In the past, we’ve highlighted some West Coast AI research labs that we think are doing some really incredible work. Now, in an attempt to look past the dominating presence of Silicon Valley, we’re turning our focus overseas and taking a closer look at some of the cutting-edge AI research labs in Europe.
Founded in 2015, the Alan Turing Institute is a fairly new research lab with a unique structure. Located within the British Library in London, it’s a national institute comprised of 13 universities and the UK Engineering and Physical Science Research Council. This structure creates an environment that promotes collaborate across disciplines. Despite its youth, it has a well-defined mission and ambitious goals: to utilize data science and AI research to create positive change in the world.
Recently, the Institute turned its focus to investigating the complicated ethics in the use of AI algorithms and data analytics for predictive purposes by police forces. Working closely with police departments, the Institute hopes to help guide the discussion about utilizing AI and data science in policing to ensure ethical outcomes.
King’s College London’s Department of Informatics offers programs for both undergraduates and postgraduates and tackles both the theoretical and practical changes caused by the emergence and growth of digital societies. Organized into five “hubs” and several groups, the lab conducts research in areas as diverse as Autonomous Systems and Urban Living.
One of Informatics’ groups, Distributed Artificial Intelligence, is dedicated to investigating the intersection of, and interaction between, artificial intelligence and society. Through its research it hopes to ameliorate bias and improve the efficiency of social and economic policies.
With a strong business orientation, the IBM Research lab in Ireland is dedicated to improving outcomes for its clients and discovering possible areas of growth in IoT/Digital Twin, Privacy, and Cloud, among others.
Recently, IBM Research has been working on automating the development of customized neural networks. Recognizing that demand exceeds the supply of data scientists capable of building neural networks, they developed NeuNetS. NeuNetS automates much of the process of building a neural network, enabling those without strong coding abilities to create AI models customized to their business problem.
IBM AI Research, an affiliated group, also offers a toolkit designed to help identify, and ultimately remove, bias in artificial intelligence. You can access the toolkit here and learn more about its background here.
With a history spanning 90 years, the Fraunhofer Heinrich Hertz Institute (HHI) is likely the oldest research lab on our list. Located in Berlin, Germany, the HHI investigates how advancements in technology will influence the future, particularly in communications. One of HHI’s groups, the Machine Learning group, focuses its research primarily on deep learning.
A recent project, MATH+, seeks to shine a light into the black boxes that have been such an integral part of deep learning algorithms. With MATH+, the HHI hopes to make it possible to understand what factors contribute to a given outcome.
Located in Cambridge, UK, Microsoft Research lab’s stated goal is to “transform the world through deep research.” To further this goal, the lab has identified six focus areas, or “themes”, All Data AI, Cloud Infrastructure, Confidential Computing, Future of Work, Game Intelligence, and Healthcare Intelligence.
Their recent publications have investigated how to apply various tools, such as reinforcement learning and deep neural nets, to solve real-world and business problems. Microsoft Research also has a number of tools, datasets, and code publicly available, which can be found here.
Even though J.P. Morgan’s AI Research team is based in New York, they have locations all over the world. With a focus on improving business outcomes and their clients, the lab is focused on using AI and Machine Learning to improve the financial service business.
A subgroup of Information Engineering, the Machine Learning Research Group at Oxford University (in Oxford, England) uses statistics to handle both information and uncertainty in a variety of research fields, including citizen science, biology, public health, autonomous intelligent systems, and animal husbandry.
Oxford ML Research is also one of the more notable research labs investigating quantum computing. Recently, Oxford University has partnered with Lockheed Martin and Nokia for a project on Quantum Optimization and Machine Learning (QuOpaL). Although, the project is interested in investigating any and all possible methods for Quantum computing, one method they are particularly focused on is adiabatic quantum optimization.
Overall, there’s some really exciting work being done in Europe, on the West Coast, and all around the world. The industry is ever-expanding, constantly pushing the boundaries of what we know to be possible, and the next major development could be from one of these research labs, or the best research you do in your garage.