Andrej Karpathy
Former Tesla AI director and OpenAI founding member known for building neural networks from scratch on video and coining the phrase "vibe coding."
Andrej Karpathy built his reputation on a mix of research and, unusually for the field, some of the clearest public teaching material available anywhere. As a PhD student under Fei-Fei LiFei-Fei LiBuilt ImageNet, the large labeled dataset that made the 2012 deep learning breakthrough in computer vision possible in the first place. at Stanford, he worked on connecting computer vision and natural language for image captioning. He was a founding member of OpenAI in 2015, then led Tesla's Autopilot computer vision team from 2017 to 2022, where "software 2.0" — his term for replacing hand-written logic with weights learned from data — became an internal design philosophy for the self-driving stack.
He's best known outside research circles for micrograd and nanoGPT: small, readable from-scratch implementations of backpropagationBackpropagationBackpropagation is the algorithm that computes the gradient of a neural network's loss with respect to every parameter, by applying the chain rule backward through the network. and a GPT-style transformerTransformerThe transformer is the neural network architecture built around self-attention, introduced in 2017, underlying essentially all modern LLMs., built live in video lectures that walk through every line rather than starting from a library. That build-from-scratch approach — feel the mechanism before you abstract it away — is the same pedagogical bet this course makes.
Karpathy returned to OpenAI briefly in 2023–2024 before founding Eureka Labs, focused on AIAI (Artificial Intelligence)AI is the field of building systems that perform tasks normally requiring human intelligence — reasoning, perception, language, and decision-making.-assisted education. He also popularized the term "vibe coding" for programming by describing intent to an LLMLLM (Large Language Model)An LLM is a large transformer trained to predict the next token on massive text corpora, then fine-tuned to follow instructions — the architecture behind GPT, Claude, Gemini, and Llama. and iterating on its output rather than writing code by hand.
See also: Fei-Fei LiFei-Fei LiBuilt ImageNet, the large labeled dataset that made the 2012 deep learning breakthrough in computer vision possible in the first place., Ilya SutskeverIlya SutskeverCo-authored AlexNet as a student, then sequence-to-sequence learning, then co-founded OpenAI and helped drive the bet that scale would produce GPT-level language models.
Learn more: Neural Networks & Backprop · Attention & Transformers · Wikipedia: Andrej Karpathy
Noam Shazeer
Co-authored "Attention Is All You Need" and pioneered sparse mixture-of-experts models, then left Google to found Character.AI before returning to lead work on Gemini.
Jürgen Schmidhuber
Co-developed LSTM in 1997, the recurrent architecture that made training on long sequences practical years before transformers existed.