PAIR

Humans have always been different from one another. While the champions of equality would glare if one mentioned the obvious inherent inequalities amongst peoples and nations, by now most of us have learned to accept the differences amongst us. In Japan that has a huge animation market a person’s expectation from AI assistants are different from the Uber driver that shuffles around Bali. Google had something like this in mind when they launched a new artificial Intelligence initiative this week to ensure that the upcoming AI program will consider the differences in human behavior and varied expectations.

The program has been designed keeping people at the center of the process. Google has announced that the PAIR initiative (People+AI research initiative) will bring together the researchers and technicians to study and redesign the ways people interact with the AI systems. The main focus of PAIR will be to understand the tricky ‘human side’ of AI. They are working to   comprehend the relationship between users and their expectations from technology and how AI can be more inclusive considering the diversity.

Users have been disappointed when Apple’s Siri failed to comprehend our commands or voices incapability in performing a certain task. To remedy such lapses the team of PAIR will also design AI systems that have greater capability in understanding human commands and also use customer feedback to improve such Artificial Intelligence weaknesses.

Since users have different needs one of the goals of PAIR will be to develop an AI that helps and assist professionals at their work. Therefore the AI systems that will be developed for assisting doctors will be different from the AI system that would be developed for farmers, musicians, professors, designers, pilots or managers. Such diversity among professionals can be the inspiration for the PAIR team to design AI systems that would be particular in its output to aid experts in their respective areas.

Two Google staff scientists ‘Martin Wattenberg and Fernanda Viegas’ will be leading the PAIR team. They have recently released two open source tools that has helped AI engineers to understand the data they use to train AI systems. These tools would have also helped the engineers that are not experts in the field of AI to get a better grip on the system.

Google has announced that the project will have external research help and consultation. They will be will be working with Prof. Brendan Meade of Harvard and Prof. Hal Abelson of MIT-both of them experts in their fields of ‘education and science in the age of AI.’

The PAIR project comes at a time when Google has clearly sets it sight on being an AI powerhouse, if not the sole producer of AI. They believe that this sector could have huge impact on bringing in more business for Google through its various processes.

The PAIR program currently has twelve full time staff members but will be having the advantage of gaining the expertise and guidance from the various units of Google. Such diversity in guidance  will surely give help it to emerge as a powerful system. Google hopes that the next AI systems will be more efficient and diverse. Just like the humans we are.

LEAVE A REPLY

Please enter your comment!
Please enter your name here