One Laptop Per Child: What Went Wrong
January 19th, 2009 by Jon Evans | 38 Comments »
In January 2005, at the World Economic Forum in Davos, Switzerland, Nicholas Negroponte announced the One Laptop Per Child project, with the stated goal of giving every poor child in the world a “rugged, low-cost, low-power, connected laptop with content and software designed for collaborative, joyful, self-empowered learning.”
Last week OLPC laid off half of its staff. Sales of its XO Laptop to developing nations are far, far below initial projections in the millions; in the third quarter of 2008 it shipped a mere 130,000 units, a trivial 2.3% of the world’s low-cost, small-screen “netbook” laptops. Meanwhile, the income from their 2008 “Give One Get One Free” drive dropped 93% from 2007. What went wrong? Any number of things, including bad timing, production delays, poor management, and superior competition. But if you ask me – and I feel bad writing this, given all the hard work and good intentions that went into One Laptop Per Child – its fundamental problems are twofold:
- It was a bad idea to begin with.
- The XO laptop is a piece of crap.
The best thing about the XO is its case: the pebbled exterior, built-in handle, and dust/water resistance (which I didn’t test) are stylish and useful. The problems begin when you open it up. It’s slow. On occasion the cursor freezes or submarines. The screen and keyboard are tiny even for a netbook. Even its vaunted connectivity is badly flawed: my XO completely failed to connect to encrypted Wi-Fi networks that worked with both my other laptops. When you do connect, its Web browser fails to show paragraph breaks on Wikipedia pages, making the world’s greatest collection of free information hard to read – on a laptop allegedly designed for education!
The XO’s Sugar interface, a “learning platform that reinvents how computers are used for education”, appears to have been designed by a team of ivory-tower PhDs and misanthropic techies sealed off from the real world since birth. For instance, type “help” into the search box that appears on top of the screen, and – nothing happens. Click on the help “Activity”, and instead of anything useful and/or context-sensitive, you get a virtual brochure that begins with pages of PR bumph extolling the greatness of OLPC and Sugar.
I played with my XO for some time, trying to imagine myself as an 8-year-old in a poor country. I like to think I’m more qualified to do this than most – I write novels for a living, and have travelled through dozens of developing nations – and I was thoroughly underwhelmed. At its best Sugar was a collection of badly executed good ideas. And I can’t help but wonder: at what point did teaching children an interface fundamentally different from the windowing UI used by all other computers seem like a good idea to anyone?
Meanwhile, the rest of the world has already lapped them. My Acer Aspire One netbook is faster, has more memory, a better screen and keyboard, connects to encrypted Wi-Fi networks, renders Wikipedia correctly, and has a user-friendly interface with many useful applications. There’s no comparison: it’s miles better, for a comparable price. As far as I can tell, the OLPC team so wanted to be revolutionaries that they insisted on reinventing everything at once, and as a result, failed everywhere. (Although to be fair they did inadvertently spur the growth of the netbook market that has since entirely overtaken them.)
But that hardly even matters, because the whole idea of distributing laptops to poor children was completely misguided to begin with. Did the OLPC braintrust think they were bringing modern technology to the Third World? They were years too late; it’s already there, in the form of the not-so-humble-any-more cell phone. Tiny villages in Africa have GSM coverage and cell-phone stalls run by local entrepeneurs. You can bank by phone from the Colombian jungle, or get market prices texted to you while fishing off the Indian coast. Mobile phones have permeated the developing world to such an amazing degree that it makes no sense to try and reproduce that existing cultural and technical infrastructure from scratch.
What Negroponte & co. should have done was One Smartphone Per Child; a smartphone is not much different from a netbook that can connect to both cell networks and Wi-Fi, and has GPS…and that you can also use as, well, a phone. But alas, OLPC suffered from a failure of imagination from the very beginning. I hate to say it, but despite all its PR glory and good intentions, it was never more than a bad implementation of a bad idea, and its eventual failure was all but inevitable.
(UPDATE: OLPC News responds.)
Another tragic victory of marketing over reality.
