Why AI must learn to admit ignorance and say 'I don't know' | New Scientist
“So many of these models… are people-pleasers,” says David Noever at People Tec, a US technology firm. “To the point ofhallucination.” …
newscientist.com 6
Covers: Hallucinations & FabricationChatbots & LLMsEducationHealthcareAI Safety & Existential Risk
Over 70 per cent of students in US survey use AI for school work | New Scientist
Many children in the US seem to be using a chatbot to help them with their work, despite therisksof errors and often in defiance of school …
OpenAI's warnings about risky AI are mostly just marketing | New Scientist
OpenAI CEO Sam Altman has warned about thedangersof AI. Chona Kasinger/Bloomberg via Getty Images. OpenAI has announced a new AI model …
Resurrecting loved ones as AI'ghosts' could harm your mental health | New Scientist
There are alsorisksthat AI hallucinations could give uncharacteristic or harmful advice to a mourner. “I think some people might see these as …
Realism of OpenAI's Sora video generator raises security concerns | New Scientist
The latest AI model capable of converting text into video is attracting praise from some AI researchers, while also raising concerns about the …
There's a 5% chance of AI causing humans to go extinct, say scientists
New Scientist In the largest survey yet of AI researchers, a majority say there is a non-trivialriskof human extinction due to the possible development of …