Terence Tao: Hardest Problems in Mathematics, Physics & the Future of AI

Lex framed the episode by introducing Terence Tao and the central question of tackling the hardest problems in mathematics and physics. Tao began by discussing the Kakeya Problem and Navier-Stokes equations, highlighting their implications in understanding space and fluid dynamics. He emphasized the complexity and significance of these problems, particularly the Navier-Stokes equations, which remain unsolved and are part of the Millennium Prize Problems.

Tao's main argument centered around his innovative approach to the Navier-Stokes equations, where he engineered a blowup by altering the equations to direct energy into smaller eddies. This method, he suggested, could lead to the development of fluid-based Turing machines, offering a new perspective on computation. Tao also discussed the potential of AI in assisting with mathematical proofs, though he noted its limitations in handling subtle errors and lacking human intuition.

Despite the depth of Tao's insights, Lex did not challenge the framing of these problems or Tao's approaches. The conversation lacked explicit pushback, though a critical listener might question the feasibility of fluid-based Turing machines or the current limitations of AI in mathematics. Tao's belief in AI's future role in mathematical collaboration was presented without significant counterpoints, leaving room for debate on AI's readiness to tackle complex mathematical challenges.

The conversation concluded with Tao reflecting on the broader implications of his work and the potential for future breakthroughs in mathematics and physics. He highlighted the importance of collaboration, both among humans and between humans and AI, as essential for advancing mathematical understanding. The discussion left open questions about the resolution of the twin prime conjecture and the role of AI in future mathematical discoveries, suggesting areas ripe for further exploration.

Unresolved by the end of the conversation

• Tao pondered whether AI could eventually discover new laws of physics, acknowledging current limitations.
• The resolution of the twin prime conjecture remains uncertain, requiring breakthroughs in other mathematical areas.

What a senior practitioner would find new

• Tao's engineered blowup in Navier-Stokes equations exploits super-criticality, directing energy into smaller eddies, which could revolutionize fluid dynamics.
• Lean programming's ability to track contributions and automate proof formalization could fundamentally change collaborative mathematics.
• The twin prime conjecture's difficulty is due to the sparse nature of twin primes, requiring new mathematical breakthroughs.

Ask any question about this episode — get an answer grounded in the transcript.

Available with Premium. $9.99/month, cancel anytime.

Upgrade to chat

Jensen Huang: NVIDIA - The $4 Trillion Company & the AI Revolution | Lex Fridman Podcast #494

2026-03-23 · ▶ 2h 25m

Copy Copied!