Aside from apparent OLPC arrogance you’ve outlined,a serious limitation related to recharging their XO darling. With an average ~6 Watts power drain,the ~20W.h battery was only good for ~3 hours use (~½ a school day). As many target regions had no mains power,recharging demanded extra batteries,costly generators or solar panels-items well beyond most schools.
The much hyped 2006-7 “crank handle” approach was near useless, as it generated only a Watt or so before user fatigue set in.
I have to disagree with your analysis, in part at least.
It began as a project to educate poor children. Calling it a “bad idea to begin with” is clearly misguided. It was never meant to be about giving poor kids laptop computers (which, I agree is a bad idea). It went wrong when it became a project to give kids laptop computers. Fault for that goes to OLPC itself, for failing to properly explain the project, for hyping the XO as a laptop, and I suppose to the tech bloggers and such for listening to them.
And calling the XO “crap” is misguided as well. Better to seperate the hardware from the software.
As you’ve pointed out, the hardware is rock solid and innovative. Some parts (keyboard and trackpad) are clearly sub-optimal (okay, they suck) if you’re trying to use the XO as a laptop, but these are necessary compromises on the original design constraints; the XO was designed for children and to be used in harsh environments.
If you really want to evaluate the XO hardware, put a laptop environment on it (http://www.olpcnews.com/forum/index.php?topic=4053.0) and see how the hardware fares against other laptops you’ve come to know.
The software (Fedora+Sugar), on the other hand, was just awful and suffered from a completely misguided approach to selecting the software they needed. It was intended to NOT be a laptop software load (so it’s no surprise it failed as a laptop) but it even failed to meet the needs of the education project as well.
In my opinion, OLPC was never ‘on’ the Free Software bandwagon; they just selected a Free Software base because they thought it would cost them less, and when the community didn’t rally to write the software for them, chaos was the result. Effectively ‘firing’ these volunteers (by partnering with Microsoft) was the final nail in the coffin.
Let’s hope that what really died here was not the OLPC mission of bringing an education to poor children, but rather the image of OLPC as developing laptop computers. That one really needs to die.
A great article. In additon, OLPC is a quasi-Charity scam, from my personal experience. I joined on the first day of 2007 Campaign. They told me to ship it before the Xmas, but didn’t ship it. After three months of lies (They always tell you they will ship it within 2 weeks), they told me they would refund $200 (not $425). I refused it. After six months of delay, I told them I would report it to Attorney General Office. They finally shipped it on August 2008. And the XO was a piece of crap. They advertised that the battery would last for 20 hours, and it lasted only 2 hours under my test.
I am very disappointed that you did not cover the major factors that derailed the project — coordinated sabotage with words and predatory pricing from Intel and Microsoft.
Revisionism is dangerous game.
I think you missed the main significance of the OLPC project that was never envisioned by Intel, MS, or OLPC. It created an entirely new computer market segment – now called netbooks. When this project was 1st announced, many scoffed at the feasibility of making a functional laptop at such a low price point. Today there is a plethora of options and models of under $400 netbooks from many different makers.
I concede that the XO and OLPC have been a sad busted experience. But the free market certainly responded to even a slapdash form of competition for developing market share.
In some ways, the OLPC could be a test-case for baseline socialization spurring increased market competition in under-served markets.
I found this article all wrong, as I say in “What Went Wrong With The Walrus’ OLPC Review?“.
Jon brings up minor issues (paragraph breaks?!) as major showstoppers and offers alternatives (cellphones) that have little practicality in classroom-based education in the developing world.
OLPC has major issues, no doubt, but failure of imagination is only attributable to this article.
I’ll admit that I don’t know enough about OLPC to verify each of your criticisms, but I’m always flabbergasted when charities or government organizations fail to contract with obviously superior private organizations. Apple, HP, Dell, Fujitsu, etc. will OF COURSE make a superior product to some startup group. Even if you have a truly revolutionary concept, why not turn to the experts?
