There are classes of problems that people want solved where modern AI tech is the only viable solution we have - and it kinda sucks that the way we got to adoption of those tools was through companies like OpenAI that were willing to just run people over to 'win'.
Makes it more or less impossible to reasonably discuss the tools, because you can't separate them from the ways they get 'weaponised' (to put it as inflammatorially as I can)
yeah I just wish the discourse was a little more informed. In particular talking about on-device models has revealed some misconceptions I think (even though a lot of the criticisms are still valid)
