Since departing OpenAI in February 2024, Andrej Karpathy has channelled his attention into Eureka Labs — an AI-native education company aimed at rebuilding how technical subjects are taught from the ground up. Karpathy's positioning is straightforward: with capable LLMs now available as patient, expert tutors, the bottleneck in technical education is no longer access to expertise but rather curriculum design and feedback loops. His widely-circulated "LLM101n" course outline — a from-first-principles walkthrough of training a small language model — has become a de facto onboarding text for engineers entering the field.
In parallel, Karpathy's open-source teaching repositories — nanoGPT, micrograd, and the "Neural Networks: Zero to Hero" YouTube series — continue to underpin how universities and bootcamps in Singapore and across Asia teach the foundations of deep learning. Local institutions including NUS, NTU, and AI Singapore have integrated portions of his materials into formal curricula, partly because the alternative — building from-scratch courses on rapidly-moving research — is impractical at university timescales.
For Singapore tech vendors and system integrators looking to upskill engineering teams quickly, Karpathy's public materials remain one of the few pedagogically rigorous starting points that doesn't require commercial licensing. Eureka Labs's eventual product launches are expected to formalise this with structured cohorts and AI-tutor pairings, with Karpathy publicly framing the long-term goal as "Starfleet Academy" for technical education.