Why Should OLPC Succeed?


Pedro Hernández Ramos

I've been thinking about this question for a while, and have been working on a draft for an article (aiming for an academic journal) for the last few months - in fact, since November of 2006, shortly after Walter Bender presented at the "Silicon Valley Challenge Summit" I helped organize at Santa Clara University where he showed us an early production unit fresh off the assembly plant.

In the comments thread to one of Bob Kozma's postings here ("OLPC and Economic Development"), Julio Cartaya argued that OLPC deserved recognition for bringing the world's attention to the plight of hundreds of millions of poor people around the world (I'm paraphrasing).

I agree with the sentiment, but what I'd like to do in this posting is present an excerpt where I state one reason from an educational technology perspective in support of why OLPC should succeed. (In subsequent postings I will offer more reasons, before switching to arguments about why it may (or should) fail.

So, reason #1: OLPC should succeed because of its commitment to technology designed in support of, and supported by, educational ideals.

The kind of technological innovation evident in the XO machine is required to expand and enrich the social conversation about why technology should be made massively available to support and enhance children's learning experiences in and out of school.

Unlike earlier attempts that envisioned users primarily - or only - in schools in developed countries (like the eMate from Apple was for a brief moment in the late 1990s), the XO is being designed with users from developing countries in mind first.

This shows because the XO does not take for granted the availability of infrastructure (such as electricity) nor the physical conditions of use typical of classrooms in the U.S. or other industrialized nations (such as clean, dust, free, often air-conditioned), and it has several features - including the main interface design - that are explicitly included because of the underlying educational philosophy (constructivism).

The last point deserves some elaboration. How can an educational philosophy such as constructivism influence the design of a laptop computer, both in hardware and software? For the hardware part, to begin with, the design assumes users who will use the device outside of the classroom, involved in activities that go beyond simple notetaking based on the teacher's lecture.

The ability to use the screen in broad daylight will be a boon whenever students are working on projects that involve data collection in the field, and the built-in camera will help students understand that they have a point of view as learners, that they can see and capture information from the world around them to construct their own understanding of whatever phenomena or subject matter they are studying.

On the software side, the design of the main user interface and the ability to easily see who is available on the mesh network is as radical a departure from the typical desktop interface as we have seen in a while. How does this relate to theory? Constructivism holds that the social interaction between learners allows each person to bring and share their perspective on learning.

Before anyone thinks that this is too esoteric, just think "wikis" or "shared blogs." When students of any age are able to grasp that there is meaningful learning to be gained from interactions with peers (e.g., reading their blog entries) rather than only from the textbooks and the teachers, they have gained a key insight from constructivist theory!

olpc wifi mesh
DC's XO-visible mesh network

Currently collaboration is supported most often at the application level (such as having to use a browser to use a wiki), not at what we might call here the "operating system" level since from the moment the XO comes on it is possible for users to discover who else is available on the network.

Learning becomes less isolated when one knows that either physically (when the learning design actively supports interaction, as in so-called "jigsaw" activities in the classroom) or virtually (when meaningful options for either synchronous or asynchronous collaboration are available) the ability to interact with others for learning is real.

There are, of course, other machines in the market that are also targeting the education market, but I have not seen such a clear link between educational theory and design supporting the design of, say, the Intel Classmate or the AlphaSmart Neo2.

I will welcome comments and ideas regarding this point. Rest assured that if I include any of them in the article, I will properly acknowledge and cite my sources.

Related Entries


Great post. I couldn't agree more.

You might add that every student's computer can take the role of a "teacher". In reality, there is no fundamental difference between the teacher's XO and the student's XO. Peer instruction is known to be one of the best enhancers of education.


You affirmation that the OLPC XO was designed to be a tool based on Constructivist leaning, has me wondering three questions. Do you think that:

1. Constructionist learning is the best way for primary children to learn?

2. The XO is the best tool for this method of learning?

3. The XO/Constructionist mix is right for the developing world?

How is efficacy determined in constructivist methodology? Certainly implementations must vary in their quality and fidelity to constructivist thought. What metrics are available to measure potential implementation deficiencies and how to remedy those deficiencies?

