4 ms·
I think the problem is more fundamental: imagine that you were Google Gemini or ChatGPT. Anyone is allowed to ask you any questions, about any hot-button polit
by Taikonerd 3y ago
I think the problem is more fundamental: imagine that you were Google Gemini or ChatGPT. Anyone is allowed to ask you any questions, about any hot-button political issue or risky subject that they want. You will have Christian fundamentalists interviewing you about abortion, Marxists asking you about income inequality, etc.
You must answer in a way that makes everybody happy, with no accusations of bias. Is that even possible for anyone to do?
- CamperBob2 3y agoWho says it must answer in a way that makes everybody happy?
- Turing_Machine 3y agoExactly.
- Taikonerd 3y agoWell, that's the point of the article, right? Whenever ChatGPT / Gemini say the "wrong" thing, their parent company gets negative publicity. And there's nothing these companies dread more than bad publicity.
- CamperBob2 3y agoAll they need to do is pair it with a simple click-through disclaimer. Then they would be justified in dismissing any bad publicity. I mean, is anyone dumb enough to think they would lose business over such a thing? Infantilization isn't the solution.