Support the news
The Massachusetts Institute of Technology is launching a new initiative aimed at producing better artificial intelligence.
The effort is expected to require hundreds of millions of dollars from private donors and corporations.
One aim of the project, called MIT Intelligence Quest, is to learn about the human mind and how it's expected to change as it increasingly collaborates with machines. Another goal: to reverse engineer the human brain and apply those lessons to developing better artificial intelligence.
Josh Tenenbaum, an MIT professor of cognitive science and computation, says right now, AI is about recognizing patterns. But that's not the only way humans learn.
"Imagine if we could build a machine that grows into intelligence the way a human being does, that starts off like a baby and that learns like a child," Tenenbaum says. "This is really the oldest dream of artificial intelligence."
Tenenbaum says this dream is contained in Alan Turing's 1950 paper introducing his famous Turing test to determine if machines can think.
"The human child is the only actually known scaling route to intelligence in the universe," says Tenenbaum. "It's the only system we know that starts off not knowing very much, it seems, and [then] knowing everything that a human being comes to know. Imagine you could build a machine like that."
Tenenbaum says now researchers are in a position to do so. Still, he predicts that this will be a decades-long quest.
MIT's initiative would bring together 200 to 250 neural and computer scientists as well as robotics researchers already on campus, but who are not necessarily collaborating with one another yet.
Click the audio atop this post for a Morning Edition interview with Tenenbaum.
This segment aired on February 1, 2018.
Support the news