Having grown up in the dawn of computers in American schools (a child of the eighties and nineties), I’m now amazed at how much substandard hardware and software we used (and worse, how much I continue to see in schools). People are suckers for “designed for eduction.” The best example of this right now (outside of schools) are educational video game systems, meant to replace your Nintendo, Microsoft, and Sony systems for little kids. Never mind that the major companies make vastly superior hardware for nearly the same price with vast libraries of software for old and young.
SmartPhones? For teaching? You’re joking, right? If not, you’ve clearly never been in a classroom. The XO might be a piece of junk, but replacing it with a smartphone isn’t the answer….
Um, a few columns ago you point out that often the most useful technology in the 3rd World is “not sexy”, then you complain about their windows designed for kids learning on slates under a tree?
This article would be more informative if it shed further light on the human-side errors (management, planning, etc.), or if you had perhaps made comparisons to similar technology development misadventures (e.g. Osborne, Lisa, 1970’s PARC).
Part of the inspiration behind the project may have been to increase kids’ literacy, and given how hard it would be to replicate Palm’s Graffiti for every writing system out there, giving kids larger-than-phone keyboards may have seemed necessary.
Having worked for a failed startup tied to some of the same folks, I can fully believe mismanagement, etc., but I can also assure you that being overtaken by mass-market tech developments is a fate shared by worthy and not-so-worthy efforts.
I agree that this was a bad idea to begin with. I’ve been in plenty of poor villages in the developing world. And when I’ve seen squalor and undernourished children, I never think: “What these people need are laptops!”
Essentially, this is where Bill Gates began his road to philanthropy. He’s a smart guy, and quickly realized that more basic needs must be met first.
I freely admit I know very little about this issue. But…
“It was never meant to be about giving poor kids laptop computers (which, I agree is a bad idea). It went wrong when it became a project to give kids laptop computers.”
If a project called One Laptop per Child wasn’t intended to give poor kids laptop computers what was it for?
Main problem = non-profit
How the hell do you define “success”? Who do you define it to? Who are the customers? What’s the return to investors? Good vibes?
Literally, literally impossible to measure “success”…
Indeed, claiming that the OLPC was a failure as measured by how many sold:
Sales of its XO Laptop to developing nations are far, far below initial projections in the millions; in the third quarter of 2008 it shipped a mere 130,000 units, a trivial 2.3% of the world’s low-cost, small-screen “netbook†laptops.
is putting the cart before the horse. Netbooks really didn’t seem to take off until the OLPC came around. Asus saw many many people interested in the size and *the price*, and Asus really failed to deliver on its own “low low price”.
Oh, and stop calling the winner of the game in the last minutes of the fourth quarter. The trick is to do so before the game starts, got some text on the OLPC written a year or more ago?
You have got to be kidding. Do you honestly believe that smartphones have the same educational potential as laptops, or for that matter, ANY educational potential at all? Their screens are tiny and their input systems exceptionally slow and unreliable. You can’t effectively browse the internet on them, you can’t compose music, draw, play with your friends, or program them – all things that can be done on a laptop.
Look, I don’t think the XO’s Sugar interface is good, and perhaps there are better laptops available at the same manufacturing cost as the XO, but saying that smartphones are a better alternative is beyond stupid.
And the XO laptop itself is not crap. It was built to be rugged and draw less power than mainstream laptops. Its LCD screen is made to be visible outdoors and has a monochrome mode to conserve power. Its backlight can be replaced with a common lightbulb. The keyboard is built so dirt can’t fall between the keys for when the laptop is used outdoors. These are considerations that you won’t find in an Acer Aspire One notebook, nor any other commercial netbook computer. You should also bear in mind that the XO laptop came first – commercial vendors (starting with Asus) jumped on the idea to sell cheap “netbooks” only after the XO laptop made its debut.
[...] article about the epic failure of the OLPC project. If you don’t know, “One Laptop Per [...]
OLPC was about the curriculum and pedagogy. The technology was just a tool. The biggest problem was the backlash against constructionism, imho, and the desire of countries to copy the trad education of their own and western cultures. As for the ‘faults’ with the laptop, I’d cite user error. I had no problems with getting my XOs on western wireless networks. It is a funky little unix box, and though I use an external keyboard with mine, that’s because I’m an adult. The bad press that they got by putting the XO in the hands of western expectations of gee wiz how fast is it meeting my preset expectations was a real killer as well.
