SMART CITIES Outtakes, Take 1: The OLPC
My editor Brendan Curry was pretty gentle, but a couple of sections of SMART CITIES never saw the light of day for various reasons. With more than a year since publication, there doesn’t seem to be much harm from sharing them now.
Here’s one about Nick Negroponte’s quixotic efforts to develop a cheap laptop for the developing world.
Connect the Children
Not everyone absorbed the lessons of the first wave of ICT4D efforts. The Media Lab’s co-founder, Nicholas Negroponte, was among them. His next foray into ICT4D, the One Laptop per Child (OLPC) project, would be the movement’s highest profile failure.
Co-founded by Negroponte and Jerome Wiesner in 1985, the Media Lab had long focused on the role of technology in learning. Negroponte also dabbled in education philanthropy, and in 1999 he presided over the opening a new school he’d financed in Reaksmy, a rural village in Cambodia’s Preah Vihear Province.[i] A kind of precursor to Lincos, sans the architectural drama of the shipping container, the school had a generator, a satellite dish, and laptops. The technology was a powerful magnet and enrollment at the school surged. Negroponte began publicly extolling the benefits of putting laptops in the hands of kids. In 2005 he launched One Laptop per Child (OLPC), a new organization devoted to the cause, and took leave from MIT to lead it.
For OLPC to achieve the universal impact implied by its name, Negroponte and his team had to dramatically slash the cost of computers, quickly earning his proposed design a catchy nickname, “the $100 laptop”. The plan was to use a cheaper components (most importantly, a new low-cost display) and open source software to power a clever, frugal new device that would eschew the bells and whistles of consumer machines. It would be sealed and ruggedized for life in the harsh physical conditions of the developing world, while a peer-to-peer wireless networking capability would allow classrooms to create their own local networks if an Internet connection was unavailable. A hand-cranked generator, like those found on emergency radios, would free the machine from inadequate and unreliable electricity grids. Negroponte’s design vision for OLPC pushed the bounds of technological possibility, captivating engineers and enthusiasts around the world. As the initial media hype faded, the OLPC team got down to work.
Immediately, the project began to falter. The first thing to break up was the laptop itself. The novel hand-cranked generator, which in early versions was built into the long joint between the laptop’s screen and keyboard, required so much force to turn that it would shear off from attachment to the thin plastic case — as was spontaneously demonstrated by United Nations Secretary-General Kofi Annan at a 2005 press conference in Tunisia.[ii] Revised designs for the generator first moved it to a tethered external attachment, but problems with the power supply continued to plague the project. The XO-1, the first production model, used a traditional plug-in AC power adapter, but a bad design put strain on the cable and poor manufacturing quality resulted in frequent breakage.[iii] In one surprising bit of luck, the laptop’s cute but fragile-looking wireless ears proved more rugged. Intended to increase roaming range and serve as access panels for audio and USB ports, they seemed prone to damage. But during testing, they were able to withstand multiple kid-high drops onto a concrete floor.[iv] Despite progress on the hardware side, the OLPC had bigger problems. As Negroponte later admitted, after a few weeks toying with it, Steve Jobs had emailed his evaluation. “The software is some of the worst I’ve ever seen… Please don’t be too mad at me…”[v]
After recovering from those early design flaws, after a slow start large orders for the XO-1 started rolling in by the autumn of 2007. Uruguay bought 100,000 units, Peru over a quarter million. Mexican billionaire Carlos Slim bought 50,000 for his nation’s schools. It was far from the 100 million units a year Negroponte had predicted in 2006, but OLPC was beginning to have an impact.[vi]
But the window of opportunity was shrinking fast as laptop makers quickly brought their own low-cost designs to market. Intel had joined the project’s board that summer, even while it competed head-on with OLPC in developing countries with a low-cost computer it called the Classmate PC. But Negroponte reportedly asked Intel not only to halt production of the Classmate, but to stop supplying chips to other laptop makers like Asus for their own low-cost designs.[vii]
Did another OLPC move made Intel reconsider its support? In the beginning, OLPC pledged that it the laptop would not be sold at retail where it might compete directly with consumer products. But as the 2007 holiday season approached, OLPC launched the “Give One, Get One” program. For a mere $399, geeks could satisfy their techno-lust, and help poor-country students as well. The OLPC site announced “you will be purchasing two XO laptops — one that will be sent to empower a child to learn in a developing nation, and one that will be sent to your child at home.” It was a clever stunt to mobilize rich world donors — more than 80,000 in all — but it seems to have backfired.[viii] Intel resigned from the OLPC board during the first week of the new year.
