# 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.

- Author: Matt Wood, Chief AI & Technology Officer at AWS
- Essays: https://mattwood.blog
- Agent instructions: https://mattwood.fyi/agents/
- Developer docs: https://mattwood.fyi/docs
- OpenAPI: https://mattwood.fyi/openapi.json
- JSON Feed (best for agents): https://mattwood.fyi/feed.json
- Site index: https://mattwood.fyi/llms.txt

## Recent items

- [Kev: Tiny Decision Models on Qwen](https://mattwood.fyi/i/33bc8661) (2026-09-21): > A family of lightweight decision models built on Qwen3.5 that you can train and run locally, inspired by Jev-like architectures.
- [Amiga Unix Desktop Revival Project](https://mattwood.fyi/i/50636e41) (2026-09-21): > A community effort to revive Amiga Unix (Amix), Commodore's 1990s System V operating system, with modern support for 68040/68060 hardware,
- [Mini-AGI: Continual Learning on 8GB VRAM](https://mattwood.fyi/i/829f42c4) (2026-09-21): > A continual learning model trained from scratch on resource-constrained hardware, using an 8GB VRAM laptop with batch-1 streaming data.
- [The Millennium Problems for Biology](https://mattwood.fyi/i/a3cc73d7) (2026-09-21): > A collection of three major unsolved biological challenges: demonstrating the laboratory emergence of life from chemical precursors, cryop
- [AX: Agentic Task Execution Platform](https://mattwood.fyi/i/a7bd465e) (2026-09-21): > AX is a platform for declaring and running agentic tasks at scale with sandboxed execution, workspace management, network policies, and mo
- [OpenCodex: Universal LLM Provider Proxy](https://mattwood.fyi/i/68c90276) (2026-09-20): > A universal provider proxy that enables using any LLM (Claude, Gemini, Grok, DeepSeek, Ollama) with OpenAI Codex CLI, App, SDK, and Claude
- [Writing in the Age of LLMs](https://mattwood.fyi/i/98250d39) (2026-09-20): > Explores how to write effectively when large language models are ubiquitous, identifying common patterns of poor LLM-generated writing, de
- [Writing in the Age of LLMs](https://mattwood.fyi/i/1cd82fe5) (2026-09-20): > Shreya Shankar examines how to write effectively when LLMs are part of the writing process, identifying common patterns of AI-generated wr
- [SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness](https://mattwood.fyi/i/1a9c4a51) (2026-09-20): > Presents SoL-Pi, a method for improving coding agents through recursive self-improvement loops that reduces token usage by 44.7-49.0% and 
- [Why You Should Almost Never Use AI to Write Anything Substantive](https://mattwood.fyi/i/0c51036f) (2026-09-19): > The author argues against using AI for substantive writing tasks, contending that writing is essential to thinking, AI-generated text cont
- [Building a Software Factory: Merging 1000 PRs in a Week](https://mattwood.fyi/i/8027eeff) (2026-09-19): > Describes how the Kiro Crew team merged 1,000 pull requests in seven days by evolving their development workflow through five stages, from
- [Apodex 1.1: Scaling Agentic Intelligence for Complex Work](https://mattwood.fyi/i/ade7c55e) (2026-09-19): > Apodex 1.1 develops AI agents capable of complex, long-horizon tasks by scaling both environment diversity (files, search, code execution)
- [Claude Code Changelog](https://mattwood.fyi/i/baf9dd75) (2026-09-19): > Documents updates and changes to Claude Code, including new features, improvements, and bug fixes for the AI-powered coding assistant. Cla
- [How To Write With An LLM](https://mattwood.fyi/i/a0e08156) (2026-09-19): > Two rules for using LLMs as copyeditors rather than ghostwriters: never use LLM-suggested phrases verbatim, and avoid taking LLM encourage
- [Jalapeño Shows Power of LLMs for Chip Design](https://mattwood.fyi/i/e9dd64eb) (2026-09-19): > Jalapeño demonstrates how large language models can be effectively applied to semiconductor chip design, showcasing the potential of LLMs 
- [Skillsync - Portable AI Context for Coding Agents](https://mattwood.fyi/i/8c00aa5e) (2026-09-17): > Skillsync is a local-first tool that makes AI coding sessions portable across different agents and teammates, allowing you to resume work 
- [LLM Classification Is Feature Engineering](https://mattwood.fyi/i/0d509160) (2026-09-17): > LLMs used as classifiers have significant limitations like poor calibration and difficulty incorporating structured data, but treating LLM
- [Link](https://mattwood.fyi/i/15f75200) (2026-09-17)
- [AI Setup Community & Sharing Platform](https://mattwood.fyi/i/9b160607) (2026-09-17): > A community platform where builders and engineers share their AI tools, workflows, and setups to learn from each other and discover how ot
- [Infinite-Parameter LLMs: Generating Weights from Live Data](https://mattwood.fyi/i/d7eb45de) (2026-09-17): > This paper proposes a hypernetwork-based architecture that generates language model weights dynamically from live interaction data rather