You sit down at the screen at nine. By ten, the model has produced a draft. You read it, you adjust two paragraphs, you send it. The work is done. Nine to ten. One hour. You used to need three.
Then it is afternoon. You try to remember what you actually thought about the problem before the model produced its draft. You cannot. There was no thinking. There was a prompt and a response and a paragraph that read well enough to ship.
This is not a future risk. This is what is happening now, in the heads of the people doing knowledge work with large language models for forty hours a week. The signs are quiet. Recall is shallower. Synthesis is harder. The cold-start gear, the one that used to fire when you opened a blank page, has been bypassed for so long that it does not fire when you ask it to. The machine has done the imagining. The mind has rented out the muscle.
Atrophy is a book about that loss, written for the people who are causing it without meaning to. It is not anti-AI. It is not pro-AI. It is a manual for keeping your cognitive fitness while you work alongside the most powerful productivity tool in the history of writing.
The argument runs in four parts.
Part I — The Diagnosis. What is actually being lost. Why generic productivity advice misses it. What the published evidence says about transactive memory, motor-encoding in handwriting, sustained-attention decay, and the offloading reflex.
Part II — The Architecture. A working theory. The mind has layers — knowledge, evidence, output, guard — and AI prosthesis pushes work toward the surface layers while leaving the deeper ones unloaded. The fitness metaphor is exact: subsystems that are not loaded atrophy. Eight cognitive domains that can and must be trained.
Part III — The Programme. Twelve weeks of cognitive load, calibrated. A weekly split. Compound and isolation reps. Foundation, building capacity, layered cognition, sustaining. The Embodied Day: four phases for working with and without the model in rotation.
Part IV — The Application. Quick wins, long-game discipline, a coda on what remains when the prosthesis is set down.
The book contains no diagrams, no charts, no infographics. The argument is that the reader's interior world atrophies when the screen does the imagining. To make that argument and then deliver pre-made images would be a contradiction. The reader builds the picture. That construction is the first reps under load.
A self-assessment battery, an exercise library, and tracking templates are included as appendices. Every claim is referenced. Every citation has been verified.
If you work with AI for a living and have noticed that something feels different in your head after eighteen months of doing so, this book gives you the diagnosis and the rehab.
Autores populares
Unknown Author (247) Barrett Williams (162) ChatGPT ChatGPT (123) Сергей Каледин (110) Lyudmil Tsvetkov (97) Sharifa McFarlane (79) Rodrigo B Santos (64) Kingston Publishing (63) YouGuide (58) animarueaidezain (51) Kurt Bai (48) Clayton Louis Turnage (47) keieisyakentoushikaken (47) Nikolay Krechet (47) NK Gosine (44) e-aizamiraiseisei (40) SHIZUOKANOSORATETSU (40) Valet WorkShop (36) Various (36) Evgeniia Naumchenko (34)