Andrej Karpathy

Andrej Karpathy

AI Researcher & Educator

About

Andrej Karpathy is an AI researcher and educator known for his work on computer vision and neural networks. He was a founding member of OpenAI's research team and later served as Director of AI at Tesla, leading the Autopilot vision team. Now independent, his educational content on neural networks has influenced countless developers entering the AI field.

Key Contributions

  • Created Stanford's CS231n course materials and later YouTube lectures that became entry points for deep learning engineers
  • Contributed influential computer-vision and vision-language research in the ImageNet-era deep-learning wave
  • Helped start OpenAI's early research group, then returned during the ChatGPT era before becoming independent again
  • Led Tesla's Autopilot vision team, pushing neural-network-heavy perception into a safety-critical consumer product
  • Popularized 'Software 2.0' and built nanoGPT, giving developers concrete code for understanding neural-network software
  • Coined and normalized 'vibe coding,' a useful phrase that also drew criticism for making AI-assisted software feel too casual

Notable Projects

Moments Timeline View all moments

Videos & Interviews

Connections

Fei-Fei Li

Fei-Fei Li

Influenced by

Co-founder & CEO, World Labs; Special Advisor on AI, Stanford

Karpathy took his Stanford PhD under Li with a thesis on connecting images and natural language — the problem ImageNet had made tractable, and the one that would reappear, transformed, inside multimodal models a decade later. He also inherited her conviction that teaching is part of research rather than a duty beside it: CS231n, which he helped build and teach, became the on-ramp for one generation of deep-learning engineers, as his lectures and nanoGPT are for the next.

en.wikipedia.org · en.wikipedia.org

Ilya Sutskever

Ilya Sutskever

Collaborated

Co-founder, Safe Superintelligence Inc.

Both are named in OpenAI's December 2015 launch announcement — Sutskever as research director, Karpathy among the founding research staff — and both eventually walked out of the building they helped fill. They mark the two exits from the same room: one toward trying to control what was built, the other toward making it legible to everyone else.

techcrunch.com · en.wikipedia.org

Richard Sutton

Richard Sutton

In contrast

Founder, Oak Lab & Professor, University of Alberta

When Sutton told Dwarkesh Patel that LLMs are a dead end, Karpathy answered with an essay: we are not building animals that learn from the world, we are summoning ghosts distilled from humanity's text — and the ghost is a practical answer to the cold-start problem the animal has never solved. Sutton holds the pure position; Karpathy the engineer's amendment. Between them runs the field's live question: is pretraining a foundation, or a ceiling?

karpathy.bearblog.dev

Theme
Language
Support
© funclosure 2025