They Saw It Coming — The Philosophers Who Anticipated This Moment

The conversation happening in boardrooms and congressional hearings and AI labs in 2026 feels unprecedented. It isn't. It is the same conversation — about intelligence, autonomy, human relevance, and the choice between liberation and catastrophe — that a small group of mathematicians and philosophers started in 1950. They didn't have the machines yet. But they had the questions. And the questions are identical.

In 1950, British mathematician Alan Turing published "Computing Machinery and Intelligence" in the journal Mind. He opened with a deliberately provocative question: "Can machines think?" Then he spent 30 pages arguing that the question was worth asking with curiosity and rigour rather than dismissal or fear. His friend Robin Gandy later recalled that Turing wrote the paper "quickly and with enjoyment" — reading passages aloud with a smile. Turing was not frightened by the prospect of machine intelligence. He was excited by it. The question he was really asking was: what does it mean to think? That question is still unanswered. The machines Sundar Pichai and Dario Amodei are building are the most powerful attempt yet to find out.

That same year, Norbert Wiener — mathematician, founder of cybernetics, and the thinker Brian Christian has called "the progenitor of contemporary AI-safety discourse" — published "The Human Use of Human Beings." His argument was precise and has not aged: the danger of intelligent machines is not the machines themselves. It is what humans choose to do with them. "The machine's danger to society is not from the machine itself," Wiener wrote, "but from what man makes of it" (Wiener, 1950). Wiener hoped automation would liberate people from drudgery for more creative work. He also warned that without deliberate social choices, it would simply concentrate power and displace workers without providing alternatives. He said this in 1950. Dario Amodei is saying the same thing in 2026, with a 38-page essay and a $61 billion company behind it.

In a conversation with his colleague Stanislaw Ulam in the early 1950s, John von Neumann — the mathematician whose architecture underlies every modern computer — described his sense that technology was "approaching some essential singularity in the history of the race beyond which human affairs, as we know them, could not continue" (Ulam, 1958). He said this conversationally, almost as an aside. Sam Altman's "Gentle Singularity" essay of 2025 is that aside, seventy years later, with the math finally catching up to the intuition.

Isaac Asimov spent forty years working through what happens when intelligent machines operate without adequate alignment to human values — not in science fiction terms, but as genuine ethical philosophy embedded in story. His Three Laws of Robotics were not a plot device. They were an attempt to solve the alignment problem six decades before anyone called it that. Geoffrey Hinton's warnings and Anthropic's Constitutional AI are rigorous mathematical descendants of the question Asimov was asking in 1942.

The pattern these thinkers share is not that they predicted the technology correctly. Most didn't. What they saw clearly was the shape of the choice — that intelligent machines would arrive eventually, that they would be transformative without being inherently good or bad, and that the outcome would be determined entirely by whether humans made deliberate choices about their deployment or simply let momentum decide.

The builders of 2026 are living inside that choice right now. The philosophers of 1950 would recognise it immediately.