In 1993, Yann LeCun, at age 32, developed the first convolutional neural network (CNN) while at AT&T Bell Laboratories. Known as the LeNet architecture, it was designed for handwritten digit recognition and became a pivotal breakthrough in machine learning.