Things didn’t go much better when the laptops reached the field. There, the XO-1 proved to be a logistical nightmare. As one analysis reported in 2010, in Uruguay — the second largest deployment of the OLPC — nearly a quarter of the 400,000 laptops were out of commission. The custom-designed keyboards and touch pads failed frequently under heavy use — one-third of reported problems in Uruguay required replacing the keyboard.[ix] During a visit to a school in Birmingham, Alabama, less than two years after some 15,000 XO-1s were distributed at a cost of $3.5 million to the city, researchers found 70 percent of students had problems with their laptops. Problems were so widespread that “in each school, some students interrupted the… survey to ask us whether we could fix their XOs.”[x],[xi]
OLPC mesmerized the tech community, but it puzzled development experts. One of the scheme’s most troubling assumptions was that poor countries would finance computer purchases through textbook replacement. The math worked on paper — the laptop would be designed for a working life of at least four years, and poor countries spent an average of $20 annually per student on textbooks, a figure Negroponte cited often in the project’s early days. When critics voiced concerns over such a sudden shift in teaching tools, Negroponte lashed out at development officials and educators whose slow, methodical approach to reforms visibly frustrated him. Calls from outside observers for pilot projects and independent evaluations of the tool’s pedagogical impact seemed a particular peeve. In one speech in 2009, he admonished them for questioning the educational benefits of laptops, which were to him as obvious as the benefits of electricity. “The fact that somebody in the room would say the impact is unclear is to me amazing — unbelievably amazing.”[xii]
In reality, few countries could seriously consider such a risky, wholesale shift from textbooks to laptops until such a plan could be proven to produce a better education. But without the textbook savings, the laptops would remain prohibitively costly. Another financial misconception involved technical support, a regular feature of technology deployments that involves a kind of Pareto rule:,hardware typically is just 20 percent of the total cost of ownership. The other 80 percent goes for service and support. But Negroponte seemed oblivious, often describing the devices as “child-repairable”.[xiii]
In the end, the actual total cost of ownership for the XO-1 turned out to be closer to $75 per year per student, nearly four times Negroponte’s original estimates.[xiv] A good chunk of the money was spent acquiring content for the devices, which in Negroponte’s plan would be downloaded from free outlets like the country’s native-language version of Wikipedia. The fears of skeptics were confirmed — OLPC was competing head-on with textbooks for scarce education funds.
Given all of its flaws, it almost didn’t matter that the $100 laptop never quite reached its promised price point. Negroponte’s original plan in 2005 was to take only orders for a minimum of one million units per country, enabling the manufacturing economies of scale that would make the $100 figure realistic. But by 2011, OLPC’s price was still nearly $200 — the XO-1 retailed for $188 in the U.S.; the current model, the XO-1.75, costs $185 for orders of 10,000 or more ($195 for 1,000 or more, and $205 for 100 or more). In total, according to the OLPC website just 2.5 million units have shipped, far short of the more than seven million Negroponte forecast in 2006.[xv] The systemic change in education he envisioned has failed to materialize, as deployments were scattered across over 40 countries, some in clusters as small as 100 units. What’s most disappointing is that the economics never really worked for the poorest countries. The vast majority of the laptops — over 80 percent at the time of this writing — were distributed in countries hardly considered poor by international standards: Uruguay, Peru, Argentina, and Mexico have all substantially reduced poverty in recent decades.
While it may never sell in numbers to make a difference for the world’s poorest, OLPC showed there was a market for a cheaper, stripped-down breed of laptop. Netbooks soon came into vogue. Priced around $300–400, they were at least half as expensive as full spec computers, and cash-strapped families of developed economies in recession snatched them up posthaste. By 2010, some 30 million were being sold annually worldwide.[xvi] Ironically, OLPC’s greatest legacy by far has been to broaden access to computing for the less well-off people of the more well-off nations.
Negroponte might be forgiven for failing to appreciate the challenge of actually deploying the laptops. As a technologist, he was primarily, perhaps inordinately, focused on bringing them to life. What is more puzzling is his apparent failure to see tablet computers coming, and the enormous impact they would have on education.
