Re:OpenClaw isn't fooling me. I remember MS-DOS
by repelsteeltje on Monday April 20, 2026 @03:32PM
(#47839391)
Interesting take. Hadn't thought of it in terms of entropy, but it's true. Almost by definition as the training proces doesn't introduce anything novel beyond scraped inputs and a randomly initialized network. From there, the stochastic generation only adds randomness (and the prompt, of course).
Re:OpenClaw isn't fooling me. I remember MS-DOS
by epr on Monday April 20, 2026 @04:00PM
(#47839753)
Generally I think this is a legitemate issue, although:
> the training process doesn't introduce anything novel
This is not always the case. A compiler, linter, proof checker, tests, etc. can all lower entropy.
This is related to my observation that for thousands of years, written text has indicated a human author - this is no longer true, and I think this is going to be very difficult for us to wrap our human brains around fully.