Frank Rosenblatt (1928-1971) was an American psychologist and computer scientist who invented the Perceptron, one of the earliest trainable artificial neural networks and a direct ancestor of modern deep learning architectures.
The perceptron
Working at the Cornell Aeronautical Laboratory, Rosenblatt built the perceptron in 1958 as a simplified model of a biological neuron capable of learning to classify simple visual patterns by adjusting weighted connections in response to errors, an early instance of the training procedure now central to Supervised learning. Implemented first in software on an IBM 704 and later as a dedicated piece of hardware called the Mark I Perceptron, complete with a 20-by-20 grid of photocells acting as a retina, it could learn to distinguish simple shapes after repeated exposure to labeled examples.
Publicity and hype
A 1958 U.S. Navy press conference and subsequent New York Times coverage described the perceptron in terms that anticipated machines that could "walk, talk, see, write, reproduce itself and be conscious of its existence," generating expectations far beyond what the single-layer architecture could deliver. Rosenblatt's own technical claims were more measured, but the gap between public expectation and the perceptron's actual, narrow capabilities contributed to a backlash once its limitations became widely understood.
The Minsky-Papert critique
In 1969, Marvin Minsky and Seymour Papert published "Perceptrons," a mathematical analysis demonstrating that single-layer perceptrons like Rosenblatt's could not compute functions such as logical XOR, which require the output to depend non-linearly on combinations of inputs. Rosenblatt had already explored multi-layer designs that could, in principle, overcome such limits, but the book's rigorous treatment of the simpler model, combined with reduced funding, is widely cited as a contributing cause of the first AI winter, a period of roughly fifteen years in which neural-network research fell out of favor relative to Symbolic AI approaches.
Death and legacy
Rosenblatt died in a boating accident on his birthday in 1971, before the eventual revival of neural-network research. The backpropagation algorithm, popularized in the mid-1980s by David Rumelhart, Geoffrey Hinton, and Ronald Williams, extended Rosenblatt's basic architecture to multiple layers and solved the very problem the Minsky-Papert critique had identified, enabling the eventual rise of Deep learning. The modern artificial neuron used throughout transformer-based large language models is, at its core, a direct descendant of Rosenblatt's original perceptron, and the IEEE's Frank Rosenblatt Award recognizes outstanding contributions to biologically and linguistically inspired computing.