Google has announced a new project aimed at improving human-AI interaction. Being called ‘PAIR’ (People plus Artificial Intelligence Research), the initiative claims to be focusing at the ‘human side’ of AI: the relationship between users and technology.
According to the company, PAIR initiative will look into the new applications and tools that AI can provide for developers and how to make it broadly inclusive. The project will also be releasing open source tools for third party users ranging from supporting engineers, researchers to everyday experiences with AI.
PAIR website claims that its goal is to do fundamental research, invent new technologies and create frameworks for design in order to drive a humanistic approach to artificial intelligence.
There are three broad sectors that PAIR is focusing at: Engineering and analysis – For easily building and understanding machine learning systems. Augmented expert intelligence – looking at helping professionals from different fields amplify their expertise. And how everyday users can be made AI inclusive, so that they can take advantage of various tools and applications.
Along with the announcement of PAIR, the company has also open sourced two visualization tools, Facets Overview and Facets Drive. The two applications are aimed at AI engineers, and address the very beginning of the machine learning process. The Facets applications will be able to give engineers a view of the data they use to train AI systems. The company claims that with Facets, engineers will be able to easily debug and understand what is being built building.
Google Brain researchers, Martin Wattenberg and Fernanda Viegas will be leading the PAIR team consisting of 12 employees. According to media reports the team will be collaborating with various institutions outside the company such as Harvard and MIT.
The company had collaborated with Facebook, IBM, Amazon and Microsoft to launch the Partnership on AI last year, and claims that PAIR will be complementing the goals of the partnership such as inclusion, open source research and transparency on AI research.