Notes on intelligence amplification, shared memory, and the brain we're building around today's LLMs.
Every serious AI product today is an attempt to bolt missing brain parts onto a language cortex. Here's the anatomy of what's actually being built — and what's still missing.
Every week you re-explain the same decisions, re-find the same context, re-discover what someone already knew. That's not a culture problem. It's a tooling problem.
Tim Berners-Lee's most underquoted line, revisited in the LLM era — and why agents change the math on superconnecting.
Vannevar Bush imagined a personal memory machine in 1945. Most of what he missed turned out to be what LLMs do well. Most of what he got right, we still don't have.
The prompt doesn't matter if the memory is wrong. A practical guide to building context layers around language models.
Niklas Luhmann's slip-box gave one man 70,000 notes that talked to each other. We're building the version where every agent you own shares the same box.
Why we still prefer Engelbart's framing — humans augmented by tools — over the race to replace them. And what that means for what we build.