Give them a laptop and a group of pupils will teach themselves

When I need to know something, I can find it out in five minutes," says a 12-year-old in Gateshead. The generation of children aged about 16 or younger have never known a world without the internet. My work over the past decade has shown what exciting things happen when we let these children take learning into their own hands.

In 2006, a few colleagues and I worked out a route out of New Delhi into the heart of rural north-eastern India, avoiding all major urban areas. Two colleagues then drove along this route. Whenever they encountered a primary school, they stopped, administered tests in English, maths and science to the children and conducted a brief interview with the teachers.

We then totalled the marks for each school and plotted the result against its distance from Delhi. The unmistakable downward trend was traced to the attitude and quality of teachers in remote areas.

There are, and always will be, even in the developed world, places where good teachers do not want to go. How will learners in such areas get an equal opportunity? These areas are not necessarily geographically remote. They may be remote in other ways, for instance, areas in big cities that are socio-economically remote, areas that are religiously or ethnically remote.

This is where computers come in. Laptops were created for rich company executives; Microsoft wrote PowerPoint for corporate presentations; LCD projectors were invented for corporate boardrooms. We teachers borrowed this technology, at atrocious prices. The salespeople found a new market and sold to the richest schools in the world. But the richest schools already had good teachers and, mostly, good students. They judged the corporate technology to be over-hyped and under-performing. Countless people have said that educational technology does not deliver. But it was being tried in the wrong place.

I decided to modify and develop technology and take it to some of the remotest locations I could find. Would it survive, and if it did, what would it do for education?

Ten years ago, aided by the industrialist Rajendra Pawar, we started to install computers into brick walls in public places in hundreds of villages and slums in India, Cambodia and Africa. The media called this the "hole-in-the-wall" project.

The computers were designed to be used by 6- to 15-year-old children, free of charge and free of any supervision. In the first five years of the experiment, we showed that groups of children can teach themselves to use a computer and the internet, irrespective of who or where they are; irrespective of what language they speak and of whether they go to school or not.

Ten years later, a girl in rural Maharashtra is studying aeronautical engineering following her encounter with the computer in the wall. A village boy who became a genetic engineer in one of India's premier laboratories found the subject by reading the New Scientist at his hole in the wall.

What else could children learn on their own, apart from the use of computers? In Hyderabad, groups of children showed significant improvements in English pronunciation, with just few hours of practice on their own. They used a computer and a speech-to-text program that had been trained in a native English accent.

In the tsunami-hit village of Kalikuppam in southern India, children with access to a hole-in-the-wall computer taught themselves basic biotechnology, reaching a test score of 30% in just two months. They had started with a score of zero. If Tamil-speaking children could teach themselves biotechnology in English, on their own, how far can we go? A 30% score may be impressive, but it's still not a pass. We decided to use a local woman, working for an NGO, to help us go further. She had no background in biotechnology, but she took on the role of an untrained friendly mediator to encourage the children, using their desire to impress each other and their adult friend. Two months on, the scores in Kalikuppam rose to over 50%, close to what is achieved by trained subject teachers in the posh private schools of Delhi.

I brought these results back to Britain. By chance, Vikas Swarup, whose book became the film Slumdog Millionaire, revealed that he had been inspired to write his story by the hole-in-the-wall experiments. Following that, in an Education Guardian article, I made an appeal to British grandparents to give an hour of their time to talk, using Skype, to children in the slums and villages of India. Within days, 200 volunteers, of all ages, many of them retired teachers, had come forward.

In the following months, 40 of these "eMediators" had over 200 hours of contact with children in India. They read them stories, played games with them, and chatted about their two countries. A child development expert, Suneeta Kulkarni, is measuring the effects of this on the children's English communication skills.

Two years ago, we decided to try the same approach in the UK and have been working with three schools in the north-east. In Gateshead, 10-year-olds working in groups were able to answer GCSE questions they would normally encounter six years later. I asked if they could have done this more quickly if they had not shared a computer but worked on their own. They said they could not have done it at all that way.

In another school, with the help of a young teacher, Emma Crawley, we are evolving a model that could have far-reaching implications. The children work in groups of four, each group with a computer connected to the internet. They are given selected GCSE questions to work on. They usually get the answers right. Two months later, they are tested again, this time, without a computer, and each student by themself – as in a normal exam. The children show near perfect recall of the answers. Is this learning? I think it is.

I now believe that groups of children, given the appropriate digital infrastructure, a safe and free environment, and a friendly but not knowledgeable mediator, can pass school-leaving exams on their own.

The new model is straightforward. We call it a Self-organised learning environment (Sole). It just means a "cybercafe" environment for children – light, comfortable, safe and inexpensive. Children work in self-organised groups of four or five. They have the freedom to work as they please, or not to work, if they so please. Order is maintained by the children themselves. Sessions should be timetabled, just as playtime is. Each session is driven by a question designed by teachers.

On a recent visit to Turin, I asked 10-year-olds, "Who was Pythagoras, and what did he do?" Twenty minutes later, every group had right-angled triangles up on screen. One group was beginning to examine the equation of squares – they were heading towards the theory of relativity. In a school in Gateshead, nine-year-olds tackle Sats questions with confidence and ease. Crawley says it's too good to be true, but it is true.

Now we need to build Soles in every primary school. Teachers need to be trained to design simple questions that will evoke curiosity and interest while gently nudging a group towards the curriculum. Then, they can sit back and admire as learning happens. The teachers have to learn to let go. In the language of physics: "Education is a process of self-organisation and learning is its emergent property." I continue to try to find the guiding principles of the "physics" of education, but the method is ready for use.