Google co-founder Sergey Brin seems to have found an effective way to overcome the hallucinations of AI chatbots — simply threaten them with physical violence.
Artificial intelligence with a presentation ChatGPT in 2022, «has already taken over the world» — no matter how hyperbolic it may sound, technology is already gradually changing and revolutionizing many areas of our lives, from entertainment to medicine or education. However, developers have not been able to overcome the main problem of AI: chatbots still «lie» in their answers — an effect that official reports have dubbed hallucinations.
At the same time, Sergey Brin, co-founder of Google (the company also has Gemini’s own AI), there is original advice on how to make chatbots work without any fancy tricks — just threaten them with physical violence, or, for example, kidnapping.
«You know, this is a strange thing. We don’t share this too often in the AI community, but not just our models, but all models tend to do better when threatened with, for example, physical violence,» he said during a recent episode of podcast All-In (via Windows Central). «But… people are weird about it, so we don’t really talk about it. Just say it: «Oh, I’m going to kidnap you if you don’t do this and that, blah, blah, blah…».
Brin co-founded Google with Larry Page in 1998 and served as president until his retirement in 2019. He returned to an active role in the company two years ago — as an advisor to help to improve the Gemini AI user experience:
«Honestly, anyone who is a computer professional should not retire now… There has simply never been a greater opportunity — a greater peak in technology».
In an interview with Business Insider Brin stated that he himself actively uses AI to optimize management tasks, such as summarizing group chats, delegating tasks, and identifying promising employees. Interestingly, he also asked chatbots who in the team he should promote:
«The AI actually singled out a young female engineer that I hadn’t even noticed. I talked to the manager, and he said something like: “You know what, you’re right. She’s been working really hard.”»
Other tech executives are not neglecting the use of AI in their everyday tasks either: Nvidia boss Jensen Huang, for example, testing chatbots as tutors and the head of Zoom sends its AI avatar to boring financial reports.
Spelling error report
The following text will be sent to our editors: