On the XO Laptop

On Education, Innovation, OLPC, And Open-Source

with 2 comments

[First published as On Education, Innovation, OLPC, And Open-Source on October 21, 2007.]

I used one of the new OLPC machines last week while attending the K12 Open Minds Conference for 2007 (k12openminds07) that was held last week in Indianapolis, Indiana. [1]

While I used the machine for only a few minutes, I came away very impressed by it, in part because of my recent thinking about the growing importance of open-source to education, and the shared sense of adventure and hope that I found at the conference.


I am a computer scientist and programmer by profession. I have a Ph.D. in Computer Science and put the food on the table by programing from the end of the my freshmanyear in college for close to four decades, until early in 2003. Most of my programming from 1996 to the end of 2002 was released by my employer, IBM, in open-source form. Since 2003 I have continued to work on open-source, but in a role that requires little programming.

I do have some experience as an educator. While in graduate school in the early 1970’s I was hired by the NYU School of Education to work for several months on a part-time basis to develop some mathematical curriculum materials for students in the “Follow Through Program” in the Atlanta Public School System. (Follow Through was an attempt to show that the success of Head Start in teaching basic skills to young children could be extended to older ones.)

I also taught a one semester senior-level elective course in the Ada Programming Language in my last four years at NYU before I went to work for IBM in 1987.

Innovation in Education

While I don’t recall that I was very good at either of these educational jobs, I did learn a lot from them.

I have also been working for much of the last year on a volunteer exploring ways to use open-source to help educators.

I’ve given a few talks on this recent work,and also had numerous conversations about it, including many at the k12openminds07 conference. I have often said on these occasions that “Education is the hardest area in which to innovate.”

Now I could give a list of the reasons that might help convince you this is so, drawing on the experiences outlined above, but on thinking about this recently I have come to appreciate, that while I have not spent my lifetime in education, I have several direct experiences in the difficulties of innovating in education, some of which go back a long way.

My first experience with the difficulties of innovating in education began in the summer after the ninth grade, in 1959. That is almost fifty years ago.

My mother knew some folks who had just started a small company called Teaching Machines, Inc., or TMI. They produced what was called “programmed learning” texts. These texts, which I now appreciate were constructed in much the same way as computer programs, were based on the work of B.F. Skinner, a famous psychologist of the era. One of the co-founders of TMI, Lloyd Homme, was a Ph.D. in Psychology who had studied under Skinner.

Because of her connections, my mother suggested I try out some of their early materials. I started with their course on Statistics. I completed it in record time,and then set a record for matching or bettering the best score that tested mastery of the materials. I then went on to do the same for all the other course materials they had prepared, as a result of which they hired me on a part-time basis. I started in proofreading, then went on to write texts on my own, add review sections to the courses written by others and so forth.

These were very exciting days, both because I found school dreadfully boring, and the work at TMI was the first time I was treated as a peer by a group of adults. It was one of the most influential experiences in shaping my life.

Those days were also very exciting in that almost everyone in the company believed TMI was on the way to redefining American education,

Few experiences in life are more exciting, more exhilarating, more “heady,’ than believing you are a member of a team that will truly change the world. TMI was one of those experiences.

On rare occasions, the team may actually change the world. Think for example of Bill Gates and Steve Ballmer and Microsoft, as well as Marc Andreesen, and many others who have achieved enormous success in the software industry.

More common is that the team fails, and the members move on to other jobs and challenges.

On other occasions, overconfidence can lead to disaster, as happened by my mother. Our landlord’s wife was also a psychologist. She caught the TMI “bug” and, believing there was money to be made, enlisted my mother in starting a small business that offered after-school instruction based on TMI’s learning materials.

That business nearly bankrupted my mother. I was fortunate in that I won a full tuition scholarship to Caltech, so she only had to worry about paying my room and board. Looking back, I doubt she had more than a few hundred dollars in her bank account when she put me on the train to Pasadena, and it took her years to recover.

TMI also failed.

Looking back, this was my first experience in the difficulties on innovating in education.

