IJsselstein, Monday, 25 February 2008 03:03:35
As you might know, I have been following Negroponte's OLPC project with more than a keen interest. The vision of having a very cheap, but powerful laptop in the hands of kids as an educational tool appeals to me strongly. However, I remain critical.
Several years ago, I send refurbished computers to schools and NGO's in the Philippines. This effort met with mixed success. These computers worked best with those NGO's that already had a reasonable infrastructure. It was somewhat useful in high-schools where teachers and volunteers where able to operate and maintain the machines. It didn't work out at all at primary schools with teachers really having no clue, and lacking the space to properly install and the means to maintain the machines. Given the cost of transport and the time it took me to refurbish machines, I decided it was not worth the effort, and stopped sending computers a few years ago.
A big issue turned out to be what to do with a computer. Basically, we want to use the computers as educational resources. That is, not to learn programming, not even to learn office tasks, but to use them for learning. For that we need educational content. A big source for educational content is the internet. Unfortunately, Rural schools in the Philippines do not have access to internet, and even providing such access will often be difficult, with the nearest land-line telephone connection many kilometers away.
Some things change quickly though: in the span of a few years, Bohol has obtained almost one hundred percent cell phone coverage, and the OLPC has gone from a dream, not taken seriously by many, to a reality.
If you look at the specifications, the OLPC has about the capacity of the refurbished computers I used to send: a 400 MHz processor with a gigabyte of storage. However, the OLPC adds a lot to that. It is more compact, energy efficient, and hardy. Its housing is made to survive tropical rural conditions, its energy consumption is modest, adjusted to places where electricity is scarce and costly, and its feature set has been trimmed down to things really needed. The historical ballast of traditional PCs and lap-tops has been pruned. Its cost, although not 100 dollars, is about the same it costs to refurbish and send a desktop computer from a western country.
And it comes with extras, such as a build in camera, a sunlight-readable screen, a very neat OS tuned for educational use, a range of valuable educational software, and last but not least, WiFi mesh networking, which makes it possible to get connectivity everywhere.
Recently, the first OLPC laptops has started to be distributed to schools in Uruguay and Nigeria, and to reviewers in western countries. The reviews vary from very positive to negative, and although I attribute some of the negative reviews to a misunderstanding of the concepts, the technology is not yet fully mature.
Putting aside the technical merits of the OLPC laptop, I remain critical of some other aspects of the OLPC project. The proposed massive scale of this project, and trying to run it through the educational ministries and other institutions of countries not exactly known for proper governance and accountability, leaves too many opportunities for large scale corruption, siphoning education money away from children into the pockets of officials. Nigeria has ranked first in the list of most corrupt countries for years, and I would be amazed to see this project function without strange malversations.
Secondly, although the laptops are intended to belong to the children, some aspects of the child's privacy seem to be badly protected. As far as I have been able to asses, teachers can have access to the child's work on the machine. The theft protection system, although innovative and very interesting from a technological point of view, keeps control of the machines centralized. This certainly is certainly a matter of concern in totalitarian countries, such as Libya, where the machine will roll out. Without proper privacy controls in the machine, it might well become the ideal surveillance tool.
Finally, although the machine is innovative, I've seen pretty little on supporting infrastructure. This varies from a distribution channel for replacement parts and service, to hubs, servers, and teacher's support, to educational content itself.
The OLPC project doesn't stand on its own. Although laughed at initially, it has demonstrated that dramatic cost reductions in lap-tops are possible. Although capacities have doubled almost yearly, the price of computers has remained quite stable, starting at about USD 500. The OLPC has broken down that barrier. The mobile computing it makes possible has all the promises to become the next disruptive technology. Competing machines, such as the Asus Eee PC have already appeared. Negroponte's response has been a bit sour, but to me this is much welcome competition. My guess is that this is just the start of a next revolution in computing technology, in which computing and communication devices will merge. Within a few years, we will probably see iPhone like devices for a fraction of the cost. Initiatives such as Google's Android and the Openmoko project promise drastic cost reductions for mobile devices. In the not so distinct future, smart phones, MP3 players, eBook readers as Amazon's Kindle, and sub-laptops may very well merge into a single mobile communication device. I have no doubt we will have cheap computing and communication devices within a few years, even in remote barrios.
