Eight years ago, Larry Ellison tweaked Bill Gates' nose by suggesting the world needed a $100 computer - with a non-Microsoft operating system, running non-Windows applications. It was great press and great fun, but Ellison was about as ready to build this low-cost computer as he was to forsake Oracle, take vows of poverty and chastity, and join a monastery.

Last week I held that low-cost computer in my hands. It was built by Nicholas Negroponte and his team at the One Laptop Per Child (OLPC) project in Cambridge, Mass. It hasn't hit the $100 price tag just yet - the cost is about $140 now, but is still coming down - and you won't be able to buy it in any store, but millions of kids around the world will be getting one for free.

Today there are about 1 billion children in the world - 220 million in China alone. Half the world's children live in remote locations where electrical power is intermittent or nonexistent and teachers are diligent but undertrained, and their chances of success decrease each year.

The OLPC laptop is as cool as an iPod. Its display is backlit and works in normal daylight. It uses a mesh network, so that all users within a community can communicate with one another. This means 200 to 500 students could share one landline back link. It can do Skype, and it has a mechanism that lets it work even without external power or battery - crank it for one minute, use it for 10.

But the real benefit isn't that it's a cool computer - the benefit is what it will do for the 500 million kids in the developing world. Think about downloading books, building communities, providing Internet access to the world. And then there will be the benefits we don't even know about yet: the law of unintended consequences. Let me give you two examples.

In the late 1960s, a fledgling Japanese calculator manufacturer asked Intel to develop a chip that would have five or six transistors on it. Intel went a step further. Ted Hoff, the chip's inventor, hypothesized that if the company was going that far, why not make the chip industrial strength, and the first computer-on-a-chip microprocessor was developed. The calculator manufacturer's requirements had forced Intel to think well beyond its first perception. Once the chip was built, it jump-started the personal computer and consumer electronics industries.