Problems with the $100 laptop

November 16, 2005 4:52 AM

Problems with the $100 laptop.With the upcoming press conference about the MIT media lab's $100 laptop, this article talks about its origin and some of the problems in trying to overcoming a digital divide. [via: del.icio.us/malaclyps/oblink]
posted by gsb (50 comments total)

Wow, fascinating article.

I've been around computers longer than modems have been common. Internet connectivity is not the first advantage I thought of, for developing countries' children. My first thought was cheaply reproduced learning materials (software and texts-on-disks). Perhaps people tend to forget, computers are useful even without internet access.

However, that is not to say that the problems discussed are immaterial. Clearly, in some locations, lack of electricity is a real issue, just to name one.

It may be arguable that many places have far more pressing issues with which to deal, than getting kids laptops. Yet I suggest that by addressing education needs, you set your feet on a path which will contribute to the resolution of the rest of the issues.
posted by Goofyy at 5:39 AM on November 16, 2005

Interesting link about children who have never seen computers before.
posted by adamvasco at 5:49 AM on November 16, 2005

You know, I've been involved with computers for most of my life, and I'm still not sure they're that much better for early learning. For LATER learning, ADULT-style learning, very much so. But the early stuff.... I think that might be better-served by just easy access to pens and paper.

I tend to think the point at which computers start becoming really useful to learning is after developing a solid basic body of knowledge... say, somewhere in the age 12 to 14 range. (likely earlier for gifted students or products of a particularly good educational system). At that point, a computer becomes, in essence, a force multiplier. But before that, I wonder if it's not as much a distraction as a benefit.
posted by Malor at 5:54 AM on November 16, 2005

This link raises valid questions about the possible disconnect between this project and the needs on the ground. There are charities that raise funds to send pencils to impoverished schools, and we're worring about wireless-enabled laptops? (I agree that they'd be sold off in fairly short order.)

Now the countries being sent pencils may be an order of magnitude more worse off than the countries that would be sent these laptops (I must confess to only a superficial awareness of the $100-laptop project), and it may well be that the infrastructure, in terms of electricity and national bandwidth, might not be an obstacle, but the concerns would probably remain.

Methinks we've been looking at technology as an educational panacea for far too long, and not just in the developing world either (my SO's school has lots of computers, but is desperately short on other supplies such as lab equipment and textbooks, for example).
posted by mcwetboy at 6:10 AM on November 16, 2005

I agree that they'd be sold off in fairly short order.

I read a case study where a goldmine wanted to increase worker productivity by giving them chickens to add to their rice and bean diet. The workers sold the chickens and the new money floating around created all sorts of unforeseen disruptions and the productivity went down.
posted by StickyCarpet at 6:21 AM on November 16, 2005

A hand crank of 6 inch (15.24 cm) length operating at 2 turns per second would require a tangential force of 11.8 pounds (5.3 kg), assuming 100% efficiency of generation and storage. This would tire a strong adult quite rapidly. It would seem apparent that the figure of 100:1 was arrived at by means other than calculation.

Heh.
posted by delmoi at 6:40 AM on November 16, 2005

$12 Laptops.
posted by Baby_Balrog at 6:46 AM on November 16, 2005

You know, I've been involved with computers for most of my life, and I'm still not sure they're that much better for early learning. For LATER learning, ADULT-style learning, very much so. But the early stuff.... I think that might be better-served by just easy access to pens and paper.

Well, using computers is a great way to learn about... computers.

I was born in 1980, and my schools were rife with computers. Apple IIs in elementary school and Macs in high school. The only thing I learned on a computer (at school) was how to type.

They were also very helpful for writing papers near their deadlines.

I also learned how to bypass crappy Mac security (and at one point lost all the data on my personal folder due to a software glitch. That was fun. Backups? What?)

Now it may be that if you hire competent people to write and design lesions you might actually be able to teach people things with computers, but I don't see that happening. Maybe in the future we'll get something figured out. (Like AI teaching software that you can ask questions too)

---

What really surprises me about this whole project is how afraid the members are of actually giving people computers. I mean, the system is going to be totally locked down. Why? Have we really reached a point that just giving someone a computer in a poor country a danger?

