I’m Evan DeLord. I started this site to share my thoughts and research on how AI behaves and how to better use it. My first exposure to agents and algorithms was doing research at UT Austin in the field of artificial life, which uses simulated organisms to study evolution. Later, I researched neural net simulations of the cerebellum to better understand the principles of the brain.
I wrote my undergraduate capstone thesis about the Big Five personality traits and how personality might extend beyond these traits into values, goals, and self-narrative. Overall I’m really interested in emergent properties of complex systems, like computation in the brain or evolutionary processes.
AI models have their own psychology and their own ways of communicating and interacting that are worth studying. We’re just beginning to understand what makes LLM-based AI tick and how to apply it. This site showcases my projects, built with Claude Code and Codex, like the autobiography series and the cabin and tower study. Coding benchmarks are important, but trying to understand the models beyond benchmarks will matter for keeping AI helpful to humans.
I’ve worked six years as an MLOps engineer at Chevron in trading. I now work as a Senior Cloud Engineer at Bosch, where AI integration is a big part of my work.
I’m also an amateur meditator and I enjoy discussing philosophy and future predictions at my local Rationalist meetup. At the moment, I’m enjoying Budapest’s unique culture and ease of travel to the rest of Europe.