One of the greatest fears the AI bros have is not „human extinction“, but that people/companies might one day just run their own models locally without giving OpenAI et al. access to their entire life and data and money on top. Because then the bubble pops.
Post by Christian Stöcker
Replies
BTW: Always a SUPREMELY bad idea to enrage gamers.
@chrisstoecker
I've been advocating for local LLMs at work for over a year.
Its interesting that it seems to be only leadership who push back, those doing day to day work being asked to use it AI are the ones agreeing with me on local LLMs.
Better for the environment and the companies costs.
I also feel the bubble is about to pop. Anthropics presumable "junk rating" to me even more so points to the pop and its all been a big money grab scam from data thieves.
@prasket @chrisstoecker Just keep in mind open-weight models aren't ethically sourced... Some people have a problem with it.
Ideally we should have models whose datasets were collected with consent but.. yeah.
@altlazybear @prasket OpenAI and Anthropic (and Meta) certainly didn't ask for consent when amassing training data. We already know that.