Kids also might be able to take online classes from home using the wireless system with these machines. Something that would be helpful if you live all dispersed.

If they really want to help these countries out, they'd figure out a way to setup the manufacture of these things locally so that the can be self-sufficient.
posted by delmoi at 6:51 AM on November 16, 2005

$12 Laptops.

Yeah, no kidding. $100 is actualy really expensive.
posted by delmoi at 6:52 AM on November 16, 2005

It would seem apparent that the figure of 100:1 was arrived at by means other than calculation.

I got a kick out of that too, delmoi. From now on, whenever someone asks me how I figured something out, I'll tell them that I used "means other than calculation." It sounds so much more dignified than "I pulled it out of my ass."
posted by Faint of Butt at 7:00 AM on November 16, 2005

Why does everyone take OLPC's high flying talk so seriously? Third world countries are the very long term goal, which they admit is unreachable at $100 per laptop. But a $100 (to $50) laptop on the desk of every first world child? Sounds like a good plan, one which should be supported. You know, we do have a "digital divide" in the first world too.

I do wish they'd drop the hand crank idea though, its just not needed in countries which can afford $100 per laptop.
posted by jeffburdges at 7:13 AM on November 16, 2005

To be tackled first: warlords "drafting' kids, water that makes people sick, blindness from lack of vitamin A, starvation, etc. etc.
posted by davy at 7:18 AM on November 16, 2005

Oh, and I forgot illiteracy.
posted by davy at 7:21 AM on November 16, 2005

My impression of computer facilitated learning is that it's not that effective because it's largely misused. When I was young, before PCs, the supposed utility of anticipated personal computing was often illustrated by the example of "balancing your checkbook". Once PCs arrived, of course hardly anyone used them to balance their checkbooks. Even now, where there is support technology and social change that makes this more attractive, still it's the minority who "balance their checkbooks" on their PCs.

But a certain kind of subject is well suited to computer learning, and it's where a dynamic virtual model of a process is used to give people a hands-on experience of something that otherwise would be completely abstract. I recall back in college when we were studying Ptolemy's Almagest (which is the geocentric model of the universe that was dominant for so long) that many students had difficulty comprehending any but the most simple motions—combinations that create retrogradations, not to mention the infamous equant, were to some incomprehensible. But I wrote a quick little program that in two-dimensions simulated Ptolemaic astronomical motions, allowing the user to vary the parameters at will and see the resulting motions traced on the screen. The tutor (what we call profs there) had me show it to the class and there were immediate "aha!" moments. When appropriately applied, computer assisted learning can be very effective. Missapplied, it's worse than ineffective.

And it's going to be more often missapplied when it's the result of ideology and not pragamtics. Felsenstein's objections to OLPC is a good example of this.
posted by Ethereal Bligh at 7:44 AM on November 16, 2005

To be tackled first: warlords "drafting' kids, water that makes people sick, blindness from lack of vitamin A, starvation, etc. etc.


I'm so glad that they tackled problems serially, ordered by importance during the US's history. It a couple years we might have universal health care.

Then we can move on to indoor plumbing, and maybe even automobiles that middle class people can afford.
posted by delmoi at 8:13 AM on November 16, 2005

davy: To be tackled first: warlords "drafting' kids, water that makes people sick, blindness from lack of vitamin A, starvation, etc. etc.

Well, to be fair, many of these problems can be most effectively solved by improving the quality and availability of education. The issue isn't that OLPC thinks laptops are more important than ending starvation, but that they think laptops are the best way improving education, which, as others have point out, is not necessarily the case.
posted by Espy Gillespie at 8:19 AM on November 16, 2005

Paul Farmer (Mountains beyond Mountains) wrote that "appropriate technology means shit for the poor."

I tend to think the point at which computers start becoming really useful to learning is after developing a solid basic body of knowledge... say, somewhere in the age 12 to 14 range. (likely earlier for gifted students or products of a particularly good educational system). At that point, a computer becomes, in essence, a force multiplier. But before that, I wonder if it's not as much a distraction as a benefit.

