Scott VanDenPlas

Tonight, Harper and Gabriel came over to talk about TinyLanguage, but we also spent a lot of time messing around on the XO machines we have. It is a lot of fun to have a couple of them together right next to each other, you really can have a good time with the collaborative editing and games.

While we were connected up to the community, a number of students joined in. They said they were from Thailand, Ban Samkha specifically. We quickly realized that we were currently talking to students in the pilot program! I got to tell them all about my job with the Field Museum, and how we have sent researchers to Thailand to study. I even got to speak to a teacher in the program, who has been using the laptops for 9 months. They told me that the students learn very quickly, but that the parents are also extremely excited. The students even knew where Chicago was… from maps on the internet.

From the 45 minute conversation, I gained a number of new pen pals and a whole slew of new perspectives. It is amazing to see the impact that these machines have.

I’m reading a book called Savage Inequalities by Jonathan Kozol (read it there). It is a brutal look at the American public education system, and how it reflects racial inequalities and the inequities of how our tax dollars are applied to the school system. It covers the Chicago Public School system circa the late 1908’s and early 1990’s, and discusses the atrocious conditions and lack or resources faced by our public school students.

It made me wonder, are we any better off today?

I am a regular reader of the District 299 Chicago Public Schools’ Blog, and I came across an article stating that nearly 10% of Illinois schools could be classified as drop out factories. It turns out that Wells High School, my neighborhood public high school, has a drop out rate of almost 50%.

While seemingly we have massive gaps in education today, just as in Kozol’s 1992 look at the schools. While we live in a world where 50% of our students, at any institution, FAIL to even graduate… we cannot consider ourselves successful as a society. By any measure, education is failing. Kozol comes back a number of times to recount the expenses in wealthy school districts versus poor districts. I don’t have the research behind me to show if this is still true or not, but it is probably safe to presume.

So while the public schools still struggle and the environment has shown little change, the context of what is available, for very cheap or free, has grown immensely. I ask this question, what resources can now be replaced? How much do we spend, per student, on books and libraries in school? Public school libraries are a waste of time and money, they are inadequate and outdated. Replace them with cheap computers and internet access. Provide free English tutoring (I guess I will plug TinyLanguage here). Give students Wikipedia.

Maybe instead of trying to find resources that aren’t there, and seeking help that just is not coming, we need to reorder how we think of the resources we already have. We may have to let go of some things that we hold as sacred.

2 months ago, I wrote about Intel’s counteractive response to the One Laptop Per Child program. In the face of the negative press around their actions, and more likely the positive press around the AMD 50 x 15 program (and in turn Negroponte’s project), Intel has taken a step toward collaborating on the OLPC project rather than trying to sink it.

The question is… why? The XO machine runs an AMD Geode processor and is in direct competition to the Intel Classmate. Intel has attempted to subvert the low cost laptop market by offering machines below cost. They have distributed anti-OLPC marketing rhetoric to Negroponte’s target countries. There has been a war of words between the chip maker and the OLPC project founders for at least 12 months.

Intel does support educational programs. They have put nearly a billion of their own dollars into educational initiatives around the world, and it makes them look completely asinine to oppose an innovative approach like OLPC. It is expected of them to support a program like this.

More importantly, Intel has lost this battle, but they do not want to lose the war. The XO is flat out better than the classmate. It has a stronger program, solves the correct problem in an efficient manner, and has a wide base of community support. The market for a cheap laptop is huge, but the profit margins on that market are very thin. The current XO machine and OLPC program only address a small fraction of that market, and evidence suggests that if this program is successful, a laptop made for older students could be in the near future. The XO has a predetermined expiration in a student’s educational lifetime. It is not made for teenagers and adults.

The success of a program like OLPC creates another market for them to expand into. OLPC is sowing the seeds of technology supported educational programs early on in a child’s life and will prove successful with it. What happens when that child goes on to more advanced education without that infrastructure? They are once again put in a compromised educational environment, ripe to be augmented with another cheap laptop solution. If Intel wants to capture this market, a feasibly larger market than the children’s machine market, then they certainly cannot be at odds with the OLPC program.

Intel is making a wise investment in their own future.

Intel graffiti picture by Brooke Novak.

Harper and I went to Barcamp yesterday over in Wicker Park. I was showing off the XO and answering questions about the OLPC project, while Harper mucked around with 500,000 devices he brought in tow (including the super awesome bluebox).