I agree with many of the other problems… which were caused by the project’s hubris, but it also changed computing forever.
[...] pointed me to this article: The Walrus Blogs » The Failure of One Laptop Per Child » World Fast Forward. I disagree with a number of the ‘failures’ as I noted in my comment on the site. But [...]
i think the real issue is that the project was not an education, nor curriculum, nor pedagogy project, but was a technology project. It was always framed as a ‘we can do this, if it can be done’ type project of putting laptops in kid’s hands. It had no real training components.
When Negroponte introduced the idea, I didn’t hear about it, but was confronted with it during a lunch conversation at a job interview. I was asked, ‘will the olpc’ 100 laptop change the world. As a political economist, I said without a doubt, no, it will not. Even if you manage to get millions of laptops into the hands of children, you cannot ensure without significant social and cultural investments that they will actually use or value such an object like we assume they would. The laptop or netbook is a very particular form of information system and the programming systems and interface systems are also not necessarily main stream, so what are we really doing by giving these things out? My basic thesis that the only real value that these had was to identify, develop and drain the intellectually gifted people of a country. If you give these to a bunch of children, some will figure it out on their own, probably around 10%, those that are natural genius’s or otherwise talented, but like all computers, those that do not have such talents end up merely using the basic interfaces to a minimal level as is required to accomplish tasks. You can easily tell the difference in patterns of use and enable those special children through programs. Perhaps I’m a cynic, but I’ve been watching computer users at the college level for some time now, from Professors to students, they all manifest roughly more or less the same patterns of use. So to me, OLPC was either a system of exploitation on that level, or perhaps a strategic exploitation in terms of westernizing cultures on another level. I thought a better idea was the Indian Simputer project, which I was sad to see fail.
[...] blogger Jon Evans has a good post on OLPC, turning its latest organizational woes into a broader critique of the idea itself. His [...]
[...] themselves. The project appears to be in a death spiral, though of course who knows for sure. This nice post characterizes why but I’ll summarize it: The XO Laptop wasn’t something anyone actually wanted to use. [...]
I think the ‘author’ (and I use the term lightly) of this article misses the point of the OLPC project and succumbs to what poster Jason refers to as “…western expectations of gee wiz how fast is it…”
Yes, it is slower than the likes of the clearly faster hardware of the Acer Aspire One, MSI Wind or the Asus EEE. But what it can accomplish is impressive, given it only has a 433 mHz processor and a small amount of RAM. It’s screen is vastly superior to the other ‘4P’ netbooks in its class.
I am an avid XO user, but I wouldn’t call myself a ‘believer’…
For me, the OLPC project has always been a bit of a ‘faith based’ mission. I waited for ~5 months to have my XO delivered during the initial G1G1 program. I had to believe it would arrive, eventually it did.
Clearly, OLPC were not prepared for the response and all of the problems associated with such a large rollout of laptops. All along I’ve been willing to put up with these shortcomings because OLPC are a non profit, doing good work.
One need only visit sites like olpcnews.com or follow the OLPC community news list to see the positive feedback from teachers and students using the XO around the world.
Granted, I’m not a huge Sugar fan myself. Most of my day to day XO use involves progams I installed on my own (with the help of other users from olpcnews.com forums).
Using Sugar on its own defeats its intended purpose. If the author had ever witnessed XO’s being used in a classroom/shared environment, he would have seen how it encourages and enables sharing and collaboration.
I think OLPC’s greatest folly was its failure to anticipate a corporation’s (i.e Microsoft’s Classmate, Intel..) willingness to compete directly with a non-profit organization in the same arena.
I applaud Nicholas Negroponte and the OLPC community for their efforts and I am also sad to see the project lose momentum.
