DeepMind And Unity Join Hands In Researching AI

A branch of Alphabet - DeepMind - is working with Unity in speeding up the process of learning and researching machine and AI (artificial intelligence). They are planning on concentrating highly in “virtual environments” especially the ones that they can use to visualize and conduct tests on experimental algorithms.

For now, the collaboration has not been revealed much. However, there has been a recent announcement that widely shows us the companies' agreement and intention.

In a press release, Deepmind's Co-founder and CEO Demis Hassabis stated his excitement to collaborate with Unity in generating virtual environments where the kind of flexible and smart algorithms can be developed and tested and them used to address the problems in real world.

This can be considered as a perfect combination. Unity is among the biggest and most famous video game developers.

A lot of studios use AI to fulfil immersive, huge worlds, especially the ones including opponents with high intelligence and challenging manner. Therefore, the medium is obviously connected with community of AI and broader machine studying, whose purpose is to address issues from autonomous vehicles to diagnoses in medical field.

The company has been working on fostering this connection using an open-source plugin namely the Unity Machine Learning Agents Toolkit. The software is free and it assists the training process on smart agents and companies with techniques in machine studying and sample settings in the engine of Unity. It is hosted on GitHub.

DeepMind comes from a developing video game background, too. For example, Hassabis that worked particularly on simulating famous gods and creatures at Lionhead Studios, dubbed Black & White before establishing Elixir Studios and generating BAFTA-nominated titles, such as Republic: The Revolution and Evil Genius. DeepMind, which was bought at $650 million in 2014 by Google, has used champion Go players and Atari games to test AI algorithms.