However, just cheap hardware is not enough.
One big issue to address: getting enough educational content for these machines.
In The Netherlands, a highly develop country, I am involved in the lobbying for open educational resources. Although the discussion of providing school books in secondary schools free of charge is high on the political agenda, the arguments to create open educational resources seem to fall on deaf ears. However, I believe in due time, the concept will take off, and looking at the Capetown declaration of open educational resources, I am not alone.
Traditionally, publishers of educational materials have kept tight control of their copyrights, allowing them to charge royalties on every copy, and keep tight control over their content. Open Educational resources are educational resources that are distributed with a very liberal license. Everybody is free to copy the material, adjust it to their needs, and republish it. Some of these licenses come with a twist, called copyleft, which allows all those things, but requires people who adopt the material for their use to share their contributions under the same terms. A model that has been very successful in growing the free Linux operating system. The Creative Commons have developed a whole set of licenses to help make this possible.
Wrapping up text-books in PDF files is not the way to go. This would reduce the OLPC to an ebook reader, and leave most of its potential untapped. We can do so much better. Basically, we will still need to have defined curricula, and a core of content, but this should be tailored to the exploratory way the internet works. Learning should not be the act of memorizing facts from a single source, but an exiting discovery of visions, ideas, and facts, as they are presented in various sources, not just looking at the whats, but also the hows and whys.
Access to pluriform resources such as Wikipedia can be put to powerful use. They teach students not just the facts, but also to filter facts, evaluate them, and understand the various levels of trust you give to sources. And they need not be just an input. With laptops, students can also be involved, and help in building these resources. A number of educational initiatives have been started in the Philippines, such as WikiPilipinas, Filipiniana.net, and my own little Project Gutenberg of the Philippines.
All that is needed is that a few dedicated people will build the tools to share and cooperate on building content, set up a core of educational materials, after which volunteers and students themselves will be able to expand the collection rapidly. Tools to make this possible are already being build, often as open source projects such as Moodle, ILIAS or ATutor.
Currently, the Philippines is running out a large, World Bank financed program to achieve text-book parity (one text book per student) in the Philippines for the first time. This USD 100 million project, however, still produces traditional paper text books. On average, a text book costs about 40 pesos (about USD 1) (including the cost of authoring, paper, printing, and delivery at the school premises.)
Of this entire project, writing the actual content is just a fraction of the entire costs. Most of it goes into distribution and paper. Both could be dispensed with if we had this same material available on-line, a good digital network to deliver them, and a reading device on which students can access it. This is exactly what the OLPC aims at.
Let us return to the 40 pesos per text book price tag. This is an amazing low price for such books. However, at this price, you can finance slightly over 4 text books per student. Assuming (optimistically) a live of 5 years for a text book. That means you can just have 20 books to cover the entire curriculum. This seems unlikely to me.
More realistic is to assume the need for 40 books for the entire curriculum, and a book life-time of 3 years. This means a total book requirement of about 300 million books, or a budget of USD 300 million for the current students' needs. This boils down to about USD 25 million annually.
Could replace those books with the OLPC?
Let have a closer look at the numbers. The complete population of the Philippines is about 91 million people. Of these, one third, or about 30 million people are below the age of 15. We have roughly 11 million pre-school children (age 0-4); 14 million children at primary education (age 5-11), and 8 million at secondary education (age 12-16) - the actual numbers of attending students are somewhat lower, with attendance rates of about 92% in primary and 68% in secondary education (based on rather old, 1995 Unesco figures). About 3 million pupils will enroll at primary school every year, and about 2 million students at secondary school.
The Philippines so far has stayed away from the OLPC project. The GILAS project, run by the Ayala Foundation aims to computerize Philippine high school, however, they still aim to build computer labs using traditional desk-top machines, and has fairly limited means. A range of similar projects do exist.
