Proponents of Nicholas Negroponte's One Laptop Per Child project (OLPC, to its friends) have high hopes. Take, for instance, Eben Moglen (long-time legal counsel for the Free Software Foundation), who in a recent speech, says its communications capabilities can help stop genocide:

"What is journalism like when every village has a video camera and is on the net? ...

What does it mean if the next time somebody starts some nasty little genocide in some little corner of the Earth the United States government would prefer to ignore, that there's video all over the place all the time in every living room?

This isn't the first time these sorts of hopes have been advanced for new communications technology. In the 19th century, serious people suggested that the telegraph alone might end war altogether. (You heard less of this sort of thing after World War I). And there were plenty of starry-eyed techies in the 1990s who would tell you with a perfectly straight face that it just wasn't technically feasible for mere, stodgy, obsolete governments to censor the internet --- right up until their friends at Cisco and Yahoo taught the Chinese how to do it.

But I digress. I really did mean to talk about OLPC, and the actual state of the project, which (as is often the case) can be better determined from release notes than manifestoes and mission statements. So, from the release notes for the "B-1" prototype build of the One Laptop Per Child X-O machine (via OLPCnews):

Enough is now present to begin to sketch the outline of where we believe the children's software should go: enabling the construction of software in which children and teachers can easily collaborate is central to our vision.

So, the software isn't there yet, but they know where it should go. Which isn't a bad thing, by itself --- everything, at some point, has to be a work in progress. But for those who tuned in late, they are planning to begin mass production of these machines for "pilot" deployments next calendar year. And a "pilot" deployment, for these guys, means not the small, carefully monitored trial in a few schools that one might expect, but deployment to all kids in an entire country --- Negroponte scoffs at smaller trials: "To do a [smaller] project is ridiculous!" (But they're going to start with a small country. That's why it's a pilot project. You don't want to get overly ambitious).

And yet for all the acknowledged technical brilliance of the people working on OLPC, versions of eductional software which are specialized for target audiences and for the machines themselves (with their unusually small screen size --- one of the cost-saving measures) don't yet seem to exist; the project's own list of shipped software is so far pretty basic, and the content more so --- the featured text is an introductory algebra "wiki-book" in English only, where even sections 5-10 of chapter two are, as I write, content-free skeletons of their hoped-for future selves. [Ed: these links added late.]

Now, it's easy to say that the risk here is low --- that the laptops are cheap, and the kids themselves will surely figure out something to do with them. (Which, by itself, is worrisome enough to folks who know what real kids actually do with computers when the adults around them are totally clueless). But while the machines may be cheap by first-world standards, a little math shows that a full deployment is a huge expense for a third-world country --- so if the OLPC projects' hopes aren't realized, and current educational programs, meager as they already are, get displaced, there is the real potential for lasting damage here.

So, I don't like the grand vision, at least not the way it's being pursued. Does that mean I don't like the project? Those little boxes are amazingly seductive. If the deployment model is all wrong, fix that. Construct a sales organization which can handle orders of less than a million a pop, accept the inevitable markup, and sell the damn things. If they're good for anything at all, the folks in the third world will very quickly figure out what, as they did with cell phones. They'll probably come up with uses that first-worlders would utterly fail to anticipate. (And in the context of OLPC, it's worth noting how most of the cell-phone users described in those articles first saw them as adults. One of the most annoying symptoms of OLPC KoolAid Poisoning is a firm, nearly psychotic conviction that children and only children will readily adapt to any new tech toys you give them. Bullshit.)

I feel badly about writing this because considered solely as a laptop project (and not the "education project" that Negroponte wants it to be instead), the thing is brilliant. There are multiple major technical breakthroughs in this machine. But what we have here is actually two projects, not one: a brilliant technology project for developing a cheap, low-power laptop, and a proposal for a government project in the third world using that laptop. And while there are things that I think government does better than anything else (like health care), this particular proposed government project is, so far, shaping up as the kind of starry-eyed, utopian boondoggle that makes libertarianism look good. For once, the libertarian solution looks good as well. Figuring out the best uses for new technology is one of the things that free markets do better than anything else --- and particularly better than government-driven top-down mandates (especially in countries where the governments aren't all that good to begin with). So, why not use markets?