The Question We Still Can't Answer

In October 1950, a British mathematician published a paper in a philosophy journal that almost no one read. He wasn't announcing a breakthrough. He wasn't unveiling a machine. He was asking a question — and deliberately asking it the wrong way. His name was Alan Turing, and the question was: Can machines think?

What almost nobody knows is that he immediately threw the question out. In the very first paragraph of that paper, Turing declared the question too ambiguous to be useful. Then he replaced it with a different one. That substitution — quiet, almost bureaucratic — is the move that made artificial intelligence possible as a science.

Turing was 37 when he wrote "Computing Machinery and Intelligence," and he was not working in a vacuum of pure abstraction. He had spent the previous decade inside Bletchley Park, Britain's wartime code-breaking operation, building machines that could crack the Nazis' encrypted communications faster than any human team. He had already watched a machine do something that looked, uncomfortably, like thinking. He had also watched the people around him refuse to call it that.

That resistance wasn't ignorance. It ran deep in Western philosophy. René Descartes had argued in 1637 that only humans possessed a rational soul — that a machine could mimic behavior but never truly understand. Three centuries later, that intuition was still the default assumption. Intelligence was seen as a spark, something irreducibly biological, something that calculation could approach but never become. Turing had a problem with this. Not an emotional one — a logical one. If you define thinking as something only humans can do, then you've made the question unanswerable by definition. You've protected the conclusion inside the premise.

So he did what any rigorous thinker would do. He changed the question. Instead of "Can it think?" he asked: "Can it behave as if it thinks — convincingly enough that you can't tell the difference?" It sounds like a retreat. It was actually an advance. For the first time, intelligence had a test, not just a definition. And a test means progress is possible.

What makes this remarkable is the cultural moment Turing was writing in. The Cold War was hardening. Computers were room-sized machines that computed artillery trajectories. The idea that one might someday hold a conversation was not mainstream speculation — it was considered faintly absurd, even by serious scientists. Turing published anyway. He included predictions about machine chess, machine learning, and machine creativity that wouldn't be validated for decades. He was, in the truest sense, writing for a future that didn't exist yet.

We're living in that future now. The systems on your phone pass Turing's test in most everyday interactions. And yet — we're still arguing about whether they think. Which tells you something important: Turing didn't solve the question. He gave us a way to make progress on it while leaving the deeper mystery intact. That gap between what AI does and what it is isn't a bug in the field. It's the founding condition.

Understanding that gap — where it came from, why it's so hard to close — is the difference between people who use AI tools and people who genuinely understand what they're dealing with.

The newsletter can give you the backstory. But the video walks you through the actual logic of Turing's move — step by step, in plain English — and shows you exactly why this 75-year-old question is more alive today than it has ever been. If you want to understand AI from the ground up rather than just keeping up with the headlines, this is the right place to start.