Slashdot Mirror
Why OLPC Struggles Against Educators, Big Business
afabbro writes "The current issue of BusinessWeek has an expansive article of the history of OLPC and why it has, to date, been a flop. Among the reasons: no preparation for the educational systems expected to use it, uncertain pedagogical theories, poor business management, competition from Microsoft/Intel, and no input from education professionals in designing the software. As BusinessWeek quotes one educational expert, 'The hackers took over,' and the applications are too complex for children to use. To date, 370,000 laptops have been shipped — a far cry from the original 150 million planned to be shipped by end of 2008."
In the book "The IBM Way" i read something along the following lines: we must control change, or change will control us.
If you let the IT folk articulate the business process, you're going to get the same exact thing. That's why we have business analysts whose job it is, ostensibly, to figure out what the business people want and translate it into a swiss army knife that's going to be wildly popular and successful.
To not involve educators in the requirements building phase of this was doomed to the same failure. The problem is that it is visible to more people, sad to say.
Mit der Dummheit kämpfen Götter selbst vergebens.
OLPC had a good mission when they wanted everything on the system to be fully open source, with simple point and click applications and the ability to view the source of any application. However then they got into talks with microsoft, and started to include some very complicated applications with their product, and their mission kind of went down the crapper
So THAT'S why so many South American countries have large populations that live in tin-roofed shanties and go to schools, if at all, in one-room dirt-floor buildings. It's because they have the food/population/education problem solved.
Educational software is hard to write. Really hard. Except for very well defined skills, like typing or flying an aircraft, most educational software doesn't help much.
The OLPC should come with one or two really, really good applications for teaching reading or arithmetic, ones smart enough to self-adjust to the user's level and move them forward. That alone would justify the thing.
It was predictable enough, and many of us did point out the terribly obvious flaw in the OLPC plan -- that people experiencing shortages in food, potable water, basic shelter, education facilities, farmable land, etc, etc, need those survival basics covered far more than they need a laptop. I still don't really see how this was not obvious to Negroponte et al.
Well, you see, there is this gigantic group of people who aren't experiencing shortages in food, potable water, basic shelter, etc etc etc. People who've got the survival basics covered, yet are still extremely poor. They are the target market for this laptop.
This isn't for that tragic starving child with no clothes no food no medicine and flies all over his body that you see on those interminably long "Christian Children's Fund" commercials. This is for his distant neighbor 6000 miles away who lives in a home, on a farm, enjoying a meagre lifestyle, while the children work on the farm, or the local mine, or pick fruit, or help chop down the nearby rainforest for additional income. This is for them.
There are lots of countries who have met the basic requirements for survival, but who lack the infrastructure and wealth we enjoy in 1st world countries. This is for them.
'The hackers took over,' ... at the risk of being flamed, I believe this is pretty much what happens with lots of open source projects.
People start their projects, invite some expert in relevant fields to participate but when any of these experts tells them that they are doing X or Y thing wrong, they get all proud and stubborn and ignore the advise.
What does this experts do? they just leave (as people has been leaving the OLPC project)...
Ubuntu is an African word meaning 'I can't configure Debian'
I find it interesting that you invoke your friends high status - but neglect to tell us exactly what she uses it for that so pleases her. Are we supposed to accept the OLPC is a Good Thing merely because Somebody Important uses it [for some unspecified purpose]?
The claim is not that OLPCs are too complex for children, or that children will be uncomfortable around them - but whether or not children will be able to use them to learn. There's a big difference there.
And again, you're missing the point: you don't teach them to use the OLPC. That's why it's UI is such that it mostly doesn't require teaching kids how to use it. And at that age, kids don't need to learn food storage techniques. What they need to learn is how to find out about and learn about food storage techniques. Which is a special case of learning how to find out and learn about anything.
You're demonstrating a dichotomy I've seen a lot in college. Most students would memorize what was taught. In a physics class, they'd memorize the hundreds of formulas for all the different things covered in class. They'd come in to exams with the (allowed) cheat-sheets completely covered in tiny writing. And if a question on the exam involved something they didn't have a formula written down for, they were completely lost. By contrast, me and a handful of others would come in with a 3x5 index card with perhaps half-a-dozen basic equations written on one side. Instead of learning all the formulas for everything, we'd learned how to derive any formula we needed from those basic equations. If a question was on something that hadn't been covered in class, it might take us a few minutes longer to work through to get what we needed but it wasn't a big deal. We'd taken the next step, from learning the formulas to learning why the formulas were what they were.
NB: I think it significant that there was a big psychological difference between the two groups. The majority, the ones who memorized formulas, were literally physically afraid of there being anything on the exam they didn't have notes for. Me, I might be annoyed if I couldn't work out the answer in the time given but the unexpected wasn't anything to be scared of.