I teach with computers and do research in that area, and I agree for the most part, except that my son taught himself to read using Reader Rabbit software when he was four.
posted by craniac at 8:30 AM on November 16, 2005

A hand crank of 6 inch (15.24 cm) length operating at 2 turns per second would require a tangential force of 11.8 pounds (5.3 kg), assuming 100% efficiency of generation and storage.
How can tangential force be measured in kg? It must be mass. Tangential force is: F=MGSin(theta) if I remember right.

So I can't see what this guy was trying to calculate. I assume you would be interested in how much force one must apply to the end of a crank of a given length (L) to generate a given amount of watts (E) at a given efficency (ef).
posted by sety at 8:39 AM on November 16, 2005

Ethereal Bligh: here's some cool software animations of Ancient Astronomy!

We should team up and create an Almagest FPP...
posted by growli at 8:46 AM on November 16, 2005

does anyone understand the 100:1 argument in that article, or can anyone reproduce the results? the requirement means, if i understand correctly, that you should only have to wind 1 minute for 100 minutes of use. how much power does something like an ipod consume in a minute? this old wired article gives around 8.5 hours for a 2Ah 3.7V battery. so that's around 1 Watt power consumption. google turns up 3W for a transmeta chip. so let's assume 5W is their target. that means the crank needs to generate at 500W for the 100:1 ratio. a 6" crank travels about a metre (suprised me, but do the maths!) per revolution. if you turn once per second you have to do 500J of work over that metre, which means a force of 500N; equivalent to a tangential mass load of 50kg. which is a factor of 10 worse than the article. are MIT assuming half a watt for power consumption, have i got something wrong, or have they made a mistake?

anyway, their numbers aren't so bad. spending a minute doing fairly exhausting work for an hour and a half of computer use is a pretty good deal. my numbers are impossible - you couldn't turn that handle.
posted by andrew cooke at 8:48 AM on November 16, 2005

Malor, I think your point makes a lot of sense, but you haven't factored in the importance of learning how to use the tool. Just like it is important to teach kids about books and libraries early on, it is important to teach computer use and computer organization. That way they will be ready to 'hit the ground running' when they get to the point when they can really use the resource - wikipedia, graphing mathematical functions, playing civilizations...
posted by Chuckles at 8:51 AM on November 16, 2005

I'm thinking that this is a good critique with some very valid points, but it should be put to the test in a pilot project somewhere to see if it plays out as predicted.

From my meager teaching experience, I tend to agree that computers won't really help these kids that much. Okay for term-paper research, but less helpful for actual classroom learning (at least for primary school kids).

Now, as a tool for economic development in these nations, I'm thinking $100 laptops might be helpful. Something you could buy with a microloan ...
posted by moonbiter at 8:55 AM on November 16, 2005

Human powered generator.
posted by Chuckles at 8:57 AM on November 16, 2005

Okay, completely off topic, but this is awesome: Human power.
posted by Chuckles at 9:01 AM on November 16, 2005

thanks Chuckles - that supports the idea that 50W is just possible through hand cranking.
posted by andrew cooke at 9:02 AM on November 16, 2005

Free Mac OS X Spurned By $100 Laptop Creators.
posted by ericb at 9:15 AM on November 16, 2005

The test of doing good is to go back in six months or a year and see what good has been done. The trouble with this plan is it requires large scale to be cost effective; they have to build huge numbers in order to meet the target price, which makes for that much bigger a boondoggle if it just plain doesn't work. If they could trial it with 1,000 or 10,000 units and then see if the computers just vanish, get broken or sold or stolen, then they'd have learned a relatively inexpensive lesson and can move on to another approach. But you can't afford to design and build 1,000 $100 computers.
posted by George_Spiggott at 9:19 AM on November 16, 2005

... and a good thing they spurned OS X, too. MacOS has more strings and hooks than Windows.* It's just that Apple gets a pass on them because they're "not evil" or something silly like that.

--
* Mac user full-time since 2/2005, and sporadically since 1987.
posted by lodurr at 9:22 AM on November 16, 2005

I read this. It says in detail some things I've been saying quietly for a long time w.r.t. this and teh Simputer.

