Webb points to China as an alternative. There, kindergarten-age students nationwide will begin studying a textbook this year that’s designed to teach students the new basics of knowledge they need to succeed in a computational future. “That’s the foundations of the kind of thinking that will allow them to work in conjunction with AI systems,” she concluded. “While everyone’s fixated on ‘Every kid must code,’ we are risking forgetting that every kid must learn to read and write.”
Charles Isbell, the executive director of Constellations and the incoming dean of computing at Georgia Tech, still sees computing education as a necessary step. “The real question,” he told me, “is: Are we interested in diversity or are we interested in integration?” Integration of women, people of color, and other under-represented voices would mean that the behavior of the entire industry would change as a result of their presence in that community. “Diversity is just membership. Integration is influence, power, and partnership.”
But integration is much harder than diversity. Isbell thinks that two separate conditions need to be met in order to accomplish it: “One is that the new folks are both capable and confident. The other is that the old folks are willing.”
Kamau Bobb, the global lead for diversity research and strategy at Google and a senior director at Constellations, isn’t so sure the tech industry is willing yet. A lot of people are involved in diversity, equity, and inclusion programs in Silicon Valley, and “those people are really committed,” Bobb told me. But their motivation is largely driven by providing access to the existing state of affairs. “They’re compelled by the argument that it just isn’t fair that more people don’t have access to the Google life—the free food and the power and the money,” Bobb said. Their goal is to get more people in the game, not necessarily to change the rules of that game. In this line of thinking, inclusion is first a problem of economic equity; any resulting social or moral benefits would just be gravy.
But for technical systems to take everyone into account, Isbell contends that representation must shift from an economic imperative to a moral one. “First you make the economic argument, and that’s where the industry is now, mostly,” he said. “Then you make the moral argument. That’s where you want to be. Until you win the second argument you haven’t won.”
In Webb’s view, that argument is unlikely ever to gain traction among big, wealthy tech companies. “A moral imperative is unlikely to motivate public companies,” she told me. Bobb agrees—Google’s focus on the “next billion users” entails a better understanding of people of color, he said, but only because the company finally understands that they represent an untapped market for advertising.