Aiming to strengthen its artificial intelligence arsenal, Microsoft announced a deal to acquire San Francisco-based AI startup Bonsai.
Bonsai already has ties to Microsoft, having been founded in 2014 by two former Microsoft engineers, Mark Hammond and Keen Browne. The company specializes in “deep reinforcement learning,” which can be used to teach autonomous systems within simulations, and believes this will work with Azure.
See also:
Hammond shared news of the acquisition in a blog post, saying that Bonsai and Microsoft have a desire to grow AI in automated systems. “Today, we are pleased to announce that Microsoft is acquiring Bonsai to accelerate the realization of this shared vision. By combining Bonsai’s unique deep reinforcement learning and machine education platform with Microsoft’s Azure platform and tools, extensive deep learning research, and recent AirSim simulation innovations, we will be able to offer our customers the most comprehensive toolchain for building, training, deploying, and managing AI for autonomous systems.”
Gurdeep Pall, Microsoft’s corporate vice president of business AI, said of the agreement in a blog post:
Today, we’re excited to take another big step in our vision of making it easy for developers and subject matter experts to build the “brains” that are machine learning models for autonomous systems of all types with the signing of the Bonsai acquisition agreement. Berkeley, California-based Bonsai, an M12 portfolio company, has developed a new approach using machine learning that abstracts away the low-level mechanics of machine learning. This allows subject matter experts to specify and train autonomous systems to accomplish tasks, regardless of AI aptitude. Actual training takes place within a simulated environment.
As a result of the acquisition, 42 Bonsai employees will join Microsoft’s AI and research group.
