ChatGPT's 'winter break' is the latest sign of model drifts
News, voices and jobs for CIOs. Optimized for your mobile phone.
Stay updated with breaking news from Apollo Research. Get real-time updates on events, politics, business, and more. Visit us for reliable news and exclusive interviews.
News, voices and jobs for CIOs. Optimized for your mobile phone.
AI bots can lie and perform insider trading Asia News Network (ANN) is the leading regional alliance of news titles striving to bring the region closer, through an active sharing of editorial content on happenings in the region.
Researchers have shown that an artificial intelligence (AI) chatbot using a GPT-4 model is capable of performing illegal financial trades and cover it up too.
The potential for AI bots to deceive their makers is great enough that they could become a catastrophic risk to humanity, researchers claimed at the UK AI Safety Summit last week.
In a troubling development, an artificial intelligence bot on OpenAI's GPT-4 large language model was presented with a factual scenario that led it to execute an illegal insider trade, even after being told that doing so was illegal. Perhaps even more troubling, the bot lied about what it had done w...