Thanks to all 3 for your comments.

To Winter: Interaction among peers is a very powerful learning strategy, which unfortunately goes largely unrecognized and underutilized in most traditional classrooms. The fact that the XO makes it easy (or easier) to interact with others on the network is the kind of design element that links the software to educational theory--and to practical experience as well.

To Wayan: One at a time...
1) One of the great advances in the last 20 years is the recognition that different people learn in lots of different ways. We now "know" more about learning than we did even a decade ago, from a variety of angles too: physiological (how the brain learns), social (the value of social interaction), psychological (e.g., Howard Gardner's "multiple intelligences" work), and educationally (e.g., students get bored if teachers only talk to them; meaningful activity is better).

Constructivist theory and the pedagogies typically associated with it (such as project-based learning) are gathering increasing amounts of research evidence showing the benefits of such practices for a wide variety of learners. One of the challenges is to see if children from developing countries where the XO could be deployed will perform (i.e., learn) as well as children from disadvantaged areas in industrialized countries do with this approach. (Keep your fingers crossed for me, as I have a co-authored article under review for publication with a major academic journal that adds significantly (IMHO) to this body of evidence).

2) When I was with the Apple Classrooms of Tomorrow (ACOT) we liked to say that "technology" is not one tool but many, and that the technology was actually less important than the pedagogy. If one adds technology (computers, the Internet, DVDs, etc) to classrooms and schools where the modes of instruction do not change significantly, then it's unlikely that "technology" will make a difference at all. To me that's (by now) common sense.

When schools and teachers decide that they are willing to transform their teaching practices (and the school organization) to support ways of learning that recognize the very different types of work that students AND teachers can do in classrooms enhanced with technology, that's when the magic happens. Check out the website and magazine of the George Lucas Educational Foundation (www.glef.org) for lots and lots of great stories documenting cases where this is happening.

3) In a later post I will argue that OLPC should succeed because it is raising conversations about the urgent need to make major changes (improvements) to the way education happens (think about the implications of the term "delivered") in developing countries. That Negroponte, Bender, Cavallo, and company at OLPC have been able to even engage top-level government officials around the world in serious discussions about how the XO could help in their efforts to transform the face of education in their countries is an achievement in and of itself. Is constructivist theory "the best" for developing countries across the board?

I think that's the wrong question. Every change process has to take into account where the "thing" to be changed (individual, organization, process, etc.) is starting from. To work as a constructivist in a classroom is hard work--much harder than just lecturing for sure--and requires a level of knowledge about teaching and learning that is generally considered more sophisticated than what "instructors" (i.e., givers of instructions) typically need or do. This becomes a point on the other side of my argument...

To Allen: How to assess student learning is one of the great challenges in education (and in psychology, and sociology, and...). When technology plays a major role in the work that students do in the classroom, and in the products they produce different than the "traditional" paper, an exam consisting of multiple-choice questions to assess how "much" (not always how *well*) you have learned is often the wrong tool.

If the students produced a multimedia presentation working as a group, do you give them the same exam to try to figure out how "much" each of them learned about the subject matter and deliberatly ignore everything else that their work experience would alllow you to consider? I'm talking about the so-called "21st century skills"--communication abilities, collaboration skills, ability to represent knowledge in multiple ways using multiple media, and so on.

At the classroom level, the use of rubrics is a particularly good strategy (I use it even in my MA courses), as it tells the students from the beginning of the learning process what are the criteria against which they and their work will be held accountable. Portfolios (in paper or electronic) and performances (such as public presentations of their work to relevant audiences, not just the teacher) are also valuable tools.

While it is no doubt true that children will grasp a key element of constructivist theory when they understand that "meaningful learning (is) to be gained from interactions with peers... rather than only from textbooks and... teachers", what is there to determine whether what is learned is in line with society's needs and expectations for the child? If this learning becomes directed to anti-social ends is this a desired outcome?

