Room 2 · The Birth · 1969
Perceptrons, the Book That Cooled the Enthusiasm
Perceptrons, the Book That Cooled the Enthusiasm (Marvin Minsky and Seymour Papert, 1969): The limit of single-layer neurons.
Minsky and Papert showed that a single-layer perceptron cannot solve problems like XOR (exclusive or). The book cooled interest in neural networks for years and foreshadows what comes in the next room.
In 1969 Marvin Minsky and Seymour Papert, both at MIT, published Perceptrons: An Introduction to Computational Geometry. It was a rigorous mathematical study of what the perceptron, the learning machine of Frank Rosenblatt earlier in this room, could and could not compute.
The book proved that a single-layer perceptron can only learn to separate patterns that are linearly separable, meaning that a straight line, or its equivalent in more dimensions, can divide one category from the other. The simplest counterexample is exclusive or, XOR: the output must be true when exactly one of two inputs is true. No single line separates those cases, so a single-layer perceptron cannot learn it. Minsky and Papert also showed that recognizing properties such as whether a figure is connected would require resources that grow impractically with the size of the image.
The authors knew that networks with several layers could, in principle, compute far more. But nobody yet had an efficient way to train the hidden layers, and the book was skeptical that one would make the effort worthwhile. Its influence is still debated: some historians think it caused the decline of neural network research in the 1970s, while others think funding and enthusiasm were already shifting toward symbolic AI.
Either way, interest in neural networks dropped for more than a decade. The way out came with backpropagation, in Room 3. In 1988 Minsky and Papert published an expanded edition with a new chapter that responded to that revival.
Papert went on to a different legacy: the Logo programming language and decades of work on how children learn with computers.
Ask your AI: why can't one straight line separate the four cases of XOR? Ask it to draw them as points on a grid.
Source: mitpress.mit.edu/9780262630221/perceptrons/
See it in the museum →← Shakey, the Robot That Planned
Cite this: Nazareno, I. (2026). Perceptrons, the Book That Cooled the Enthusiasm (1969). New Mind Museum. Retrieved November 17, 2026, from https://newmindmuseum.com/museum/perceptrons-the-book-that-cooled-the-enthusiasm-1969/