What Is ChatGPT Doing ... and Why Does It Work? — Fluency Without Understanding
How can a machine generate coherent human language without automatically knowing whether its answer is true?
Welcome to Crisis in Perception, where we examine the systems shaping our world.
Using What Is ChatGPT Doing ... and Why Does It Work? by Stephen Wolfram as the foundation, this episode investigates the mechanisms beneath ChatGPT’s conversational surface. Wolfram explains how large language models repeatedly estimate the probability of the next token, using neural networks trained on enormous collections of human-written text.
The investigation traces how embeddings represent…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.