In 2012, Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton presented AlexNet: eight layers trained on GPUs with ReLU, dropout, and data augmentation, winning ILSVRC by a wide margin.
The result convinced industry that deep learning was not a lab fad. Google, Facebook, Baidu, and startups accelerated hiring and compute.
AlexNet did not invent CNNs, but it did invent the cultural moment: since then, “more data + more compute + gradients” became the dominant heuristic.