In April 2010, Apple stunned the world by selling nearly 500,000 iPads in the first week after it launched (starting at $499 in the United States). It was the death knell for netbooks, sales of which would peak that year and rapidly decline. OLPC was clearly taken by surprise, and scrambled to release a concept design for its own low-cost tablet a few weeks later. With his broad experience and understanding of the deep currents of thought in computing, Negroponte shouldn’t just have seen tablets coming — he ought to have pushed OLPC and the industry there sooner. Tablets open a whole range of possible interactions in the classroom that laptops never could. Groups can gather more easily around them, and they can be casually passed between students and teachers. Combined with gesture and language recognition, they allow a whole new way of interacting with computers and information. As Negroponte himself reminded a group gathered for MIT’s 150th anniversary in 2011, 50 percent of the world’s megacity inhabitants are functionally illiterate. A huge opportunity to bring computing to the poor had opened, but OLPC was still just building cheaper versions of 1990s laptops.
Even as OLPC struggled to catch up, Negroponte hadn’t seemed to learn from the setbacks and disappointments of the launch of the XO-1. Rather than fix the deployment model, rethink the focus on technology as a cure-all, or bring in educators to explore other kinds of classroom technologies that might have a greater impact, the new tablet represented another shot at what seems to have been the real goal all along — creating a masterpiece of frugal technology design. The new tablet, the XO 3.0, was unveiled in January 2012, once again with a proposed price of $100.[xvii]
[i] “What If Every Child Had A Laptop?,” CBS News, Last modified February 11, 2009, http://www.cbsnews.com/2100-18560_162-2830058.html.
[ii] Ken Banks, “Intel, OLPC Affordable Laptop Bout Only Hurts Users,” PCWorld, Last modifed May 5, 2008, http://www.pcworld.com/article/145500/article.html.
[iii] Mark Warschauer and Morgan Ames, “Can One Laptop Per Child Save the World’s Poor?,” Journal of International Affairs, Fall/Winter 2010, 64(1):41.
[iv] “Antenna reliability,” The One Laptop Per Child Wiki, Last modified November 10, 2007, http://wiki.laptop.org/go/Wireless#Antenna_Reliability.
[v] Nicholas Negroponte, “Nicholas Negroponte: Steve Jobs — Influence, Not Influenced.,” Forbes, Last modified October 5, 2011, http://www.forbes.com/sites/velocity/2011/10/05/nicholas-negroponte-steve-jobs-influence-not-influenced/.
[vi] David Talbot, “$100 Laptop Program’s New President,” Technology Review, Last modified May 2, 2008, http://www.technologyreview.com/news/410072/100-laptop-programs-new-president/.
[vii] Tom Krazit, “Intel leaves the OLPC after dispute,” CNET, Last modified January 3, 2008, http://news.cnet.com/8301-13579_3-9839806-37.html.
[viii] Peter Glaskowsky, “How successful was OLPC’s ‘Give One, Get One’ program?,” CNET, Last modified January 20, 2008, http://news.cnet.com/8301-13512_3-9854461-23.html.
[ix] Warschauer and Ames, “Can One Laptop Per Child Save the World’s Poor?,” 41.
[x] “Education Initiatives,” City of Birmingham, Alabama, [n.d], accessed September 29, 2012, http://www.birminghamal.gov/pdf/education.pdf.
[xi] M Warschauer, S R Cotten, and M G Ames, “One Laptop per Child Birmingham: Case Study of a Radical Experiment,” International Journal of Learning and Media, 2012, 3(2):68.
[xii] “Nicholas Negroponte at IADB: Lessons Learned and Future Challenges” One Laptop Per Child Talks, Last modified September 15, 2009, http://www.olpctalks.com/nicholas_negroponte/nicholas_negroponte_lessons_learned_and_future_challenges.html.
[xiii] TK
[xiv] TK
[xv] Alice Rawthorn, “A Few Stumbles on the Road to Connectivity” The New York Times, Last modified December 19, 2011, http://www.nytimes.com/2011/12/19/arts/design/a-few-stumbles-on-the-road-to-connectivity.html?_ .
[xvi] Charles Moore, “Netbook Shipments Skyrocket, Apple Still Missing the Boat,” GigaOm Blog, Last modified February 17, 2010, http://gigaom.com/apple/netbook-shipments-skyrocket-apple-still-missing-the-boat/.
[xvii] TK
