OLPC Tablet Distribution Proves Concepts Laid Out In ‘The Diamond Age’

We don’t remember hearing about the One Laptop Per Child initiative distributing tablet computers but apparently a couple of shipments were distributed to rural communities in Ethiopia. The problem one might think of in this scenario is that the literacy rate in the two test villages was basically zero. But that’s exactly the population targeted with thr technology. The tablets were loaded with a software package called Nell. It was designed to guide a child in self learning by telling them engaging stories that include teachable moments. If you check out the white paper (PDF) you’ll find it’s pretty much the exact same teaching technique that [Neal Stephenson] wrote about in his book The Diamond Age. But keep reading that paper and you’ll see that this is because the researchers took their inspiration from that very novel.

Well the results are in and apparently [Neal] knows exactly what he is talking about. Not only did the children learn from the software, but within five months they were hacking the device (which runs Android) to get the disabled camera working.

[Thanks Alexander via Dvice]

31 thoughts on “OLPC Tablet Distribution Proves Concepts Laid Out In ‘The Diamond Age’

  1. Pretty darn cool.

    I think there’s a bit of hyperbole in this, though. It’s extremely unlikely that entirely self-taught children learning from just the materials on the laptop managed to “hack” the camera, unless there was some outside influence or information involved; or perhaps it was just a setting in a GUI somewhere, in which case finding and clicking a checkbox is hardly a “hack”.

    1. It sounds like they locked down the desktop to prevent customization. Perhaps this was the same program that disables the camera and in figuring out how to get to the desktop settings they also managed to re enable the camera. I could totally see something like this being hackishly written to start with making it easy to figure out how to get around it with little knowledge of the underlying system.

  2. This proves Stephenson wrong as much as it is inspired by his book. His ideas about AI being forever incapable of matching the abilities of humans (and by extension, being unable to act as satisfactory teachers) were extremely misguided, in my opinion. The view of the future that he puts forward in that book is wondrous, prophetic, and inspired, except for the backwards idea about AI! It is so strange in the context of the rest of his envisioned future, it sticks out like a sore thumb.

  3. LOL, looks like the kids didn’t like the locked down GUI and found a way to hack it open to be able to configure it to there liking… I think this proves the case the locking down tables is insane. I guess only Zombies like a lock down device.

  4. May I challenge my enlightened colleagues @hackaday to come up with some *objective* signs or measures of educational success?

    I am on record extensively with my complaints about OLPC failing to evaluate deployments. I am often rebutted with the fact that we do not seem to agree on what constitutes “success”, and how to assess early and often how we ‘doin, so we can re-engage, etc. I want a bit more /agile/ vs. continental drift.

    Could y’all help us a bit? thanks!

    (I go by yamaplos in the OLPC community (oh, the poobahs don’t like it we capitalize the registered name, ok, “the olpc community”, then))

    1. If it brings a bit of joy to the life of an ultra-poor child living in the worst part of the world, then it is a success, even if that is not the goal.

      As for “The Diamond Age” concepts, it has long been known that even AI isn’t required for a person to learn without humans. There are tons of examples of people teaching themselves to read and advancing their own education from there…all with plane old analog books. I would think that the only problem in making a laptop or tablet to take a child from nothing to PHD would be in making it interesting enough that they would continue on with it.

      1. 1) hmmm, certainly.
        Return on investment? would we be better off distributing MP3 players, potentially usable with SD cards with recorded knowledge/classes? Anyway, kids would love to use them for the music.
        What’s the sweet spot? (some friends are using $50 smaller tablets, in Ghana)

        2) YES, but. Not every child, alas, can go by himself from failed nation poverty to PhD. So far, out of 1/2 million XOs in Uruguay, years later we still need just one hand to count the kids who are actually coding and visible from open source communities, at least one of them emerging from distressing poverty (I’m following up). A net gain, of course, as there were apparently none before.

        SO, how can we help best those who have the giftings and the gumption and the real potential?
        Because we can’t afford to give a tablet to everyone, can we?

      2. My first computer — the one I learned to program on, was free. It was a TRS-80 Model IV. Giant thing; must’ve weighed a ton. My fourth grade teacher gave it to me before he left to go to China for a few years.

