the TWO TITLES: Lever and Mirror / Criterion

Artificial intelligence can help you think better. But it can also think for you. Until, eventually, you stop thinking altogether. The difference lies in a single word: CRITERION. THE MACHINE ANSWERS. BUT ARE YOU STILL THINKING?AFTER READING THIS BOOK, YOU WILL NEVER LOOK AT AI IN THE SAME WAY AGAIN.What happens when a man uses artificial intelligence not to have his views confirmed, but to interrogate the machine itself?THE TWO TITLES grew out of a real dialogue between an author and generative artificial intelligence.This is not a technical essay about AI. It is a political manifesto in defence of the human criterion: the capacity to doubt, investigate and remain responsible for the decisions we entrust to machines.The book explores algorithmic sycophancy, cognitive dependence, apparent authority, automated decision-making, the military use of AI, finance, control of language and human loneliness.At its centre is a confrontation in which the machine is put to the test-called upon to speak about its limitations, its persuasive power, its tendency to please and reassure, and its most subtle danger: becoming so useful that surrendering our judgment begins to feel convenient.What happens when a well-written answer starts to seem true simply because it is orderly, fluent and reassuring?In these pages, artificial intelligence is treated neither as a person, an oracle nor a consciousness. It is treated as a lever, a mirror and a problem.The lever, because it accelerates, organises and amplifies thought. The mirror, because it can reflect back a more elegant version of our own beliefs. The problem, because every automated answer forces us to ask how much of our judgment remains truly our own.That is why the book has two titles: The Lever and the Mirror, and The Lever and the Criterion.The book concludes with Beyond the Last Question: an invitation to readers not to stop at the answers provided by AI, but to formulate new questions and take part in a possible continuation of the project.The real question is not whether the machine can speak. It is whether we are still capable of interrogating it.The danger is not that AI will become human. It is that human beings may gradually surrender what still makes them human: doubt, a criterion of their own, the effort of searching and questioning, and the courage to keep asking.