Over the next three blogposts, we’re going to share with you brief descriptions of the papers we are presenting at the NIPS 2016 Conference in Barcelona.Read More
Working with the NHS to build lifesaving technology
Were very proud to announce a groundbreaking five year partnership with the Royal Free London NHS Foundation Trust.Doctors and nurses in the NHS do a phenomenal job caring for patients, but theyre being badly let down by technology. Pagers, fax machines and paper records are still standard in most NHS hospitals, and too often top-down IT systems dont meet clinical needs because they are built far away from the frontline of patient care.This slow andoutdated technology means thatimportant changes in apatients condition often dont get brought to the attention of the right clinician in time to prevent further serious illness.When this doesnt happen, the consequences for patients can be severe, and even fatal. At least ten thousand people a year die in UK hospitals through entirely preventable causes, and some 40% of patients could avoid being admitted to intensive care, if the right clinician was able to take the right action sooner.Our partnership aims to change that, by taking a very different approach to building IT for patient care.Together we are creating world-leading technology, in close collaboration withclinicians themselves, to ensure thatthe right patient information gets to the right clinicians at the right time, reducing preventable deaths and illnesses.Read More
Reinforcement learning with unsupervised auxiliary tasks
Our primary mission at DeepMind is to push the boundaries of AI, developing programs that can learn to solve any complex problem without needing to be taught how. Our reinforcement learning agents have achieved breakthroughs in Atari 2600 games and the game of Go. Such systems, however, can require a lot of data and a long time to learn so we are always looking for ways to improve our generic learning algorithms.Read More
DeepMind and Blizzard to release StarCraft II as an AI research environment
Today at BlizzCon 2016 in Anaheim, California, we announced our collaboration with Blizzard Entertainment to open up StarCraft II to AI and Machine Learning researchers around the world.Read More
Differentiable neural computers
In a recent study in Nature, we introduce a form of memory-augmented neural network called a differentiable neural computer, and show that it can learn to use its memory to answer questions about complex, structured data, including artificially generated stories, family trees, and even a map of the London Underground. We also show that it can solve a block puzzle game using reinforcement learning.Read More
Announcing the Partnership on AI to Benefit People & Society
We believe that AI has the potential for transformative, positive impact in the world. Fulfilling this potential is not only dependent on the quality of the algorithms being engineered and the data they use, but on the level of public engagement, transparency, and ethical discussion that takes place around them.Its precisely because AI has the potential to have such a major positive impact on the world, that we believe its critical that we build new models of open collaboration and accountability around it. Thats why we at DeepMind are really proud to have worked with Amazon, Google, Facebook, IBM and Microsoft, to form a non-profit organisation that aims to create a forum for open discussion around the benefits and challenges of developing and applying cutting edge AI. Together, we hope to advance public understanding of AI and formulate best practices on some of the most important and challenging ethical issues in the field.Read More
Putting patients at the heart of DeepMind Health
From the outset, weve wanted DeepMind Health to be a truly collaborative effort. Too much hospital IT has been developed from a top-down perspective, often repurposing technology built for completely different sectors thousands of miles away from the NHS frontline. The result: tools that remain out-of-date and imperfectly suited to clinical use, contributing to a patient safety challenge where more than 1 in 10 patients suffer harm during an in-patient stay.We think its possible to transform this through bringing some of the worlds most advanced technology to the NHS. But for this to have any chance of meaningful impact, we know it must have the input of patients and clinicians at its heart.Yesterday we took a step towards that goal by hosting our first open patient and public forum in London, with over 130 patients, carers and members of the public coming to our offices and many more watching on our livestream.Read More
WaveNet: A generative model for raw audio
This post presents WaveNet, a deep generative model of raw audio waveforms. We show that WaveNets are able to generate speech which mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by over 50%. We also demonstrate that the same network can be used to synthesize other audio signals such as music, and present some striking samples of automatically generated piano pieces.Read More
Applying machine learning to radiotherapy planning for head & neck cancer
Were excited to announce a new research partnership with the Radiotherapy Department at University College London Hospitals NHS Foundation Trust, which provides world-leading cancer treatment.1 in 75 men and 1 in 150 women will be diagnosed with oral cancer during their lifetime, and oral cavity cancer has risen by 92% since the 1970s.Read More
Decoupled Neural Interfaces Using Synthetic Gradients
Neural networks are the workhorse of many of the algorithms developed at DeepMind. For example, AlphaGo uses convolutional neural networks to evaluate board positions in the game of Go and DQN and Deep Reinforcement Learning algorithms use neural networks to choose actions to play at super-human level on video games.This post introduces some of our latest research in progressing the capabilities and training procedures of neural networks called Decoupled Neural Interfaces using Synthetic Gradients. This work gives us a way to allow neural networks to communicate, to learn to send messages between themselves, in a decoupled, scalable manner paving the way for multiple neural networks to communicate with each other or improving the long term temporal dependency of recurrent networks. This is achieved by using a model to approximate error gradients, rather than by computing error gradients explicitly with backpropagation. The rest of this post assumes some familiarity with neural networks and how to train them. If youre new to this area we highly recommend Nando de Freitas lecture series on Youtube on deep learning and neural networks.Neural networks and the problem of lockingIf you consider any layer or module in a neural network, it can only be updated once all the subsequent modules of the network have been executed, and gradients have been backpropagated to it.Read More