But I don't say these things too loudly, because though none of these projects will accomplish their stated goals, they will nevertheless accomplish important things.

Sometimes the peripheral benefits are more important than the primary ones.
posted by lodurr at 9:24 AM on November 16, 2005

On the cranking issue, that kind of rotating repetitive movement isn't very good for you -- just look at the skeletal problems they've found in the preserved bodies of prehistoric kids who grew up having to grind grain all day. It wouldn't put the price up much to make a simple foot-pedal lever/pulley crank arrangement for it, which would put the burden on larger muscles and transfer the repetitive movement to larger, more durable joints moving in the way they were designed to. Never happen, of course.
posted by George_Spiggott at 9:28 AM on November 16, 2005

I think that human powered link must be measuring power at the electrical terminals, there would be a further loss when stored - NiMH batteries have a 66% coulometric charging efficiency (pretty low!). I can't imagine getting the $100 laptop down to 5W total consumption, under 20W would be realistic, maybe 10W - 3W for a very efficient CPU, similar for associated on board electronics, and similar again to run the display.

On the other hand, 100:1 would be a fantastic use:charge ratio. I think it would still be practical at 10:1.

And I guess I should have linked to the previous discussion on this topic initially: One Laptop per Child.

Finally, a much more practical and inspiring project, linked several times in that thread (not that there is anything wrong with the laptop idea): Hole-in-the-Wall Computing. If this search is anything to go by, the Hole-in-the-Wall project has never seen a FPP, it should!
posted by Chuckles at 9:30 AM on November 16, 2005

Damn, that could have been much more coherent...
posted by Chuckles at 9:35 AM on November 16, 2005

A lot of this critique illustrates the weakness of requirements-driven design. The project advances based on the requirements, not on a real design; the requirements are (right now) unrealistic, and include some problematic components, and some people take that as damning the project.

It's baby:bathwater.

People lack vision with regard to this sort of thing. I mean, they see the original vision, but they don't see the spinoff benefit. The Simputer, for example, could serve as essentially an open-source reference platform for kickass handhelds. Instead, it's niche, and struggling, because people take the spec too literally.

The real power of an idea like this isn't $100 laptops -- it's $500 laptops, with similar capabilities and less subsidy, and minus some of the more egregious requirements. That's how you seed the market. Hell, if I could get something lighter and smaller than my powerbook, with fewer capabilities but portable file formats, i'd be all over it. (I use standard file formats all the time, anyway, even for email.)

on prev: chuckles, don't sweat it, good link effort. Hole-in-the-wall is a great example of throwing things against the wall [pun intended] to see if they stick. Looking back at what I've written, I'm thinking that the problem isn't lack of vision so much as lack of willingness to try more than one thing. People get married ot the specific vision, and they don't expand their view to include other versions of it.
posted by lodurr at 9:42 AM on November 16, 2005

Pedal power.

At the rate I play on the Internets I'd have legs like oak trunks and a waist like a wasp.
posted by davy at 9:51 AM on November 16, 2005

Perhaps one the intended uses of the $100 laptop can be found in the final question of Marc Stiegler's final exam:

You live in North Korea. Three days ago the soldiers came to your tiny patch of farmland and took the few scraps of food they hadn't taken the week before. You have just boiled the last of your shoes and fed the softened leather to your 3-year-old child. She coughs, a sickly sound that cannot last much longer. Overhead you hear the drone of massive engines. You look into the sky, and thousands of tiny packages float down. You pick one up. It is made of plastic; you cannot feed it to your daughter. But the device talks to you, is solar powered, and teaches you how to use it to link to the Web. You have all the knowledge of the world at your fingertips; you can talk to thousands of others who share your desperate fate. The time has come to solve your problem in the most fundamental sense, and save the life of your daughter.

posted by AccordionGuy at 10:04 AM on November 16, 2005

I think you're right, Malor. So does Todd Oppenheimer in this Atlantic article (via Arts & Letters), which most people here have probably read and discussed. Oppenheimer deals with the delusion that a computer is a transformatory learning tool. Another problem is that a computer, apart from being a learning tool, is also a toy. That's distracting. Boys + distracting toys does not equal good concentration early on.

