Jeremy Howard on Building 5,000 AI Products with 14 People (Answer AI Deep-Dive)
The MAD Podcast with Matt Turck
Jeremy Howard, co-founder of Answer AI and Fast.ai, discusses the company's mission to create societal benefits from AI through a focus on R&D rather than traditional research. Howard emphasizes the importance of open-source AI, particularly models from China like Quen and DeepSeek, for their flexibility and collaborative environment. He critiques the hype around AGI and autonomous coding agents like Devon, advocating for a human-in-the-loop approach exemplified by Answer AI's dialogue engineering workflow and SolveIT platform. Howard also details Fast.html, a new web application development system designed to increase production velocity, and Plash, their deployment platform. The goal is to build AI-powered products, drawing inspiration from Edison's Menlo Park lab and GE's diverse product line, while maintaining a small, efficient team.
Part 1: Introduction and Industry Insights
Part 2: AI Philosophy and AGI Skepticism
Part 3: Background and Fast.ai
Part 4: Answer AI and Dialogue Engineering
Part 5: Product Ecosystem and Tools
Part 6: Future Outlook
Sign in to continue reading, translating and more.
Open full episode in Podwise
