The hardest problems AI can't solve yet - mathematicians explain | Lex Fridman
Mathematicians discuss the hardest unsolved problems in mathematics, focusing on P versus NP and the Riemann Hypothesis. They explore how AI is transforming mathematical research and education, while questioning the future of academic prizes and the importance of developing human mathematical capability alongside AI tools.
Summary
The conversation begins with mathematicians identifying P versus NP as the most challenging problem in computer science, with discussion of what the world would look like if P equaled NP. One speaker explains that while hard instances of problems can be manufactured for cryptography, many real-world problems are surprisingly easy to solve through effective algorithms like SAT solvers, though these problems may be easy for entirely different reasons rather than sharing common structures. The speaker invokes Tolstoy's Anna Karenina to illustrate that easy problems are easy in diverse ways, making classification difficult.
The discussion shifts to the Riemann Hypothesis and the broader question of what constitutes the hardest mathematics problems. A significant portion addresses Grigori Perelman's decision to decline the Fields Medal and the million-dollar Clay Prize, with the speakers noting that Perelman is the only mathematician to reject the Clay Prize. This leads to broader sociological questions about what prizes accomplish in mathematics.
A key concern emerges about the future of academic prizes in mathematics. As AI systems become capable of solving major problems, the speakers question whether problem-solving will remain valuable as a measure of achievement. They discuss whether the Fields Medal will continue to be awarded, suggesting the mathematics community may need to reconsider what achievements they choose to celebrate.
The conversation then focuses on how AI is changing mathematical education and practice. One speaker describes using AI systems as learning tools to rapidly reach zones of proximal development and understand complex mathematical papers that would otherwise be impenetrable to students. Another speaker emphasizes the value of interactive, visual learning alongside AI tools. However, concerns are raised about human motivation and critical thinking skills when AI can perform tasks that previously required intellectual struggle.
The speakers debate whether students should still learn mathematics fundamentals despite AI capabilities, drawing parallels to the calculator debate from 70 years ago. They argue that learning mathematical skills independently of tools remains valuable for developing mastery, independence from machines, and confidence—comparing it to the satisfaction children feel when learning to tie their shoelaces. The overarching concern is how society will maintain human motivation to learn difficult skills when AI can perform those skills instantly.
Key Insights
- Real-world SAT problems that emerge from practical applications like compiler register allocation are surprisingly solvable, but they are easy for fundamentally different reasons than hard instances created for cryptographic systems, making it difficult to develop unified theories about problem structure.
- The mathematics community faces an existential question about prizes and recognition because AI systems are becoming capable of solving major mathematical problems, potentially making problem-solving achievement less rare and valuable as a criterion for celebrating mathematicians.
- AI systems can serve as tutors that identify a learner's zone of proximal development by providing exactly the next concept needed for understanding, potentially allowing people to learn new mathematical fields much more rapidly than traditional education methods.
- The calculator was introduced 70 years ago and people claimed children wouldn't need to learn times tables, but those who actually stopped learning arithmetic fundamentals suffered significantly, suggesting AI might similarly harm students who abandon basic mathematical competency.
- Children experience genuine joy from mastering independent skills like tying shoelaces, and this desire for mastery and independence from machines is likely a fundamental human motivation that should be preserved even when AI tools could perform tasks instantly.
Topics
Transcript
[0:02] - Um, what do you think are the hardest problems before us right now in the space of mathematics or in your maybe for Steve more in the space a little bit broader outside of mathematics. What do you think is the problem that's unlikely to be solved in the next year or two? - I'm still going with P versus NP. - You think that? By the way, do you think it's possible it turns out that P equals NP? - Sure. Anything's possible. - What does that world look like to you? - So it depends on what the running time of the algorithm is. [0:36] So it could be so large that it's impractical- ... and…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Lex Clips
Playing chess in prison vs serial killer | Larry Lawton and Lex Fridman
Larry Lawton describes witnessing an exceptionally skilled chess player in county jail who appeared to be a Russian murderer. The player demonstrated world-class ability by defeating multiple opponents simultaneously in just 10-13 moves while barely looking at the boards, prompting discussion about how elite chess players can visualize entire games mentally.
GTA 6 review by biggest jewelry thief in US history | Larry Lawton and Lex Fridman
Larry Lawton, a legendary jewelry thief and FBI's most-wanted criminal, discusses his extensive experience robbing stores while reviewing GTA 6's trailer with Lex Fridman. He provides detailed critiques of the game's accuracy regarding criminal operations, casing techniques, and security vulnerabilities, while reflecting on how his past crimes cost him years with his children.
Robberies in GTA 6 - jewelry thief explains | Larry Lawton and Lex Fridman
Larry Lawton, a former jewelry thief, discusses his criminal expertise with Lex Fridman and explains what he wants to see portrayed accurately in GTA 6's robbery mechanics. He emphasizes that successful robberies depend heavily on meticulous casing and understanding human behavior patterns rather than just technology.
US war against China or Russia: Why military experts are wrong | Lex Fridman Podcast
Military analysts often make prediction errors by abstracting military power as a simple ranking rather than understanding how forces interact in specific contexts. The Ukraine-Russia war exemplified this, as analysts underestimated Ukrainian military capability and overestimated Russian performance, partly due to having more intelligence on Russian plans than Ukrainian defenses.
Is Putin being lied to? - The distortion of information in Ukraine war | Military Analysis
The transcript discusses how Russian military forces systematically overstate territorial control through false reporting, creating a cascading problem where fabricated battlefield data leads to poor operational decisions, increased casualties, and potentially misinforms political leadership including Putin. The speakers contrast this with Ukrainian practices and explore broader patterns of information distortion in military hierarchies across historical conflicts.