Microsoft Research AI for Science

AI for Science is a global organization of researchers and engineers dedicated to advancing an exciting new research frontier at the intersection of machine learning and the natural sciences. Specifically, deep learning has the potential to have a transformational impact on natural sciences by dramatically improving our ability to model and predict natural phenomena. Read more about the team in our overview.

We seek to drive major advances with lasting influence in the tools for scientific discovery through the use of machine learning, with a focus on ‘fifth paradigm’ scenarios. Through these advances, we aim to empower real-world impact on some of the most pressing problems facing society including climate change, green energy, sustainable materials, and the discovery of new drugs.

The fifth paradigm of scientific discovery uses machine learning to accelerate the simulation of the fundamental equations of nature. These equations describe natural phenomena with exquisite precision over vast ranges of length and time scales, but are computationally extremely expensive to solve. The fifth paradigm complements and amplifies the first four paradigms.

Learn more about the fifth paradigm in our Microsoft Research Blog.

Jim Gray, a Turing Award winner, and former Microsoft Technical Fellow, characterized the historical evolution of scientific discovery through four paradigms: (1) empirical, (2) theoretical, (3) computational, and (4) data-intensive.

We are an active member of the scientific community, through peer-reviewed publications, participation in key conferences, and collaborations. We hope to deliver scientific advances that have lasting value. You can read more about our published work, people and projects.

The work that the AI for Science team is doing has the potential to accelerate the solution of fundamental equations by multiple orders of magnitude and to facilitate more rapid discovery of new materials and molecules with direct relevance to addressing some of society’s most pressing challenges. Microsoft seeks to empower individuals and organisations by creating tools and services to enhance scientific discovery. Moreover, research for new materials and catalysis mechanisms is directly relevant for Microsoft’s commitment of becoming carbon-negative by 2030 (opens in new tab), while research on biology and drug development can directly support Microsoft’s activities in health and life sciences.

Together with our partners, collaborators, and customers we aim to facilitate real-world impact at a global scale on some of the most pressing problems facing society including climate change, green energy, sustainable materials, and the discovery of new drugs.

Our team works on a number of exciting and ambitious projects, in a highly collaborative and multidisciplinary way. Here is a small selection. More can be found on our projects tab.

• Graphormer, a deep learning package that allows researchers and developers to train custom models for molecule modelling tasks, such as materials science, or drug discovery
• Generative Chemistry (opens in new tab), a collaboration with Novartis (opens in new tab), where together we are empowering scientists with AI to speed up the discovery and development of break-through medicines.

We view strategic collaboration with external partners as key to collectively addressing some of the most pressing challenges facing our society and scaling the impact that we can have. We have no one-size-fits all approach to external partnerships, but instead engage in a small number of highly strategic relationships, such as our collaboration with Novartis on the use of AI to speed up the discovery and development of medicines.

The global team encompasses world experts in machine learning, quantum physics, computational chemistry, molecular biology, fluid dynamics, software engineering, and other disciplines who are working together to tackle some of the most important real-world challenges facing humanity. They are based in sites in China, the US, the UK, Germany and the Netherlands. You can find open roles here.

Reach out to contactai4science@microsoft.com to discuss specific scenarios. We will review any enquiry on a case-by-case basis. However, the number of such engagements will be very limited and aligned to our strategic priorities.

Microsoft Research offers support for PhD Scholarships via our annual PhD Fellowship Program.

Yes, we welcome applications for internships from those in machine learning, the natural sciences, and software engineering. Keep an eye out for open internships here.

Yes, we welcome applications from Master’s and PhD students.