December 29, 2008

Below the Capstone

The irony of the Internet and mobile communications, which link us together in what village idiot Tom Friedman calls a "flat" world, is that the promise of them is so poorly kept. Much has been made of mobile phones--with or without Internet access--as a "leapfrogging technology. And some businesses have leapt at that opportunity: Visa, for example, looks at the microcredit market and salivates like a trained dog. But there's a blind spot the size of Africa for most handset manufacturers, it seems. Samsung only lists South Africa for the continent on its website. Motorola files Africa under "other" on its net sales figures with 8% of the total (a category that I suspect also includes large chunks of South America). Apple--well, the iPhone's been subsidized in a lot of countries now. But I've got a feeling it's priced a bit out of the range of many consumers in the Global South. Like its competitors, its business model is flashy connections for rich people.

The one company clearly not ignoring the bottom of the pyramid is Nokia, which sells hundreds of millions of phones every year on a range from Symbian superphones like the N and E series to bottom-line models aimed at "emerging markets." Nokia doesn't get a lot of respect in the US, partly because its marketing is stunningly bad and partly because its S60 devices are not typically offered by American carriers. But the company sells hundreds of millions of simple, durable phones to developing markets, markets which Nokia specifically targets (features like multiple phone books, for instance, aren't particularly useful here but make a lot of sense where the phone may be shared or rented out). They're doing a lot of research on sustainability and use--Nokia researcher Jan Chipchase's blog is a fascinating glimpse into real-world uses of mobile tech. The company funds a number of projects for mobile innovation. And they've recently shown signs of trying to pitch themselves as a service provider for people in lesser-developed economies: their Ovi portal just rolled out a free e-mail app aimed at first-time users that requires no computer for set-up or use at all.

At the same time, Nokia recently announced a program it calls "Life Tools", a set of informational and educational services powered by SMS instead of GPRS (or, as we might expect in the USA, EDGE and 3G packet data). To start, the program will distribute market and weather information to farmers, and it will include English language training. A section for ringtones and astrology is also included, although it's obviously given less emphasis in the press release.

Although top-down technology projects like this rarely work, in my experience, this is honestly a pretty impressive attempt at addressing development problems with mobile tech. It's affordable, reliable, and relevant to the needs of customers (something which I'm not convinced OLPC ever understood). Design in Africa says that "Nokia shows how its done once again" while Kiwanja's Ken Banks is a bit more reserved. Banks raises a couple of good cautionary points that had also occurred to me--but let's come back to that in a minute.

First, let's look at what Nokia's doing right here. At the most basic level, they've admitted that there are parts of the world outside of the G20, which (as I noted above) is surprisingly rare. It's particularly surprising considering that the bottom of the pyramid (or BoP, as practitioners call it) is where the next big consumer segment will probably emerge. I mean, if you sell to the developed countries, these have become a pretty saturated market. Everyone's already got a phone here. That means you have to compete on shiny features and high-powered hardware, which is a turbulent place to be--just ask Motorola, who have been coasting on the RAZR for four years now, to ever-decreasing results.

Not only that, but the developing world's an excellent place to foster innovations that might also be profitable in rich countries. As Ethan Zuckerman is fond of saying, if it works in Africa, it'll work anywhere. It's easy to see how Nokia might be hoping to roll what it learns about building cheaper, sustainable, and more reliable electronics into its higher-end projects.

Software is less likely, of course, to move from lesser-developed countries to their rich counterparts. But give Nokia credit for embracing (sometimes, at least) open source technologies and relatively open ecosystems. All kinds of hardware can be attached to their phones, from keyboards to sensors to Wiimotes. They're simple to extend and program using Python, Javascript, Java, or C. The browser is WebKit, and they've got a port of Apache, believe it or not. You can install software on the phone without a computer, straight from the web or via SMS link. And no-one's acting as a gatekeeper or bottleneck: while the Symbian Signed initiative is a major misstep (one that will likely be remedied when the OS is open-sourced in 2010), in my opinion it's still better than a walled garden app store. In other words, although they may seem primitive to Americans, they're relatively open, cheap platforms for innovation and problem-solving.

