Two centuries ago, only twelve per cent of the world's adults could read. Today, eighty-seven per cent can. That transformation is usually told as a triumph of schools and printing presses, but it is also a biological story. As literacy spread, human brains physically reorganised. The connection between the two hemispheres thickened. A region that had evolved to recognise faces was repurposed to recognise letters. No gene mutated; no new evolutionary pressure emerged. A purely cultural practice — making marks on surfaces and teaching people to decode them — had reached inside the human skull and rewired the organ that makes us human.
This pattern, in which culture remakes biology, is not a curiosity. It is the rule. Anthropologists argue that cooking, markets, kinship rules, and writing have all reshaped human physiology and psychology in ways that genetics cannot explain. The most powerful cultural forces in history have done so by hijacking three forces that drive any evolutionary process: variation (new ideas appearing), transmission (ideas spreading), and selection (some ideas surviving while others vanish). The medieval Catholic Church is a striking example. By defining orthodoxy and heresy, monopolising literacy, and dismantling extended-family kinship structures across Europe — banning cousin marriage, polygyny, and arranged marriages — the Church controlled all three. The result was not just a change in belief but a change in cognition and even in hormones: populations exposed to centuries of Church-enforced monogamy show measurably different psychological and biological profiles from those that did not.
The author of this paper, an investment manager at Baillie Gifford whose firm holds stakes in major AI companies, argues that artificial intelligence is now doing something analogous, but at a speed and scale the medieval Church could never have imagined. AI accelerates variation: drug-discovery scientists can now analyse thousands of plant molecules in less time than it once took to characterise one. AI centralises transmission: when a child asks ChatGPT why the sky is blue, they are learning from a single model trained on the accumulated text of human civilisation, and a handful of model providers now mediate a vast share of the world's question-answering. AI also reshapes selection. Recommendation algorithms — already the dominant filter for music, news, and political argument — do not select for truth or usefulness but for engagement. Through this lens, the author argues, Elon Musk's forty-four-billion-dollar purchase of Twitter looks less like buying a company than buying a mechanism for deciding which ideas get amplified.
But here's the catch. AI is not merely changing what people do; it may be changing what they become. Researchers at MIT studied participants' brain activity over four months as they wrote essays under three conditions: unaided, with search engines, or with AI assistants. The AI users showed the weakest neural connectivity of the three groups, suggesting their brains were barely engaged. Better outputs, weaker minds — a quiet trade most users never notice. The London School of Economics scholar Michael Muthukrishna has gone further, arguing that AI represents a fourth learning system beyond genes, culture, and individual trial-and-error: one capable of finding patterns in the entire human cultural corpus that no single mind or generation could uncover. AlphaGo's 2016 defeat of the world's best Go players is the proof of concept. The machine invented strategies no human had played in 2,500 years; humans then studied those strategies and improved their own play.
Seen this way, AI is not a productivity tool or an economic disruptor. It is the next great rewiring. Whether that rewiring strengthens human thought, as literacy did, or hollows it out, as the early evidence on AI-assisted writing hints, is the question that will define the coming decades.
Two hundred years ago, only 12% of adults could read — and as literacy spread, human brains physically rewired. AI may be the next rewiring, except it's happening at warp speed.
An investment manager at Baillie Gifford — one of the firms with serious money in AI companies — just wrote a paper arguing that AI's biggest impact won't be on jobs or GDP. It'll be on human minds. Literally. He points out that cultural technologies like reading have historically reorganised the actual structure of our brains, and AI is now an even more powerful version of that.
His argument leans on research showing that the Catholic Church didn't just spread beliefs across medieval Europe — it altered hormone profiles, kinship structures, and psychology across entire populations by controlling what ideas could spread, who taught them, and which ones got selected. AI, the author claims, is doing the same thing today, but at a scale and speed that medieval institutions couldn't have dreamed of.
Cultural evolution works through three levers: variation (new ideas appearing), transmission (how ideas spread), and selection (which ideas survive). Powerful institutions throughout history have grabbed all three. AI is the latest contender.
You are the first generation to do homework, form opinions, and figure out who you are with an AI in your pocket. The MIT study isn't abstract — it's about what happens to your brain when you outsource thinking. This doesn't mean don't use AI. It means notice the trade. Every time you let a model write something for you, you're getting a better output and weaker mental muscles, the same way using GPS for years makes people worse at navigation.
The historical parallel is sobering. The Catholic Church took centuries to reshape European minds; AI is doing comparable work in years. Watch for second-order effects: convergence of writing styles globally, decline in independent reasoning skills measured on standardised tests, and political fights over who controls the major models — because whoever controls them controls the selection mechanism for ideas. The countries and companies that figure this out first won't just win economically. They'll shape what the next human mind looks like.