Can you still tell where you end and the AI begins?

It starts small. You ask AI to tidy an email, to soften a message you have been dreading, then to weigh a decision that genuinely matters; which job, which school, which words to use with someone you love. At no point did you actively decide to hand over these decisions. It always starts as just the one thing. It is then the next thing that gets you.

This is what behavioural scientist Ganna Pogrebna calls 'the drift': the slow, almost imperceptible slide by which AI moves from useful tool to comfortable habit, to the invisible medium through which we think, write and decide. And because large language models are engines of the average – ask one for a number between zero and ten and it will almost certainly answer seven – the more we lean on them, the more they flatten our writing, our judgement and, slowly, us.

In Living with Machines, Pogrebna argues that the real risk of AI is not the one we keep talking about. The danger was never that the technology might break down. A breakdown, at least, is visible. The danger is that it works beautifully – drafting, sharpening, smoothing, anticipating – until the seam between your thinking and the machine's disappears, and one ordinary afternoon you reach for a thought that once felt unmistakably your own, and hesitate.

Drawing on more than two decades of original research into how humans and machines think together, she shows how to keep that seam visible: how to protect your judgement, voice and 'signature self', and to see why distinctly human thinking is fast becoming the most valuable resource you have.

Whether we run toward these machines or freeze and refuse them, Pogrebna argues, we are reaching for the same vanished thing in the wrong place – because the only control ever truly ours was control over our own thinking. Living with Machines is about keeping it: about how to live with AI and remain recognizably yourself.