Forcing probabilistic models to engage in determinatic processes is perhaps the most fantastic form of thumb twiddling we have yet invented.
BackwardsPuzzleBox
joined 2 years ago
The very idea of using GPT models to create datasets is such a mind-numbing, dumb incestuous decision to begin with. Essentially the 21st century version creating a xerox of a xerox.
In a lot of ways, it's kind of heralding the future enshitification of AI as dabblers think every problem can be automated away without human judgement or editorialisation.
Except it's not a brain. It's, at best, a tiny portion of a brain's processing, hence the move toward multi-modal models.
It's a bit like trying to do poetry with your visual cortex, or math using your autonomic system. I mean, more strength to you, but while you can use a hammer on a screw, it doesn't mean you should.