The summer Math fell to the machines...

The summer Math fell to the machines...

August 19, 2026 • 5 min
📺 Watch Now

🤖 AI Summary

Overview

This episode explores the groundbreaking advancements in artificial intelligence (AI) and its unprecedented impact on the field of mathematics. From solving decades-old conjectures to making progress on some of the most famous unsolved problems, AI is reshaping the landscape of mathematical research and raising questions about the future of the discipline.

Notable Quotes

- Over the last 3 weeks, AI has killed more open math problems than the entire human race managed in the previous decade.

- Terence Tao...stood up at the International Congress of Mathematicians and warned of a crisis in the foundations of mathematical values and practices.

- During his run, Jared asked Claude to take a real stab at proving it, and so it did...burning 31 million output tokens.

📉 The Decline of Human Dominance in Mathematics

- AI has rapidly solved numerous longstanding mathematical problems, including an 87-year-old conjecture and a 30-year-old graph theory problem.

- OpenAI's models have tackled 10 open problems, including improving a general bound on sphere packing that hadn't advanced since 1978.

- Anthropic's AI made significant progress on the Riemann Hypothesis, a 167-year-old problem with a $1 million prize for its solution.

📜 The Leiden Declaration and Ethical Concerns

- In June, 16 researchers from 15 universities issued the Leiden Declaration, urging the establishment of ethical guidelines for AI in mathematics.

- Despite the call for guardrails, the global response has been minimal, raising concerns about the unchecked influence of AI in academic research.

🧮 AI's Role in Solving Mathematical Conjectures

- AI disproved the Jacobian conjecture, a problem in algebraic geometry that had been unsolved since 1939.

- GPT-5.6 solved the dense Garg-Gommans conjecture, a 30-year-old graph theory problem, by identifying a counterexample with just seven nodes and nine edges.

- Amateur mathematicians and students are also leveraging AI to solve open problems, further democratizing access to advanced mathematical research.

🤖 The Rise of AI-Driven Mathematical Research

- OpenAI's next major model not only solved multiple open problems but also provided formal proofs verified by Lean, a mechanical proof-checking system.

- Anthropic's AI, Claude, coordinated 60 sub-agents, executed 2,400 shell commands, and generated 31 million output tokens to make progress on the Riemann Hypothesis.

- These developments highlight the potential for AI to not only assist but also lead in mathematical discovery.

⚠️ The Future of Mathematics in an AI-Driven World

- Mathematician Terence Tao expressed concerns about a crisis in the foundations of mathematical values and practices, as AI challenges traditional methods of discovery.

- The rapid pace of AI advancements raises questions about the role of human mathematicians and the ethical implications of relying on machines for foundational knowledge.

- The episode underscores the need for a global conversation on how to integrate AI into mathematics responsibly.

AI-generated content may not be accurate or complete and should not be relied upon as a sole source of truth.

📋 Video Description

Railway is the smoothest way to deploy software. Get $20 in free credits - https://railway.com/?referralCode=fireship

AI has killed more open math problems in the last few weeks than the entire human race managed in the previous decade. Let's dive in.

#coding #programming

Want more Fireship?

🗞️ Newsletter: https://bytes.dev
🧠 Courses: https://fireship.dev