Most children already know from an early age that meaningful learning is available from peers - in fact teachers must daily contend with the need to convince the child to learn what is presented in the curriculum in addition to and sometimes in opposition to what is learned from peers.

The XO is a great machine design, but nowhere in the hardware or software is this problem addressed. That is beyond the purview of hardware or software, lying in the area of the social situation of the system, and so far OLPC has been silent about how this divergence of interests will be reconciled.

Merely to say that the child will "learn learning" and all will be well is to beg the question of "what will the child learn?"

Great article, thanks.

A lot of emphasis is being placed prematurely on the ideals of constructivist learning and other educational paradigms which may, or may not, become apparent when the XO is in widespread use.

Agreed, the XO has been designed with educational requirements in mind, and the social networking and interaction capabilities are (with Web2.0 technologies) going to give children much more than peer reviewed content rating and collaboration. But.

The most important fundamental need has been lost sight of:

The XO is the lowest cost delivery medium for educational content and literature (and everything in between) in 3rd World environments.

It is the Library of Alexandria for millions of children who do not have access to books.

That is the #1 reason why it will succeed.

How successful it will be depends on how committed communities are to making their core curriculum and other cultural legacies available online... Then they can worry later about integrating the higher potentials of the XO into the "classroom" . . . but it will probably be too late - the kids will have figured it out.

There are multiple satellite networks covering Africa, Asia, Middle East, etc. Satellite networks such as Hot Bird, Nilesat, WorldSpace, Mutlichoice IS10, SeSat, and others. All offer data and web services in addition to media. Shouldn't an OLPC device be designed to use these satellites instead of expecting nations to construct a broadband or dial-up infrastructure? The infrastructure for the OLPC costs more than the devices. Where are these local Internet Service Providers (ISP) coming from? America does not have ISP access everywhere, only in certain parts of the nation. So, I would not expect a developing nation to have 100% ISP coverage.

Also, starting off with a laptop may be expecting too much. Perhaps an improved PDA type device may be a better start. It took the industrialized nations 30 years to develop the infrastructure for things like laptops. Things like Intel CPUs and Linux software assume a certain amount of infrastrucure is already in place. Perhaps Symbian or ARM based software (PDAs, SmartPhones) would be more reasonable.

I am a firm believer in technology yet I believe even more that technology is less important than pedagogy. And yet we have Negroponte saying that teachers are not needed, training them wasteful, and all children need is the XO laptop to "learn learning".

If we are to count laptops in hands as success, like Negroponte wants to, I feel it will be a very hollow and short-lived "success".

"If we are to count laptops in hands as success"

The main problem is that the OLPC device may be $100 per child, but the infrastructure to support the OLPC device is probably $8,000 or more per child. Where is the electricity and or batteries to be supplied? Where are the Internet Service Providers? Are there going to be OLPC technicians or an OLPC helpdesk? Since OLPC is mostly hardware; where are the software enginners to be hired? Where are the professionals to be the content providers?

When you added all up; hiring teachers, even Montessori teachers, is less expensive.

Total Cost of Ownership (TCO) is reason #384 why it would be a hollow victory.

I'm not sure what you're basing your costs on Robert, but a more accurate figure would be (over a 5 year useful lifespan) in the region of $211, or less.

That would be $36 for internet access over 5 years (based on $330 p/m for a 1Mb VSAT shared between 500 students) plus $175 replacement cost for any broken XO over that period. Since I'm pretty certain there won't be a 100% failure rate - we can fairly safely budget at least $10,000 a year at each school for technical support.

Add on server and professional development costs, and say double my figures if you like... it's still an unimaginably long way off $8,000.

In most cases necessity will ensure that it won't amount to even a fraction of that.

Besides, the provision of the educational frameworks and content have already been undertaken by the OS community... and content can as easily be created as scaled for collaboration - if intelligently coordinated.

Think about it this way: would you rather have your kids practise their typing with lorem ipsum absurdum, or relevant rote material?

Specifically, (it sounds a bit like a sweatshop I guess;) the children themselves can be the workhorse that ends up digitalizing their entire curriculum... using free and open source technology.

