AlphaGo defeated a world Go champion with Move 37 - a maneuver so unconventional it looked like intuition.
That was my personal trigger. I wanted to know how AI — neural networks, gradient descent, and backpropagation — may produce intuition. AlphaGo learned the game first by studying expert players, then by playing millions of games against itself. The next generation skipped the human step entirely and started with random moves.
The digital mind doesn't just reflect its training data — it reshapes it, with biases and reasoning we can't fully trace.
A decade later, AI is our all-day co-pilot. We write emails, code apps, and design our future with AI. And we are replicating a winning move without understanding the reasoning behind it.
In a world of brilliant automated intelligence, our role is to stay awake: Asking why the move works, not just play it.