Above all, Nokia actually seems to be interested in being a better global citizen, not just a better market player, which is one of the reasons I'm such a huge fan of the company. Granted, purely in business logic, being pro-development and pro-environment makes sense for a company: it's more profitable selling phones to people who possess money and health. Still, Nokia's entry into development still makes me nervous. I want to give it the benefit of a doubt and hope for the best, but caution still seems advisable for several reasons.

First of all, let's be frank: Western corporations don't have a great track record in development. I'll never forget what one African said at a conference I helped run while at the Bank: "These countries are not poor," he admonished the development institutions in attendance. "They were impoverished." And indeed, the history of interference from large companies has not been a happy one, from the controversy surrounding Nestle's infant formula to the use of military forces by Shell in Nigeria. Moreover, the influence of multinational companies on graft and bribery should not be discounted: while it is certainly a good governance issue, those who pay the bribes--and reap the benefits--need to be held accountable as well, if not more so.

There's a genuine reason, then, to be cautious whenever a company announces that it's making a major push into lesser-developed economies. The exploitation and damage that could be done--that has been done--is immense. And even with the best of intentions and smart people behind a project, the possibility for unforeseen consequences wrecking havoc among those least able to defend themselves is unsettling. Take it from someone who used to be a peon at the World Bank: having smart, well-intelligent personnel is no guarantee of success (or anything else, for that matter).

Ken Banks raises another cautionary point: Nokia's setting itself up to potentially control a massive amount of the information technology space, including most of the vertical infrastructure (services, software, and hardware). How far will they go with that kind of power? As Banks speculates, could they also integrate mobile payment systems, capturing that market? Will they partner with local services, or will they steamroller over them? The ability to integrate and share innovations is a powerful one--for development, for Nokia's bottom line, or both. If the future is built on proprietary Nokia tech, there's potential for lock-in rivalling anything that Microsoft, Google, or Apple could have ever hoped for.

I don't care for the fact that so much of this comes down to trust. Do I trust Nokia to keep doing good? Can we trust a company to act honorably when they've got real incentives to cut corners for profit? Can they be trusted to remain open when it counts? Noticeably, Nokia's open source initiatives are oriented a bit more toward their high-end customers than their emerging market phones. Optimally, I think we'd all like to see development from the bottom up, with local companies profiting. At the very least, competition would be good--and Nokia could be seen as simply working for competitive advantage in a marketplace soon to boom. As I've said, there are perfectly valid, harmless reasons for them to be there.

But as long as they're the only big handset manufacturer who's even really thought about selling to the BoP, instead of using low-end to build mindshare in developed countries, we won't really know the answers to those questions. That strikes me as a troubling state of affairs. There are worse companies to be taking these steps, I think, and yet that's small comfort given what's at stake.

May 15, 2008

Things Fall Apart

This letter from former OLPC security director Ivan Krstic is well worth reading. Krstic lays out the history of Nick Negroponte's "constructionist" educational failures, the managerial issues that have doomed deployment of the OLPC laptops to client countries, and the basic philosophical problems that continue to dog the machine's Sugar UI system.

There's also some tug-of-war in there about the conflicts between Linux, Windows, and evangelism on both sides, but I don't think that's nearly as interesting: it's fodder for the watchers on the sidelines, but not (entirely) why this project has failed.

As one of the people who predicted, early on, that it would indeed fail dramatically, I won't deny that it's a little satisfying. But mainly it's just sad. Poor countries paid good money for these machines. They bought into the Negroponte's hype. Well-meaning Americans did the same, hoping to benefit the lesser-developed nations through the Get-One-Give-One program. Those resources could have been used more effectively, and now they won't be.

It's worth pointing out, just to be clear, that the problem here was not technological. I don't think anyone doubted that the laptops could be built, or that they would be very cool toys. The OLPC has been a failure because it literally assumed that given the technology, learning would follow. It assumed that the laptops could form their own self-healing infrastructure, separate from any existing institutions. No need for a network: we've got mesh wifi! No need for power: we've got hand-cranked generators! No need for support: we're open-source!

No need for teachers: the kids will teach themselves.

But teaching is not something that comes from a box. It's part of a larger system, an institution built out of people. It's really hard work. You have to work hard to teach badly, and even harder to teach well. You can teach using technology. But technology on its own, as a general rule, is not in and of itself an education.

