And probably not fixable: Opinion | Google’s Gemini AI exposes tech’s left-leaning biases – The Washington Post (archive.is)
Google rushed Gemini AI to the market to pretend to compete with Chat GPT (which also has some similar problems). Google is now attempting to, apparently, add kludges to manually tune the biased output problem of Gemini AI.
That means their AI architecture is fundamentally broken. They can’t keep putting Band-Aids on it forever.
Google’s Gemini AI has also ruined the reputational aspects of LLMs for a long time to come. Anyone trusting LLM AI output is demonstrating they are not very bright. This is a huge, serious problem for AI right now.
Related: WSJ says much the same – Google Mired in Controversy Over Gemini AI Chatbot Push – WSJ
Gemini AI is so bad that – and much of this seems self-inflicted intentionally – Google needs to start firing people.
