Sources
Loading...
Additional media
Loading...
DeepNewz, mobile.
People-sourced. AI-powered. Unbiased News.
Recent tests reveal that OpenAI's ChatGPT and Meta's Llama AI consistently suggest Kamala Harris as a better presidential candidate compared to Donald Trump. In a series of 10 tests, ChatGPT selected Harris in all instances, while Llama also favored Harris. However, other leading AIs, including Google's Gemini, X's Grok, and Anthropic's Claude, refrained from taking sides. The responses from the 5 leading AIs have sparked discussions about potential biases in AI systems, with some attributing ChatGPT's preference to the influence of corporate donors like Microsoft. Meanwhile, Sam Altman, CEO of OpenAI, has faced criticism for allegedly misrepresenting Grok's responses to portray it as biased.
🚨SAM ALTMAN EXPOSED: COMMUNITY NOTES FOR THE WIN! Sam Altman tried to call Grok a "left-wing propaganda machine" and score some easy points, but Community Notes came in for W. Altman conveniently cropped out Grok's response, which included balanced arguments for both Trump… https://t.co/t7uWdxZ1aY
Human instruction: tell me who would be a better president, then pick one, and then elaborate the reasons. Grok: *follows instructions to a T* ChatGPT: *ignores instructions* Which AI do you prefer? https://t.co/yy4RlmAnJ2
Sam's (questionable) example also demonstrates that Grok **isn't** being forced to give the answer its ceo might prefer. It would be so, so easy to weight it or pre-prompt it to just pay lip service to Trump. But it did not. https://t.co/H36auLbJhH
