Attention
Can you stay with one hard object?
The score asks whether you can still read, write, or solve without reaching for a model, a feed, or a search bar when friction appears.
The Anti-AI Brain Score
Ten questions. Two minutes. One number from 0 to 30 that changes how you use your tools.
2,487 taken so far
Based on the diagnostic from Anti-AI Brain
Research: MIT · Microsoft · Princeton · UPenn
Score diagnostic
The Anti-AI Brain Score is a ten-question diagnostic for AI cognitive debt. It measures how much of your attention, memory, reasoning, and decision-making has moved from your own nervous system into the machine you use every day.
It is not a medical diagnosis. It is a protocol starting point: a fast way to see whether AI is still acting as a tool, or whether it has become the first place your brain goes when the work gets difficult.
Attention
The score asks whether you can still read, write, or solve without reaching for a model, a feed, or a search bar when friction appears.
Memory
It checks whether you remember the thing itself, or whether your working memory has learned to store only where the machine can retrieve it.
Reasoning
The diagnostic separates Socratic AI use from answer outsourcing: whether you think before prompting, or prompt before thinking.
Decisions
It measures whether AI is helping you falsify a decision after you make it, or quietly making the decision while you approve the output.
Read your result
Higher is better. The number is a snapshot of cognitive agency under AI pressure: how much human friction remains before the machine enters the loop.
AI is no longer assisting your work; it is carrying the first draft of your cognition.
You still have resistance, but the default route is delegation when the work gets uncomfortable.
You use the machine with some discipline. The next gain is making the rules explicit and repeatable.
You think before you prompt and use AI as a sparring partner, not a replacement nervous system.
After the quiz
Use the result to choose the first circuit to repair, then run the 30-day protocol in The Anti-AI Brain. The book starts with the same thesis as the score: AI is not the enemy. Undosed AI use is the risk.
Original data
Completed quiz results now write to a separate anonymous aggregate. The aggregate records score bucket, tier, weakest circuit, strongest circuit, completion date, and country-level Cloudflare header. It does not store email, IP address, name, or raw answer text.
Once the sample is large enough to be useful, the public research update becomes the Anti-AI Brain Score Index: which circuits fail first for daily AI users, how the distribution changes over time, and which reader cohorts recover fastest after the 30-day protocol.
Score FAQ
No. A low score is not a clinical diagnosis and does not say anything permanent about intelligence. It says your current AI habits are probably doing more cognitive substitution than cognitive support.
Use it on Day 0, Day 14, and Day 30 of the protocol. The useful signal is not one number; it is whether the number rises after you restore reading, handwriting, silence, strategic play, recall, movement, and Socratic AI use.
No. The target is dosed AI use. You keep the machine, but move it later in the thinking sequence: first draft from you, pressure test from AI, final decision from you.
A high score means you still have cognitive agency under pressure. The protocol turns that into a durable operating system so future tools do not quietly move the baseline.