There is a particular glint in Danny Hillis’s eyes when he talks about machines, not the machines of industry or utility, but those that think. He has spent a lifetime not merely imagining intelligence but constructing it, piece by piece, in the circuits of supercomputers, in the mechanics of storytelling, and, more recently, in the quiet patience of a clock meant to last ten thousand years.
I have photographed Hillis many times over the past decade, often in the context of his work with the Long Now Foundation and its great monument to deep time, the 10,000-Year Clock. But my fascination with him predates even that. He is one of those rare figures whose mind seems permanently stretched between the past and the future, between the raw mechanics of the world and the ephemeral nature of thought itself.
When Hillis founded Thinking Machines Corporation in 1983, he was chasing an idea that most of his contemporaries considered impossible: that computers should function not like oversized calculators but like living brains. He built the Connection Machine, a supercomputer of unprecedented parallelism, a system designed to process information not linearly, but in a way that mimicked the neural networks of biological cognition. It was a machine that did not simply compute; it thought. On his door at Connection Machines read a sign that said; “I want to build a machine that will be proud of me.”
For a moment, it seemed the future might arrive early. Scientists at Los Alamos used the Connection Machine to model nuclear reactions, biologists used it to map the complexities of protein folding, and artificial intelligence researchers saw in it the beginnings of something profound. And then, as is often the case with the most ambitious ideas, it collapsed under the weight of history. The company faltered, AI entered one of its long winters, and Hillis moved on. But the dream remained.
At Disney, where he became Vice President of Research and a Disney Fellow, Hillis turned his attention to a different kind of intelligence, not the cold logic of computation, but the warm, human intuition of storytelling. He worked on interactive experiences, machines that could shape narratives, immersive worlds that responded not just to inputs, but to emotion and imagination. If a machine could think, could it also dream?
Yet Hillis has never been a man satisfied with a single future. His work with the Long Now Foundation, where he designed the 10,000-Year Clock, feels like an answer to an entirely different kind of question: not how machines can think, but how they can remember. The Clock, buried deep in the mountains of West Texas, is a slow machine, an enduring machine, a monument not to speed or power but to patience. It is built to measure time beyond the scale of empires, to remind future generations, whoever they may be, that they are part of something vast.
I remember standing with him at the Clock’s site, tracing my fingers over the cold metal of the gears, feeling their immensity. Hillis placed his hand on one of them and smiled. “This,” he said, “is a machine that waits.”
And perhaps that is what all of his machines have been, in some way, waiting. Waiting for intelligence to emerge, waiting for understanding to catch up. Thinking Machines, the Connection Machine, even the experimental storytelling of Disney, they were never just tools. They were invitations to imagine something beyond ourselves.
The question of thinking machines is not simply whether they will surpass us, whether they will become smarter, faster, more autonomous. It is a question of what we choose to build, and why. Hillis has spent his life designing machines that do more than compute. They provoke. They challenge. They endure.
And perhaps, in the end, the most profound thinking machines are not the ones that replace human minds, but the ones that teach us to think beyond our own limitations.
