CHAPTER 1
Redesigning the Social World as if by Accident

Imagine a parallel human world with a global government whose policymakers were given complete authority to redesign how information and communication flows around the planet. The year: circa 1990. The immediate aim: to ensure the freest possible flow of content from all possible sources to all possible destinations. The long-term goal: to move away from a world where communication was monopolised by large institutions such as media companies, enabling more democratic societies.

The early 1990s was the time, in the US particularly, when, after years of multiplying TV channels, there was restlessness for a different type of media: more decentralised, more open to individual production. Let ‘prime time’ become ‘my time’ announced MIT professor Nicholas Negroponte in his book Being Digital; let’s unleash the potential of the ‘Daily Me’ had written Stewart Brand, creator of the Whole Earth Catalog a few years before.2 In the early 1990s the internet was starting to become a feature of everyday life, and a period of major innovation in the internet’s design followed, with the World Wide Web protocols being published in April 1993.3 A new world of information seemed to be opening up.

What happened next was so bizarre that it makes more sense, initially, if retold in fable form. We are talking about a major step-change in the organisation of social life, or even, as one commentator put it, an evolutionary shift from one state of things to another completely different state of things4 – or, as I put it in the introduction, the design of a new space of the world. What actually happened quickly erased earlier forms of civic and social computing, and it happened without any democratic consultation. Technological elites just took on themselves the task of ‘re-engineer[ing] society’.5 Let’s pretend, for a moment, that everyone knew what they were doing.

How We Came to Have the Internet We Have (a Fable)

Let’s suppose our imaginary policymakers met and decided on some simple steps, designed to unfold over time, that they must authorise in order to transform the world for the better. In reality, events unfolded more chaotically, but let’s pretend for the moment there was a plan, that what unfolded were steps which as a sequence made sense. There were seven steps in all.

First, create a large number of computers. And give special attention to optimising for a key early feature of computers, which was to archive traces of the actions they performed: all computers track their performance and create a complete record how they are used.6 Optimise for this feature, and you have created a world with lots of little performance archives which record, in data form, what happens on the device on which they are recorded.

Second, connect up those computers with each other, allowing them to exchange information and making it easier for computers to find information which other computers are ready to exchange. We know this step historically as the expansion of the internet, in the late 1980s and early 1990s, from a small, closed network to an infinitely expanding space of connection. This space was made massively more productive for information retrieval by Tim Berners-Lee’s design of the World Wide Web protocols. The result: a much wider space of connection where all computers could exchange information and the asymmetrical extraction of information by powerful surveillant computers from countless other computers was possible.

Third, shift this vast space of connection over from public to private commercial ownership and control. This involved a number of complex technical moves, for example, encouraging the proliferation of commercial web browsers, of which the first was Mosaic, launched in 1993. The most important step was the decision to move the management of the internet’s underlying infrastructure from public to private commercial hands. The closed computer network that Vint Cerf had known (called ARPANET) was in 1990 absorbed into NSFNET, an infrastructure owned by the US’s National Science Foundation. Yes, there was still at that time a vision of building an ‘information superhighway’ whose role would be basically public and educational. But all that ended when the businesses that serviced the technical side of NSFNET insisted on offering their services direct to commercial network providers. In 1995 the publicly oriented NSFNET was retired, and all that was left was a commercialised internet infrastructure focused on delivering commercial, not public, goals.7 Over time, other decisions were taken, at every level of the infrastructure ‘stack’,8 to let market forces unfold: that’s why, for example, today we have a world where commercial corporations own the 400 undersea cables that carry 99 per cent of the world’s digital communications,9 and commercial property developers own the key connection points that hold the internet together as an inter-net.10 This ‘privatised’ model became the liberal model that underlay the internet almost everywhere in the years that followed.11

Once this key shift of ownership had occurred, market priorities took over completely. Rising numbers of people and daily interactions joined the emerging information space. Major voices celebrated the freedom of ‘cyberspace’ from political interference, including John Perry Barlow in his famous 1996 Declaration of the Independence of Cyberspace.12 Already some marketers saw the potential for using the internet’s surveillance capacity to get better informed about what their customers did, even when not in the act of buying something: devices like the cookie were invented to capture our activities online,13 and suddenly, at a stroke, the transcripts of our computer use became open territory for commercial exploitation.14 Yet in our fable we are still in the 1990s: the era of US internet providers like CompuServe and AOL, when many still got their internet connection via dial-up connections working across phone lines, while in the UK the internet was only slowly becoming available at home.

Then a fourth step: to embed connected computers much more deeply in everyday life. How? By massively expanding the number of computer-based devices that could connect us online and also capture our activities. Such devices could be of any sort; they could be a desktop or laptop, but, in time, mobile phones would be as good, provided they were connected to the internet. This began with the iPhone’s launch in 2007 and the launch of Google’s Android operating system in 2008. Once so many devices were connected, you had a space with vast numbers of interconnected performance archives which could record what had been done with each of those devices. Luckily for internet entrepreneurs, only a few people realised that this new connected space was potentially a vast surveillance system: at the time most of our activity online was not yet being surveilled.

By then a fifth and decisive step was also unfolding: to allow the market to generate new ways of bringing this vast new information flow into focus. For example, search engines to help people find the online information they wanted (Google had some years before developed a search engine to rival then market leader, Yahoo, while also launching Adwords in 2000 to streamline its service to advertisers). And new spaces where we could be social: what we now know as social media ‘platforms’. The early social media were by no means all advertising-focused: Myspace quickly was, which was why it was bought by News Corporation, but Orkut created by Google wasn’t. But those early platforms fell away when a new competitor, Facebook, grew to a huge scale in the early years of the last decade. Even Facebook, however, was not at first focused exclusively on generating advertising through every means possible. This started to change from 2008. when Sheryl Sandberg was appointed from Google to be Facebook’s chief executive.15

Sixth step – and here the full potential and danger of the fifth step emerged: allow those who run search engines and platforms to operate them according to whatever business model they want. The first five steps had unfolded without developing any generally recognised way of securing payment for content or services online. But the development of platforms and interfaces cost a lot of money, costs which private corporations paid for. So, why not (the sixth step) allow search engine and platform owners to gather unlimited information about those who used their services and collaborate with marketers who aspired to do just the same and might be willing to pay for that opportunity? Why not allow data gathered for one purpose to be freely reused for any other purpose? Why not let businesses build algorithms to evaluate, predict and, in the long-term, push ‘engagement’ with their services?16 Indeed, why not let businesses, in pursuit of more engagement, track any aspect of user behaviour they want? And, finally, why not allow – indeed encourage – those platforms to proliferate across the world, insisting that other countries put nothing in their way, so sweeping aside rival local platforms? If you put your trust...