🧠 Our brain doesn’t even fit in our brain

⏩ TL;DW

You simplify, then you forget the cut

You have to simplify to think. You drop details, draw a line around the problem, and accept a few small lies so you can hold it in your head. The spherical cow (the physics joke where you turn a cow into a sphere so the math works) is the honest version: one crude cut of a messy animal so you can do the math.

The trap is when you forget you made that cut. The slice that helped you can then trick you into overgeneralizing. Say it enough times and it hardens into “how reality works”.

Science makes the world make sense to us. Proving the world is simple was never the job.

When “like” turns into “is”

Brain talk spreads in short comparisons: tubes, “animal spirits” (old idea of invisible fluids in the nerves), telegraph wires, computers, prediction machines. Each retelling drops what doesn’t fit the story: glia (brain support cells that are not neurons), chemistry, bodies, the world outside the skull.

Repeat the slice often enough and the comparison stops being a comparison. The brain gets sold as a computer.

What everyone is talking about

Your mind-story follows whatever tool is on the bench, the metaphor fashion of the moment.

Same move as heat: fluid you pour, then particles jiggling. Same world, new label. Knowledge always needs some physical thing to live on.

Today, the bench is an LLM, a large language model trained on huge written archives. The “prediction engine” brain story starts to sound like fact. The training library it learned from comes back as smooth insight.

What do you really need an answer for?

“How does the brain work?” sounds simple; it isn’t. Ship of Theseus: rebuild the ship plank by plank and you still argue whether it’s the same ship. The answer depends on what it’s for: a mechanism → research, a story → funding, a roadmap → product, reassurance → easing worry. If you don’t spell that out, you sell the wrong answer with a face full of certainty.

AGI (artificial general intelligence) turns the volume up and narrows the question to “are we on track?” and “is this inevitable?” Instead of asking whether the metaphor fits, you get benchmarks: numbers that sound like certainty until the model leaves training and stumbles.

Before you repeat a frame, say it out loud: what it’s for, and what it leaves out.

📑 Chapter Guide