by my reckoning there are opposing factions that use hexbear. the ones that are opposed to ai and llms unconditionally and the ones that see it as a tech that should be completely embraced so long as it’s chinese. for my part it’s a technology that while useful in principle was invented for oppression’s sake and will only be applied in the most anti-human way but it must be put to use whether we like it or not because it has applications that have been made by the us government against the left already. we must recognize its purpose and its potential, then get the better of it like anything used to attack us
i ask because i don’t think it’s been asked before and it would be useful in defusing conflicts i’ve seen and participated in, and because time is of the essence for it to be hashed out
The ideal AI is small, runs on device, and has a very specific purpose. Hyperscaling to create AI god is bad. A model that can accurately translate between languages and can run on a decade old smart phone is ideal. In the case of LLMs, the less you actually dialog with it the better. Ask it a question, get a result. If the result wasn’t good, start over.
Your opinion on AI will largely depend on your job and what job the AI has done. I’m a programmer. To me an AI is either a super proficient gig worker with absolutely zero real world experience at best, or fully automated human error at worst.
My personal experience is that AIs ability to translate is significantly overblown. It’s good, but not meaningfully moreso than traditional translation. As with all topics, it’s just good at convincing people when it comes to facts they have no way to verify.
IIRC, google translate has been LLM based for years.
Yep, and it has since been doing a fuckton lot more ‘guessing’ than people realise.