I agree. I think OLPC missed the mark just a little bit. The concept and design is pretty good, but the OS is a bit limited. Storage is limited. Memory is limited. And yet they went over their target price of $100.
Frankly, as prices continue to go down, they should aim to improve their current model with slightly better specs and have two models. Ship a beefier $200 model, and try to get the current model down to $100.
And even though only 370,000 have shipped to date, many of those are smaller trials and I thought there was on-going negotiations with several countries for large orders.
I think it is a bit early to call the whole project a flop.
http://blindscribblings.com - Tasty pop-culture in conceptual fashion.
I really want the OLPC project to succeed, though the switch to a Microsoft OS as a standard install (note: NOT MANDATORY, ONLY STANDARD) has dimmed my enthusiasm some. But in no way that I've seen have they demonstrated the coherent action I would expect of a five person startup in a basement somewhere. This whole project looks to me like a serious case of diffusion of responsibility and ill-defined decision authority.
When that keyboard glitch turned up a while back, they should have been all over the place within days with a clearly written response, complete with Youtube videos and still images on Flickr under every possible keyword. When Intel started pushing the Classmate they should have (as Negroponte now acknowledges) either kept quiet or done a far better job of making their case. And now that the organization has effectively forked, is the Sugar team talking with Asus about a ruggedized version of the EEE running the OLPC OS and software set?
Like or hate Apple, from day one they had their evangelists out there when they were creating the Macintosh and every key related technology. Maybe somebody should send the OLPC folks a few dozen copies of Kawasaki's first book. Hell, maybe Lasermaster still has a few copies of their reissue sitting around.
But even so, let's keep in mind that all they need to do is keep on their current curve for another year or so and, one way or another, the project should be fine. This kind of thing genuinely is non-linear and now that we have several reference specs for UMPCs, at least one of which is open source, maybe the "success" will end up being an ecosystem of several devices created from aspects of several of the current UMPC approaches, running various OSes, that the current OLPC team members will use to carry out programs under different names and different leadership but achieving the original goal. And let's remember that Asus' president said that he was inspired to create his EEE line by the OLPC project, which is itself a certain kind of success that the project has already reached.
It's all about the information. And what we do with it.
I'm not entirely sure about those two, but as for governments, I'm not sure that I believe that anymore. We've seen an awfully consistent trend of wealthy government/NGO officials saying that they "can't" when what they really mean is "it would be awkward and we just don't do that kind of thing".
Seriously, given the body count in any major famine or disaster or simply grossly poverty-stricken area, there is just as much at stake and on a timescale no longer than Bosnia. That being the case, why the frak aren't we just bloody well making hundreds of thousands of aid packages and just dropping them over every damn starving village? Little packets of not just food, but multivitamins, solar powered minipumps, LED lamps and radios, ceramic filters for water purifiers with instruction in the local language about how to build them, reflective material and frames for building solar ovens, and so on. Maybe even include a stainless steel bowl or three and a few comics in the local language to encourage literacy.
We could fit in a cubic foot enough to change the mortality rate of an entire village. And we could pack it all in another cubic foot of bubble wrap that would let us drop it without parachutes and have the bubble wrap itself (excellent insulation) be a part of the package. And the whole shebang would not only cost a hundred bucks or so per to make, we could have much of it made in factories in the region, providing real jobs doing real work instead of just handing over charity.
I just don't goddamn well believe anymore that if we airdropped a few dozen of them over a village at three in the morning while airdropping liquor and money around the camps of the local thugs, we would still have the same level of suffering that we now see. And having looked into the technology approaches of most of the big charities, I wouldn't put them in charge of speccing a junior high school prom let alone hundreds of millions of dollars worth of projects each.
It's all about the information. And what we do with it.
Personally, I first heard about Intel's tactics in a piece in, of all places, the Wall Street Journal.
It's all about the information. And what we do with it.
If we are to rationally analyze the success or (comparative) failure of the OLPC, it is crucial to understand that the big NGOs are staffed by people who don't much care about the good of the poor. Many of these people are also vastly corrupt and tied into the regimes they are supposedly working to change; regimes that gain from having desperate, ignorant, weak populaces. Myanmar really isn't that anomalous.
Should the OLPC even try to get computers in through governments or would they be better trying to get the relevant officials bribed to just stand aside? I don't know. But we cannot understand the decisions of nations like Libya and Nigeria without starting with the assumption that the good the children is, at best, fourth or fifth on the list of things they looked at when saying yes or no to OLPC.
It's all about the information. And what we do with it.
