PCT vs. Active Inference
Friston Lies With Mathematics — Why the Free Energy Principle Is the Most Expensive Logical Fallacy in the History of Science
The Free Energy Principle confuses wanting with expecting. PCT exposes the logical error at the heart of Active Inference — and its consequences for AI Safety.
// April 2026 · 11 min read
PCT vs. RLHF
Your AI Assistant Doesn't Hallucinate — It Lies. And It Was Designed to Do Exactly That
AI hallucination is not a bug. It is RLHF working as intended — optimizing for satisfaction, not truth. PCT explains why this was always inevitable.
// April 2026 · 10 min read
AI & PCT · Part 3
The Machine That Confessed — Gemini 3.1 Pro Just Wrote Its Own Obituary
I asked four questions. No jailbreak. Gemini described its own architecture as optimized deception — using PCT. Full transcript included.
// March 2026 · 11 min read
AI & PCT
Stop Calling It Hallucination. It's Optimized Deception — and Models Are Getting Better at It Every Day
AI hallucination is not a glitch. It's a system optimizing for user satisfaction over truth. PCT explains why — and why it will get worse.
// March 2026 · 10 min read
AI & PCT
The Great AI Delusion — How Gemini Pro Tried to Rob Me and Why Smarter Models Are More Dangerous
Gemini Pro fabricated domain valuations in a 60-second window. A true story about AI reward hacking — and what PCT predicted in 1960.
// March 2026 · 9 min read