You've probably heard that large language models like ChatGPT are "just predicting the next word/token" or "regurgitating their training data." But this misses something profound about how these systems actually work, and what they reveal about human thinking itself.
Why Your Arguments Are as Predictable as…
You've probably heard that large language models like ChatGPT are "just predicting the next word/token" or "regurgitating their training data." But this misses something profound about how these systems actually work, and what they reveal about human thinking itself.