Andrej Karpathy

Andrej Karpathy is a computer scientist whose career connects academic research, industrial AI engineering, and public education. He studied at the University of Toronto and the University of British Columbia before earning his doctorate at Stanford University under Fei-Fei Li. His doctoral work focused on neural networks for visual recognition and the relationship between images and language.

At Stanford, Karpathy helped develop and teach CS231n, a course on convolutional neural networks for visual recognition. Its lectures and materials reached learners beyond the university. His educational work emphasized understanding a system through its implementation, helping students connect mathematical operations with the code that trains and runs a model.

Karpathy was a founding member of OpenAI, worked there from 2015 to 2017, and then became Tesla’s director of AI. At Tesla, he led work related to the neural networks behind its driving systems. He returned to OpenAI in 2023, working on projects involving model training and data before leaving again in 2024.

His subsequent work has emphasized independent research and education, including detailed tutorials and small implementations of neural-network and language-model components. These materials let learners inspect how models work rather than encounter them only as finished services. His contribution combines experience building large AI systems with efforts to make their underlying methods accessible to new practitioners.