> Machine learning technologies like the Bergamot translation project offer real, tangible utility. Bergamot is transparent in what it does (translate text locally, period), auditable (you can inspect the model and its behavior), and has clear, limited scope, even if the internal neural network logic isn’t strictly deterministic. This really weakens the point of the post. It strikes me as a: we just don't like those…
It's not necessarily close minded to choose to abstain from interacting with generative text, and choose not to use software that integrates it. I could say it's equally close minded not to sympathize with this position, or various reasoning behind it. For me, I feel that my spoken language is effected by those I interact with, and the more exposed someone is to a bot, the more they will speak like that bot, and I do…
I try to be polite just to not gain bad habits. But, for example, chatGPT is extremely confident, often wrong, and very weasely about it, so it can be hard to be “nice” to it (especially knowing that under the hood it has no feelings). It can be annoying when you bounce the third idea off the thing and it confidently replies with wrong instructions.
Anyway, I’ve been less worried about running local models, mostly just because I’m running them CPU-only. The capacity is just so limited, they don’t enter the uncanny valley where they can become truly annoying.