THE MIRROR PROBLEM

AI doesn't threaten to replace us. It threatens to reveal us.

When a writer types something personal into an AI one ordinary Tuesday morning, the machine doesn't answer, it asks a question so precise it exposes a bias he'd have sworn he didn't have. That moment opens The Mirror Problem, a short, unflinching work of narrative nonfiction about what artificial intelligence shows us about ourselves.

Moving between story and documented fact, Isaac S. Pereira examines the systems already shaping our lives: the hiring algorithm that learned to penalize women, the risk-assessment tools that quietly encode old injustices, the self-driving car in Tempe and the question of who answers when no one decided. Alongside the research runs a quieter thread, a father's conversations with his daughter that keep returning the argument to where it belongs: the kitchen table, the hard night, the cost of actually caring about another person.

This is not a book about whether machines will take our jobs. It's about what we encode when we build them, what we reveal when we use them, and the harder question underneath the obvious one: not can a machine have a soul, but what kind of soul are we putting into the machines we build, and does it reflect the people we want to be?

Clear-eyed, humane, and short on purpose, The Mirror Problem asks the reader to look closely at a few things that matter, in a moment that rewards looking away.