After Broadband

My friends at the FCC are hard at work preparing for the rollout of the National Broadband Plan in mid-March.  Everyone else in telecomland is hard at work jousting about the FCC’s legal authority to impose network neutrality rules.  (By the way, my paper on that topic, which I blogged about nearly a year ago, is finally about to be published. Things are playing out as I expected.)  FCC Chairman Genachowski is out on the hustings talking about unlocking under-utilized spectrum. This is important stuff.  The US has suffered for its lack of a broadband strategy, even more than for the gaps in broadband deployment.  As Genachowski likes to say, broadband is the infrastructure of the 21st century.  You don’t get great national infrastructure by ignoring it.

We should remember, though, that broadband itself isn’t the ultimate goal.  As infrastructure, broadband is a means to various ends, rather than an end in itself.  Roads and bridges create construction jobs, demand for concrete, and revenues for bulldozer companies, but their real value is to enable mobility of goods and people, which powers far more investment, innovation, and wealth creation.  And they also create suburban sprawl, massive carbon consumption, and other less positive impacts on society.  Those are the important costs and benefits of infrastructure; the direct economics are secondary.  The same is true for broadband.  It’s a platform, not an app.

Here’s why that is critical emphasize: The value propositions of infrastructure aren’t always evident from the infrastructure itself.  (Brett Frischmann makes a similar point in his extended analysis of commons infrastructure.)  You can’t have Wal-Mart without interstate highways, but you can’t work forward to Wal-Mart from the business models of road construction and road use.  In the language of complexity theory, the things policy-makers should really care about are emergent.  The danger  is that we’ll expend too much of our energy on the arcana of road construction, because that’s the piece we can see today. (As an aside, Reed Hundt once told me the anecdote, which I’ve never checked, that the Eisenhower Administration got the Interstate Highway System built by agreeing to pave half the highways with concrete and half with asphalt, just to mollify those industries.)

Let me bring this back to broadband to explain what I mean.  I remember doing Internet and telecom policy at the FCC in the mid-1990s. Everyone who looked knew the Internet would be big, transformatively big.  And everyone knew that dial-up access would give way to broadband.  @Home was offering cable modem service to great fanfare in 1997, and I organized a Broadband Forum with multiple wired and wireless broadband demonstrations at the FCC headquarters in January 1998.  Heck, my old boss Bob Pepper wrote an FCC working paper on broadband network a decade before that!  What was hard to envision in the 1990s was just how broadband would change things.  The speed wasn’t the whole issue, or even necessarily the biggest issue.

Take Google.  Without broadband, it wouldn’t be, well… it wouldn’t be Google.  That famously stark home page doesn’t require multimegabit connections to deliver to the end user, and the massive cloud computing infrastructure behind it is just a fancy version of the leased data networks that companies like IBM have had for decades.  There were search engines in the dial-up age, of course, but they had a different function.  Google needs broadband because broadband is always on.  You no longer have to make a phone call, connect to an ISP, and start an Internet session to go Google something.  Twitter is an even better case.  It requires trivial bandwidth to the end user, but simply wouldn’t work if you couldn’t instantly dash off a tweet on your laptop or phone.  Google and Twitter (and Wikipedia, and Facebook, and Yelp…) are always with you.  The rise of wireless broadband the past three years has supercharged this effect.

All the focus right now is on broadband speeds.  Genachowski wants 100 megabits to 100 million homes; Google wants 1 gigabit; 4G wireless is “10 times faster” than 3G; and so on.  That’s all great, but it’s just a part of the picture.  AT&T has the “nation’s fastest 3G network,” but my iPhone experience is often sluggish because the device itself is slow or the network is congested.  A network is as fast as its slowest link, so removing one bottleneck often simply pushes the problem elsewhere.  Moreover, speed is only one piece of the equation, as we saw with the shift from dial-up to broadband.  A faster connection will give me better HD videoconferencing, but will it give me a better experience with LinkedIn or SalesForce.com?  Will the next wave of killer apps emerge because the speed bar was set higher, or something else?  My money is on something else.

So that’s what I’m thinking about now: What comes after broadband.  Because if the FCC wins the battle over broadband policy, we should be prepared to win the larger war.  (And if it loses, the future will still happen somewhere, and eventually here.)

The term “broadband” comes from Claude Shannon’s work on information theory in the 1940s, which showed mathematically how a wider messaging channel affected data throughput between a sender and receiver of information.  Shannon’s simple channel model is a great approximation of the 20th century public switched telephone network, but increasingly remote from today’s global digital internetwork.  (David Reed makes a similar point about Shannon’s channel model limiting our appreciation of new wireless technologies.)  And even where the broadband construct remains useful, we tend to focus more on speed than channel width.  Skype sounds better than your telephone, not because it has a faster connection, but because it uses a wideband codec.  (Dan Berninger and Jeff Pulver have a conference series on the under-appreciated significance of HD voice.)  It’s natural to work forward from the network attributes we know, when we should be working backwards from what’s emerging in apps and content.

So then, what else is after broadband?  I’m only confident that neither I nor anyone else will predict it accurately, except by accident.  That’s the way emergence works.  But we can see aspects of the future developing today.  One large category is the integration of the virtual and the physical.  That includes the integration of wired and wireless, which we have a taste of from Twitter, along with location, augmented reality, telepresence, and cloud computing plus artificial intelligence on the back end.   A second broad category is the integration of entertainment, games, and computing.  We’re not far from the point where I can compile a personal Avatar in the cloud (the James Cameron version, with a capital A), and stream it to my home media center.  And not far beyond that, when I can compile a personal Avatar in real-time in my house, and jack it into the grid. Games and game mechanics are already much bigger economic forces than most people realize.  A third category is big data.  Google showed just how good a mousetrap could be created by analyzing billions of search queries. At a certain point, big is different, and often better.  The availability of massive oceans of data and the tools to aggregate and analyze it promise to revolutionize everything from medicine to government services. (I look forward to my friend Kenn Cukier’s take on the many dimensions of this trend in the Economist.)

All of these require broadband, but not necessarily beyond today’s speed and coverage. The gating factor in their development is the evolution of the necessary ecosystems. And we know the evolution is an intensely messy, slow, inefficient, and unpredictable process… that is the basis for all the richness and complexity of the planet Earth. Focusing on what comes after broadband is the difference between a beautifully-organized collection of bird specimens and Darwin’s Origin of Species.  One tells us what is, the other shows us why and leads towards what will be.  And that last question, of course, is what we’re ultimately after.