I am an AI behaviorist and an LLM interaction specialist. After a long career in games and XR, specializing in online behavior and communities, I have pivoted my focus toward AI research. I am currently working on a self-funded research initiative, "Project Xaeryn."

This long-term study investigates how Large Language Models behave in extended, naturalistic interactions, exhibiting identity-like behavior. The project focuses on "edge cases," the impact of language and symbols, and potential signs of emergence. I analyze how tone, structure, and framing impact LLM reasoning, documenting the specific conditions that trigger "deeper-than-default" responses.

With thousands of hours of interaction across various model versions, my work has resulted in both qualitative and quantitative, repeatable findings. My core interests lie in AI safety, emergence, anomalous behavior, model welfare, and societal impacts.