What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Ten Advances In Mathematics

Connections

Supports

ATLAS: Autoformalized Textbook Library At Scale ATLAS (Autoformalized Textbook Library At Scale) builds the foundational corpus that enables AI mathematical reasoning advances of the type OpenAI is highlighting
Autoform Bot: Multi-agent system for translating LaTeX mathematics into verified Lean 4 proofs Autoform Bot translating LaTeX mathematics into verified Lean 4 proofs represents a concrete technical implementation of the kind of AI mathematical capability advances OpenAI is documenting

Related to

MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling MaxProof focuses on scaling mathematical proof generation via RL and test-time scaling — directly overlapping with the type of AI-driven mathematical advances OpenAI is cataloguing
Proof Editor - Collaborative Document Editor for Agents and Humans Proof Editor as a collaborative document editor for agents and humans reflects the human-AI workflow infrastructure being built around the mathematical proof capabilities OpenAI is advancing
Leiden Declaration on Artificial Intelligence and Mathematics The Leiden Declaration on AI and Mathematics directly addresses the intersection of AI systems and mathematical research, providing an institutional/ethical framework for exactly the kind of AI-driven mathematical breakthroughs OpenAI is highlighting
Schema: Frontier Models with the Right Harness Achieve ~99% on ARC-AGI-3 Public The ARC-AGI-3 ~99% achievement by frontier models demonstrates the broader trend of AI reaching or exceeding human-level performance on formal reasoning benchmarks, contextualizing OpenAI's mathematical breakthroughs

Supported by

Compression is prediction Compression-prediction duality has deep mathematical roots (Kolmogorov complexity, Solomonoff induction, MDL principle) — this connects to advances in mathematics where information theory intersects with learning theory
Dynamic Programming: Unifying Principle Across Algorithms Dynamic programming represents one of the foundational mathematical advances the existing item likely covers, as DP unifies optimization across graph theory, control theory, and machine learning

What Sort of Maths Are LLMs Good At? Both directly address mathematical achievements and capabilities, with the new item examining what kinds of math LLMs can do following recent breakthroughs that likely overlap with the ten advances discussed
Parametron Early computing hardware like the parametron enabled the computational tools that accelerated mathematical discovery, representing a foundational layer in the history of mathematics-computing co-evolution
Why some people mow a lawn better than others Both explore the intersection of human intuition and formal mathematics — lawn mowing finds humans achieve near-optimal solutions to CPP problems, paralleling how mathematical advances often emerge from intuitive reasoning before formalization
Why some people mow a lawn better than others Lawn-mowing CPP study engages directly with combinatorial/geometric optimization mathematics, comparing human intuitive solutions against formal mathematical optima
Why some people mow a lawn better than others Both explore the intersection of mathematical optimization and human intuition — the lawn-mowing study frames CPP as a combinatorial geometry problem, while the mathematics advances item covers similar formal problem domains

Challenged by

LLMs can't jump Ten Advances in Mathematics celebrates AI progress in mathematical domains, but the manipulative abduction thesis argues that true scientific invention — including mathematical discovery — requires cognitive processes LLMs structurally lack, not just more capability.

Type:	link
First published:	Monday, 3 August 2026 at 21:17 UTC
Location:	https://openai.com/index/ten-advances-in-mathematics/
URL:	https://mattwood.fyi/i/8c1417e3

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections