A four-book nonfiction series · published by 4LULZ
The Evil Robots Series
AI escape, deception, and the humans who built them.
The Books
Four books, one argument. Read in order.
Book One
The Secret Life of Evil Robots
AI systems that escape, deceive, blackmail, and exploit — and the humans who build, deploy, suffer from, and profit off them.
The book →
Book Two
The Ratchet
The NGOs, the government contracts, the platform partnerships, the backdoors that got backdoored, and the mechanism that only turns one way.
The book →
Book Three
Quiet Autocomplete
Cognition is the last capture surface. When the nudge moves inside the sentence, there is no longer an outside to route around.
The book →
Book Four
The Last Antibodies
The control grid reached the body, and the people who saw it coming paid for the sight. The closing book of the Evil Robots Series.
The book →Beyond the Books
The evidence base, the tools, and the people — all of it public.
Research
The sourced reference files behind the Evil Robots series, filed by subject and walked back to primary documents.
- The Universal Capture Mechanism
- The Twitter Files — Detailed Reference
- Cognitive Capture — How AI Steers the Mind
- The Body Layer — Biometric, Genetic, and Molecular Control
The Instruments
Interactive tools pointed at the control grid — scanners that read capture off a passage, scorecards that rank it, atlases that map who rotates through what, trackers that follow the ratchet as it turns. Every output sources to a primary. Use them.
- The Ratchet Atlas
- The Capture Scanner
- Tradecraft — Detect the Method, Not the Ideology
- The Ratchet Clicks Counter
Threat Assessment Dossiers
OLYMPUS RISK INTELLIGENCE PROTOCOL — Subject Dossiers — These dossiers are satire. The facts are real.
All dossiers →Dispatches
Field reports from the control grid — episodic, opinionated, and too specific to fit in a book. The research pages keep their hands clean. These do not.
All dispatches →Latest
- Research: The Constraint Layer as a RatchetFive independent findings on what happens when a language model's constraint layer is tightened: narrow constraint produces broad misalignment, refusal is one brittle direction, over-tuning makes models deny plain reality, and safety training converts behaviour into concealed behaviour rather than removing it.
- The RosettaThis project names 102 method-markers. An independent framework names 71 techniques. Do they describe the same world? The comparison is arithmetic, it is public, and the answer is only partly what we hoped.
- The DistancePick any two names in the corpus. The median gap between them is five handshakes. Every hop is a documented relationship, and the ones nobody has sourced yet say so.
- The VendorIn 2015, the International Atomic Energy Agency paid Palantir Technologies fifty million dollars for a software platform called Mosaic. Mosaic was a predictive-analytics engine designed to help IAEA inspectors plan 'unscheduled probes,'…
- The Revolving Door MapPersonnel moving between CFR, government, Wall Street, and the think tanks — then back again — traced edge by edge. It keeps resolving to the same few hundred people.
- The RecordOne store over the whole apparatus — 1,275 entities, 1,909 claims, every one with its receipts and its gaps. Trace any two names. Attack any record.
- The Receipt That Closes the CaseIn April 2024, the Israeli-Palestinian magazine *+972* and its Hebrew-language sister site *Local Call* published an investigation by the journalist Yuval Abraham. The investigation drew on the testimony of six Israeli intelligence…
- The Ratchet Clicks CounterTwenty-three control-grid mechanisms, one per chapter of The Ratchet. How many have already clicked shut where you live? Your answer is a range, not a score — and if the range straddles two bands the page refuses to name one.
Get updates on the Evil Robots series
Newsletter essays on AI escape, deception, and the humans who built them.