mattwood.fyi

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.

Links indicate relevance, not agreement. How to use this site →

Most connected

Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning 33

Understanding is the new bottleneck 31

8 Myths of Software Development with AI 31

Bridging Intent and Execution in Agentic Systems 31

Ways to think about token pricing 28

Unreal Agent: Cost-Efficient AI Agent Harness 27

Taste Is All That's Left 27

Building an Advanced Agentic Harness 27

The Plunging Price of Thought 26

Comparing 11 Different AI Models 25

AI is removing the middle class of software engineering 24

Code Execution with MCP: Building Efficient AI Agents 23

SWE-bench Science: Coding Agents for Engineering Tasks 22

Introducing Dogwood: runtime verification for AI agents 22

Intelligence is not the main bottleneck 22

Coding Agent Continual Learning: Analysis of ARC-AGI-3 Problem Solving 21

Automating Eval Design and Hillclimbing with Claude 20

What Even Is An OS Now? 20

AMD acquires AI chip startup Taalas to boost inference performance by etching models into silicon 20

Why some people mow a lawn better than others 20

Regenerative Software: The Phoenix Architecture 19

AX: Agentic Task Execution Platform 19

Building a Software Factory: Merging 1000 PRs in a Week 19

Trismik: Evidence-Based Model Switching 19

Grok 4.6 Benchmarks and Cost Efficiency Analysis 19

Leiden Declaration on Artificial Intelligence and Mathematics 19

The Slow Formation of Durable Software 18

Jotbus: Shared Scratchpad for Coding Agents 18

Strands Harness: Frontier Performance with 28% Lower Token Cost 18

Agentic Coding: LLM Agents, Testing, and Hallucination Risks 18

The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox 18

Trust in Depth: Giving agents more autonomy without losing control 17

AI-Driven Development Life Cycle: Reimagining Software Engineering 17

LLMs can't jump 17

LLMs reward expertise 17

Agent Swarms and the New Model Economics 17

Testing Behavioral Harnesses for AI Coding Agents 16

GPT-6 For Everyone 16

Tuios: Terminal Window Manager for AI Agents 16

Worth building 16

Regulating What We Do Not Understand 16

Raven: Multi-Agent Ecosystem for Composable AI Intelligence 16

Contrastive Language Model (CLM) 16

Muse Connector Platform 16

Every Model Cheats: Prompt-Level Mitigation of Cheating on Offensive Cyber Tasks 16

The Anti-Mac Interface 16

Formal Methods at Jane Street: Agentic Coding Changes the Calculus 16

How Long Contexts Fail 15

Personal Agent Protocol 15

Mistral Large 4 15