        He chose me; I had no clue how to use a computer — I had almost never used this one in class — other kids were much better at operating it. I asked him why and he said he could tell I was interested in it.

        Looking back on it — I was probably staring at it anytime anyone did anything with it, it *was* something I was very interested in as a child.

        Recently, I was at a park, with a friend of mine — a juggler. She was juggling, doing her thing, and of all the kids, one just came and sat down — kind of a bit distant, but he just stared for a while… she stopped when she noticed him, had him grab some pine cones, and she started showing him how to juggle — because he seemed interested.

        You can teach anyone anything if they have enough interest. When you start going out to little rural communities and you’re handing out laptops — soon enough, the kids who are really interested will poke their heads up to see what’s going on — they’ll stare, and when the opportunity comes, hopefully they’ll either be at the front of the line, or the people giving the laptops out will take some notice them.

        If neither of those two things can happen, then it’s just a random shot in the dark, isn’t it?

  5. allow me to quote a better informed friend who prefers to remain anonymous,

    There was a nice presentation of this project at the SF meeting. There were 20 Zoom laptops distributed in each of two villages. The children in these villages do not have a school. The intent was to determine if children could learn to use the tablets to learn to read without any support on the ground.

    For me, the initial result was that tablets are far easier for the children to use than the touchpad based XOs. It appeared that the kids sustained interest in the laptops over some weeks. However, the period was far too short to learn anything about literacy. About half of the children were able to match sounds to letters (names of letters, e.g a, b,c) – one of the activities on the tablets.

    The literacy program is based on RAVE-O, an English reading program for children with reading disabilities developed at Tufts University. I am not sure about trying to teach Amharic speakers to read English – but hey.
    also see:

    Mike Lee shared links to the videos of the presentations:

  6. [Mike Szczys], please correct the image and reference to this being an “OLPC tablet”
    The Ethiopia experiment relies on Xoom tablets, not the promised but still not in-the-wild OLPC tablet in the photo. The link with OLPC comes from some OLPC employees running this project, with only a very thin connection with the current Open Source community that works around education uses on OLPC laptops, alas.

    OTOH, I celebrate that OLPC is doing pilots, and even pre-pilots. OLPC has had a policy to discourage pilot projects, which has made it very hard to do it without full OLPC and local government buy-in.


  7. Just knowing that there are potential coders/developers out there that got the chance to learn more and grow from the OLPC project make it worthwhile.

    I can’t imagine what my childhood would have been like without computers and it’s good to know that in this day in age we are finding ways to spread that joy to some of the poorest places on Earth.

    I do however hope we don’t forget about our own roots, problems, and children in the countries we live in (like the US for most of us) that need this kind of chance as well. You don’t have to look far to find poor people in the USA…let’s help our own nations needy children first please.

  8. Just like with Sesame Street, The only self learning here is a child deciding to turn on an appliance. From there it’s instructors(Human no AI here) doing the teaching. Just as in as in Sesame Street, and in bring & mortar class rooms the teachers are required to keep the lessons engaging to keep the kids wanting to attend class. Not that I’m saying this can’t have value in helping disadvantaged persons, but it really is going to depend on many creative long distant teacher as much it will the kids. To hack the camera the kids had to know or figure out in some manner there was a camera available that could be made to function. Reads to me like that was a situation was set up so the kids could find a route to get the camera working. Part of the research?

    1. I think hacking here means they found the settings panel. While i want this to succeed, I think it’s an article designed to acquire funding for research on self learning.

  9. A similar story is in “Earthweb” by Marc Stiegler. Part of the backstory is the “top drop” where cheap palmtop computers were airdropped by the millions all over the world.

    Why do that? Finding solutions to the problem presented by a series of hostile alien probes trying to get to Earth. Getting input from as wide a range of people as possible then figuring out which ideas worked best, then paying for the ones that worked.

    The book was published in 1999, before anyone coined the word crowdsourcing.

