entity · entities/OpenClaw.md
OpenClaw
The early-2026 paradigm (and tool) for maximally autonomous personal agents: run an agent 24/7 on a dedicated Mac mini/studio, interface with it via Telegram, and let it drive itself with "heartbeats." Now widely seen as overshooting the useful level of autonomy.
Name captured as transcribed from the podcast; treat the spelling as provisional pending a primary source.
What it was
The "first blush" of agent excitement in 2026: Mac minis (and Mac Studios) "running 24-7 agents, doing everything you could possibly imagine." The interaction model was minimal — give the agent a set of tools/info, then check in via Telegram. It used heartbeats: recurring, timed self-reminders to keep progressing against goals.
Why it faded
- The autonomy "wasn't actually accomplishing the things that people wanted."
- It was "good for absolutely burning tokens" — costly in the Token Shortage era.
- Maintenance burden on individual owners drove Every's move away from personal agents (Personal vs Team Agents).
Early adopters migrated: Matt Schumer wiped his OpenClaw Mac mini for an "always-on dev box" with Codex Mobile.
Connections
- Agent Autonomy Spectrum — OpenClaw marks the mega-autonomy end of the spectrum.
- Token Shortage — its token appetite became a liability.
- Codex, Claude Code — the semi-synchronous harnesses people moved toward.