The Hackable Human Animal

To hack a system is to understand it well enough to manipulate it from the outside, exploiting how it actually works rather than how it presents itself. For most of human history the individual person was unhackable in this sense not because of any magical free will but because no external party had enough data or computing power to model the biochemical and psychological machinery of a particular person. Yuval Noah Harari argues that the convergence of biological knowledge and machine learning is dissolving this protection: a recommendation engine that has watched millions of hours of your attention does not need a theory of your soul to predict which stimulus will capture you next, and prediction—not understanding—is the engine of manipulation. The dangerous asymmetry is that the entity trying to move you can now model your desires better than you can examine them, which dissolves the epistemic foundation of the liberal