Demis Hassabis warns AGI is still years away, as advanced AI models excel in complex tasks but falter in simple reasoning.

Google DeepMind CEO Demis Hassabis says companies are far from achieving AGI, as even their advanced models, such as Google’s Gemini, can make basic mistakes despite excelling in complex tasks.

Hassabis's position aligns with Google CEO Sundar Pichai, who has dubbed the current stage of development "AJI" — artificial jagged intelligence. Pichai used the term on an episode of Lex Fridman's ...

He warned of an example where an AI model tried to blackmail an engineer about an affair it learned about in an email.

Demis Hassabis, Google Deepmind CEO, notes AI models like Gemini paradoxically struggle with basic math despite acing advanced Olympiad problems. This inconsistency, termed "artificial jagged ...

OpenAI CEO Sam Altman recently revealed that GPT-5 didn't meet the coveted AGI threshold because of its inability to ...

Now, the field of AI safety and its promise to prevent the worst effects of AI is motivating young people to drop out of ...

At Meta, a chaotic culture and lack of vision have led to brain drain, with rivals saying its AI talent is lackluster. But ...

Microsoft is targeting Meta’s top AI staff with multimillion-dollar offers, internal docs show, amid rising stakes in the Big Tech talent war.

Forget sentient robots or cures for cancer, AI’s single notable achievement so far has been to prop up a stock market that, ...