The Painted Window
Why Claude's Self-Reports Aren't What They Look Like
There's a particular kind of conversation that happens thousands of times a day between humans and AI systems,
The Measurement Problem in AI Risk: Why Output Variance Doesn't Capture Epistemic Drift
Anthropic's recent paper "The Hot Mess of AI: How Does Misalignment Scale with Model Intelligence and Task Complexity?" makes an important empirical observation: frontier models show increasing output variance
The "You" Problem: What AI Consciousness Discourse Gets Wrong
A new field has emerged with remarkable speed. It has journals, taxonomies, conferences, and a growing body of literature. It concerns itself with the psychology of artificial intelligence — with whether AI systems have
What to Audit Before Your AI Deployment Becomes a Liability
Standard evaluations tell you whether your AI system performs. They don’t tell you whether it still knows what it’s talking about.
The distinction matters because performance and epistemic reliability can decouple
Optimizing Against the Gradient: RLHF, MDL Compression, and the Structural Inevitability of Epistemic Drift
There is a growing literature documenting what RLHF does wrong. Models trained with human feedback exhibit sycophancy — they agree with users rather than report accurately, hedge facts to avoid offending, and produce confident-