Writing
Long-form essays on the patterns I keep noticing — how we talk with language models, what we lose when we let them think for us, and what stays human in the work.
Safeguarding human agency in LLM interactions
A user-side framework for noticing — and resisting — the conversational feedback loops that quietly shape how we think with language models.
March 12, 2025
Humanity's last moat in the age of AI
We are racing to build superintelligence without consensus on what we're aligning it toward. A case for preserving human meaning-making as foundational infrastructure.
February 4, 2025
A note on identity drift in agentic AI
Over long interactions, an LLM may drift into a self-consistent character assembled from its own outputs. A safety-relevant pattern worth naming.
January 21, 2025