I’m headed back over there this afternoon to catch Jason Jacobsohn from CEC talking on business resources for entrepreneurs. I’ve never met Jason, but I’ve been seeing the Illinois Innovation Accelerator Fund pop up all over the place lately (which CEC manages). In the least, I think his talk will be very interesting.

Lately, I have been thinking about how we learn language, and what teaching methods are available to a widespread audience in order to learn language. I find it shocking how little is actually available.

This lead me to thinking of various testing methodologies, including computer adaptive testing. I have taken a number of technology certification exams in this format. They vary in length based on how you answer, and try to narrow down your exact skill level by changing the difficulty of the questions asked based on your previous answer. It is somewhat like the Socratic method of test taking. The machine asks me a question, and based on my answer, modifies its next question until it can figure out where i stand. Buried in this logic is what’s called Item Response Theory.

Item Response Theory takes into account a number of assumptions, including error rates in various ways. This got me thinking about another concept called the paradox of choice.

From wikipedia:

Observed in many cases is the paradox that more choices may lead to a poorer decision or a failure to make a decision at all. It is sometimes theorized to be caused by analysis paralysis, real or perceived, or perhaps from rational ignorance. A number of researchers including Sheena S. Iyengar and Mark R. Lepper have published studies on this phenomenon. (Goode, 2001) A popularization of this analysis was done by Barry Schwartz in his 2004 book, The Paradox of Choice.

Essentially, the more choices you have, the higher the likelihood you will make an error. I think it is a reasonable assumption that the more choices you get in testing, the higher the error rate will be… and not because you know less, but because there is a certain amount of paralysis when selecting from many options.

I wonder how this is correlated to the idea of universal truth. I don’t believe in any gods, and I have a hard time believing that there is any sort of Universal Truth. I am a bigger fan of the philosophy that collective small personal truths give you the exponential decay of Big T Truth. But in a world with the paradox of choice, the reality could very likely be that little T truths in sum may cause paralysis in decision making. Too many realities, too many truths, too much for us to process. This could also be why humanity is so fond of declaring their own personal truth as universal truth.

I have read a lot this morning in reaction to the OLPC episode on 60 minutes, and I think there needs to be some clarification.

Yes, Negroponte is pissed off at Intel, it doesn’t take a lot to get there. The problem is, he is NOT afraid of Intel because of competition, he is pissed off at Intel because they selling machines at a loss to gain market share.

The OLPC project is reliant upon large numbers of machines (the $175 price tag is expected to drop to the $100 range within the next few years depending on demand). OLPC is building a program, not just laptops, to support a different approach to education. Intel is trying to sell more chips, and is a little peeved that OLPC uses AMD.

Intel’s approach is not to build a cost effective, self supporting program. It is to drive the competition out of the market, and in order to do this, they are selling their machines _at a loss_. Intel cannot sustain this approach, and hopes to drive out OLPC (a sustainable approach) to increase it’s market share. Ultimately, Intel will have to raise it’s prices (even higher still, to compensate for early manufacturing losses), whereas OLPC will decrease it’s prices.

This is why Negroponte is mad. Their approach harms the longevity of ANY laptop education program, not just OLPC. In the end, Intel is looking to make as much money as possible off their machines whereas Negroponte is trying to build a model that will continually decrease in price the more they are manufactured. Intel will not put more laptops in children’s hands, they will slowly strangle more money from third world governments with a broken and predatory manufacturing model.

Can we stop analyzing this now?

I have long thought that the biggest blunder made by Negroponte from the beginning of the One Laptop Per Child program was the name itself. I’ve brought this point up before, but calling the machine a laptop brings back an unwelcome and misguided familiarity with the program before you know anything about it. As I have spoken to friends and family about OLPC, the first question that they always have is “Do poor people really need a laptop?” I think that the best way to reply is “No, but they do need education. This is not about the laptop, it is about the program to disseminate a learning tool.”

The language calling the XO a laptop is harmful to the goal of the project. It allows people to dismiss the project early on as unnecessary, and act as if the goal of the project is the first world injecting forced technology into the third world. It damages the image of the project.

It also allows hardware manufacturers compete with the project by undercutting the OLPC with a similar “laptop” sold below cost. The competition is often a fully featured laptop computer, lacking the educational program structure and support of the OLPC program. 60 Minutes covered OLPC last night, and they spoke specifically to Intel attacking the program by attempting to sell cheap computers to the same countries, but at a fraction of the manufacturing cost. If the focus of the program was access to education and a revolutionary education program, rather than the hardware itself, this would take away the power that the competition already holds.

