Neil Thompson: What drives AI model development—scaling laws or hidden innovations?

TL;DR

MIT CSAIL researcher Neil Thompson examines whether a 'secret sauce' exists in AI model development or if progress follows predictable scaling patterns. The talk likely explores the relationship between compute, data, and architectural innovations in driving model capabilities.

• •Investigates whether AI model breakthroughs follow hidden principles or predictable scaling laws
• •Addresses the role of compute, data, and architecture in model development
• •Relevant for understanding AI research trajectories and investment implications

Generated with AI, which can make mistakes.