I didn't predict these exact failures when I wrote about the project a couple years ago. I was much more inclined to think in terms of capacity development. But in many ways, I'd argue that the lack of attention to capacity on the project's part is intertwined with its lack of attention to other basic social and pedagogical issues. Just as OLPC wasn't interested in building electrical or informational infrastructure, they weren't really interested in the workings of an actual classroom, or the problems of equipment distribution. Just as they simply assumed that their technology would make capacity development irrelevant, they figured that deployment and education would simply self-organize under the force of the OLPC's philosophical mandate.

Unfortunately, the real world doesn't particularly care about your philosophical mandate--just ask Marx.

August 14, 2007

No Teacher Left Behind

It is a little funny to watch the OLPC team run through their "world-changing laptop" schtick in the keynote presentation this morning, followed two sessions later by a Yahoo! team that points out that American students are as wired as anyone in the entire world, and yet one-third of them don't graduate from high school.

But if they just had a "constructionist" "school-in-a-box" with cute green plastic antennae...

Right.

April 11, 2007

NGOTube

I remember reading on Ars Technica that there are Youtube variations popping up all over the place now. There are a couple for porn, of course, but there's also GodTube (for fundamentalists) and RuTube (for Russians). I'd like to propose another, if it hasn't already been done: there needs to be a free streaming-video host for activists and small non-governmental organizations (NGOs) in developing countries.

From my perspective at the Bank, video streaming's going to increasingly be a big deal. It's getting faster and cheaper to shoot video nowadays, especially if you stay digital instead of going back and forth to tape. And if you don't need to edit, uploading to Youtube or Google Video is a one-step process. Once the video is there, it can be embedded anywhere you want, and the only payment required is a small amount of branding.

Even an organization like WBI has a very limited streaming capacity. We max out at about 300 low-bandwidth users. In Brussels, we blew through that almost immediately. And we're the World Bank: smaller organizations almost certainly can't afford to be hosting video for very many people. For those NGOs and civil society actors, something like Google Video or Youtube is a godsend. There's a tremendous potential for raising money and awareness through online video, but it also gives a tremendous amount of power to the services that host this material. And while Google has promised to "do no evil," for NGOs that are aware of how the company caved in China, that may not be terribly reassuring.

The Internet Archive may be one solution. But it doesn't allow embedding Flash, and it's not completely immune to political pressure, as when it bowed to the Church of Scientology. So we probably need something else. Sure, it's possible for an organization to purchase web space or rent from a provider, but why restrict activism only to those who have credit cards and reliable Internet access? Why force them to reinvent the wheel?

My favorite social movement in recent memory, the Otpor youth movement against Milosevic in Serbia, accomplished a lot of its work through grassroots activism and smart media stunts. It turns out that they also channeled large amounts of funds from Western donors who were captivated by their oddball non-violent resistance. Both the activism and the fundraising might be made much easier in other countries with a trustworthy Internet vehicle: an NGOTube.

Update: Colleagues note that some groups are already working on this--WITNESS, which documents human rights abuses, will be launching the Hub sometime this year.

November 16, 2006

Fit to be Thai'd

Thailand has pulled out of the One Laptop per Child program. Apparently the new military government thinks it's a bit silly.

Not that I expect a lot, but the immediate response from the Ars forum is, shall we say, bleak:

Right. It's not that the critics have valid objections (and don't just take my word for it, this is also a nice post from a WorldChanging contributor). It's that we're just ignorant Americans. That's a nice bit of rhetorical judo for a program that basically amounts to rich technocrats selling hardware to the lesser-developed countries without any pedagogical or infrastructural support. But then, the line that "we don't need a plan now, we're fast-paced venture capitalist geniuses with a secret plan" has long been an excuse of the program.

July 21, 2006

Pied Piper

CDM reports that the One Laptop Per Child program now has some of its groundbreaking educational software available for screenshots.

It's a music sequencer.

Good to know they're concentrating on the parts of development that are really important. I'm sure clients in Nigeria, Thailand, India, and China will feel much better about their $100 million investment.

