AlphaGo Zero: Starting from scratch

Artificial intelligence research has made rapid progress in a wide variety of domains from speech recognition and image classification to genomics and drug discovery. In many cases, these are specialist systems that leverage enormous amounts of human expertise and data.However, for some problems this human knowledge may be too expensive, too unreliable or simply unavailable. As a result, a long-standing ambition of AI research is to bypass this step, creating algorithms that achieve superhuman performance in the most challenging domains with no human input. In our most recent paper, published in the journal Nature, we demonstrate a significant step towards this goal.Read More

Strengthening our commitment to Canadian research

(French translation below)Three months ago we announced the opening of DeepMinds first ever international AI research laboratory in Edmonton, Canada. Today, we are thrilled to announce that we are strengthening our commitment to the Canadian AI community with the opening of a DeepMind office in Montreal, in close collaboration with McGill University.Opening a second office is a natural next step for us in Canada, a country that is globally recognised as a leader in artificial intelligence research. We have always had strong links with the thriving research community in Canada and Montreal, where large companies, startups, incubators and government come together with ground-breaking teams, such as those at the Montreal Institute for Learning Algorithms (MILA) and McGill University.We are delighted that DeepMind Montreal will be led by one of the pioneers of this community,Doina Precup, Associate Professor in the School of Computer Science at McGill, Senior Fellow of the Canadian Institute for Advanced Research, and a member of MILA. Doinas expertise is in reinforcement learning – one of DeepMinds specialities – which is critical for areas such as reasoning and planning.In her new position, Doina will continue to focus on fundamental research at McGill, MILA, and DeepMind.Read More

WaveNet launches in the Google Assistant

Just over a year ago we presented WaveNet, a new deep neural network for generating raw audio waveforms that is capable of producing better and more realistic-sounding speech than existing techniques. At that time, the model was a research prototype and was too computationally intensive to work in consumer products. But over the last 12 months we have worked hard to significantly improve both the speed and quality of our model and today we are proud to announce that an updated version of WaveNet is being used to generate the Google Assistant voices for US English and Japanese across all platforms.Using the new WaveNet model results in a range of more natural sounding voices for the Assistant.Read More

Why we launched DeepMind Ethics & Society

At DeepMind, were proud of the role weve played in pushing forward the science of AI, and our track record of exciting breakthroughs and major publications. We believe AI can be of extraordinary benefit to the world, but only if held to the highest ethical standards. Technology is not value neutral, and technologists must take responsibility for the ethical and social impact of their work.As history attests, technological innovation in itself is no guarantee of broader social progress. The development of AI creates important and complex questions. Its impact on societyand on all our livesis not something that should be left to chance. Beneficial outcomes and protections against harms must be actively fought for and built-in from the beginning. But in a field as complex as AI, this is easier said than done.As scientists developing AI technologies, we have a responsibility to conduct and support open research and investigation into the wider implications of our work. At DeepMind, we start from the premise that all AI applications should remain under meaningful human control, and be used for socially beneficial purposes. Understanding what this means in practice requires rigorous scientific inquiry into the most sensitive challenges we face.Read More

The hippocampus as a predictive map

Think about how you choose a route to work, where to move house, or even which move to make in a game like Go. All of these scenarios require you to estimate the likely future reward of your decision. This is tricky because the number of possible scenarios explodes as one peers farther and farther into the future. Understanding how we do this is a major research question in neuroscience, while building systems that can effectively predict rewards is a major focus in AI research.In our new paper, in Nature Neuroscience, we apply a neuroscience lens to a longstanding mathematical theory from machine learning to provide new insights into the nature of learning and memory. Specifically, wepropose that the area of the brain known as the hippocampus offers a unique solution to this problem by compactly summarising future events using what we call a predictive map.The hippocampus has traditionally been thought to only represent an animals current state, particularly in spatial tasks, such as navigating a maze. This view gained significant traction with thediscovery of place cells in the rodent hippocampus, which fire selectively when the animal is in specific locations. While this theory accounts for many neurophysiological findings, it does not fully explain why the hippocampus is also involved in other functions, such as memory, relational reasoning, and decision making.Read More