Scientific advances, real-world benefits

WaveNet generates realistic human-sounding speech that reduced the gap between computer and human performance by over 50%, when it was first introduced. It now powers the voice of the Google Assistant.

Our technology is helping doctors diagnose acute kidney injury (AKI) up to 48-hours earlier than current methods. With early detection, patients get better preventative care, avoiding invasive procedures, and reducing costs.

We worked with Google Health, Northwestern University, Cancer Research UK and Royal Surrey County Hospital to develop an AI system that can better identify breast cancer in X-rays across populations and systems.

AlphaStar is the first AI to reach the top league of StarCraft II without any restrictions. Understanding the potential and limitations of open-ended learning like this is a critcial step towards creating robust systems for real-world domains.

AlphaZero learned to play three famously complex games, becoming the strongest player in history for each. Learning entirely from scratch, it developed its own distinctive style that continues to inspire human grandmasters.

A great challenge in AI is building flexible systems that can take on a wide range of tasks. Our Deep Q-Network (DQN) made progress on this goal when it learned how to play 49 different Atari games using only raw pixels and the score as inputs.

The differentiable neural computer (DNC) can use its external memory to answer questions about complex structured data, such as stories, family trees, or a map of the London Underground.

While becoming the first computer program to defeat a professional human Go player, AlphaGo taught the world new knowledge about perhaps the most studied and contemplated game in history.

The Generative Query Network (GQN) allows computers to learn about a generated scene purely from observation, much like how infants learn to understand the world.