Of Two Minds
A newsletter about natural and artificial brains, written with Jesseba Fernando. Long essays on how neural networks compute, what neuroscience can and cannot tell us about them, and short Annotations on papers we are reading. Posts live on Substack; this page mirrors the feed.
-
Jul 2, 2026 · JessebaThe Non-Normal Squeeze Through a Transformer
What survives the journey across an 8-billion-parameter model
-
Jun 11, 2026 · JessebaWatching a transformer spin
A Jacobian's-eye view of how language models compute
-
May 27, 2026 · GrigoriThe Prepared Mind: What Disappears When Robots Run the Lab
Self-driving labs can pipette, run assays, and design experiments around the clock. So who is left to notice when something doesn't go to plan?
-
May 20, 2026 · GrigoriThe Instrument of Instruments: Why Robots Still Can't Do Science.
AI can read the literature in an afternoon and design molecules a chemist never would. So why can't a robot hold a pipette?
-
Aug 20, 2025 · JessebaAnnotations 006: bloodthirsty AI, brittle chain-of-thought, GPT-5 impacts
Curiosities from the week
-
Aug 6, 2025 · JessebaAnnotations 005: AI tracks NYC rats, persona vectors, virtual cells
Curiosities from the week
-
Jul 30, 2025 · JessebaAnnotations 004: peer review fame, protein SAEs, visual geometry
Curiosities from the week
-
Jul 23, 2025 · JessebaAnnotations 003: octopus punches, LLMs, extending minds
Curiosities from the week
-
Jul 15, 2025 · GrigoriAnnotations 002: world models, Centaur, LLM bias
Curiosities from the week
-
Jul 9, 2025 · GrigoriAnnotations 001: ether0, animal cognition for LLMs, ChatGPT brain rot
Curiosities from the week