It's also nice to see that their wiki still does not address the serious capacity concerns that I summarized from numerous critics a while back. Because for an education project to learn from the experts as opposed to attacking straw men--well, that would just be crazy.

This whole thing makes me so furious that I am honestly just about speechless. And we all know how hard that is to accomplish.

May 26, 2006

Let Them Eat Cake, Part II

I see that the One Laptop Per Child program has been revealed. Since I didn't get any takers for the op-ed I wrote a few months back, here's my perspective on it:

Let Them Eat Cake
By Thomas Wilburn

Before the guillotine gave Marie Antoinette her final haircut and took a bit too much off the top, the legend says that she responded to news of a bread shortage with the words "If they have no bread, let them eat cake." There is, in truth, no evidence that Antoinette actually said any such thing, but the line is too good to pass up. It's the perfect encapsulation of the rich who do not understand--nor want to understand--the poor.

At the recent Microsoft Government Leaders Forum in March this year, Bill Gates made disparaging comments about the One Laptop Per Child program, which aims to create a cheap laptop for third-world schoolchildren as an educational tool. He noted that its screen would be difficult to share, that there is no money allocated for support or networking, and that the applications were not yet developed. Since he is Bill Gates, despite his own substantial work in fighting malaria and other diseases, he was immediately set upon by the tech-savvy. How dare he try to disparage the work of the project's leader, former Wired columnist and venture capitalist Nick Negroponte?

But Gates was right: "Let them eat cake" should be the motto of the One Laptop Per Child (OLPC) program. It's a selfish bit of techno-utopianism that will do more harm than good. To understand why, we are going to have to spend some time with some basic development economics.

When organizations like the World Bank talk about learning programs for developing countries, they often use the term "capacity development." That's a fancy way of saying that communities should be taught how to develop themselves so that they're not reliant on outside parties for knowledge and technology. Before the problem of capacity became clear, development projects often failed as soon as the sponsor left, because they took knowledgeable personnel away with them, like a miniature brain drain. Development efforts still struggle to train workers and initiate a sustainable process--one that will continue after the rich sponsor has moved on.

The OLPC program, in theory, is designed as an educational tool. Of course, the project itself has not always been clear on what kind of education it plans to offer. Although originally pitched as a way for students to learn about technology, the primary justification now being offered is as a textbook replacement, most likely in a Wiki-style format. Putting aside worries about the mutability of electronic textbooks in the hands of potentially corrupt governments or corporations, what happens to the existing textbook industries for those countries?

Or even the existing IT industries, which are probably struggling already? Negroponte's plan won't employ labor, manufacturing, or software from the host countries. It simply dumps technology into the market, without taking into account the harm that could do to local companies. Instead of working within local markets to build a telecommunication infrastructure, the OLPC program calls for a spontaneously-generated, non-standard "mesh network" that will join the laptops to each other, and eventually to the Internet. Even assuming that this mesh network can overcome its technical problems and actually introduce students to the World Wide Web, it bypasses the wider population completely. It ignores the opportunity to build a real IT infrastructure, and the capacity to use it.

Finally, the problem of energy. OLPC has given up on hand-cranked laptops, because it's impossible for a child to generate enough power. That means that the machines must run on the local power grid. To most developed countries, that still sounds satisfactory because we have access to electricity 24/7. Yet there are 1.6 billion people out there who have no electricity at all, anywhere. Of the rest, many don't have constant access--they might have it for a few hours a day, or there might be regular rolling brownouts. Again, the laptop does nothing to address the problem of building electrical capacity. In fact, it completely skips over the problem--and treats that as a selling point, instead of the missed opportunity it really represents.

As with the other criticisms of the project, the real question becomes: isn't there a better place for that money to be spent? Wouldn't Negroponte be doing more for the developing world if he helped to build an electrical and information infrastructure with cheap recycled computers? Or even, as the Gates Foundation has done, poured money into disease research and vaccination programs? Let's not forget that the eventual target of the OLPC project is the very poor--people who, in many cases, live on less than a dollar a day. $100 spent on a laptop is money not spent on food, clean water, clothes, and medical care.

