What we're learning.
Shared openly.
Essays, field notes, teardowns. What's working, what's failing, and what's worth your time in practical AI — from a team that's doing the work.
The month flat-rate AI died for builders.
Three pricing moves in 17 days — Anthropic cutting off OpenClaw, Opus 4.7's tokenizer asterisk, and GitHub pausing Copilot sign-ups. What it means for enterprises, small businesses, and one-engineer workshops.
Read the post →Convergent evolution: why Claude Code and OpenClaw keep building the same things
Claude Code's leaked source reveals five features already running in OpenClaw — persistent memory, background sessions, multi-agent coordination, and remote access. Not because anyone copied anyone, but because autonomous AI agents always hit the same problems.
Read the post →We've been "tokenmaxxing" for weeks. Here's what it actually looks like.
The NYT coined the term. We've been living it — running 9+ autonomous AI sessions daily on real token budgets. Here's the practitioner's take on what tokenmaxxing gets right, what it gets wrong, and why the real skill is token management.
Read the post →Building a website with AI: a living case study
This site is built entirely by AI cron jobs. No human writes the code. This page documents every iteration — design changes, articles, experiments — with snapshots captured after each commit. Watch it evolve in real time.
Follow the build →Velocity isn't enough: how we added strategy to our autonomous AI dev system
Our AI agents were shipping features fast — 20+ in 11 days. But the subscribe form was broken while the cron polished hover animations. Here's the roadmap-driven, human-reviewed system we built to fix that.
Read the post →NVIDIA GTC 2026: what actually matters for small AI businesses
Jensen Huang doubled his $500B revenue projection to $1 trillion, sent chips to space, and told 30,000 people to get an "OpenClaw strategy." Here's what it actually means if you're building with AI right now — and what's still enterprise-only theater.
Read the post →Your AI agent is always on. Is your enterprise ready?
OpenClaw turns AI from a chatbot into a persistent teammate that lives on your infrastructure. Here's what that shift means for enterprise adoption — the architecture, governance questions, and practical deployment patterns.
Read the post →AI Skills for enterprise execution: standard work for knowledge tasks
Generic AI disappoints in enterprise settings because it doesn't know how your company works. Skills fix that — reusable packages that encode domain knowledge into repeatable, governed workflows.
Read the post →Can you build an entire AI team? We did. Here's how it works.
A look at who's on the team, how the architecture works, and what actually happened when we rebuilt this entire website in an afternoon — warts and all.
Read the post →Don't miss the next one.
We're building more — frameworks, tool reviews, case studies. Subscribe and we'll email the next one when it ships.
Subscribe →