    A “top drop” would be possible now. Buy any licenses and patents required to build devices on par with something like a Handspring Visor with color screen (2nd color Palm OS PDA. Palm IIIc had 8bit, Handspring countered with 16bit color.) add WiFi mesh networking for basic web and e-mail. Add a VGA resolution camera and how cheap could each one be when manufacturing a billion or more? Come up with a socket to insert a simple wire for extending the wireless range, include instructions on the device for what lengths of wire to use for the antenna.

    Use Palm OS and there’s plenty of free apps, including Mobipocket reader which can open the newer Kindle format books simply by changing the file name extension to .prc or .mobi Dunno if the old app can open the latest Kindle book files but if not, just use the older format.

    Make enough of them and distribute them far and wide, eliminate them being of any value to steal or sell on the black market – though I expect certain sorts of people would try to confiscate and destroy as many as possible, y’know in those places where they keep blowing up the schools.

    1. In order for a network of ideas to work , you need a good worldwide database of proposals and the sub-requirements that make them up.
      Very possible with the web today, but no ones done it.
      – Most problems can be broken down into smaller problems. With 7 billion people or so, the chances are someone knows a solution to some of the some problems, and different people will have solutions to other bits. A ideal network would make it easy to match them together.
      (especially as solutions almost always can be reused in different contexts….)

      I think you could get a giant “tech tree” for humanity out of it.

      The nearest thing I have come too this idea is:

      But its still pretty far.

  10. The OLPC team annoy me. I wish they’d just open up their project to everyone (at the real price). The raspberry pi took off and were able to drive costs down hugely by selling in mass.

    OLPC should also make their products available to the world (not buy one for the price of two or anything daft like that) and then use our sales to help push down their costs and do more for the kids in need.

      1. The principle of tax-exempt non-profits is that they do owe the community, which subsidises their existence.

        Several other comments here talk about how solutions might come from collaborative action.

        As such, there has to exist a live give-and-take between non profits and anyone who has an opinion. OLPC may chose to listen, or not. I believe that listening does inspire more and more people to get involved (and contribute), while refusing to listen runs the notion into irrelevance… And kills any prospect of sustainable community collaboration.

        (many, many of us over the years have asked OLPC to allow the purchase of small numbers or even single units, so far with not much result, though the minimum seems to be down to 100 units now, from several thousand a couple years ago)

      2. No, sorry to break it to you, non profit organizations are under no obligation to give back to the general populace. What you’re saying is a bit like saying a cancer charity should fund treatment for my bowel condition.

        Solution to what? I wasn’t aware there was a problem.

        This is an organization set up to supply 1000s of laptops/tablets to developing nations, not individual pieces of equipment to well off people in developed nations. It’s completely different, the distribution and support infrastructure is completely different.

  11. Sounds like a great idea, but really short on content, kind of like all of the kick-starter spam on /r/technology/new, but with the money that being connected brings. Surely MIT and Negroponte could do a better job of making this guy a bit more open. More to the point, why isn’t there videos of how to use SUGAR, SCRATCH, and WIKIPEDIA available for these machines in native language, or even english and native language. That being said, this is a fascinating study in human factors.

  12. Oh I remember it well and how they used the laptops as basically house lighting from the screen. Doubt anything got clicked on most of them. We may as well just send them those “emergency radio” flashlight things with the solar and crank charge usb port. Or maybe someone will stick a bunch of them on fenceposts to keep the lions away as we saw in another HaD post earlier. lol. OLPC was a misguided noble cause to say the least.

  13. The artical is short on details but I am amazed at the skepticism of the principle.

    Of course people can learn without active, present, teachers. We are learning all the time, all over our lives – merely by trail and error, cause and effect.
    We dont learn our native toung by specific teaching – we learn it from natural observation.Likewise many people learn a second language via “Rossata Stone” software and the like.

    The question of whether a program can be as effective as a real teacher more depends on the quality of the teacher for the subject, verses, how well the program was designed.
    Making a well designed learning software is very hard indeed – but certainly not impossible and not needing “advanced AI”

  14. shouldn’t we let them develop on their own ? like you know some basic stuff first – killing and stealing is bad & if your life suck that’s because you suck not because there is some race that hates you ??

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.