Notes
Writing on cognitive drift
What the research establishes, what it does not, and what any of it means for the way you actually work. Nothing here is generated — which would be an odd thing to do on this particular subject.
17 October 2026 · 5 min read
Why I built a tool to measure my own cognitive decline
I could not tell which judgements in my own work were mine any more. This is the engineering account of what I built to find out, why the measurement is weaker than I would like, and why I published its flaws.
10 October 2026 · 4 min read
The 300-word test: measuring your own drift in ten minutes
A free, self-administered check on whether AI delegation has changed how you think. No account needed. Four measurements you can compute yourself, and what each one does and does not tell you.
3 October 2026 · 4 min read
Not all offloading is harmful: which delegations actually cost you
Cognitive offloading is as old as the shopping list. What changed is which part of thinking gets absorbed. A concrete test for telling the delegations that free you from the ones that hollow you out.
26 September 2026 · 4 min read
How to use AI without losing your critical thinking
The useful question is not how much AI you use — it is which difficulty you are removing. A practical framework for keeping the thinking that develops you while delegating the work that does not.
19 September 2026 · 4 min read
Cognitive debt: the MIT finding, explained without the hype
Cognitive debt is the accumulated cost of outsourcing mental effort — each delegation saves time now and leaves a learning deficit that compounds. What MIT measured, why the metaphor holds, and where it breaks.
12 September 2026 · 5 min read
Is AI making us dumber? What the research actually shows
MIT found reduced neural connectivity in LLM-assisted writers. A 666-participant study found critical thinking declines alongside AI use. Here is what those results establish, and the larger claim they do not.