But then, money becomes an uncomfortable question when the project's organizers state that they will not be providing these for free, but will be asking host governments to buy the laptops with a minimum commitment of 1 million machines. $100 million dollars, to be spent on an unproven and so far unseen technology, in the developing world where more pressing problems abound, may go beyond foolish. The money being used, once it becomes a matter of large-scale government funding, is not Nick Negroponte's to waste. It is being taken from the poor themselves. It is difficult to believe, given the choice, that this is how they would choose to spend it--on cake, rather than the bread and essential services that they really need.

March 21, 2006

Let Them Eat Cake

I'm working on a draft to be submitted to various outlets, should my superiors here at the Bank allow it, about the One Laptop Per Child program. Suffice to say that I think it, like the quote attributed to Marie Antoinette, is a grave and harmful misunderstanding of the development process. I'll put my draft up here if it turns out that no-one will take it, but in the meantime the following two links provide a pretty good overview of the problems involved.

First, there's a series of blog posts by NGO head Lee Felsenstein, a computer pioneer who now works in developing infrastructure. Start here and then work forward through the posts using the menu at the top. He lays out problems with the hardware and the shady implementation so far--like the constantly shifting specs for these machines. He also pitches a telecenter as a better solution, which I would endorse--although at this point it would be best to disclose that I work closely with GDLN, the Bank-funded initiative for distributed learning, which does implement some of these solutions.

Second, an article linked from the Fonly Institute blog comments, from a Peruvian professor. It talks about the humanitarian and prioritization aspects of this project, which to me are even more convincing. But then, consider that these laptops won't be given to host governments. They're to be sold, with a minimum commitment of 1 million units--$100 million dollars on machines that may or may not have any positive effect at all. Perhaps the motive here is more profit than altruism. Nick Negroponte is one of the original dot-com investors, after all. And there are far more proven ways to help people with $100 million.

Finally, I will only say that the phrase "capacity development" should be ringing alarm bells all over this project. It is almost as if it were specifically designed not to build any useful capacity for developing countries. As far as I'm concerned, that's the most dangerous aspect of the plan--that it entails tremendous expense at the expense of other initiatives, both in monetary and opportunity costs.

June 16, 2005

Thinking Globally

I've been doing research on Global Public Goods, or GPGs, at the Bank these last couple weeks. GPGs are a paradigm for development work, and they have a lot in common with the Tragedy of the Commons. Remember that? Originally used by Garrett Hardin in a 1968 article, the phrase reminds us that when people share a public good for which responsibility is diffuse, they will tend to use it to its maximum capacity without bearing the full negative cost of their actions. For example, polluters who dump into the ocean do so because they reap all the benefits (cheap disposal of toxins) without claiming all the costs (the polluted water is "shared" with the rest of the ecosystem).

Global Public Goods are a bigger implementation of this idea. The International Task Force on Global Public Goods defines it this way:

In other words, GPGs are a way of addressing the commons on an international basis when individual countries won't. My task has been to figure out the cost of not addressing these goods, like pollution and disease, plus the costs of wider issues that we're lumping in with the GPGs, like migration, trade barriers, and technical standards.

The implications of GPGs when tied to their costs could be, I think, a powerful tool for getting the public interested in development. Right now, your average American couldn't care less about disease in Africa. On the other hand, malaria alone costs Africa millions of dollars every year in lost productivity, tardy medical care, and the continued spread of a disease that should have been wiped out long ago. That price is passed on to the US through disease, increased cost of keeping the problem from spreading, and losses for businesses who have to risk interacting with those countries. I personally think that development should be a moral and ethical responsibility, but those who are on the fence may be swayed by a more economic argument.

Speaking of economic arguments, have you ever taken a look at remittances, the money sent back overseas by immigrants to various countries? A paper by Jens Reinke and Neil Patterson of the IMF states:

The Bank notes that these remittances may go up to more than $200 billion by the year 2020--that's more than total official development aid! I'm not convinced, personally, that immigration can be shown to have hurt our economy (note that 1/3 of the current $90 billion goes to Latin America, probably from unskilled labor that US citizens aren't particularly interested in performing), but this certainly seems to show that it helps.

What other issues could be framed as a global public good? Can this be useful in encouraging progressive social policy?

Future - Present - Past