Essay
The Best Agents Become Invisible
The best tools eventually disappear. Not because they become unimportant, but because they become ordinary. Personal agents will follow the same path.
Read essayWriting
Essays on AI systems and autonomy.
Essays on AI systems, autonomy validation, and the craft of experimental programs.
Essay
From GTC to OpenClaw, Personal Agents Get Real
This month, GTC made personal agents feel less speculative. The headline was still infrastructure, inference, and the broader AI stack. But when NVIDIA put Build-a-Claw at GTC Park and described it as a "proactive, always-on AI assistant" reachable through your preferred messaging app, the more interesting signal was distribution. AI is moving from chat tabs into persistent agents you can message from anywhere.
Read essayEssay
The Agentic Test Spine
Over the last few weeks, both OpenAI and Anthropic have pushed hard on enterprise agents and long-running workflows. To me, this reinforces the shift from prompting to testing. If your agent can run longer, touch more systems, and take more autonomous actions, weak tests stop being a quality issue and become an operational risk.
Read essayEssay
Autonomy Is a Feedback Loop
I used to talk about the while (true) agent loop as a design pattern. After reading Anthropic's engineering write-up, I now treat it as a proven operating model. They used an autonomous loop to help build a real C compiler, and the result was not a toy demo. It compiled substantial code, including Linux, across multiple architectures.
Read essayEssay
Compound Engineering
Great, rapid, reliable engineering is rarely about single breakthroughs. It is about designing systems where ordinary work compounds quietly, predictably, and in your favor, increasingly accelerated by agentic workflows.
Read essayEssay
Simulated AI Agent Stock Trading
Replayable stock-trading sim where LLM agents decide buy, sell, or hold and two agents start to collude.
Read essayEssay
Less Prompting, Better Results
As models improve, shorter prompts outperform over-specified instructions.
Read essayEssay
Beyond the Loop
Extending the feedback loop beyond a single session and into longer horizons.
Read essayEssay
Closing the Feedback Loop
Reliability returned when tests stayed non-negotiable and feedback stayed continuous.
Read essayEssay
When Duration Became the Constraint
The next step was not exploratory. It extended the same system across a broader surface area.
Read essayEssay
The Constraint
The problem was well defined, and the constraint was execution.
Read essayEssay
Holiday Note: What's Next in AI
Holiday note on long-running agentic workflows, orchestration, and resilient progress.
Read essayEssay
Managing AI Like an Engineer
Why AI needs engineering-grade constraints, tests, and guardrails to deliver real leverage.
Read essayStay in touch
Subscribe for new essays and project updates.
