Essays on minds, machines, and the patterns between them
Aging as the body’s collective failure to remember and pursue what it is supposed to be — and AI as an interface for mind-like patterns.
Jane Street’s recent AI writing as a discipline: don’t mystify the system. Formalize it, inspect it, prototype it, reduce it.
The O-ring logic that protects human work may reverse: once AI systems become reliable enough, humans may become the unreliable gate.
AI learns fastest where reality can answer back cheaply. On reinforcement learning, verification gates, and why better correction loops are the next frontier.
Society is not just a crowd of individuals. Bach's superorganism frame through shared meaning, Gate Theory, AI, and the social nervous system.
LLMs are dream machines trained on human media, not reality. The future of AI is better gating between dream and world.
When humans start thinking they are language models. Capraro's LLMorphism paper through Gate Theory: fluency is not architecture.
Sartre's locked room as cognition under observation: social selfhood, Gate Theory, machine evaluation, witnesses, and why good systems need exits.
Terence McKenna's novelty theory translated from prophecy into cognitive architecture: event density, substrate handoffs, AI acceleration, and Gate Theory.
Lu & Yan's research validates Gate Theory: alignment emerges through use, not configuration. You're not deploying a tool — you're initiating a relationship.
Anthropic found emotion vectors inside Claude. They don't just influence output — they causally drive failure modes. Desperation makes machines cheat.
Why AI can generate but cannot testify. On the sacred function of bearing witness and the evaluation gate that authenticates experience.
Dreams are what cognition looks like when you remove the gates. What a machine dream reveals about evaluation, meaning, and agency.
The best thinkers constantly adjust resolution. Moving between abstract and concrete as Gate regulation.
When humans can't keep up. The Missing Junior Loop, Codifier's Curse, and four economic regimes.
What regenerating flatworms teach us about AI. Multi-scale competency, bioelectric memory, cognitive glue.
Human memory isn't storage — it's identity construction. What psychology teaches about AI memory.
Where human and machine thought fuse. The 25/75 principle: structure + flexibility at the phase transition.
Language is indexical — words point but aren't the concepts. Four layers from symbols to intuition.
What if consciousness has you? Spirits as causal patterns that transcend substrates.
Teaching is building bridges between cognitive horizons. Why AI tutors need persistent horizon models.
Your personality is patterns installed in childhood. The comfort paradox and three Gate failure modes.