open endedness

Now

Last updated February 2026


I'm currently thinking about open-ended search in multi-agent systems — the idea that the most interesting outcomes emerge when you optimize for novelty rather than any fixed objective.

Reading: The Beginning of Infinity by David Deutsch (re-read), various papers on quality-diversity algorithms, and too many AI safety threads.

Building: This site, slowly and deliberately. Also experimenting with agentic workflows for research synthesis.

Thinking about: Trust as a protocol. What it means to delegate cognitive work to systems you can't fully inspect. The difference between legibility and transparency.

Location: Somewhere with good coffee and unreliable weather.