Google's 'PAIR' project wants AI and humans to get along

From Engadget - July 10, 2017

The PAIR team will be led by Google Brain researchers Fernanda Vigas and Martin Wattenberg and its 12 full-time staff members will also work with researchers outside of Google, such as Harvard University professor Brendan Meade and MIT professor Hal Abelson. Its mission sounds similar to another group Google helped launch last year -- the Partnership on Artificial Intelligence to Benefit People and Society -- which is a collaboration with Facebook, Microsoft, IBM and Amazon. PAIR, according to Google, will be complementary to the Partnership on AI, which is also working towards goals like open research platforms and transparency in AI studies. Both initiatives appear to have some overlaps with other like-minded groups, like OpenAI, as well.

Two efforts already in the works are open-source tools that improve how datasets are viewed. Facets Overview and Facets Dive give developers details on their datasets that can help them see how their training data are falling short, which could lead to issues in AI bias. The effects of how training data can lead to biased AI were made clear in Microsoft's racist and sexist Tay AI.


Continue reading at Engadget »