Loading slide
Rosenblatt showed the Perceptron simple shapes: letters, squares, and other patterns fed in through the camera.
At first, it made mistakes.
Then the dials moved. The mistakes went down. The machine got better.
Rosenblatt thought this was the beginning of something much bigger.
He started making big claims about where it could lead. In a 1958 newspaper report, the Perceptron was described as the first step toward machines that might one day rival the human mind itself.
Some people were excited. Others thought he was going much too far.
Many researchers dismissed the idea: a machine turns some dials, and one day that leads to intelligence?
But the machine did work.
Only on a narrow task. Only with simple shapes. But it learned from examples. And for a while, that was enough to make people believe.
Did you know?