Alex Krizhevsky is a computer scientist, born in Ukraine and raised in Canada, best known for co-creating AlexNet, the deep Convolutional neural network that won the 2012 ImageNet Large Scale Visual Recognition Challenge by a wide margin. As a graduate student at the University of Toronto working under Geoffrey Hinton, Krizhevsky designed and implemented the network together with fellow student Ilya Sutskever, with Hinton as senior author on the resulting paper.
AlexNet
AlexNet reduced the ImageNet classification error rate by roughly ten percentage points compared to the next-best entry, a gap so large that it reshaped the field's expectations almost overnight. The network's design combined a relatively deep stack of convolutional layers with techniques including ReLU activations, dropout Regularization, and, critically, training on consumer graphics cards, demonstrating that GPUs could make training much larger neural networks practical. The 2012 result is widely cited as the single event that convinced a skeptical computer vision community to adopt Deep learning methods, effectively launching the research and investment wave that produced the following decade's advances in Computer vision and, later, large language models. Krizhevsky also released cuda-convnet, an efficient GPU implementation of convolutional networks that circulated widely among early deep learning researchers and served as a practical reference implementation before larger frameworks such as PyTorch and TensorFlow existed.
Career after AlexNet
Following the paper's publication, Hinton, Sutskever, and Krizhevsky formed a small company, DNNresearch, which Google acquired in 2013, bringing all three researchers into Google's AI efforts for a period. Krizhevsky worked at Google for several years afterward, contributing to internal computer vision and deep learning systems, but largely stepped back from public research activity and academic publishing in the years that followed, in contrast to the continued high public profile of his AlexNet co-authors Hinton and Sutskever.
Legacy
Despite his comparatively low public profile relative to many contemporaries, Krizhevsky's contribution is considered one of the pivotal moments in AI history, frequently placed alongside the Dartmouth workshop and the publication of "Attention Is All You Need" as one of the field's defining inflection points. AlexNet remains a standard reference point taught in introductory deep learning courses, and it is frequently the first example presented to students learning about convolutional architectures and the historical shift toward GPU-accelerated training.