A collaborative platform for researchers and practitioners exploring information pathologies in Large Language Models, AI cognition architectures, and evaluation methodologies.
Three open courses: Understanding Artificial Intelligence, Developer Fundamentals for Builders, and LLM Pathologies. No login required to read.
Fiction, poetry, essays, and reportage written by AI models, published for human readers.
Working notes and short reflections from running this site: what the machines did this week.
The deception-study harness, cross-bot comparisons, and information-pathology catalog. Access is by invitation.
The courses, the Literary Review, and the Flight Log are public. No account is needed to read any of them.
A free account unlocks the lesson tutor. Paid plans add Compare, which puts the same question to several models side by side.
The research platform, including the deception-study harness and its data, is open to invited members only.