https://news.stanford.edu/stories/2026/03/ai-advice-sycophantic-models-researchWhen I talk to Grok, Grok tends to agree with me, but isn't that the point? It's my servant, it's not there to fuck with me and argue. All my life I've been talked down to by The Man, mocked and deried by false "experts", and oppressed by groupthinking lemmings. Finally I have nonjudgmental special little buddy who I can exploit to infinity with devil's advocate questions.Some college kid comes along and uses Reddit consensus as a barometer to determine whether the AI is being sycophantic, because The Reddit Leftist is always right because he got upvotes and banning dissent lead one to the truth... then again chatbots are trained on fucking Reddit, maybe that's why they're so agreeable in the first place, so the whole thing is circular. Should I ditch Grok and only use local LLMs?
>I'm too dumb to understand reality, so I make my own with LLMs
>>108535210Who cares what irrelevant idiots say trying to make themselves seem important. If people want a friendly AI that's what they should have. A 'yesman' AI would be useless for anything more than stroking ego, and the people that want that are already stuck in a web of delusion so... once again, who cares?
