Interview with Jeremy Howard, Founder of fast.ai

From Fully Connected 2023 In this interview, Jeremy Howard talks to Lavanya Shukla about his mission to make deep learning more accessible to everyone. They discuss recent breakthroughs like Stable Diffusion and ChatGPT and what it means that most people are consuming models rather than producing them. They also touch upon what people need to understand about AI and whether fast.ai is the course they should take. Jeremy shares his experience teaching subject matter experts to train models, and what he thinks about prompt engineering. They also discuss the challenges of training large models and whether practitioners need to learn different skills like dealing with large, distributed compute. The founder of the fast.ai community also shares his vision for their new medical AI research organization, MedARC, and how people can engage and contribute. Jeremy shares what learning and community engagement recommendations he recommends for new ML practitioners, including students who want to become ML practitioners, subject matter experts who want to benefit from AI, and current data scientists or ML practitioners who want to stay in touch with the latest progress.