Review of The Internet Trap: How the Digital Economy Builds Monopolies and Undermines Democracy
Princeton University Press, Princeton, NJ, 2018
256 pp., illus. 7 b/w. Trade, $29.95
In The Internet Trap Hindman argues that we in fact live with two internets: the imagined internet and the real internet. The ‘imagined internet’, where a flattened digital commons made possible by technology that – through digital copying – is not limited by physical restraints either of reproduction or distribution – offers a democratised space in which the marginalised can become the centre, and small tech companies run from garages can topple giant competitors. The other is the ‘real internet’, which the book sets out to describe and explain. This is an internet that is monopolised by two major companies, Facebook and Google. It is also an internet where recent legislative changes in the U.S. now mean that ISPs can charge more for greater bandwidth, meaning that no longer are all online publishers equal in terms of the accessibility of their content. It is an internet where local news is in effect non-existent, erased by the economic realities of economies of speed and scale – and where even the largest news sites, such as the Huffington Post, cannot justify the resources needed to maintain aspects of their web presence independently of the infrastructures provided by the Facebook platform. In the face of this distinction between the ambitions and rhetoric of the imagined internet and the current situation, The Internet Trap sets out to offer some explanation of the current online monopolies, the forces that maintain them, and finally the extreme difficulty, perhaps impossibility, of dislodging them.
As with the recent acceptance by media and public of the dangers to both privacy and democracy posed by contemporary web technologies, any informed reader might recognise the manner in which digital technologies are often represented – as immaterial, ubiquitous and egalitarian– as inaccurate. However, Hindman takes this critique further by furnishing us with an analysis that shows the extent to which digital technologies have far more in common with the large scale, physical smoke-stack industries that they were purported to replace. Energy and resource hungry, running on monolithic data centres, it makes more sense to consider the services offered by businesses like Google or Facebook as a continuation of the industries of the past, and just as subject to familiar economic principles such as economies of scale. Applying principles of economic theory that might usually be considered more suited to ‘real life’ markets and industries, Hindman argues for understanding the internet’s story as one of continuity rather than revolution and situates our current predicament in the wider context of media history. The text favours topics such as local new sites over the more futuristic and fashionable current developments, such as cryptocurrency and the dark web, but also offers a useful and up-to-date grounding in many developments in web technology including accessible explanations of how things like algorithms get made (by people it turns out).
The central theme of the book is understanding how the web’s form gives rise to concentrations of power. One of Hindman’s examples, personalisation, is a key tool for ensuring that a site maintains its web traffic. In chapter 3 he tells the story of the emergence of personalisation algorithms through a close examination of their development. Hindman uses the documentation of an open competition that was conducted by Netflix to push the state-of-the-art in the recommendation algorithms used on its film and television streaming website. Over three years the company orchestrated a fierce competition between a number of international teams, involving many leaders in the field. The result was the discovery that the accretion of multiple small improvements in making meaningful predictions of what different people would prefer led to the most effective, if extremely complex, algorithms. However, as Hindman points out, what also emerged for the competitors and company was the realisation that despite the complexity and mathematical sophistication of any algorithm, it was the quantity of available data about content and users that always in resulted in an advantage. The conclusion? Effective personalisation requires massive data sets and resources, both in terms of computing power and the workforce able to implement the hardware and software resources that can carry it out. The result is simply that larger sites that monopolise a market and combine the most visitors with breadth of content are placed to always dominate because they can accumulate the required mass of user data to make personalisation valuable to their users in a way that smaller site will never be able to do. Hindman’s thesis: The internet consistently favours those who have scale on their side.
The Internet Trap makes use of economic theory that the digital revolution was supposed to turn on its head. It makes these ideas accessible to a wide readership but also includes an appendix with the formulas and approaches used. The result is a reading of the contemporary web and its politics that feels refreshingly grounded and cleanly argued. The book offers a useful manual to understanding the web in its current form and offers a lucid sense of the accretion of power that has taken place online, and the efforts that will be needed to resist it. In the end the titular ‘trap’ seems far less hyperbolic than we might want to think. However, it also does a good job of connecting back to the bodies of literature and theory that have examined the characteristics of emerging media from earlier times, print, broadcast and electronic, all of which made the same, apparently irresistible move from open and undefined emerging media to reified monopolies.