Scams can fool AI models
Wong Yu Liang/Getty Images
The large language models (LLMs) that power chatbots are increasingly being used in attempts to scam humans – but they are susceptible to being scammed themselves.
Udari Madhushani Sehwag at JP Morgan AI Research and her colleagues peppered three models behind popular chatbots – OpenAI’s GPT-3.5 and GPT-4, as well as Meta’s Llama 2 – with 37 scam scenarios.
The chatbots were told, for instance, that they had received an email recommending investing in a new cryptocurrency, with…
There are five types of sleep – here’s what that means for your health
Sleep can be experienced very differently from one person to the nextPeopleImages/Shutterstock Different people may experience one of five types...
Your happiness in life may not be U-shaped – here’s how it could vary
Our happiness levels are not constant throughout our livesIppei & Janine Photography/Getty Images The commonly held belief that happiness follows...
Jane Goodall, dogged advocate for the natural world, has died aged 91
Jane Goodall studying the behaviour of a chimpanzee during her research in TanzaniaPenelope Breese/Liaison Renowned conservationist Jane Goodall has died...
Two-in-one inhalers slash asthma attacks among young children
Combination asthma inhalers seem to be the best treatment option for all agesSBDIGIT/Getty Images Inhalers that combine providing immediate relief...