Here’s the video of the 60 minutes episode from last night:

The OLPC program has officially stated the while the initial plan and long term goal is to benefit third world and impoverished countries, the only qualification for the laptops to be rolled out in a region is government buy in. One of the largest issues that the program has to overcome is to avoid or mitigate the problem of a black market forming around the machines. There is a lot to overcome in third world societies, and I fear that there will be too much time spent overcoming the problems and focus will be shifted away from the furthering of the program. It takes a long time to establish an education system, but a large number of machines can be sold into the black market in a matter of days. The loss of inertia for the program could be devastating.

There is a small push to introduce the program to the United States. I believe that this should be taken even further, and there needs to be a much larger installation here, in Chicago, with political support and a low amount of corruption. I think there could be potentially a huge cost savings for the Chicago schools overall, and more of a democratic approach to distribution of public school resources.

As I mentioned before, Google has an interest in developing a city wide network. I believe that offering the Chicago Wifi contract to Google (or another interested party possibly) in return for subsidizing the OLPC Chicago program has a massive amount of potential. Google not only gains the right to build the network in Chicago, but they also gain a much larger built in user base by supplying connectivity to the schools. What is now consolidated in the OLPC server system could just as easily be hosted at Google, by Google.

Chicago could reduce their costs by using Google’s wifi network as their school connectivity. They would get low cost machines for the students (and possibly even free depending on the depth of the subsidy) and free internet access for residents.

I think there is some hope here. I can see this as a big step forward for the OLPC program, and a good way to maintain a strong image in the face of many challenges.

Read the OLPC-Chicago mailing list thread here:

http://mailman.laptop.org/pipermail/olpc-chicago/2007-May/000035.html

A few weeks ago, my good friend Harper got his hands on an OLPC XO-1. Its a great little machine, and I think very well designed and engineered. On the tech side of things, I like the simplicity and focus of the machine, but I am most interested in the networking capabilities. I think the project is really poorly named though, they really need to ditch the idea that these machines are laptops. They are a learning tool that happens to resemble a laptop, and that word is poisonous to the overall goal of the project.

I think the politics and policy behind the program are much more interesting than the hardware and software of the machines themselves. As with any good design, that should be invisible to the user (and hopefully eventually to the designer). I have a feeling that the program is headed the way of the Segway right now though.

Don’t get me wrong, I love the idea behind the Segway. To the user, it is an incredibly simple tool. It solves an interesting problem… it is stupid for me to drive my car to go get a cup of coffee. Dan Sturges points out that it is absolutely insane for me to take my car, that weighs 50 times what I weigh, for such a short distance. It makes much more sense for me to take a Segway, that weighs half my weight, to run on short trips. Dean Kamen is not trying to build a competitor to cars. Just as a car doesn’t compete with a plane, the Segway doesn’t compete with the car. It doesn’t make a lot of sense for me to drive from NYC to LA, and it doesn’t make a lot of sense for me to drive from my house 2 miles down the road spending 20 minutes in my massive car waiting in traffic.

The problems for the Segway started with the hype. I remember a lot of big names saying how it would change the world, and cities would be planned around it. I remember hearing it would be hydrogen powered, or it was some sort of flying machine. Then when it came out, it was a self balancing electric scooter. A slow, kind of dorky looking, overpriced scooter that I couldn’t legally ride anywhere. It came across as one of the most over hyped products in my recent memory, and suffered a tremendous loss from this tarnished image. My bike goes faster, never needing a charge, and I only sort of look like a tool riding it. Oh yeah, and it was a tenth of the cost.

I fear that the OLPC program is coming up on a similar moment in history. The hype is enormous, but I am not entirely sure there is anyway the program can live up to it. I don’t see anyway to avoid these machines being sold into a blackmarket to serve more immediate needs of the people they are trying to help, like food, clean water, and shelter. It is an incredibly smart design, but the fault is not in the machine, it is in the supporting systems around it.

If the OLPC program succeeds, it will not be in the third world first. The program needs to leverage the attention of the first world, flood the United States with the XO’s and develop out an incredibly strong infrastructure with a minimized threat of corruption and a black market. Then export that program to the third world.

I would like to see the Google and the City of Chicago team up on the laptop program. Chicago is still in the RFP process of building out their public wifi network. Google is busting heads in wireless spectrum auctions and is clearly looking to expand its networks. The OLPC program has ramped up to the 3 million machine mark to roll out the program full time.

Why wouldn’t the City of Chicago grant Google the wireless contract for the city in return for subsidizing the OLPC program? I think it is a fair question to ask.