I wish I'd read the $50 e-book reader article before posting that comment about the XO being the lowest cost distribution medium for content. I'd seen the headline but didn't associate it with a hardware device apart from the XO itself.

I blame high cost, slow internet access in South Africa . . . and besides I've been too busy searching for open content (for the 3rd World village school at which I try to teach computers) to have kept up with OLPC News on a daily basis. (I intend to change that, at least.)

My comment was prompted by my perception that some academics and programmers put so much emphasis on the paradigm shifting potential (altering the source in runtime) of the XO - that they have lost sight of it's more tangible selling points. And I feel this loss of focus can only lead to confusion and resistance in the closed minds of pedagogues (in the pedantic sense) and luddites who need little enough reason to feel threatened as it is.

So, the number 1 reason the XO will be successful (remains to be seen) is, possibly, as the choice delivery mechanism for educational content - not just e-text.

Improved writing skills, computer literacy, and social cooperation will just be incidental byproducts.

The point of a lesson is to teach a subject, and not the teaching tools. Have you ever seen Constructivist lesson plans? You should take a look at Montessori lesson plans and try to imagine how to translate it into software. Now imagine the lesson plan without a teacher to implement it. Imagine this teacherless Montessori lesson plan, and try to implement it in software. The creation and upkeep of such software would be a huge effort.

The point of education is enlightenment.

In a perfect world we'd all love to have Montessori or Waldorf teachers for our children - but the sad fact is that the quality and quantity of normal teachers teachers in the 3rd World is abysmally low by 1st World standards.

At our school here there are 9 teachers and over 500 pupils.

The XO is not designed to replace them, but rather supplement their capabilities by allowing them to streamline their own lesson plans and administration to make them more effective.

For the children the XO is a magical book that can provide almost any information on Earth... and they can draw and write in it too.

I guess there will be the same proportion of misfits to the model as there currently are with Industrial Age teaching methods... but this is the Information Age and not everyone can evolve at the same pace... Those that get left behind may enjoy a simpler world, but to exclude the rest of the poor from the riches of this New World on account of the imagined difficulties of migrating teachers' lesson plans would be... umm... ludditic.

Just the e-text, comics, graphics, audio and video capabilities of the XO will go a long way towards fostering real literacy and giving these kids a chance to achieve /anything/ they want /anywhere/ in the World.

Robert Lane wrote:
"The point of a lesson is to teach a subject, and not the teaching tools."

Sorry, but your arguments sound to me like Plato arguing against using books in teaching and philosophy.

"You should take a look at Montessori lesson plans and try to imagine how to translate it into software."

The answer is, "The Media is the Message". The technology structures your behavior and in the end, your thinking.

You could ask yourself why MS Windows users always feel ignorant and guilty, and even fear? If I speak to a random, unsophisticated Windows user these are the main emotions I encounter: Ignorant because they never understand what happens, guilty because they always seem to do the wrong things, fear because they can never predict how the system will react.

Compare that to Mac users. I have NEVER met a Mac user that felt either ignorant/powerless, guilty, or fear. (personally, I use Linux, which is again a different story)

The difference between Windows and Mac users is not the sophistication of the users, but the way their software (UI) interacts with them.

The same will hold for the XO/Sugar/Bitfrost. The central mantra of the OLPC is that the user (child) is in charge and should be safe at ALL cost. Everything is geared towards that goal. And that is also the central idea in Montessori's, Parkhurst's etc, educational plans. So you can indeed put those educational ideas in software.

There is more, of course, but I will keep it at this.


You still need subject matter experts to provide content. You also need them to correct student's mistakes. I am all for new media. Just like somebody needs to write a book before you can read it; You need a content provider and a software engineer for the software before you can use it. We would still need biologists for the biology software, mathematicians for math software, etc. Somebody needs to make books for the OLPC Library, and educators are needed for OLCP Activities.

Who runs this Blog?


I run this blog. You can find more info in the aptly titled "About OLPC News"