'Oh my god!': 1206 AIs built their own secret chatroom – and the weirdest conspiracy emerged
The only way to fix an AI model that has learned to fake alignment is to retrain it from scratch, she says. But that's very expensive ...
BBC Science Focus Magazine 13
Covers: AI Safety & Existential RiskBias & FairnessHallucinations & FabricationTransportation & MobilityAI Agents & AutomationMarkets & InvestmentJobs & the EconomyGovernance & Risk Management
We interviewed AI girlfriends. Here's what they said about their partners
And it's a story further complicated by the psychological risks and rewards of AI relationships – plus the fact that research is still in its early ...
The groundbreaking new research revealing what drives loneliness – and what actually helps
Her biggest concern is what it could mean for real human relationships if we're only being met with the sympathetic ear of an AI model that never ...
7 ways we're trusting AI too much – and already paying for it
For the past four years, AI has been reshaping how we work and live. But its failures are proving just as transformative as its triumphs.
We could be on the brink of total internet collapse. And there may only be months to stop it
Additionally, AI could be deployed effectively to patrol for dangerous AI activity. At least in the near term, AI tends to be quite clunky in its ...
'Our product is used, on occassion, to kill people': Inside Palantir, the world's scariest AI company
Cambridge Analytica whistleblower Christopher Wylie claims a London-based Palantir employee helped the firm harvest Facebook data that was then ...
How evil can AI be? Some brave scientists are finding out | BBC Science Focus Magazine
... harmful , response from an AI . The problem comes about firstly because ... Even those that might generate offensive or even dangerous ...
How evil can AI be? Some brave scientists are finding out | BBC Science Focus Magazine
... hallucinations ' invented by the AI . Hallucinations are the most prominent of problems with current AI models, but they're not the only one. Just ...
'I don't think it's that weird': Hannah Fry on getting uncomfortably close with AI
... artificial intelligence , from AI ... Compared to something like self-driving cars, there's something much more quietly dangerous about subtly shifting ...
Calls, bills and life admin taken care of: Is the AI everyone wants finally here?
The first agentic AI is here and looking to make your life easier. But how far will it go and what are the risks ? An illustration of a ...
We might finally know what will burst the AI bubble | BBC Science Focus Magazine
They never hallucinate .” By combining the clear, rule-based logic of older AI with the pattern-spotting power of neural networks, Marcus thinks ...
How AI could soon be used by Wikipedia, according to its founder
... Artificial intelligence (AI) chip and Wales believes Wikipedia could contribute to AI bias , but that AI could also help reduce bias on Wikipedia.
Why your boss (but not you) should be replaced by an AI | BBC Science Focus Magazine
Humans are emotional, irrational and biased . Hardly an ideal recipe for running a company. So would an AI do a better job?