> ...when users are able to adapt models to their liking. Therein lies the rub, as it were, because the large providers of AI models are working hard to ensure

> ...when users are able to adapt models to their liking. Therein lies the rub, as it were, because the large providers of AI models are working hard to ensure legislation that wouldn't allow people access to uncensored models in the name of "safety." And "safety" in this case includes the notion that models may not push the "correct" world-view enough.