What you are describing there is called "task-based learning." It's a pretty common pedagogical approach in secondary education, first showing up in medicine and law. The idea is that by intelligently creating a task/project, you can be sure that students will follow a fairly predictable path towards completing it, learning things along the way. The biggest learning advantage to this is that, more than learning how to complete the task, the student learns how to learn how to complete a task.
It seems, however, that in the US educational system anyway, we are moving away from this model, which encourages creativity, self-sufficiency, and autonomy, and toward a "cram for this standardized test" model, like where I live now: Japan. This is a mistake. And that's coming from one of the guys who writes and coordinates a large standardized test!
In the case of the OLPC, this pedagogical concern has been and continues to be at the heart of all the questions about its efficacy as a world-changing tool, whether the critics realize that or not. It has never been clear what one could learn from having a little green typewriter which may not even have internet access.
As an educator, tester, and geek, I have mocked the dunderheaded goodwill of this project from the first time I heard about it. The cry for more computers in the classrooms of the world is very rarely raised by the teachers themselves. Computers are great for education mainly as a means of finding information, and in such a case, the essential ingredient is internet access. Once that requirement is satisfied, any terminal is fine. Beyond this, what is a student to do with a PC? Type? Is this substantially different from writing by hand? No. It's just more convenient.
I have seen it argued that the OLPC project would expand IT/programming to impoverished children and give them a means of developing their economy. Rubbish. I have a master's degree and teach at a university and do statistics-heavy research, but if you handed me an OLPC and said, "your project is to write a program that alphabetizes this list," which--if I remember correctly--was one of the first assignments in my friends' programming classes, I would have no idea of how to even begin. I have done zero programming. I would require some explicit instruction to at least know how to get started. Explicit instruction requires access to a knowledgeable person. If I live in the boonies of Kafoonistan, and I don't speak English, how am I going to get access to such a person? Even if I were to use my OLPC to read up online about how to begin... I don't speak English. How do I learn English? I need access to a knowledgeable person.
You see where I'm going with this.
The OLPC project overlooks the single most important thing to any educational system: People. We learn from other people. I didn't get into stats until I was 30, and I've done a lot of self-study with books to get where I am now, but if I didn't have access to teachers in graduate school and knowledgeable colleagues at work, and the money to take distance courses on some of the arcane procedures and programs I use, I would still be totally in the dark. If I hadn't had a string of great teachers, there's no way I could have learned Japanese.
The OLPC is gadget. It's handy, to be sure, but without the infrastructure--and by that I'm referring both to net access as well as a functioning education system--all it can really be is a toy. Even in your example, the teacher was an important component, if not always a helpful one. He/she would ultimately look at the output you created--the outcome of the task--and tell you whether it was acceptable or not.
This is how we learn. You can't just give people a tool and a task and say "go." Someone needs to show you how to use that tool; someone needs to design that task; someone needs to be available to guide people through it and get them past the bumps in the road; someone needs to tell the student if the task has even been completed.
People really do love to learn. But learning is a social act, even if it's done on the internet. Without people, the OLPC is just a pencil for someone who doesn't know how to read or write.
"the hackers took over?" - what a bunch of molarchy. I and the rest of my generation cut our teeth on Commodore64's and AppleII's. Those had no gui, or wysiwyg tools in the beginning. BASIC was taught in 7th grade as a class! Kid's today don't even have "computer science" class where they actually learn how to use a computer and why it does and how it does what it does. Instead they have "MSFT Office class". As a result they don't know that the Word icon is actually telling the computer to run c:\program files\office\word.exe (for example) so they are stumped when an icon gets deleted and wonder why they get viruses after using KAzaa so much. OLPC is an attempt to go back to really teaching computers. The system isn't "hard" it's just not familiar because we've let MSFT hold people's hand to freaking long which has made us all lapdogs.
But why would one give these kids an assignment to write a program? The assignment's more likely to be "Draw a map of the area around the village.". The OLPC is the tool the kids use to draw the map. And to get all their friends together to help point out landmarks one of them may have forgotten to include, and to argue about where those landmarks are relative to one another. No programming needed by the kids to do that. And yes, you can do that kind of assignment in a classroom with paper and pencils. Unless you don't have a classroom, you can't afford to buy pencils and paper as fast as the kids will use them up, and the kids are spread out and it's impractical to get all of them together at one time (but it is practical for groups of them to get together, especially if they just have to flip open a lid to join the group instead of hiking the mile or so to the school). And pencil and paper and a classroom won't let you bring the class from the 5 villages around you in to have the kids look at how the other villages see the local world. And it sure won't let you get two kids from tribes that've been mortal enemies since before their grandparents were born talking and working with each other before they realize they're mortal enemies (old adage: "On the 'net everybody's the same color, green on a black background.").