Ep. 367: What if AI Doesn’t Get Much Better Than This?
Deep Questions with Cal Newport
00:00
Reality Check on AI Advancements
This chapter explores the gap between expectations and the reality of AI advancements, particularly in the case of GPT-5. It discusses the concept of scaling law in AI, the evolution of language models from GPT-2 to GPT-4, and the subsequent challenges faced by leading firms in achieving anticipated performance improvements. The narrative highlights a shift towards specialized post-training techniques aimed at optimizing AI models for specific tasks, moving away from simply increasing model sizes.
Transcript
Play full episode
