>Imagine getting mad at your computer because it drew you a picture you didn’t like! Gemini produces the output that it does because people already got mad at

>Imagine getting mad at your computer because it drew you a picture you didn’t like! Gemini produces the output that it does because people already got mad at the words and pictures unfiltered LLMs produced. It's bizarre to throw out this jab at critics of Gemini, while simultaneously thinking it is unremarkable or even good that AI "Ethicists" spent a lot of time and resources making sure it drew pictures they did like because there is a "well-documented bias in training data" (and earning salaries for it to boot). But all of a sudden... >I’m not sure that “how did this happen?” or “why did this happen?” are all that interesting or enlightening questions compared to something like “well, what did you want the computer to do?” ...OP thinks it's dumb people care. Yeah it's not like there is a very clear pattern of ideological capture, doublethink, motte-and-bailey tactics and character assassination of Damore-like figures behind all this. Unless... >while I’m sympathetic on a visceral level to the concern that Gemini or ChatGPT might be spewing thousands of words of hate on command ...it involves something classified as "hate" in which case a "visceral" response would be perfectly understandable. So which is it? >I can’t really even come up with situation where Gemini’s refusal to say that Hitler is worse than Elon Musk has some terrible downstream effect. It's very simple. AI systems are already being used for official and non-official purposes. Cameras installed for explicitly limited purposes are already being used outside their original mandate. Police authorities have been caught multiple times using facial recognition like Clearview even if it's explicitly disallowed. In such a world, having naked ideological bias in general models is very dangerous. It's so dangerous that it turns into an immediate PR nightmare for the company involved, because there is no more fig leaf to hide behind. Their AI proudly and openly says the quiet part out loud, colors it in, and adds lovely sparkles to boot. The proper response isn't to go "lol u caremad", it's to take a step back and ask exactly the question OP claims is irrelevant: how the fuck did it get to this point, who is responsible, and why are the people who are supposed to guard against exactly this sort of thing excusing it or endorsing it?