Sugata Mitra gave street kids in a slum in New Delhi access to a computer connected to the Internet, and found that they quickly taught themselves how to use it. This was the moment he says he discovered a new way of teaching.

He calls it the grandmother technique, and it goes like this: expose a half dozen or so kids to a computer, and let them have at it. The only supervision required is an adult to listen the kids brag about what they learn. It’s the opposite, he says, of the disciplinary ways of many parents—more like a kindly grandmother, who rewards curiosity with acceptance and encouragement. And it is a challenge to the past century and a half of formalized schooling.

Since this first experience in 1999, Mitra has been working to extend the notion of self-organized learning to address the needs of poor children, especially in developing countries, who have little or no educational resources. He is convinced that school children can teach themselves just about anything—that they can achieve educational objectives without formal direction. For these kids, formal education, at least as practiced in the U.K., where he is professor of educational technology at Newcastle University, is of little help.

His ideas, however, have implications for formal education in the west, too. Mitra doesn’t have kind words for English schooling, which he says is better suited to the needs of the British empire than the age of Twitter. England ran three quarters of the globe through a vast bureaucracy that relied on the ability of clerks to write letters and tally spreadsheets by hand. Competency in reading, writing and arithmetic was paramount, and formal classroom teaching was the best way to instill the three Rs. But as the tools of education have changed radically, schooling hasn’t. The British system, he says, “was a phenomenal achievement, but it's out of date. It's not needed.”

The question is, what is needed—or what will be needed in the future? Mitra thinks self-organized learning will be an important part. “There may be 10 different ways to do this. I believe I have touched on one of the ways.”

Last night Mitra won the $1 million TED Prize for his work. He will use the money to establish a lab in New Delhi that will put his ideas of a "School in the Cloud" to the test. The lab will be set up as a kind of cyber café, where 48 kids at any one time can go to learn English, considered in India to be key to any child's future. Volunteer “grandmothers”—retired school teachers, for the most part—will participate via Skype to lend guidance. The cyber café will serve as a lab to see how self-organized learning can be scaled globally. “I want to see if this is feasible,” he says. “What are the technical problems, what are the management problems? If it works, we'll have a technique that will level the playing field, and that is the big missing piece.”

Self-organized learning is potentially disruptive to traditional education in the west, and in talking about it Mitra has alienated some teachers. For now, he's keeping to the developing world, and to the teaching of English.

His long-term ambitions go further, however. “My agenda," he says, "is to see how far this can go.”