OLPC smackdown…
Jon Evans, a writer for The Walrus magazine, brings the smackdown on the One Laptop per Child project: Meanwhile, the rest of the world has already lapped them. My Acer Aspire One netbook is faster, has more memory, a better screen and keyboard, connec…
I agree that there are faster machines out on put the market than the OLPC’s XO, but you must also consider the purpose of the machine and who will be using it. For you, small, light and fast are your most important concerns. Energy consumption, and kid-friendliness were probably not your ideals. Features like the rubber membrane keyboard, large touchpad, easy-to-find controls and the ability to switch to a monochrome display were all put in place with a target audience in mind. I wouldn’t want to sit and use the XO as my personal laptop either, but in places where they are likely to get dirty, and electricity is inconsistent this machine has some advantages. I would rather see the Asus eee PC get used as a netbook, though. But, OLPC was about more than just making a computer – they are a non-profit geared towards getting more kids to use this technology. That’s what they should stick to.
Just stumbled upon this. My brief thoughts.
1) The OLPC project is far from over.
2) I am amazed (first) and appalled (next) at the incredible short-sightedness of people who comment here. This includes the author.
3) Smartphones is a misnomer. People are smart, not the phones they use.
4) Smartphones are definitely not good for learning.
5) I am a teacher.
6) I live in the “third world”.
7) I get electricity from biogas and biodiesel, but not the grid.
8) OLPC is the best thing that technology has done for educating children. Ever.
9) Talk is cheap. Work is not.
10) Go, do some work.
[...] talking mobile! Will mobile phones with keyboards replace laptops? Was the OLPC doomed from the start? Why don’t any of these ICT4D projects seem to be run by Africans, anyway? Ushahidi! Web 2.0! [...]
[...] is remarkable how polarised the discussion of the OLPC has become. You either love it or hate it. I have already said my own say about the value of the OLPC but there is one point that I [...]
[...] been a veritable bevy of blog posts and rebuttals lately debating what went wrong with the OLPC and what sort of device [...]
[...] so far tanked, they just had to let go half of their work force. Jon Evans rips OLPC a new one in this article in The [...]
One element missing here is reviews from the target audience! A friend of mine did the buy one get one free deal. She gave up on it when it wouldn’t connect on her network and gave it to me to play with.
A few days later I brought it to a New Year’s Day dinner; all adults and one nine-year-old. I showed the nine-year-old how to turn it on and left her alone. She spent hours laughing, smiling, and showing the adults what she had been able to create. I’m no educator, but maybe we should ask some more kids what they think? (Kid’s without desktops/PlayStations/Xbox’s.)
[...] Then it will all go away… In Uncategorized on January 31, 2009 at 7:02 am Here’s an interesting elegy to the One Laptop Per Child project, the much-lauded initiative of Nicholas Negroponte to bring computers to the world’s poor. [...]
[...] Evans of The Walrus didn’t like the OLPC. He thought it was the wrong tool for the job — badly engineered, improperly envisioned, and [...]
I always thought the initiative would be more appropriate if named “one laptop per university student in a developing country.” Often, these universities are under-resourced and – even if the laptops are not able to connect to the internet – they could be used as textbooks (if materials could be put on CDs/DVDs and loaded onto the laptops). I understand the merits of bringing computer technology to the youngest, but a more realistic and purposeful goal may be giving them to those at the tertiary level… given the challenges with money and internet-connection the laptops seem to be having.
I’m about to go read OLPC’s response to this but I say you have revealed some pretty cool ideas. OLPC failure bringing about an eventual netbook revolution? That’s just awesome.
The sugar interface was unimpressive to begin with, why they choose to invest money into that with so much more work to do in ensuring the larger program was successful and sustainable.
I do think you could have mentioned something about competition contributing to the failure though. It’s not like they 100% independently drove themselves into the ground by themselves.
[...] – The Walrus chimes in here, perhaps a little harsh for a project with such good intentions, but Jon Evans makes a valid argument. [...]
Energy consumption, and kid-friendliness were probably not your ideals. Features like the rubber membrane keyboard, large touchpad, easy-to-find controls and the ability to switch to a monochrome display were all put in place with a target audience in mind.
[...] many to list without copy-and-pasting the entire article. On a side note, does anyone remember the One Laptop Per Child campaign? [...]