#connectionism

Question 1: Single choice

mediumAI-checkedAI Foundations & History

Minsky and Papert's 1969 book Perceptrons stressed what single-layer perceptrons cannot do, and funding for neural networks dried up. Why did this not end neural-network research for good?

Multi-layer networks escape those limits, and backpropagation later made them trainable1
The book's proofs about single-layer perceptrons were later shown to be wrong2
Single-layer perceptrons turned out to overcome the limits when given much more training data3
Symbolic AI adopted single-layer perceptrons as its standard reasoning engine4

Timed test