Stanford aims to soften AI's impact with new institute

The rise of artificial intelligence is raising a hornet's nest of complex, nuanced and scary issues for society in general and the broad technical community developing these powerful platforms, in particular. The Stanford Institute for Human-Centered Artificial Intelligence (HAI), which will be housed in a new 200,000-square-foot building at the center of the campus, aims to push AI in a more humane direction.

A Washington Post story describing HAI's kickoff says that the founders of HAI realize that to have an impact it has to go big—in terms of budget and names. Bill Gates will deliver the keynote at its initial symposium tonight.

The genesis of the organization is a conversation in 2016 between AI pioneer and former Google Vice President Fei-Fei Li and John Etchemendy, a philosopher and former Stanford provost. Other big names involved in HAI are former Yahoo chief executive Marissa Mayer, Google Executive Chairman Eric Schmidt, and LinkedIn co-founder Reid Hoffman. The goal is to raise $1 billion for grants to create technology and to sign some of the talent that now is in the private sector.

Both business and academia are starting to pay attention to AI ethics. For instance, computer vision, a key element of AI, doesn't do as well in recognizing the faces of people of color. Voice recognition struggles with nonstandard English accents, and algorithms aimed at predicting parole violations are "rife with racial bias," the Washington Post says. HAI sees the importance of addressing these issues while the technology still is relatively immature.

Elon Musk has issued warnings about AI, calling it "humanity's existential threat."