At least speling punktuation and grammer have improved since the Internets, id hate to loose that.
posted by QuietDesperation at 10:23 AM on November 16, 2005

growli, that's a great page—thanks so much for bringing it to my attention.

I don't think the equant simulation is as helpful as it could be, though. My guess is that it will be helpful in clarifying some things to people who already basically understand it. People seem to have a very difficult time grokking the equant.

My description of the equant is to ask someone to imagine a circular model train track, a ball that rolls on the track with a horizontal hole in it for a rod acting as an axle around which the ball turns, and a motor in the center which pushes the rod. The ball is free to move inward or outward along the rod. Imagine this system in motion. Now imagine moving either the track or the motor off-center. What happens? The ball continues to trace the same circular path on the track as before, but now it also moves inward and outward on the rod and slows down and speeds up, respectively. The center of the path the ball traces is still the center of the circle, of course. But the center of the angular motion of the ball around the track is the location of the motor. That location and the circle it implies are the equant.

By the way, it's my personal but not-uncommon opinion that the equant is the straw that breaks the geocentric camel's back. I don't think that a "why?" is intuitively required by the presentation of motion in a circle, nor even motion in a circle around a circle. But motion in a circle around a circle with a rate of angular motion centered elsewhere? Why? Even so, I don't really think this inelegance is sufficient to make an Earth In Motion palatable, nor comprehensible the great distance to the "fixed stars" such that there is no observable parallax. That's just crazy talk.
posted by Ethereal Bligh at 10:30 AM on November 16, 2005

Ethan Zuckerman's writeup. Well thought out and reasoned with plenty of external links.
posted by dhacker at 10:51 AM on November 16, 2005

oh, i skipped a paragraph in their argument. they have a factor of 5 difference because they assume 1W consumption and an extra factor of 2 in the equivalent load on the handle because they turn twice as fast.

so we do agree on the physics. sorry about that.
posted by andrew cooke at 10:52 AM on November 16, 2005

How-to hand crank power your iPod /off-topic
posted by eddydamascene at 11:00 AM on November 16, 2005

This is an extremely well-thought-out article (with a few caveats), and makes some serious and thoughtful objections to the goals, the scale, and the structure of the project. That said, I still support it because of the disruptive potential it has -- even if it fails in some of the ways Felsenstein outlines.

I really can't wait to see what comes out of it. Even if just 0.1% of the laptops end up creating a master programmer, it may be worth it, because imagine what problems they'll be motivated to solve. To date the PC has largely been used to solve problems of want, problems of first world peoples. I wonder what third world peoples with problems of need will use it for? And I wouldn't presume to decide in advance that they don't want it. Even if you say that the funds might be more effectively spent, I think that multipliers will kick in and the value of the project will be many-fold what is put into it.

I will say here that the kind of collaborative, creative projects I hope that this will spark are notably not the way the project itself is being promoted and designed. I'm not entirely sure why that is, except that Negroponte is very goal-oriented and believes this is the fastest way to his goal. What he's designing, though, is a more corporate, top-down project that seems less likely to spark the creativity that I would want. That troubles me.
posted by dhartung at 11:25 AM on November 16, 2005

I couldn't find a link, but I remember there being a conference of billionaires about bridging the digital divide i the third world... it was like Gates, Ted Turner the 3Com guy and a couple other dot-billionaires.

Ted Turner basically called the rest of them out and accused them of wanting to lock in marketshare for their own companies... and went on to say that the first priorities should be clean water, electricity, plumbing, basic medical care, etc.

The internets can come after that, was his basic thing.
posted by illovich at 11:28 AM on November 16, 2005

Well thought-out article? Not when it contains a condescending hypergeneralization of this magnitude:

In developing societies children are perceived to have a place in helping the family advance, not in racing ahead and leaving the family behind.

Ah, yes, that monolithic developing world with its rigid and homogenous value system and approach to child-rearing.

