Intel's departure from Nicholas Negroponte's low-cost laptop program was inevitable as the workings of Moore's Law - but not necessarily a bad thing.
Getting the world's biggest chipmaker onboard was an important step for OLPC because of its market and political clout. Intel is to the chip sector what Nokia is to handsets, only more so. Just as really only Nokia has the scale to make very cheap phones profitably, only Intel has the ability to create a real global business out of making processors that will drive cheap laptops.
But it was always an uneasy partner. Understandably, Intel wasn't happy about pushing laptops containing product from rival AMD, though it claims that wasn't an issue.
The chip giant resigned from One Laptop Per Child (OLPC) project in early January, claiming it had been asked to dump its own laptop initiative, Classmate. Nicholas Negroponte, the MIT professor behind OLPC, denies that. Indeed, from OLPC's point of view, he points out that the more companies making low-cost chips the better.
Whatever the official explanation, it's hard not to believe the real reason is that Intel thinks it can do a more profitable job from making cheap notebooks itself. If that's the case, we all wish them luck.
The mobile industry would argue that the best single device for closing the digital divide is the cellphone. It would be right. The world has nearly three billion mobile phone services, compared to one billion PCs. The mobile doesn't require its owner to be literate, costs as little as $30 and provides connectivity worldwide.
So that's that settled. But the OLPC is not about closing the digital divide, but narrowing it at a point where people can jump across for themselves.
Which comes first‾
In particular, it's about education. The starting point for the OLPC project is that children are evolved to learn in their own way rather than in the recent tradition of sitting at a desk in front of a teacher and blackboard.
But that runs into established educational wisdom and is liable to provoke quotes such as this one from the Nigerian education minister: 'What is the sense of introducing One Laptop per Child when they don't have seats to sit down and learn; when they don't have uniforms to go to school in, where they don't have facilities‾'
The logic is irresistible, then one remembers this is Nigeria, an oil-rich kleptocracy whose education system is known best for its production of semi-literate fraudsters. It turns out that the minister is in discussions with Intel and Microsoft about their low-cost computers. So he's not against laptops in classrooms per se, just the terms.
Of course, the OLPC is just one part of the mix of educational investment, along with spending on school buildings and fittings and teachers.
The biggest challenge for OLPC though is not to produce a $100 laptop - the cost is now down to $188 - but to overcome inertia in education thinking. Teachers are not natural early adopters of technology, and are hardly likely to accept that children can learn just as well without them.
Children are certainly excited by internet-connected laptops. They are stimulated them into thinking, playing and problem-solving.
For all that perhaps, ultimately, they are not useful educational tools for children in developing countries. Perhaps there is a far better economic return on teachers and books and school desks. Perhaps a local assembly model is better and fairer.
But an aggressive Intel challenging OLPC in the low-cost market expands the opportunity for all of that to be tested. Intel's departure from OLPC could lead many others to follow.
The Classmate is in trial in a handful of Asian classrooms.