News for hackers. Stuff that matters.
Re:Gemini: Google's most capable AI model yet by freedomben on Wednesday December 06, 2023 @10:49AM (#38545426)
Absolutely I do. Internally they have some incredible stuff, but the leadership is terrified of letting normies try it out because of the (real or perceived I don't know) damage to the brand that would happen if it said something racist or misogynist, etc.
Re:Gemini: Google's most capable AI model yet by KaoruAoiShiho on Wednesday December 06, 2023 @10:53AM (#38545483)
No way, that's what they want you to think. The idea that Google would be behind technologically would be an embarrassment they can't handle. The 3.5 level gemini pro is probably just as capable of saying racist or misogynist stuff so there's no reason why they're allowing that to be public while the "GPT-4 beating" Ultra is hidden if it's just because of that. More likely Ultra is just not as good as these benchmarks indicate and they still need some time to improve it.
Re:Gemini: Google's most capable AI model yet by Workaccount2 on Wednesday December 06, 2023 @11:09AM (#38545756)
Google can hardly put a picture of a white male on their website. They're so deep in the swamp of virtue signalling it's a miracle they haven't yet drowned.
Re:Gemini: Google's most capable AI model yet by satchlj on Wednesday December 06, 2023 @11:15AM (#38545868)
Google has many photos of white males on their website wtf: https://about.google/google-in-america/