The next experience came several years later when I was asked to travel to the University of Illinois in Champaign-Urbana to evaluate a project called “PLATO,” as my advisor Jack Schwartz had been asked by the NSF to lead an evaluation team of PLATO.

My first visit to PLATO came around 1977. I flew to the campus with Ed Shonberg, who had recently joined the NYU faculty. We were met at the airport by two of the key people on the project, Paul Tenczar and Bruce Sherwood. They immediately took us to the PLATO lab.

We arrived around 7PM and stayed until 3 or 4 in the morning. It was probably the most exciting evening of my professional career.

PLATO was a project to create a new way of education based on computers. Only those who were part of the project or had a chance to observe it in action can fully appreciate the revolutionary promise it offered. Suffice it to say that within less than twelve hours I first observed laser displays, chat rooms, highly interactive educational materials, a new way of writing educational software, sophisticated computer games, incredibly fast response times, and many more things I can’t recall right now. Just trust me, it *really* was that good.

The team believed as solidly that they were going to change the world as did I while still in high school working at TMI.

PLATO, like TMI, ultimately failed, I think mainly because it was ahead of its time in that the personal computer wasn’t yet on the scene and so PLATO required very expensive hardware.

That was my second experience in the difficulty in innovating in education.

The third experience has come over the last quarter-century, the era of the personal computer. It’s one we all share.

Ask yourself, “What are the three major advances in education that have occurred during the era of the PC, advances that truly reshaped the way we teach?”

Answer, at least in my view, “None.” To paraphrase our good friend Holmes, who can be found in my recent “Dr Watson” posts, “The dog did not bark.”

OLPC (One Laptop Per Child)

I have been waiting for close to three decades to see an application of computer technology to improve education that was comparable to PLATO in its potential impact.

I finally think the search has ended. That would be last week, when I first had a chance to use a real OLPC.

I will write more in future posts to expand on this, and perhaps convince you to share the excitement I feel about this innovation.


As I said earlier, I have been waiting a long time to see an application of computer technology that can truly improve education.

I have been working in open-source for most of the last decade, and my recent work on exploring how open-source can help improve education has given my confidence that meaningful innovation will soon be upon us, and I take the OLPC project as the first real evidence that my confidence is merited, and that open-source will not join TMI and PLATO as failed efforts.

Though OLPC is still in its early days, I do think it will not fail, and that it will prove to be an enormous success.

The main reason for my optimism is that OLPC is not too early, not ahead of its time, in promoting a new technology.

OLPC is a remarkable piece of hardware design. The prototypes already available serve as a proof point that computers worthy of use in the classroom can be produced for under two hundred dollars. OLPC is particularly notable in that was designed to work in the hostile environments that are found in developing countries, where electric power and internet connectivity are not readily available.

But the main reason for my confidence is that OLPC is based entirely on open-source software, so that the cost of production is essentially solely the cost of producing the hardware. With the application of of Moore’s law, the use of commodity components, and the reduction in production cost as more and more OLPC machines are produced, the hardware cost will inexorably go down … and down … and down.

I believe that another kind of commoditization will help improve OLPC.

As OLPC becomes more widely deployed, there will be greater and greater interest in the underlying open-source software. This will have many effects. I expect that it’s only a matter of time until that software is as well-supported as is the Linux kernel itself. For example, assume two million OLPC’s are known to be in use, and that a serious security issue is then discovered. I have no doubt that an army of programmers would immediately form to track down and fix the problem, providing a level of support that IBM or Microsoft can only dream of providing.

Also, once a critical mass of OLPC’s are deployed, there will be a real incentive to develop new, high-quality applications. It is one thing to write a KDE addon or a Mozilla plugin that may be used by perhaps a hundred thousand users. It will be a whole new ball game once developers know that a good piece of code can find its way onto OLPC’s in use throughout the world.

What’s the critical mass? I don’t know, though I would venture it is less than a million OLPC’s, and more likely just a few hundred thousand.

I’m also extremely confident because of a property of open-source that I think only I and my fellow open-source developers fully appreciate.

