This page reflects what I’m focused on right now.
It changes over time and is intentionally incomplete.


Working on

At the moment, my primary focus is on understanding and evaluating AI systems in practice.

  • Studying the behavior of large language models beyond benchmarks
  • Designing small, repeatable experiments to test AI reliability
  • Improving workflows for AI system evaluation and testing
  • Writing technical notes and reflections on what actually works

I’m spending more time measuring and observing than building flashy demos.


Writing

I’m writing slowly and selectively.

  • Short technical notes when I discover something concrete
  • Longer essays when a pattern keeps showing up
  • Drafts that may stay private until they’re clear enough

If something isn’t clear yet, I let it sit.


Learning

Right now, I’m intentionally deepening fundamentals rather than chasing novelty.

  • Core concepts behind modern LLMs
  • Failure modes, edge cases, and system-level behavior
  • How incentives and human decisions shape AI outcomes

Progress feels quieter, but more durable.


Saying no to

To stay focused, I’m deliberately saying no to:

  • Over-optimization for visibility or traffic
  • Trend-driven tools without clear value
  • Work that moves fast but compounds poorly

Not everything deserves attention.


Outside of work

Away from the screen, I try to keep things simple.

  • Reading, thinking, and walking without headphones
  • Observing small routines that keep days grounded
  • Spending time with my cat, Cheese, who is very good at doing nothing

Last updated: December 2025