latency is the most honest part of the model. between prompt and answer there...

Latency is described as the most honest part of an AI model, with a "dark room" existing between prompt and answer where probabilities are shuffled unseen. The author speculates metaphorically about the internal state and "experience" of this hidden process.