Neil Kandalgaonkar

hacker, maker of things

Thoughts about Outernet

Outernet is a bold concept to broadcast a stream of data to the whole world via satellite. Here’s the founder’s vision video, complete with inspiring soundtrack.

This all sounds really cool until you realize they are talking about a 100MB/day stream. One single stream for the whole world, that amounts to a handful of ebooks. And the Lantern can’t even read them; you need something else like a laptop or mobile device anyway.

I did a little tweetstorm about them - here they are with some expanded comments.

As Tanenbaum said: “Never underestimate the bandwidth of a station wagon full of tapes hurtling down the highway.” If what we want is a library of e-books and Wikipedia content, I bet I could seed the world with thumb drives for much less than the price of one satellite launch.

But wait! Outernet isn’t just a static set of e-books, it also updates… at 100MB/day. Of content you don’t choose. I can’t imagine what I would do with such a service, and I doubt the founders can either, because they handwave about other people’s problems.

In the video, there’s a quick flash of those bulletin boards people post, to find people after a natural disaster. So there would be a trickling broadcast of such info, every time there’s a disaster? It is a very big world and there are lots of disasters every day. Is the appropriate response to distribute specialized devices, and then have everyone fight for attention in a trickle of data? We really can’t think of anything more localized or cheaper?

The anti-censorship claim is for me, the most interesting one. On the other hand, as their founder even points out, the same thing is happening right now with ordinary satellite TV.

So the big difference is that we have a mostly static data set, slowly updating. So when the Wikipedia article about the local Arab Spring movement updates, everyone gets the same information. There are interesting political possibilities once a large group of people can act together, synchronously, outside of a censorship regime. However, is everyone buying these devices in anticipation of a revolution, or because they want to educate themselves?

Naive people have this model of technology that works like this – there was no possibility of doing a thing, and then some genius invented the thing, and then boom, it changed the world. It almost never works like that. Instead you usually see a few people trying to do something new in a crappy and difficult way, with the most advanced technology of the past iteration, at a high cost. Then some genius comes along and productizes that, and then it becomes economical for the whole world.

So if a slow-updating Wikipedia was really useful to activists or resisted censorship in some important way, I think we’d see people carrying around clandestine wiki diffs on motorcycles. Maybe that’s already happening but I’m skeptical.

The worst possibility is that this is a roundabout way of getting the world to fund their ambitions of creating a satellite ISP. So perhaps the development-tech angle is all a sham.

But let’s think of it as if it was a sincere effort to make tech for the developing world, or the parts of the world subject to state-based censorship (which is everywhere now, but who’s counting)

Bootleg CDs and DVDs for sale by sonyaseattle

I find there’s a very damaging tendency to think of development tech as operating by different rules. As if the intended recipients weren’t intelligent or part of the global economy and didn’t have the same needs you or I would. Like, we know we need two-way fast internet, but surely those benighted poor bastards would be happy with a one-way, 100MB/day internet, that makes them dependent on some organization with headquarters in the USA? Because there’s no way they can copy DVDs or thumb drives and they have never heard of BitTorrent. And it’s not like regular cell phone service isn’t exploding already in the developing world.

Another annoying thing about development tech projects is that they act as if tech was an act of will, and not part of a global supply chain. One Laptop Per Child had a similarly bold vision. And after years of struggle managed to invent netbooks about one year before the market did anyway Except they did it with pricier devices, experimental interfaces, and mesh networking that never really worked well. And then left everyone stranded when they disappeared. We would deride any for-profit CEO who delivered a product of such low quality.

Oh, and by the way, one of the justifications of the OLPC was that they would host static Wikipedia content, too.

At first glance, Outernet seems to combine the flawed vision of the OLPC with the high cost and low agility of a space-based project. Such projects tend to be lapped by more prosaic tech, iterating quickly.

On the anti-censorship front, just today, a technology that the developing world really has embraced – WhatsApp – announced end-to-end encryption.

I’m not arguing that the market will always provide or that we shouldn’t have foundations or activist tech projects thinking about these issues. But there are ways of doing that with available tech on the ground, where the locals define project goals, where we are mindful to remain open to iteration and experiment. And then there are very ways of spending kajillions of donor money on stuff that makes a great backdrop for a TED talk, indulging the fantasies of our technocratic elite.

Unfortunately, the incentives for developed-world non-profits will always tend to the latter.