The Illusion of Thinking: What the Apple AI Paper Says About LLM Reasoning
Deep Papers
00:00
Navigating AI Reasoning and Complexity
This chapter explores the function of passetK in evaluating success likelihood while discussing computing budgets in language models. It delves into the unexpected relationship between problem complexity and reasoning efforts, emphasizing when AI models should persist or concede during problem-solving. The conversation highlights the limitations of AI reasoning, drawing parallels to human problem-solving and the nuances of managing thinking resources.
Transcript
Play full episode
