Open research of the highest quality. Over 100 publications, including three Nature papers.
Applying our research for real-world impact in health, science, energy and more.
Deep science meets start-up energy and focus.
Committed to AI that benefits the many, not just the few.
Milestones
Latest News
View The BlogReinforcement learning with unsupervised auxiliary tasks
Our recent paper “Reinforcement Learning with Unsupervised Auxiliary Tasks” introduces a method for greatly improving the learning speed and final performance of agents. We do this by augmenting the standard deep reinforcement learning methods with two main additional tasks for our agents to perform during training.
DeepMind and Blizzard to release StarCraft II as an AI research environment
Differentiable neural computers
Announcing the Partnership on AI to Benefit People & Society
Working at DeepMind
Job OpportunitiesOur world class team consists of many renowned experts in their fields.