DeepMind’s approach to AGI safety and security splits threats into four categories. One solution could be a “monitor” AI.

AI can now play the bestselling game Minecraft — and accomplish a multi-step task without being trained. A study published on ...

This work has identified four possible types of AGI risk, along with suggestions on how we might ameliorate said risks. The ...

Microsoft AI VP Nando de Freitas acknowledged Google's DeepMind for its new AI models while addressing restrictive employment ...

Google DeepMind has published an exploratory paper about all the ways AGI could go wrong and what we need to do to stay safe.

Though the paper does discuss AGI through what Google DeepMind is doing, it notes that no single organization should tackle ...

Google DeepMind shared a safety report with its perspective and warnings on AGI, but it didn't convince everyone.

Google DeepMind executives outlined an approach to artificial general intelligence safety, warning of "severe harm" that can ...

Dreamer is a Google AI model that learned how to play Minecraft and find diamonds on its own, without prior training.

DeepMind predicts artificial general intelligence (AGI) by 2030, necessitating new strategies to prevent potential threats to ...

The danger isn’t that the assistant is smarter. The danger is that we start to mistake intelligence for authority—and allow ...