Sunday, June 26, 2011

TEDx IIT: No Shortcuts

I'm about 3/4 of the way thorough Dr Laura Hosman's OLPC talk, Technology for Development: No Shortcuts". She's speaking directly from experience in Haiti, Senegal, and the Solomon Islands. And it's brilliant and honest about what it takes to make a program with a tangible, sustainable impact. We need to send this to everyone who applies for OLPC laptops.

What can we do to help her convince laptop-hungry countries to put sustainability first?

Monday, June 20, 2011

The most difficult language

In school, we were reminded constantly of how lucky we were to be native speakers of English. You see, English is the most difficult language. Not so, I was told in Uruguay. Spanish is more difficult, but only the third in complexity. There was disagreement about the top two. I suggested Chinese, and sure, that was more difficult. What about Rutooro, spoken in Uganda? Or Tuvan, spoken by nomads in Russia? You would be hard-pressed to find speakers of those languages at the legendary linguistics schools of Columbia and Yale. Doesn't that make them more difficult?

Now that I'm learning a new language for work (no, not a new computer language, a people language) I'm thinking about how languages work. Like any programmer, the edge cases fascinate me the most. Fluency in American Sign Language is almost interpretive dance: YouTube music video

And if you can't whistle, this language would be outright impossible: Video

When we're so focused on digitizing books and making culture into a mass media commodity, we forget that computers don't come with every language in mind. The language which I'm using at work was only added to Unicode in 1999, meaning before then it was impossible to type it on a computer. My smart phone shows it as blank boxes. When 90% of our languages are expected to disappear this century, being visible on a cell phone, and being printable on a Kindle could be deciding factors for an entire culture. There are arguments for a survival of the fittest / translate or perish attitude, but something about that has never rung true for me. A computer scientist usually uses the best language for the job, because there are both general and specific traits that make languages stand out. Just as different people-languages change our way of describing and perceiving life. There are many great dinosaur comics that explain this better than I can.

Saturday, June 18, 2011

Of maps and medicine

There was a great TED talk posted this month about technologies which make present-day doctors look like trepanners-the guys who did brain surgery with stone tools. Medicines formulated for your DNA, inkjet printers making kidneys, nanobots, that sort of thing. watch on TED

Although genetics undeniably work, chances are small that I'd encounter it unless I was very ill and had an especially thorough doctor. Daniel Kraft, the speaker, has paid a great deal to get details about his DNA, indicating some health probabilities but not any obvious tune-ups. From a practical perspective,I think the only advice you could give to a med student today is, "this looks important; keep an eye on it". How else can you prepare someone for a career in a field you expect to be reinvented?

When you're gazing out a bus window at block after block of unswept shanty towns, on your way to talk about sensors and mapping, any technology has that same distant feeling. Explaining to someone that I use satellite photos to get directions or plan the growth of a city, that I can translate between languages that I've never spoken-in the end neither of us can say what is possible, and when it's going to arrive in their neighborhood. I'm convinced, we have to spread mini examples of the technology. Just as the first computers were oddities for our museums and labs, why not bring hackable phones and id chips and robots into discussions about education and development. And not in the context of what awesome gadgets we have, but with a message that all of it is constructable, malleable. If a kid in Peru or Philadelphia is good at math, they probably could write a Facebook app much more easily than writing an essay in a foreign language. I don't know that people see that, or that the world's average Facebook user sees it as any more accessible than a fighter jet engine. I don't expect Facebook to go out of its way to make that possible. Instead, I think it sets a challenge for everyone developing technology that kids use in schools. How easy is it to open the cover of your program ? Can an accomplished hacker of your program move on to rewire the real world?

When I was a kid, I reprogrammed a race car game to cause complete havoc - cars with the mass of black holes shooting off into space after a collision, objects with super powerful springs in their center, bounding from one track to the next, even a track which I could win every time because I knew the AI would trap itself. There aren't comparable experiences with video editing and map making and hardware hacking, at least not yet. There ought to be a Project Euler for these media/maker skills, like "trick my printer into writing backwards" or "disguise a city map as a leaf in this photo". Little things that take clever hacks to solve.

