The pitfalls and promise of Internet.org and Project Loon

My first career, before I got into technology, was in international development. I spent two years as a Peace Corps health and water sanitation volunteer in south Cameroon, followed by several more working as a project manager for a number of USAID-funded health workforce programs in South Sudan, Kenya and Vietnam. I still maintain a lot of ties to the development community, and remain personally aware of what the “digital divide” really means and looks like for the majority of the world today.

It is thus with keen interest that I’ve kept tabs on tech industry “goodwill” initiatives for the developing world – in particular, Facebook-led Internet.org and Google’s Project Loon. These two projects have many noble aims in common: broadly, in Facebook’s words, bringing internet connectivity to “the two-thirds of the world’s population that doesn’t have it.” If you watch the high-gloss YouTube marketing videos for each respective project, it’s hard not to get a little starry-eyed at the gee-whiz technology they employ.

Unfortunately, the understandable enthusiasm around these projects has totally outpaced their feasibility. The keystone components of each – exotic new airborne platforms to beam internet down upon the earth – are sheer technologist fantasy. Improving connectivity in the developing world is happening, and several “new billion” users are waiting to access the digital world – but very little of that will have anything to do with either Google or Facebook. Here’s why.

It’s estimated that there will be three billion humans with internet access by the end of 2014. (That’s up from two billion in 2010 and and one billion in 2005 – the pace is accelerating.) For the majority of internet users today, and certainly going forward, fixed connections (PCs) are not the first or primary experience of internet connectivity – it’s mobile devices, primarily smartphones.

Device costs have steadily come down in every region of the world, as have connection fees and the cost of data. Much has been written about how many developing countries are leapfrogging expensive, last-generation copper wire altogether and building out widespread wireless infrastructure instead. Indeed, Facebook’s own research shows that 80-90% of the world’s population today lives in areas covered by 2G or 3G networks.

If you’ve never spent much time in a developing country, let alone in a rural, isolated village in one, it’s hard to convey just how incredible this is. Suffice it to say, radical the democratization of knowledge it offers will completely change the world – indeed, it already is.

How do you connect the world?

Through Internet.org and Project Loon, both Facebook and Google (respectively) have offered similar, if not competing, solutions to address how to connect the remaining 60% of the world’s population that does not currently use the internet.

Internet.org – This is actually a consortium of companies (Ericsson, Nokia, Qualcomm, Mediatek, Opera and Samsung, but led/fronted by Facebook) that offers a multi-pronged strategy, including:

industry collaboration to reduce the cost of connectivity by developing cheaper devices and reducing the amount of data that apps consume
negotiating “zero-rate” deals with telecom providers in developing countries to provide free access to a new Internet.org app
developing new “network extension” technology to provide access to new users – drones for some users and special orbital satellites for others

Project Loon – a product of Google’s secretive “moonshot” lab, this is a comparatively more focused project limited to developing networks of specially outfitted balloons that will float in the stratosphere while “beaming down” internet connections.

The technology in either project is really pretty amazing – solar-powered autonomous devices that can stay afloat high in the atmosphere for months, if not years, at a time? Wow. Go watch some of the videos on either site – they’re awesome.

That said, they’re also a totally impractical way to roll out internet connectivity to the vast majority of the world.

A cold bucket of water

Not all of the solutions offered above are bad ones – as I’ll get to in a moment. Unfortunately, the good is pretty strongly outweighed by the ridiculous, and it’s certain that far more is being spent on the latter.

Anyone who tells me that fleets of specially designed, experimental airborne platforms that beam down data connections – either to ground-based repeater stations or to individual devices themselves – are an easier or cheaper option than using widespread, commoditized, proven technology available off the shelf today has a very difficult case to make. Facebook’s own research, as mentioned above, shows that 80-90% of the world’s population already lives in areas covered by existing mobile networks. Natural growth of network coverage is likely to grow that number even further. Enhancing those existing networks – as, say, has happened in every part of the world with data-heavy connections today – is an obvious and more cost-effective strategy right off the bat. Indeed, some countries are jumping right to widespread 3G coverage from almost nothing at all.

In addition to the cost-ineffectiveness of the drone/balloon platform, you have the practical one. How many governments do we think will warm up to the idea of Facebook drones flying above their airspace beaming down unregulated, ungovernable connections to their citizens? (Would any Western government allow that? Ha!) How would those airborne platforms’ bandwidth capacities scale? If one Loon balloon can support, say, 1,000 3G connections, how will it support 10,000 connection attempts? How about when those original 1,000 suddenly begin demanding more bandwidth? Sure, the connection technology can be enhanced, but at the same rate and scale as demand? I doubt it.

Most people who work in international development soon get used to playing the role of wet rags to others’ enthusiasm. They wind up having to explain to excited would be do-gooders why their charity or NGO idea is either impractical or actually counterproductive and harmful (ex. TOMS or “1 Million T-shirts for Africa”). Unfortunately, most of these bad charitable plans go forward anyway, which is how you get dozens of young Americans flying halfway around the world to amateurishly build houses in Africa for a week, when the cost of their airfare alone could have done so much more.