The Philippine government is involved in two controversial programs that are supposed to improve connectivity and on-line education in the Philippines. First is the National Broadband Network project. This project, worth over USD 300 Million was suspended September 2007 over irregularities and corruption accusations. The Department of Education Cyber Education Project, to be paid for by a USD 650 Million loan from China, also proposes to provide computer hardware to schools on a large scale (over two-thirds of the budget), however, when they touch the subject of content, the proposal appears to be about televised lectures through satellite connections, something I certainly do not think as an optimal use of the possibilities current technology offers (for a fraction of the price, schools could drop recorded lectures on YouTube, as the for example the University of Berkeley is doing). This project too is stalled in political turmoil.
In general, I am critical of large scale government investments. They offer just way too many opportunities for corruption, waste, or just plain mismanagement on a grand scale. Whenever possible, smaller scale programs, aimed at a high impact, and mobilizing grass-roots resources work better. If we build up a core curriculum of educational materials, publish them for free, combined with licenses that enable wide-scale sharing and co-development, that could truly revolutionize education.
An example of such a small scale project is Tony Ranque's eLearning center, he is currently operating in Jagna, Bohol. With limited resources he is trying to set up a facility that both adds social value, and can be commercially self-sufficient -- in a landscape currently dominated by internet cafés doubling as gaming shops.
The current World Bank text book program would only buy about half a million OLPC laptops, that is, just one for every 44 students! With my estimated required book budget of 25M dollars annually, you can by 125.000 laptops, one for every 16 pupils starting high school. To give a new OLPC laptop to every student at age six would require a whopping 200M dollars annually. To do this in Bohol alone would cost 2.5 million USD per year. Given the stalled projects cited above, to believe somebody would come up with such amounts for education in the Philippines is dreaming. That is except if we can convince parents themselves...
Can we convince Filipino parents that investing USD 175 or PHP 7000 in a laptop make sense? It depends. Available services and content will be key here.
One thing amazed me the dramatic speed in which cell phones have been adopted in the Philippines. In the span of a few years, coverage has gone from fairly limited to near 100% nationwide, and nowadays you can see cell phones everywhere. Even poor farmers have them. Cell phones are not expensive in the Philippines, with cell phones being available for less than 1000 pesos (USD 25), and text messages costing less than 1 peso (2 cents) each. Still this is a considerable investment for a farmer, with income levels sometimes less than 2000 pesos per month per family. However, it helps considerable: with a single text message, he can verify whether it is cheaper to sell his produce to a local middleman or in the city. With a single text message, he can save himself a two-hour walk to an office or shop to verify something is available, etc. These things speed up the local economy and productivity considerably, so make it worth every peso.
In the coming years, I think the vision of the OLPC will take off, although not necessarily as an invasion of little green laptops...
Since I believe content is key, I will be working on content.
I am very happy to see my wishes answered even before I make them. When researching for my story, I already came across various educational tools, and now I am being pointed to more projects in this area. Already starting is Bayanihan Books, a project to do exactly what I think is needed in the Philippines. From other sources I know similar projects are in the making.
So, while the government is still very much in trouble with corruption surrounding its large scale projects, such as the ZTE broadband deal, small scale private initiative is running around them on all tracks.
One little snag may hurt the Bayanihan Books project: its proposal to use the creative commons attribution/share-alike/non-commercial license. The non-commercial part of it puts a very wide prohibition on using materials in a commercial context. Although it looks nice at first, the non-commercial clause has a number of unintended side effects, which will harm its purposes more than they help stopping "free riders" from abusing your content.
The main problem with the NC clause is that it will make the materials incompatible with the existing and growing body of other open content resources. Second, it might rule out use of such educational resources in schools run for profit, which make up quite a large group of schools, which may otherwise be in a good position to give back materials under the share-alike scheme.
Finally, if you want to use the NC clause to keep your own option of commercialization open, you will have a hard time. First, the work will be available for free. Second, if the community, in spite of the NC clause, picks up your project and contributes anything substantial, you yourself will also be bound by the NC clause on that contribution, as the copyright on that contribution will remain with the contributor.