Demis Hassabis on the web: AI Agents will cause big disruption.
Demis Hassabis: “I think there’s going to be a very interesting phase in the next few years on the web and the way we interact with websites and apps and so on."...
Training Language Models to Self-Correct via Reinforcement Learning - DeepMind paper
Google DeepMind's SCoRe method uses reinforcement learning to train language models to effectively self-correct through iterative revision, significantly improving their accuracy and reliability....
Agents Thinking Fast and Slow: A Talker-Reasoner Architecture - DeepMind research paper
In this paper DeepMind presents a new architecture for AI agents based on the concept of "thinking fast and slow."...