Any argument that assesses the problem by treating the whole of the developing world as a single, known value and not an incalculable variable couldn't possibly begin to grasp the transformative power that putting cheap communications tools in the hands of poor children might have.

There are surely technical and logistic wrinkles in a plan this ambitious, but I think the spirit of it is great. In international development circles, many of the biggest successes have come when the recepients of aid are given tools to use to decide which problems need fixing and how instead of (as davy for example would have it) parachuting in like a First World Triage Unit to apply bandaids to the (perceived) worst problems.
posted by gompa at 11:54 AM on November 16, 2005

gompa, I agree with you on that last point. I think projects like this should go ahead, for reasons I've already stated. (dhartung put it better with the term "disruptive"; I hate that term, but the common meaning of it encompasses what I'm after and I think he means what i mean.)

But what i think is very interesting about projects like "hole in the wall" is that they do something like what you're talking about: They put it out there, and they see if people use it. And they do. So let's find ways to do more of that.

It's not either-or, it never is. There is more than one problem, there is more than one solution.
posted by lodurr at 12:18 PM on November 16, 2005

Just give everyone a leapfrog.
posted by smackfu at 1:44 PM on November 16, 2005

Oh, and I forgot illiteracy.

Don't forget over population. Many of these countries have more people than they can support even on a good day. Throw in military juntas, ethnic cleansing, droughts, plagues like HIV/AIDS, etc etc and I don't see how a computer is anything other than a liability to the poor child unfortunate enough to be mugged or killed by someone who wants it.

Books, pencils, papers, chalkboards and most importantly teachers - - this is the way most of the world has been taught for hundreds if not thousands of years.
posted by wfrgms at 2:51 PM on November 16, 2005

The leapfrog that smackfu suggests as an alternative to the $100 laptop is an "intelligent" pen that uses micro-coded paper as input and speech as output. Very, very cool. Reminds me (somewhat) of the primer from Neil Stephenson's The Diamond Age.
posted by zanni at 2:59 PM on November 16, 2005

As an inhabitant of the 3rd world, can I say fuck (almost) all of you? For the condescension part, the physics discussion I have no qualms with.

In developing societies children are perceived to have a place in helping the family advance, not in racing ahead and leaving the family behind.

That got to me too. Right, everybody not in the U.S. or Western Europe can't wait to sell their kids into slavery, and of course helping them have a better life has never even occurred to them.

I read a case study where a goldmine wanted to increase worker productivity by giving them chickens to add to their rice and bean diet. The workers sold the chickens and the new money floating around created all sorts of unforeseen disruptions and the productivity went down.

Which proves, of course, that poor people are too stupid and evil to be helped, so why the fuck even try?

The world's poor and middle class children will figure out their own uses for computers, thank you very much, without all your handwringing about the 'unforeseen disruptions'. And hopefully kick your asses.
posted by signal at 6:25 PM on November 16, 2005

"The real power of an idea like this isn't $100 laptops -- it's $500 laptops, with similar capabilities and less subsidy"

The iBook is only $999. Somehow a $100 laptop isn't that impressive to me. e-Machines are $350 or something. What would impress me would be a $10 laptop.
posted by muppetboy at 7:36 PM on November 16, 2005

muppetboy, i cringe -- I thought $500 was too high as I wrote that, but I wrote it anyway. Silly me. My point was lost, though, I see (because I didn't make it well).

And it was this: If you can create a platform that serves commercial ends, you can ramp up volume a lot faster. That creates a bigger installed base, and thus a bigger target for developers -- both propreitary and open-source.

The payoff for corporations is that they don't have ot do basic research on the hardware -- or they can share it. It worked for x86. (Vis that $350 laptop, which is basically a Wintel brick, minus the Win.)

This is what I thought would happen with Simputers. It's a pretty cool platform. I thought companies would use the Simputer as a reference design, then build smaller-faster-more-expensive versions for upscale and "first world" markets. They could subsidize the low-end. Maybe that's happening and I just don't know that market well enough, but I don't think so. Instead, it's being sold only as an upscale consumer device, and only through one vendor, AFAICS.
posted by lodurr at 12:41 PM on November 17, 2005