Of course, providing connectivity to the developing world is not the only application of these platforms. Post-disaster and emergency connectivity could be a realistic one. And, of course, the folks at Google X and Facebook are far from dumb, so a part of me suspects that they must understand why these solutions won’t really work for the scenarios they’re marketed for. Airborne platforms can have other interesting uses – like real-time mapping – which could be part of an ulterior motive.

Okay, fine, it’s not all bad

Criticisms of airborne connectivity platforms aside, some of the other strategies that Internet.org present actually make some of sense – primarily around reducing cost. But it’s a pretty mixed bag.

The cost challenges to bridging the digital divide are about two things: devices (i.e. smartphones) and data. Internet.org proposes to reduce the cost of smartphones by, among other things, supporting open source device standards and innovation to bring down the cost of production. I don’t personally see a lot there – smartphone costs, like most other kinds of consumer electronics, have crashed in price anyway as the underlying technology has become completely commoditized and demand has risen.

Reducing the cost of data, on the other hand, is more interesting. The most promising idea the organization has proposed is zero-rate deals for the Internet.org app, which just went live last week in Zambia. AirTel Zambia customers will have basic, rate-free access to the app, which includes 13 very useful services (including Wikipedia, Google Search and AccuWeather – as well as Facebook itself). Zuckerberg himself has described this approach as an “onramp to the internet,” designed to entice users (particularly those in the growing African middle class) to consider spending more of their discretionary income on internet connectivity.

Of course, zero-rate deals are very controversial in some quarters, but the major social networks – Facebook andTwitter – have been aggressively expanding them for years in individual markets. Zero-rating does present a potential problem for local competitors, of course – it’s hard to compete with “free” – and national regulators must weigh those costs against the benefits of more widespread access to information.

Finally, one of the more confusing projects underway is a Facebook-Ericsson collaboration called the “Innovation Lab” that offers developers the ability to test their apps in network conditions and on devices simulating those found in emerging markets. Sounds great… except that the Innovation Lab isn’t located in Lagos, Nairobi, Mumbai, Guangzhou or Lima. It’s at Facebook HQ in Menlo Park, California, where most developers are not exactly focused on catering to users whose ARPU is a fraction of those found in Menlo Park, California. And I can’t envision lots of Ivorian, Kenyan or Indonesian developers requesting the Innovation Lab’s help in testing apps for the devices/networks they themselves use.

A Xiaomi beats an OLPC every time

One tragic flaw of the history of international development is its utter failure – often, even, resistance – to learn from past mistakes. The tech industry’s dreams of connecting the “bottom of the pyramid” are no different. An interesting case study that’s particularly relevant here is the “One Laptop Per Child” debacle, which began with the lofty ideal of developing and distributing basic, affordable laptops for children throughout the world for educational use.

To say their plan didn’t exactly work out would be an understatement. A consortium of industry partners, including Intel, Red Hat, Google, eBay and Nortel, were never able to develop the new device form factors cheaply enough. Distribution was a mess. There was very little demand by educators. But ultimately, what really killed OLPC was that substantially better and cheaper devices – laptops, tablets, smartphones and the like – were developed in the meantime by industry for the consumer market. No one needs an OLPC laptop when you have Xiaomi.

(While in my old job back in 2008, I actually got to play with an OLPC laptop for a day. Even then, I was unimpressed by its features compared to a relatively high-end mobile phone then available in most any African market.)

Could a non-profit build a cheap laptop with a bunch of industry partners? Sure. But was that really the most practical, let alone cost-effective, way to democratize technology? Not at all. So it is with the Internet.org and Loon projects. Device costs are sinking just as network coverage throughout the world is expanding. We do not need new mobile devices or flying ISPs, and the millions – more likely billions – being spent on them could be much better put to use on less sexy, but more impactful, efforts like seed financing for emerging market startups, subsidizing bandwidth expansion or laying new fiber optic cable.

Solutionism doesn’t lead to development

One constant source of frustration to many working in development is the constant proposal of new technological fixes to old problems. It seems like every week, the Gates Foundation or MIT’s Poverty Lab comes up with some new whiz-bang “solution” to a niche problem in developing countries: “next generation” toilets, “improved” cookstoves,playground water pumps, and many, many more.

To be clear, these are all cool gadgets that address serious problems. But few, if any, of these gadgets will ever see widespread use, because the problems themselves are not fundamentally technological in nature. The basic tools needed to address many issues of people living in poverty have been with us for many decades, and further technological improvement will probably not be direct causes for reducing that poverty. Poverty is a complex phenomenon more directly attributable to political, economic and sometimes cultural forces than anything else, and only insofar as technology changes those dynamics will it actually affect people’s lives.

(I personally subscribe to many of the theories of development put forth by Professor Bill Easterly at NYU, and strongly recommend his books The Tyranny of Experts and The White Man’s Burden if you want to learn more.)

Greater connectivity in developing countries will not solve poverty, but it might help a little bit, and even that much is worth working for. Unfortunately, I fear that the tech titans involved are throwing enormous sums of money (that are otherwise unheard-of in development circles) towards “solutions” that will never really work. Connectivity is spreading rapidly in the developing world no matter what Facebook or Google do.

A more interesting question to begin preparing for might be: what will those next billion do once they’re online?