Friday, June 10, 2011

Khan Academy Follow-up

Reading my earlier post on Khan Academy you might think I'm opposed to it, but just a few days later I found myself defending it when its brand new history section was getting some undeservedly zealous scrutiny from the paper-based media. I still believe in the core point that Khan is promoting, and, importantly, I think the lessons are informative.

My main concern is the real differences that exist in technology and cultures around the world. Khan Academy was famous in the US and adopted in some schools before Khan's TED talk-even before Spanish translations had even been started. You can either believe that no one had heard about it, or, that it was harder to accept. Look at how many teachers were afraid about being replaced by their OLPC laptops. What will they think?

My ideal would be taking our existing content server, Pathagar, beef it up with multimedia support, then send it out with a diverse library of local content, classic English kids books, and Khan Academy for subjects not covered by most laptop content, like basic math and biology

The frustrating thing is, we're so close to making this reality. Just grab the right hardware and software, a few interested people, and we could solve this in a week. The reason no one can make that type of commitment is that we're having trouble finding enough free and open content to make it useful to the school. Khan Academy could be what we needed to tip the balance and provide a full content library. The main problem then, as I described a couple of days ago, is localization. Who's going to step up and develop this, and what schools are going to be the guinea pigs? (ps if it's Peru, don't call them guinea pigs)

Tuesday, June 7, 2011

Khan and the XO

I've been reading along with the excellent ideas in the olpc community about sharing Khan Academy videos with schools around the world. Ever since my report on Khan Academy for Plan Ceibal, I've been thinking about the best ways to put aside some difficulties and make this a reality. Still, there are some major red flags to consider:

Translation-longer than you think

Khan Academy videos are narrated, and animated, in English
Due to how they are animated, some translators are redoing the entire videos to make the written content match and sync to the narration. Is this still Khan Academy, or a volunteer inspired by Khan? I can't say. But it means that a quality translation, to haitian creole for example, would take a lot of skilled and tedious work, and subsequent translations could be just as difficult.

Secret sauce

The selling point in olpc, and the larger education community, seems to be Khan Academy's unique progress tracking and problem review system. It's the central point into research from the Gates Foundation and other key players. What we won't be seeing in the near future is an opened up version of this system, no matter how cool it would be. Like Apple's strategy to keep the iPhone exclusive, Khan Academy recognizes that their structure doesn't just help them stand out from the rest-it's their unique blend of content, careful explanation, and integration with the education progress system We can attempt to duplicate it on the school servers, but Khan won't be letting just anyone experiment with the platform and risk losing the polished image that education research has of his project.

XO vs Video

I have a lot written in my notebook about this, but I'll just say: nightmare.

Education culture and content

First off, there's precious little content in Khan Academy below algebra. This could change, but it'd take a concentrated effort by teachers. If you're just going to translate it afterwards, you might as well make the content yourself.

Second, you'll notice that Khan Academy is a little weird to explain to people. From a traditional perspective, we should be using the internet to listen to experts and lecturers, not some one on YouTube... We've reached a meritocracy in some ways in the us, to the point which we accept Khan Academy on reputation and results. This kind of thinking is not universal. It's also very focused on the individual-one of the reasons we trust in Khan is that he's invested a great deal of his time and energy, which makes him less of a stranger than our actual professors. Does all of this translate across cultures?

I think we need to look at khan academy in the context of other education crazes such as Schoolhouse Rock. Disruptive and interesting, but not necessarily the solution. Whatever happened to constructivism? Lets not give up on individual creativity yet, please.

TEDx Montevideo-Plan Ceibal

I haven't had the chance to watch the full video, but the head of Plan Ceibal gave a TEDx talk last week in Montevideo

Video, in Spanish, is online on YouTube