

  • Get FREE 10/day: transcripts, summaries, chats
  • Chat with videos, export text & PDF
  • $1 free API credit for RAG, chatbots & research

Free forever plan • All features unlocked

App screenshot