If you read the earliest post in this blog you will find an account of a panel discussion I attended in February, 2006. It was part of the IT Academy, a program supported in part by IBM, intended to attract high school students to careers in technology, or to at least educate them about technology. Some of those students labor under great economic disadvantage.

One of my colleagues on the panel was an LTC colleague and Linux strategist. He noted that IBM estimates it takes about a billion dollars to create a new hardware platform. IBM has some experience in this area, including the creation of the BlueGene supercomputer and the Cell chips that power the gaming platforms of Microsoft and Sony.

But chips alone are of little value. They need software to do useful work.

The open-source community has been hard at work for over two decades to create that software, so that now there is a complete “stack” of solutions available in open-source form, ranging from the code to be found in a BIOS chip that is executed in the moments soon after a computer is turned on to bring it up as a useful platform, all the way of to the latest BlueGene/p supercomputers that are the fastest supercomputers in the world. [3]

All that code is available at no cost. It is there for the taking. It would cost several billion dollars to reproduce it, to write an equivalent software stack starting from scratch.

That is why I am so proud to have played a role in creating the current open-source stack. For over the last two decades part of the programming community has created a tool in the form of software that is worth several billion dollars.

I can think of no other community that has donated so much work of such quality to the world at large.

What do actors give back? Baseball players? The NFL and the New England Patriots? [4]

It is open-source programmers who give back.

I have written several posts that have explored the question, “What is open-source all about?”

Here is what I think it is all about to many of the developers who have written it, myself included.

Each of us does this as best we can, adding a line of code here, a patch there, starting some projects that succeed, others that fail, all the time trying to help each other out, and training newcomers in what it means to be a part of this community.

We also deal with a lot of crap, primarily by trying to clean up our code and help others clean up theirs. We also deal with all the nonsense from the media, the analysts [5], the software companies that actively oppose open-source and in at least one case have said that “open-source is a cancer.”

We also deal with disputes within our community: GPL versus Apache, how to revise the Linux scheduler, KDE or Gnome, vi or emacs, how to fend off the FUD sent our way by others, and all that.

We do all this happily, though it is often unrewarding, a nuisance, and a distraction from our real work, writing good code in the hope that others may find useful.


We know many people use open-source every day, often without even knowing it. That is nice, but it’s not the real reason.

We do it because every so often some people want to do something that may really help make the world a better place and they need software to help them.

We don’t know just what software they may need, so we write as much of it as we can.

I know some of those people. They include the people who are working on the Sahana Project, and the people who are help making Ubuntu our best shot yet at bringing Linux to the masses.

And, on rare occasions, though I trust these occasions will become more common in the years ahead as open-source gets better and better, we are reminded that it is worth every needless distraction, every bug that we had to fix, every minute we wasted reading Slashdot or lwn.net instead of coding, or twittering away, or even blogging.

We got our most recent reminder this past week. It was sent our way by the over 300 attendees at the K12 Open Minds Conference.

Almost all of them were educators. And, courtesy of the OLPC that Mako brought along, they got to see what we can do on their behalf.

Keep on coding — it does matter.



1. “k12openminds07” is the recommended tag for writing about this conference. Use it.

2. That is why I take such pride in that I have paid my own way since the end of my freshman year in college, beginning with my good fortune in securing a well-paying summer job at the Air Force Weapons Lab in Albuquerque.

3. I’ve managed proposals in both these areas in my job at IBM. That’s one reason I love it so. It gives unique insight into what IBM is doing in the open-source arena.

4. See Injury on the field — now to the commercial

5. I don’t include the Redmonker’s in this group. They understand open-source.


Written by daveshields

December 11, 2007 at 6:37 pm

Posted in xo-laptop

2 Responses

Subscribe to comments with RSS.

  1. […] would have to be tried, even though some would fail, for that is the nature of true science. See On Education, Innovation, OLPC, And Open-Source for some more thoughts on the difficulty of doing innovation in […]

  2. […] met Robert when Ed and I made a visit to Champaign-Urbana to review Plato, as described in my post On Education, Innovation, OLPC, And Open-Source. I also worked with Robert on several implementations of Spitbol; see for example Computer History […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: