From the Walkley Public Affairs Conference on Social Media, Sydney, 6 May 2008.
I.
One of the things I find the most exhilarating about Australia is the relative shallowness of its social networks. Where we’re accustomed to hearing about the “six degrees of separation” which…
I.
Everybody talks about the weather. It happens in Singapore, where the weather never changes much, and in Melbourne, where four seasons unfold over the course of an afternoon. Why? It comes down…
I.
Sydney looks very little different from the city of Gough Whitlam’s day. Although almost forty years have passed, we see most of the same concrete monstrosities at the Big End of town, the same…
I.
When I was a young man, I was obsessed by computers. I remember perfectly the first time I sat at a keyboard – at a “line printing” terminal, which had an endless sheet of paper spooling…
Case One: Lists
I moved to San Francisco in 1991, because I wanted to work in the brand-new field of virtual reality, and San Francisco was the epicenter of all commercial development in VR. The…
I. The Wheels Fall Off the Cart
In mid-1994, sometime shortly after Tony Parisi and I had fused the new technology of the World Wide Web to a 3D visualization engine, to create VRML, we paid a visit to the University of Santa Cruz, about 120 kilometers south of San Francisco. Two UCSC students wanted to pitch us on their own web media project. The Internet Underground Music Archive, or IUMA, featured a simple directory of artists, complete with links to MP3 files of these artists’…
Chaos
The world has changed. The world is changing. The world will change a whole lot more. We lucky few, we band of coders, bear witness to the most comprehensive transformation in human communication since the advent of language. We are embedded in the midst of this transition; we make it happen with every script we write and every page we publish and every blog we post and every video we upload. For that reason, it’s hard to see the forest for the trees. No wonder it looks so crazy and chaotic.
In the mid 20th century, American philosopher H. Richard Neibur wrote that the first question of ethics is not, “What is right?”, but rather, “What is going on?” This arvo, before we retire to the Shelbourne for drinks and conversation, I’d like to take you on a tour of our very peculiar present. Something’s happening that is so unexpected, most of us don’t even know it’s going on.
Confusion: Three Billion
We begin on the shores of the Indian Ocean, in the south Indian state of Kerala. For at least a thousand years the fishermen of Kerala have sailed their sturdy dhows to sea, lowered their nets, prayed to their gods, and – if their prayers were heard – hauled in a bountiful catch. Fully laden, the fishermen set their sails to shore, to any one of the many fishing villages and fish markets which dot the Kerala coast. The selection of a port is done more or less at random, so throughout all these thousand years too many boats pulled into one port, leaving the markets oversupplied, and the fisherman selling their catch at a loss, while another market, just a few kilometers away, has no fish for sale at any price. This kept the fishermen poor, and the markets consistently either oversupplied or undersupplied.
From 1997 through 2001, as India’s rush to industrialization gathered momentum, several of India’s mobile telecoms firms strung the Kerala coast with GSM towers. GSM is a radio signal, and travels in line-of-sight, which means that, out at sea, the signal can reach 25 kilometers, the point where the curvature of the Earth blocks the view of the shore.
GSM handsets cost a month’s wages for a Kerala fishermen – imagine if a handset here cost four or five thousand dollars. (Even my Nokia N95 didn’t cost that much.) Yet, some wealthy fisherman, somewhere in Kerala, bought a GSM handset and took it to sea. At some point during a fishing voyage that fisherman had some communication with the mainland – perhaps a trivial family matter. But, in the course of that communication, he learned of a village going wanting for fish, at any price. So he made for that port and sold his catch at a tidy profit that day. The next day, perhaps, he called into shore, talking to fish sellers to the various ports, and learned which market needed fish the most – and was willing to pay for it. So it began.
Fishermen form a tight-knit community; while they might be secretive about their favorite spots to fish, they all trade technique with one another, and – within a very short period of time – all the other Kerala fishermen had learned of the power of the GSM handset, and each of them brought their own handset to sea, made calls to the markets, and sold their catch for a tidy profit. Today, the fish markets in Kerala are only rarely oversupplied with fish, and are almost never undersupplied. The network of fish sellers and fishermen have created their own bourse, a marketplace which grows organically out of an emergent web of SMS and voice calls which distribute the catch efficiently across the market. The customers are happy – there’s always fish for sale. The fish sellers are happy – they always have fish to sell, and at a good price. And the fisherman are happy – and earning so much more, these days, that a GSM handset pays for itself in two months’ time.
None of this was predicted. None of this was expected. None of this was anything but shocking to the legion of economists who are now studying this unprecedented phenomenon. To our Western eyes this doesn’t even make much sense. We think of mobile phones as a bit of bling, a technological googaw that makes our lives a bit easier – something that removes the friction from our social interactions. In the age of the mobile, you’re never late, just delayed. You can always call to say you’re sorry. (Or text to say you’ve broken up.) While they can be useful in our economic lives, they’re hardly necessary – and, given that the boss can now reach you 24 hours a day, wherever you are on Earth – they’re often more of a pain in the arse than a blessing. But at the end of the day they’re extraneous. Nice, but non-essential.
Except they’re not.
Study after study is confirming something that many were already beginning to suspect: the very poorest people on Earth – the five billion of us who earn less than a few thousand dollars a year – can benefit enormously from pervasive wireless communications. It seems counterintuitive – why would a subsistence farmer in Kenya need a mobile phone? As it turns out, that farmer – and farmers in Nigeria, and Bangladesh and Peru – will phone ahead to the markets, and learn where their produce will bring the best price. Left to their own devices, human beings with things to trade will create their own markets. When mobile communications enter the mix, their ability to trade effectively increases enormously.
Those who serve the poor – microfinance institutions like Bangladesh’s Grameen Bank – have real experience of the power of mobiles to help the poor. So many of Grameen Bank’s loans went to finance mobile handsets that they recently founded their own telecoms firm – Grameen Phone – to provide services to the poor. None of this is charity work – all of these are profit-making enterprises; but it turns out that helping the poor to communicate is one of the most effective ways to help them to improve their economic effectiveness.
That, too, wasn’t predicted by anyone. After all, don’t the poor need schools, clean water, inoculations and transparent governments? Yes, certainly they need all these things, but they also need the tools that let them help themselves. Near as anyone can tell, a mobile handset pretty much tops that list of tools. And although this singular discovery is nearly unknown in the Western world, the poor of the world know it – because they’ve been snapping up mobiles in unprecedented and unexpected numbers.
Sometime in the next 30 days, the telecoms firms of the world will have reached a new milestone – three billion subscribers. About ten percent of that number are customers who have multiple accounts, but – somewhere in the middle of 2008, half of humanity will own a mobile handset. In just a decade’s time, we’ll have gone from half the world never having made a telephone call to half the world owning a phone. Unprecedented. Unexpected. But, given what we now know, perfectly natural. And it’s not slowing down. It took a decade to get to the first billion mobile subscribers, four years to get to the second billion, and eighteen months to get to three billion. In a year, more or less, we’ll hit four billion, then things will begin to slow, as we reach the ranks of the desperately poor, the two billion who earn less than a dollar a day. Yet these are precisely the people who would most benefit from a mobile. Expect to see some big campaigns in the next few years, from Oxfam and World Vision, asking you to buy mobiles for the poor.
Nokia looked at the curves, figured out what’s going on, and created a mobile handset targeted directly at the emerging markets of the world – the Nokia 1100. It’s cheap, simple, has predictive text for just about any language with more than 10 million speakers, and – in the four years since its introduction – they’ve sold well over 200 million of them. By comparison, Nokia sold twice as many 1100s as Apple sold iPods – in half the time. The most successful consumer electronics device in history, the 1100 is the Model T of wireless networking. Put an 1100 in someone’s hands, and they’ll use it to improve their life. It’s as simple as that.
And – what’s really interesting here – these farmers and fishermen and spice traders and so forth didn’t need an eBay to help them trade. They don’t need fancy services – and wouldn’t use them. They only need to be connected to other people. That in itself is entirely sufficient. People come fully equipped to provide all the services they need. Nothing else is required. Five thousand years of civilization have seen to that. We know how to organize our own affairs – and can do so without any assistance. But now we can do so globally and instantaneously. That’s not a power restricted to the billion richest of us; it’s now within reach of half of us, and improves the lives of the poor far more than it helps us. Our innate capacity for self-organization, now extended and amplified almost infinitely, has itself produced some unpredicted and unexpected effects.
Discord: The Center Will Not Hold
In the Jurassic Era of the Internet, before the Web was more than a few hundred pages in size, and still mostly run off a series of servers in Geneva, John Gilmore, who co-founded SUN Microsystems before going off to found Cygnus Support and the EFF, recognized an inherent quality of networks: they promote the sharing of information. This was codified in what I (only half-jokingly) call Gilmore’s Law:
“The net regards censorship as a failure, and routes around it.”
At the time Gilmore made this statement, he was talking politics. Gilmore is a political animal – many of you probably know of his long-running tangle with US Homeland Security over the free right to travel within the States without having to display ID. And, for many years this aphorism was interpreted as a political maxim – that political censorship of the net was essentially impossible.
As we all know, the Chinese have tried, with their “Great Firewall of China”, but even they’ve given up. Just two months ago, Wang Guoqing, the Vice-Minister for Information in China was quoted as saying, “It has been repeatedly proved that information blocking is like walking into a dead end.”
At around the same time as that shock admission of failure, Senator Coonan introduced the Government’s latest attempt to appease its conservative base by locking down the Australian Internet, because, well, “Won’t somebody please think of the children?” Turns out that’s just what the children were doing – it took a 16 year-old Australian boy 30 minutes to crack through that filter, and another 40 minutes to crack it again, after the filter was “upgraded.”
In that same week, a fifteen year-old in the United States got his hands on a copy of Harry Potter and the Deathly Hallows, photographed the entire text, bound it up as a PDF, and uploaded it to the Pirate Bay so that tens of thousands could use BitTorrent and download their own copy – four days before the much-hyped simultaneous international release.
Gilmore, it seems, wasn’t thinking broadly enough. He assumed that censorship necessarily has a political dimension. It doesn’t. Censorship can be driven by a wide range of motives: some are political, some are moral, some are cultural, and some are economic. In the end, it doesn’t matter. All censorship inevitably encounters Gilmore’s Law, and loses. The net finds a way around it.
Before we get all hippy-dippy and attribute agency to something that we all know is really just a collection of wires and routing boxen, we need to clarify what we mean when we use the word “net”. The wiring isn’t the network. The routers aren’t the network. The people are the network. We had social networks ten million years before we ever had a telephone exchange; we carry those networks around in our heads, they’re part of the standard “kit” of our cortical biology. We have been blessed with the biggest and best networking gear of all the hominids, but we all share the same capability. The social sharing of information has played a big part in the success of the hominids, and, in particular, human beings. We are born to plug into the network of other human beings and share information. It’s what we do.
But just now we’re facing increasingly frequent collisions between Gilmore’s Law and old-fashioned and time-tested ways of the world. We’ve long known that there are no secrets in a small town; now that same law of interpersonal relationships are being applied to businesses, to governments, to institutions of every shape and description. Consider these examples:
What’s happening here? What is it about the network that makes it so potent? Simply this: the network, in every form, is anathema to hierarchy. The network represents the other form of organization, not a contradiction of hierarchy, but, rather, a counterpoint to it. I’ve rewritten Gilmore’s Law to reflect this:
“The net regards hierarchy as a failure, and routes around it.”
For the fifty-five hundred years of human civilization, hierarchy has always had the upper hand. Now the network, amplified by all those wires and routers, is stronger than hierarchy, and battle has been joined. But this isn’t going to be some full-on Armageddon, a battle between the Empire and the Alliance; this is the Death of a Thousand Cuts. The network is simply kicking the legs out from under hierarchies, everywhere they exist, for as long as they exist, until they find themselves unable to rise again. What it really come down to is this: we are assuming management of our own affairs, because we are now empowered to do so. It doesn’t matter if you’re a maize farmer in Kenya or a video producer in Queensland; these mob rules apply to us mob.
Unexpected. Unprecedented.
In a future which looks increasingly like the present, there is no center anywhere, no locus of authority, no controlling power ordering our daily lives. There are no governments, no institutions, no businesses that look anything like the limited liability enterprises born in the Netherlands five hundred years ago. Instead, there are groupings, networks within the network, that come together around a project or ideology, a shared sense of salience – meaning – for that group. The product of that network could be Wikipedia – or it could be al Qaeda. Buy the ticket, take the ride.
And it’s not over yet. The network hasn’t finished changing, and it hasn’t finished changing us.
Bureaucracy: Collapse and How to Profit From It
To recap: we know where we are, and we have some idea of what is really going on. But enough of philosophy: let’s play!
But. Well. One more thing…
Although the network has done a tidy job of disassembling the hierarchies of the world, there is still one hierarchy which remains stubbornly resistant to change, which retains its top-down, command-and-control hierarchical model of authority – and has for well over a hundred years. Telcos.
I find this endlessly ironic: the firms which created the network are somehow immune to the effects of the network. And, in consequence, so are the networks themselves. In fact, you can look at any of the networks – telephone, broadband, or wireless – and see in them the physical embodiment of hierarchy. It’s curious. It’s damned interesting. It’s also over.
Four months ago, a small startup in Silicon Valley named Meraki (Greek for “doing it with love”) for unveiled a cute little device, a wireless router that they simply named the Mini. Inside it has a RISC CPU running a custom version of LINUX which handles all of the routing tasks. That’s where it gets interesting. You see, Meraki have pioneered a new technology known as “wireless mesh networking”. You can power up a Mini in anywhere you like, and if there’s another Mini within distance – and these devices can reach nearly half a kilometer, outdoors – it will connect to it, share routing information, and route packets from one to another – all without any need to configure anything at all. Add another, and another, and another, and all of a sudden you’ve created a very wide area WiFi network. Only one of the Minis needs to be connected to the Internet as a gateway; the others will find it and route traffic through it. The Minis are small – and they’re also cheap. For just $49 dollars US, you can order one complete with an Australian wall wart. That’s cheaper than most access points out there, and because of the mesh networking, it does a whole lot more.
But what does the Meraki Mini have to do with the end of the telcos? Just this: a mesh network is a network that’s been subject to the corrosive effects of a network. There is no center anywhere. There’s no hierarcy or preferred route. There’s no gatekeeper anywhere. You can have one gateway, or twenty. You can have one mesh node or a thousand. Just throw another mesh node into the mix, and it’ll all work seamlessly. And mesh networks scale: the dynamics of a network of a thousand mesh repeaters aren’t substantially different from a network with ten. Packets still find their way, with minimal delay.
What this means is that we all have the capability to create our own large-scale, low-cost wireless networks within our grasp. Meraki is already proving this in San Francisco, where Google and Earthlink had been fighting the telcos for years to get a city-wide free wireless network installed. Last week, Earthlink pulled out – they just couldn’t fight the politically power of AT&T. Meanwhile, since February, Meraki has been offering free Meraki Minis to anyone in San Francisco who wanted to donate a little of their own broadband to a free municipal WiFi network. Lately that network has been growing by leaps and bounds – no easy feat in a city which effectively broken up by a series of large hills. The “Free the Net SF” project already has almost 14,000 users – that’s nearly triple the number two months ago – and hundreds of nodes. It is proof that us mob can seize control of the spectrum and use it for our own ends.
That’s fine and dandy for San Francisco, but what about here in Australia, where we’re suffering under a decade-old peering agreement which makes us pay and pay and pay for every bit we take out of the cloud? Which costs us tens of dollars an hour if we want to use a public WiFi hotspot, or, in the case of the Sydney Convention Centre, $800 for an hour’s access? (That was the quote Maxine received when I asked if we could have public WiFi during my talk.) Internet access in Australia has always been about bending over and taking it like a man.
Or at least it was.
But for the past thirty five minutes, you’ve all been bathing in WiFi, which I’m providing to all of you, free of charge. Here’s how I did it: my Nokia N95 connects to Vodofone’s HSDPA network at a couple of megabits per second. That’s piping through the Bluetooth connection of my mate David’s MacBook Pro, which is Internet Sharing the Bluetooth connection out to his Ethernet port. That Ethernet port is connected to a Meraki Mini, which, in turn, is talking to three more Meraki Minis scattered throughout the auditorium. You’ve all got good signal, and (I hope) plenty of bandwidth to blog, or check email, or whatever you might want to do when I get boring.
But here’s the kicker – it’s all running off batteries. The Meraki Minis only use three watts, so I built some simple power supplies for them. The N95 and the MacBook Pro already have their own batteries built into them. The whole thing is good for at least four hours of fun before someone needs to go find the mains. And, because it’s both entirely battery powered and entirely wireless, I can drop it anywhere in Sydney. Were we out-of-doors, I could probably cover a square kilometer, with just these four Minis. Of course, you can always add a few more. Or a thousand more.
Ok, Mark, that’s nice, you might be saying. That’s kind of cool. But big deal. We don’t own Meraki Minis – and we don’t really plan on buying one. That’s fine, and it doesn’t matter at all. You see, a mesh network node isn’t hardware device. It’s software which runs on arbitrary hardware. You can mesh network WiFi. Or Bluetooth. Or infrared, if you wanted to be perverse. It’s software. Which means that every laptop in this room is potentially another mesh network node, listening to the traffic and passing packets along. Consider the density of laptops and desktops (equipped with WiFi adapters) in Sydney, or Melbourne. Now imagine them as nodes within a vast mesh network. That’s where we’re going – and it’s just a software update away.
When I originally composed this section of the talk, I was going to make a prediction: because mesh networks are just software, and because my Nokia N95 has built-in WiFi, I predicted we’d soon see mesh networks for mobile phones. But I don’t need to make that prediction: a Swedish start-up, TerraNet, came out of stealth mode two weeks ago to announce they were doing precisely this. With their software, the mobile doesn’t even need the carrier’s wireless network. Mobiles simply route packets between themselves until they reach their destination. You wonder why the wireless telcos fought so hard and so long to keep WiFi out of mobiles? Was it just to prevent VOIP? Hardly. The telcos have known about mesh networking for a long time. And they know it spells their doom. So watch now, as the network frees itself from the authoritarian forms of those most hierarchical of organizations, the telcos.
But I said it was time to play. And it is. It’s time to put the mob rules to work for you. Because you all need to earn a living. But this world we’re entering is so chaotic, so accidental and unplanned for, everything we believe to be absolutely true is about to be severely tested.
ONE: The mob is everywhere.
There are very few places left on Earth where you can’t receive a text. Ulaanbataar to Timbuktu, Tierra del Fuego to Vladivostok, the network is truly global, and now encompasses the majority of humanity. It’s interesting to note that within the same year that half of humanity is urbanized, half of humanity will have a mobile handset. That’s not coincidental; they’re two sides of the same process. Just as we’ve been lured out from our villages into the vitality and opportunity of the city, we’re being drawn into the unexpected and unpredictable global mob.
TWO: The mob is faster, smarter and stronger than you are.
William Gibson put this much more elegantly when he wrote, “The street finds its own use for things, uses its manufacturers never intended.” No one set out to create arbitrage markets for the fishermen of Kerala; that’s something that emerged from the mob. SMS was meant to be used for emergency messaging; now the world sends several billion texts a day. Just add mobiles, and you get a mob.
You can’t push a mob any more than you can push a rope; you can pull them, lure them, and, if you’re very lucky, dazzle them for a moment or two, but then, inevitably, they’ll move along. That’s bad news for anyone building web sites. The world of mob rules isn’t about sites; it’s about services, things that the street uses and permutes indefinitely. The idea of web sites dates from a time before the network ate hierarchy; sites are places where you go and follow the rules laid down by some information architect. Well, there’s no way to enforce those rules. The first Google Maps mashup didn’t come from Google. Or the second. Or the third. Or the hundredth. Google resisted the mashup. Claimed mashups violated their terms of use. Mashups come from the mob, the street finding its own use for things. The mob pushed on through; Google bowed down and obeyed. The most powerful institution of the Internet era, pushed around like a child’s toy. Ponder that.
THREE: Advertising is a form of censorship.
The Web of 2007 is a house built upon sand. Nearly everything online hopes to fund itself through some sort of advertising and sponsorship. Advertising is a demand that you pay attention – a demand which can no longer be enforced. But the mob doesn’t like advertisements; it either ignores them or actively filters them away. In just the last few weeks, certain sites have been blocked to Firefox because it frequently incorporates the AdBlock extension. That’s upset some institutions which built their business model on the delivery of ads – demanding the attention of the mob. But the mob doesn’t like that. Even worse, for those who are raising a hew and cry about the “theft” of their precious content, the more they scream, the more they thrash about, the stronger the mob becomes. Consider: filesharing has only grown more pervasive despite every attempt of every copyright holder to bring it to heel. Each move has been met with a counter-move. There is no safety in copyright, nor any arguing with the mob. Music and movies are freely and broadly available, and will remain so into the indefinite future. Sadly, we’re now seeing that same, sorry battle repeated in double-time as advertisers – and those dependent upon them – assert an authority they no longer possess.
FOUR: The mob does not need a business model.
But what about your precious business models? How do you get paid for all this work you’re pouring into your projects? I have to be honest with you: the mob simply doesn’t care. The mob doesn’t need a business model. Heck, the mob doesn’t even need all this lovely wireless technology. If we took the mobiles away from the Kerala fishermen, they’d develop something – semaphores, mirrors, smoke signals – to maintain the integrity of the network. Once networks are created, they can not be destroyed. Networks are intrinsically resilient against all sorts of failures, and they’ll simply find a way to route around them. So if your business goes tits up because you built it around an economic model that is not viable in the era of mob rules, it will make no difference – the mob will simply route around you and find another way to do it.
So forget your business models, and remember the golden rule, as expressed by Talking Heads, in the song “Found a Job”:
“If your work isn’t what you love, then something isn’t right.”
If you – you folks in this room, who have the mob in your hands, who play with it as if it were a toy – if you don’t wake up in the morning completely possessed by the knowledge that what you’re doing is simply the coolest thing ever, you need to quit that job and find another. You need to reach into that bucket of dreams and ambitions and pull something out to share with us mob, something that will dazzle and excite us. It might only do so for a moment, but, in that moment, your social stock will rise so high that you’ll never have to worry about putting food on the table or paying the mortgage. You may not retire a millionaire, but you’ll certainly never go hungry. The mob is a meritocracy – admittedly a very perverse and bizarre meritocracy – but it is the one place where “quality will out”. Quality only comes from the marriage of craft and obsession. You have the craft. Embrace your obsessions. You will be rewarded.
FIVE: Make networks happen.
I need to leave you with one concrete example of how this is all going to work, and for this example I’ve selected the last bastion of authority and hierarchy – after everything else has dissolved into the gray goo of the network, one thing will remain. It won’t be government – that’s half gone already. It’s medicine. Medicine is very nearly the oldest of the professions, and has been a closely held monopoly for half a thousand years – closer to a guild than anything resembling a modern profession. Why? Medicine is guarded by the twin bulwarks of complexity and mortality: medicine is rich and deep body of knowledge, and, if you screw it up, you’ll kill yourself or somebody else. While the pursuit of medical knowledge is conducted within the peer-review frameworks of science, that knowledge is closely held. That leaves all of us – as patients – in a distinctly disempowered position when it comes to medicine. But that is all going to change.
In twenty years’ time, one in four Australians will be 65 or older – and I’ll be one of them. There is no medical authority big enough to deal with such a mass of gerontology; the system will be overloaded, and it will begin to collapse. Out of that collapse, we will see those of us who grew up within the Network Era – and I’m among the oldest of that generation – begin to work the network to our own ends. We will not be alone. There will be tens of millions of us – first in the West, then throughout the world – who will be facing the same problems, and searching for the same answers. We might not get to live forever, but we’ll want to die trying. So we’ll set to work, creating a common base of collective intelligence – think Wikipedia, but with a depth of medical knowledge that it doesn’t even begin to explore – together with strong social networking tools that embeds us deep within a network of experts – who may or may not be “board qualified”. I’ll probably come to expect that my GP and other specialists are members of this network – peers who share their expertise, not experts pronouncing solutions. And this network will never leave me; in fact, it will probably watch every move I make, every breath I take, every calorie I eat, and every heartbeat. It sounds Orwellian, but I will want this – because I will see it as a profoundly empowering form of surveillance. In other words, my wellness becomes a quality of my network.
This is not a website. This is not WebMD or Healtheon or a cancer support group, or anything that looks like anything we’ve seen yet. This is a self-organizing quality of the mob, painfully aware of their own accelerating senescence, and fully empowered to do something about it. And it represents an enormous opportunity for you. In just the last paragraph I’ve dropped a half a dozen strong business ideas onto you; but they’re so different from how we’re thinking about the network today that it will probably take some time to work it all out. But the mob won’t wait forever. Remember: it is smarter and faster and stronger than you. You can try to get in front of it, and get picked up by it – I’ve given you more than enough clues to do that – or you can get run down. That choice is yours. But if I’ve learned anything from my study of mob rules, it’s that the future lies in making networks happen. If you do that, there’s a place for you with us mob.
Aftermath
We live in increasingly interesting times. Half of humanity has suddenly dropped in – uninvited and unannounced – crashing our private party, eager to participate in an exploration of the possibilities of human communication. Whatever they want, they’re going to get. That’s the way things work now. Fortunately, they want what we want: better lives for themselves and their families. How they get it – that’s in their hands. We can assist them, but they don’t really need our help. That mob will work it out for themselves. And in the process, everything will change for us, as well.
Journalist Norman Cousins wrote, “Wisdom consists of the anticipation of consequences.” Sound advice, particularly in an time when everything is fluctuating out of control. We can’t know what to do – there’s too much uncertainty and potency in us mob for that – but we can know what not to do. For now, that will have to be enough.
Still, there is one thing I can recommend: have courage and keep moving. Standing still is not an option. The world has changed. The world is changing. The world will change a whole lot more. Good luck.
The adoption of mobile phones by fishermen and wholesalers was associated with a dramatic reduction in price dispersion, the complete elimination of waste, and near-perfect adherence to the Law of One Price. Both consumer and producer welfare increased.
When A Thousand Plateaus was published, a quarter-century ago, it shook the foundations of philosophy. Much of the “postmodernism” which cultural conservatives sneer at comes from the pages of the that text. (This reaction is perfectly in keeping with the recognized conservative tendency to bow to authority, and demonize anything that represents a threat to that authority.) Yet, although the text presented a sort of “map” of a territory free from the schizimogenic qualities of authority and hierarchy, Deleuze and Guattari were philosophers, not revolutionaries: they did not present a battle plan to manage the transition from hierarchy to milieu. As it turns out, that roadmap proved unnecessary. It’s not that the ideas within A Thousand Plateaus were fruitless, but rather, at just the time both philosophers passed from the world, the rhizome rose and subsumed us all into its milieu. Where is this rhizome? All around us, now: pervasively, wirelessly, instantly accessible to nearly half the planet. The rhizome is the network.The middle is by no means an average; on the contrary, it is where things pick up speed. Between things does not designate a localizable relation going from one thing to another and back again, but a perpendicular direction, a transversal movement that sweeps one and the other away, a stream without beginning or end that undermines its banks and picks up speed in the middle.
I have some shocking news to report: it hasn’t turned out that way. Yes, Wikipedia is still growing, but - for at least the last year - the rate of growth has dramatically slowed down. The acceleration is actually negative. Wikipedia’s growth is slowing down. Why did this happen? Just a few weeks ago Wikipedia passed two million articles in English (all these figures concern the English-language version of Wikipedia), and yes, it will grow for some time into the future. But the growth of articles in Wikipedia should be steadily accelerating; it should be growing faster as it grows bigger. It was certainly doing that for several years. What’s changed? Is it possible that there are only two million topics of interest to the English-language users of Wikipedia? That seems unlikely, if only because Wikipedia is the outstanding example of the power and beauty of the miscellaneous. Yes, all the major topics have been covered, but there’s absolutely no way that two million entries can begin to explore the depth of human experience. It’s inconceivable that this is all there is to say about Life, Culture, the Universe and Everything. Nor do I believe it likely that we have “crossed the chasm” into the downward slope - which would imply that four million article entries would pretty much represent the sum total of the English-language experience.
Introduction
“The net interprets censorship as damage and routes around it.”
– John Gilmore
I read a very interesting article last week. It turns out that, despite their best efforts, the Communist government of the People’s Republic of China have failed to insulate their prodigious population from the outrageous truths to be found online. In the article from the Times, Wang Guoqing, a vice-minister in the information office of the Chinese cabinet was quoted as saying, “It has been repeatedly proved that information blocking is like walking into a dead end.” If China, with all of the resources of a one-party state, and thus able to “lock down” its internet service providers, directing their IP traffic through a “great firewall of China”, can not block the free-flow of information, how can any government, anywhere – or any organization, or institution – hope to try?
Of course, we all chuckle a little bit when we see the Chinese attempt the Sisyphean task of damming the torrent of information which characterizes life in the 21st century. We, in the democratic West, know better, and pat ourselves on the back. But we are in no position to throw stones. Gilmore’s Law is not specifically tuned for political censorship; censorship simply means the willful withholding of information – for any reason. China does it for political reasons; in the West our reasons for censorship are primarily economic. Take, for example, the hullabaloo associated with the online release of Harry Potter and the Deathly Hallows, three days before its simultaneous, world-wide publication. It turns out that someone, somewhere, got a copy of the book, and laboriously photographed every single page of the 784-page text, bound these images together into a single PDF file, and then uploaded it to the global peer-to-peer filesharing networks. Everyone with a vested financial interest in the book – author J.K. Rowling, Bloomsbury and Scholastic publishing houses, film studio Warner Brothers – had been feeding the hype for the impending release, all focused around the 21st of July. An enormous pressure had been built up to “peek at the present” before it was formally unwrapped, and all it took was one single gap in the $20 million security system Bloomsbury had constructed to keep the text safely secure. Then it became a globally distributed media artifact. Curiously, Bloomsbury was reported as saying they thought it would only add to sales - if many people are reading the book now, even illegally, then even more people will want to be reading the book right now. Piracy, in this case, might be a good thing.
These two examples represent two data points which show the breadth and reach of Gilmore’s Law. Censorship, broadly defined, is anything which restricts the free flow of information. The barriers could be political, or they could be economic, or they could – as in the case immediately relevant today – they could be a nexus of the two. Broadband in Australia is neither purely an economic nor purely a political issue. In this, broadband reflects the Janus-like nature of Telstra, with one face turned outward, toward the markets, and another turned inward, toward the Federal Government. Even though Telstra is now (more or less) wholly privatized, the institutional memory of all those years as an arm of the Federal Government hasn’t yet been forgotten. Telstra still behaves as though it has a political mandate, and is more than willing to use its near-monopoly economic strength to reinforce that impression.
Although seemingly unavoidable, given the established patterns of the organization, Telstra’s behavior has consequences. Telstra has engendered enormous resentment – both from its competitors and its customers – for its actions and attitude. They’ve recently pushed the Government too far (at least, publicly), and have been told to back off. What may not be as clear – and what I want to warn you of today – is how Telstra has sewn the seeds of its own failure. What’s more, this may not be anything that Telstra can now avoid, because this is neither a regulatory nor an economic failure. It can not be remedied by any mechanism that Telstra has access to. Instead, it may require a top-down rethinking of the entire business.
I: Network Effects
For the past several thousand years, the fishermen of Kerala, on the southern coast of India, have sailed their dhows out into the Indian Ocean, lowered their nets, and hoped for the best. When the fishing is good, they come back to shore fully laden, and ready to sell their catch in the little fish markets that dot the coastline. A fisherman might have a favorite market, docking there only to find that half a dozen other dhows have had the same idea. In that market there are too many fish for sale that day, and the fisherman might not even earn enough from his catch to cover costs. Meanwhile, in a market just a few kilometers away, no fishing boats have docked, and there’s no fish available at any price. This fundamental chaos of the fish trade in Kerala has been a fact of life for a very long time.
Just a few years ago, several of India’s rapidly-growing wireless carriers strung GSM towers along the Kerala coast. This gives those carriers a signal reach of up to about 25km offshore – enough to be very useful for a fisherman. While mobile service in India is almost ridiculously cheap by Australian standards – many carriers charge a penny for an SMS, and a penny or two per minute for voice calls – a handset is still relatively expensive, even one such as the Nokia 1100, which was marketed specifically at emerging mobile markets, designed to be cheap and durable. Such a handset might cost a month’s profits for a fisherman – which makes it a serious investment. But, at some point in the last few years, one fisherman – probably a more prosperous one – bought a handset, and took it to sea. Then, perhaps quite accidentally, he learned, through a call ashore, of a market wanting for fish that day, brought his dhow to dock there, and made a handsome profit. After that, the word got around rapidly, and soon all of Kerala’s fisherman were sporting their own GSM handsets, calling into shore, making deals with fishmongers, acting as their own arbitrageurs, creating a true market where none had existed before. Today in Kerala the markets are almost always stocked with just enough fish; the fishmongers make a good price for their fish, and the fishermen themselves earn enough to fully recoup the cost of their handsets in just two months. Mobile service in Kerala has dramatically altered the economic prospects for these people.
This is not the only example: in Kenya farmers call ahead to the markets to learn which ones will have the best prices for their onions and maize; spice traders, again in Kerala, use SMS to create their own, far-flung bourse. Although we in the West generally associate mobile communications with affluent lifestyles, a significant number of microfinance loans made by Grameen Bank in Bangladesh, and others in Pakistan, India, Africa and South America are used to purchase mobile handsets – precisely because the correlation between access to mobile communications and earning potential has become so visible in the developing world. Grameen Bank has even started its own carrier, GrameenPhone, to service its microfinance clientele.
Although economists are beginning to recognize and document this curious relationship between economics and access to communication, it needs to be noted that this relationship was not predicted – by anyone. It happened all by itself, emerging from the interaction of individuals and the network. People – who are always the intelligent actors in the network – simply recognized the capabilities of the network, and put them to work. As we approach the watershed month of October 2007, when three billion people will be using mobile handsets, when half of humanity will be interconnected, we can expect more of the unexpected.
All of this means that none of us – even the most foresighted futurist – can know in advance what will happen when people are connected together in an electronic network. People themselves are too resourceful, and too intelligent, to model their behavior in any realistic way. We might be able to model their network usage – though even that has confounded the experts – but we can’t know why they’re using the network, nor what kind of second-order effects that usage will have on culture. Nor can we realistically provision for service offerings; people are more intelligent, and more useful, than any other service the carriers could hope to offer. The only truly successful service offering in mobile communications is SMS – because it provides an asynchronous communications channel between people. The essential feature of the network is simply that it connects people together, not that it connects them to services.
This strikes at the heart of the most avaricious aspects of the carriers’ long-term plans, which center around increasing the levels of services on offer, by the carrier, to the users of the network. Although this strategy has consistently proven to be a complete failure – consider Compuserve, Prodigy and AOL – it nevertheless has become the idée fixe of shareholder reports, corporate plans, and press releases. The network, we are told, will become increasingly more intelligent, more useful, and more valuable. But all of the history of the network argues directly against this. Nearly 40 years after its invention, the most successful service on the Internet is still electronic mail, the Internet’s own version of SMS. Although the Web has become an important service in its own right, it will never be as important as electronic mail, because it connects individuals.
Although the network in Kerala was brought into being by the technology of GSM transponders and mobile handsets, the intelligence of the network truly does lie in the individuals who are connected by the network. Let’s run a little thought experiment, and imagine a world where all of India’s telecoms firms suffered a simultaneous catastrophic and long-lasting failure. (Perhaps they all went bankrupt.) Do you suppose that the fishermen would simply shrug their shoulders and go back to their old, chaotic market-making strategies? Hardly. Whether they used smoke signals, or semaphores, or mirrors on the seashore, they’d find some way to maintain those networks of communication – even in the absence of the technology of the network. The benefits of the network so outweigh the implementation of the network that, once created, networks can not be destroyed. The network will be rebuilt from whatever technology comes to hand – because the network is not the technology, but the individuals connected through it.
This is the kind of bold assertion that could get me into a lot of trouble; after all, everyone knows that the network is the towers, the routers, and the handsets which comprise its physical and logical layers. But if that were true, then we could deterministically predict the qualities and uses of networks well in advance of their deployment. The quintessence of the network is not a physical property; it is an emergent property of the interaction of the network’s users. And while people do persistently believe that there is some “magic” in the network, the source of that magic is the endlessly inventive intellects of the network’s users. When someone – anywhere in the network – invents a new use for the network, it propagates widely, and almost instantaneously, transmitted throughout the length and breadth of the network. The network amplifies the reach of its users, but it does not goad them into being inventive. The service providers are the users of the network.
I hope this gives everyone here some pause; after all, it is widely known that the promise to bring a high-speed broadband network to Australia is paired with the desire to provide services on that network, including – most importantly – IPTV. It’s time to take a look at that promise with our new understanding of the real power of networks. It is under threat from two directions: the emergence of peer-produced content; and the dramatic, disruptive collapse in the price of high-speed wide-area networking which will fully power individuals to create their own network infrastructure.
II: DIYnet
Although nearly all high-speed broadband providers – which are, by and large, monopoly or formerly monopoly telcos – have bet the house on the sale of high-priced services to finance the build-out of high-speed (ADSL2/FTTN/FTTH) network infrastructure, it is not at all clear that these service offerings will be successful. Mobile carriers earn some revenue from ringtone and game sales, but this is a trivial income stream when compared to the fees they earn from carriage. Despite almost a decade of efforts to milk more ARPU from their customers, those same customers have proven stubbornly resistant to a continuous fleecing. The only thing that customers seem obviously willing to pay for is more connectivity – whether that’s more voice calls, more SMS, or more data.
What is most interesting is what these customers have done with this ever-increasing level of interconnectivity. These formerly passive consumers of entertainment have become their own media producers, and – perhaps more ominously, in this context – their own broadcasters. Anyone with a cheap webcam (or mobile handset), a cheap computer, and a broadband link can make and share their own videos. This trend had been growing for several years, but since the launch of YouTube, in 2005, it has rocketed into prominence. YouTube is now the 4th busiest website, world-wide, and perhaps 65% of all video downloads on the web take place through Google-owned properties. Amateur productions regularly garner tens of thousands of viewers – and sometimes millions.
We need to be very careful about how we judge both the meaning of the word “amateur” in the context of peer-produced media. An amateur production may be produced with little or no funding, but that does not automatically mean it will appear clumsy to the audience. The rough edges of an amateur prodution are balanced out by a corresponding increase in salience – that is, the importance which the viewer attaches to the subject of the media. If something is compelling because it is important to us – something which we care passionately about – high production values do not enter into our assessment. Chad Hurley, one of the founders of YouTube has remarked that the site has no “gold-standard” for production; in fact, YouTube’s gold-standard is salience – if the YouTube audience feels the work is important, audience members will share it within their own communities of interest. Sharing is the proof of salience.
After two years of media sharing, the audience for YouTube (which is now coincident with the global television audience in the developed world) has grown accustomed to being able to share salient media freely. This is another of the unexpected and unpredicted emergent effects of the intelligence of humans using the network. We now have an expectation that when we encounter some media we find highly salient, we should be able to forward it along within our social networks, sharing it within our communities of salience. But this is not the desire of many copyright holders, who collect their revenues by placing barriers to the access of media. This fundamental conflict, between the desire to share, as engendered by our own interactions with the network, and the desire of copyright holders to restrain media consumption to economic channels has, thus far, been consistently resolved in favor of sharing. The copyright holders have tried to use the legal system as a bludgeon to change the behavior of the audience; this has not, nor will it ever work. But, as the copyright holders resort to ever-more-draconian techniques to maintain control over the distribution of their works, the audience is presented with an ever-growing world of works that are meant to be shared. The danger here is that the audience is beginning to ignore works which they can not share freely, seeing them as “broken” in some fundamental way. Since sharing has now become essential quality of media, the audience is simply reacting to a perceived defect in those works. In this sense, the media multinationals have been their own worst enemies; by restricting the ability of the audiences to share the works they control, they have helped to turn audiences toward works which audiences can distribute through their own “do-it-yourself” networks.
These DIYnets are now a permanent fixture of the media landscape, even as their forms evolve through YouTube playlists, RSS feeds, and sharing sites such as Facebook and Pownce. These networks exist entirely outside the regular and licensed channels of distribution; they are not suitable – legally or economically – for distribution via a commercial IPTV network. Telstra can not provide these DIYnets to their customers through its IPTV service – nor can any other broadband carrier. IPTV, to a carrier, means the distribution of a few hundred highly regularized television channels. While there will doubtless be a continuing market for mass entertainment, that audience is continuously being eroded by a growing range of peer-produced programming which is growing in salience. In the long-term this, like so much in the world, will probably obey an 80/20 rule, with about 80 percent of the audience’s attention absorbed in peer-produced, highly-salient media, while 20 percent will come from mass-market, high-production-value works. It doesn’t make a lot of sense to bet the house on a service offering which will command such a small portion of the audience’s attention. Yes, Telstra will offer it. But it will never be able to compete with the productions created by the audience.
Because of this tension between the desires of the carrier and the interests of the audience, the carrier will seek to manipulate the capabilities of the broadband offering, to weight it in favor of a highly regularized IPTV offering. In the United States this has become known as the “net neutrality” argument, and centers on the question of whether a carrier has the right to shape traffic within its own IP network to advantage its own traffic over that of others. In Australia, the argument has focused on tariff rates: Telstra believes that if they build the network, they should be able to set the tariff. The ACCC argues otherwise. This has been the characterized as the central stumbling block which has prevented the deployment of a high-speed broadband network across the nation, and, in some sense that is entirely true – Telstra has chosen not move forward until it feels assured that both economic and regulatory conditions prove favorable. But this does not mean that the consumer demand for a high-speed network was simply put on pause over the last years. More significantly, the world beyond Telstra has not stopped advancing. While it now costs roughly USD $750 per household to provide a high-speed fiber-optic connection to the carrier network, other technologies are coming on-line, right now, which promise to reduce those costs by an order of magnitude, and furthermore, which don’t require any infrastructure build-out on the part of the carrier. This disruptive innovation could change the game completely.
III: Check, MateAll parties to the high-speed broadband dispute – government, Telstra, the Group of Nine, and the public – share the belief that this network must be built by a large organization, able to command the billions of dollars in capital required to dig up the streets, lay the fiber, and run the enormous data centers. This model of a network is an reflection in copper, plastic and silicon, of the hierarchical forms of organization which characterize large institutions – such as governments and carriers. However, if we have learned anything about the emergent qualities of networks, it is that they quickly replace hierarchies with “netocracies“: horizontal meritocracies, which use the connective power of the network to out-compete slower and rigid hierarchies. It is odd that, while the network has transformed nearly everything it has touched, the purveyors of those networks – the carriers – somehow seem immune from those transformative qualities. Telecommunications firms are – and have ever been – the very definition of hierarchical organizations. During the era of plain-old telephone service, the organizational form of the carrier was isomorphic to the form of the network. However, over the last decade, as the internal network has transitioned from circuit-switched to packed-switched, the institution lost synchronization with the form of the network it provided to consumers. As each day passes, carriers move even further out of sync: this helps to explain the current disconnect between Telstra and Australians.
We are about to see an adjustment. First, the data on the network which was broken into packets; now, the hardware of the network has followed. Telephone networks were centralized because they required explicit wiring from point-to-point; cellular networks are decentralized, but use licensed spectrum – which requires enormous capital resources. Both of these conditions created significant barriers to entry. But there is no need to use wires, nor is there any need to use licensed spectrum. The 2.4 GHz radio band is freely available for anyone to use, so long as that use stays below certain power values. We now see a plethora of devices using that spectrum: cordless handsets, Bluetooth devices, and the all-but-ubiquitous 802.11 “WiFi” data networks. The chaos which broadcasters and governments had always claimed would be the by-product of unlicensed spectrum has, instead, become an wonderfully rich marketplace of products and services. The first generation of these products made connection to the centralized network even easier: cordless handsets liberated the telephone from the twisted-pair connection to the central office, while WiFi freed computers from heavy and clumsy RJ-45 jacks and CAT-5 cabling. While these devices had some intelligence, that intelligence centered on making and maintaining a connection to the centralized network.
Recently, advances in software have produced a new class of devices which create their own networks. Devices connected to these ad-hoc “mesh” networks act as peers in a swarm (similar to the participants in peer-to-peer filesharing), rather than clients within a hierarchical distribution system. These network peers share information about their evolving topology, forming a highly-resilient fabric of connections. Devices maintain multiple connections to multiple nodes throughout the network, and a packet travels through the mesh along a non-deterministic path. While this was always the promise of TCP/IP networks, static routes through the network cloud are now the rule, because they provide greater efficiency, make it easier to maintain the routers, diagnose network problems, and keeps maintenance costs down. But mesh networks are decentralized; there is no controlling authority, no central router providing an interconnection with a peer network. And – most significantly – mesh networks now incredibly inexpensive to implement.
Earlier this year, the US-based firm Meraki launched their long-awaited Meraki Mini wireless mesh router. For about AUD $60, plus the cost of electricity, anyone can become a peer within a wireless mesh network providing speeds of up to 50 megabits per second. The device is deceptively simple; it’s just an 802.11 transceiver paired with a single-chip computer running LINUX and Meraki’s mesh routing software – which was developed by Meraki’s founders while Ph.D. students at the Massachusetts Institute of Technology. The 802.11 radio within the Meraki Mini has been highly optimized for long-distance communication. Instead of the normal 50 meter radius associated with WiFi, the Meraki Mini provides coverage over at least 250 meters – and, depending upon topography, can reach 750 meters. Let me put that in context, by showing you the coverage I’ll get when I install a Meraki Mini on my sixth-floor balcony in Surry Hills:

From my flat, I will be able to reach all the way from Central Station to Riley Street, from Belvoir Street over to Albion Street. Thousands of people will be within range of my network access point. Of course, if all of them chose to use my single point of access, my Meraki Mini would be swamped with traffic. It simply wouldn’t be able to cope. But – given that the Meraki Mini is cheaper than most WiFi access points available at Harvey Norman – it’s likely that many people within that radius would install their own access points. These access points would detect each others’ presence, forming a self-organizing mesh network. If every WiFi access point visible from my flat (I can sense between 10 and 20 of them at any given time) were replaced with a Meraki Mini, or, perhaps more significantly, if these WiFi access points were given firmware upgrades which allowed them to interoperate with the mesh networks created by the Meraki Mini – my Surry Hills neighborhood would suddenly be blanketed in a highly resilient and wholly pervasive wireless high-speed network, at nearly no cost to the users of that network. In other words, this could all be done in software. The infrastructure is already deployed.
As some of you have no doubt noted, this network is highly local; while there are high-speed connections within the wireless cloud, the mesh doesn’t necessarily have connections to the global Internet. In fact, Meraki Minis can act as routers to the Internet, routing packets through their Ethernet interfaces to the broader Internet, and Meraki recommends that at least every tenth device in a mesh be so equipped. But it’s not strictly necessary, and – if dedicated to a particular task – completely unnecessary. Let us say, for example, that I wanted to provide a low-cost IPTV service to the residents of Surry Hills. I could create a “head-end” in my own flat, and provide my “subscribers” with Meraki Minis and an inexpensive set-top-box to interface with their televisions. For a total install cost of perhaps $300, I could give everyone in Surry Hills a full IPTV service (though it’s unlikely I could provide HD-quality). No wiring required, no high-speed broadband buildout, no billions of dollars, no regulatory relaxation. I could just do it. And collect both subscriber fees and advertiser revenues. No Telstra. No Group of Nine. No blessing from Senator Coonan. No go-over by the ACCC. The technology is all in place, today.
I bring up this thought experiment to drive home my final point: Telstra isn’t needed. It might not even be wanted. We have so many other avenues open to us to create and deploy high-speed broadband services that it’s likely Telstra has just missed the boat. You’ve waited too long, dilly-dallying while the audience and the technology have made you obsolete. The audience doesn’t want the same few hundred channels they can get on FoxTel: they want the nearly endless stream of salience they can get from YouTube. The technology is no longer about centralized distribution networks: it favors light, flexible, inexpensive mesh networks. Both of these are long-term trends, and both will only grow more pronounced as the years pass. In the years it takes Telstra – or whomever gets the blessing of the regulators – to build out this high-speed broadband network, you will be fighting a rearguard action, as both the audience and the technology of the network race on past you. They have already passed you by, and it’s been my task this morning to point this out. You simply do not matter.
This doesn’t mean it’s game over. I don’t want you to report to Sol Trujilo that it’s time to have a quick fire-sale of Telstra’s assets. But it does mean you need to radically rethink your business – right now. In the age of pervasive peer-production, paired with the advent of cheap wireless mesh networks, your best option is to become a high-quality connection to the global Internet – in short, a commodity. All of this pervasive wireless networking will engender an incredible demand for bandwidth; the more people are connected together, the more they want to be connected together. That’s the one inarguable truth we can glean from the 160 years of electric communication. Telstra has the infrastructure to leverage itself into becoming the most reliable data carrier connecting Australians to the global Internet. It isn’t glamorous, but it is a business with high barriers to entry, and promises a steadily growing (if unexciting) continuing revenue stream. But, if you continue to base your plans around selling Australians services we don’t want, you are building your castles on the sand. And the tide is rising.