3 ms·
The thesis of this article seems to assume that models will be built to serve what the user wants, but, as we've seen with the Gemini fiasco, AI providers are m
by yifanl 3y ago
The thesis of this article seems to assume that models will be built to serve what the user wants, but, as we've seen with the Gemini fiasco, AI providers are more than happy to silently inject things into prompts, so I have very little confidence in this assumption.
- make3 3y ago[flagged]
- ahtihn 3y agoA company making an extremely racist tool by trying not to be racist? It's completely intentional, why would you assume it's accidental?
- make3 3y ago[flagged]
- CobrastanJorji 3y agoYour premise denies your conclusion. If the company is "trying not to be racist," then it cannot "intentionally" be extremely racist because that's what intentional means.
- yifanl 3y agoI just used the most visible example I could think of, the specific choice wasn't particularly relevant.
- make3 3y agoI'm white myself, but people focusing on the one tiny thing that could be seen as racism against whites but is maybe accidental when there's so much extremely serious non-white racism everywhere every day since forever is just of incredibly poor taste
- CaptainFever 3y ago(Lighthearted) It's quite funny actually, I'm not white but I thought the Gemini situation was kinda fucked up. We're like opposites!
- mvdtnz 3y agoMaybe the user is more likely to remember a blockbuster story from 2 weeks ago than a much less well known story from 6 years ago. Your impugning racist motives on the parent comment is really nasty.
- make3 3y ago[flagged]