Towards the Platform Society

After an exciting workshop day, we're now starting AoIR 2016 proper with the opening keynote by José van Dijck from the University of Amsterdam. She begins by noting the work of Tarleton Gillespie on the politics of online platforms, which has been very influential in Internet studies in recent years. Internet platforms are now intricately interwoven in a technical, commercial, and social ecosystem, with a number of leading platforms serving as the major gateways to that ecosystem.

But new platforms are constantly emerging, to systematically connect people to things, ideas, and money. These platforms penetrate all aspects of our public and private lives; in any major area these platforms are important gateways to information and connectivity. A platform in this context is an online site that deploys automated technologies and business models to organise data streams, economic interactions, and social exchanges between users of the Internet, José suggests. They are therefore not simple facilitators, or stand-alone objects, but are intricately connected to each other.

The platform ecosystem that the comprise is an assemblage of online platforms, governed by its own dynamics and operating off a set of mechanisms that is inscribed in its architecture. Amongst these mechanisms are datafication, which translates activities into information that is transformed into value by algorithms; commodification, which represents a platform's business and governance models and defines the way that datafied information is transformed into value; and selection, which determines how data flows are filtered by algorithms and bots to allow for personalisation, rankings, reputation, etc. Many such mechanisms are implicit, and it is crucial for research to make these explicit (not least also in order to hold them to account).

At the core of the ecosystem are platforms such as Google (or now Alphabet), Apple, Facebook, Amazon, and Microsoft. These lead the market, which therefore is not a level playing field; they gain value because of the volume of users they are able to attract, especially through network effects. These have also led the platformisation of an increasing number of industry sectors, transforming their underlying architecture – from information industries to car sharing and beyond.

In the car sharing sector, a company like Uber is now a major player, but a number of for-profit and not-for-profit companies have also emerged. In news, Huffington Post and Buzzfeed have emerged as major players, but the key platforms themselves are also increasingly active as news aggregators and publishers. In health and fitness, Google Fit and Apple Health are major hubs of the industry now. In local content, a wide variety of neighbourhood apps have emerged – some local, some transnational; some independent, some as major businesses. And in education, there has been a strong process of platformisation especially in the context of Massive Online Open Courses (MOOCs). Many other industries could also be mentioned here.

Sector-specific platforms in just about any sector are crucially interrelated with the 'big five' platforms that underpin what José calls the platform society: a society in which social, economic and interpersonal traffic is largely channelled by an (overwhelmingly corporate) global online platform ecosystem that is driven by algorithms and fuelled by data. This builds on other, previous concepts such as participatory culture and the sharing economy, but moves on from the participatory euphoria associated with of the early days of 'Web 2.0' and shifts from an emphasis on sharing between individuals to one on sharing data with these major platforms. It also highlights some of the personal discomfort that arises from such sharing in all its guises.

There have therefore also been protests against the drivers of this sharing economy – such as Uber or airbnb: these have been seen as disrupting existing business models as well as local communities, and state authorities have been trying to develop more effective legislation to address them. airbnb hosts are disrupting the local rental market, especially in major destination cities like Amsterdam and Berlin; Uber evades relevant licencing and safety rules, and has poor relations with its own drivers. They do not take the responsibility for anchoring public values in their own companies.

There are therefore platform promises and paradoxes: they claim that they offer personalised services while contributing to the public good, and thereby that they are better able to do so than conventional public services; but in order to do so (and to do so more easily and more cheaply) bypass important government regulations. They also offer cheap or free services while obscuring the private interests and private gain that are involved in operating these services.

How do we anchor collective, public values in a platform society, then? There is a distinction here between governance of and governance by platforms. Governments themselves are mostly concerned with legislating the governance of platforms – but there are other values that are increasingly governed by the platforms themselves, with little involvement from legislators, because these values are hidden in the mechanisms, dynamics, and architecture of the platform ecosystem (which also keeps changing rapidly).

Who rules the platform society, then? Who are the stakeholders responsible for the organisation of platform society (or platform societies around the world)? Involved here is a contested range of market, state, private and public stakeholders at local, national, and international levels, but the interactions between them are not neatly balanced as in some Habermasian fantasy, but complex and often agonistic.

In the health sector, for instance, there is now a substantial number of health and fitness apps available from the various app stores; these monitor a wide range of health data. Such data flows are part of a broad push towards the datafication of users and their activities, and are now redirected to connect users and companies directly. These data flows are constantly revised, and also redirected towards third party companies, some of which claim to be 'not just for profit'. Private gain in these contexts is hiding behind the personal and public good argument, and there is increasing third-party pressure to use these apps – for instance, insurance companies have begun to offer discounts to users who do use fitness apps.

Governments are meant to regulate these apps, but it commonly takes months to test the accuracy and safety of one single app; testing authorities thus simply cannot keep up with the development of this sector. There are also significant questions over who owns the data generated by these apps. Regulation is further complicated by the different rules that apply across various legislations; there is a certain amount of forum shopping by app developers in order to ease their passage through the approvals process.

This has also led to the transformation of some companies in this industry; these have moved from being medical device manufacturers to positioning themselves as data producers, because their (now networked) devices are now also major data generators. The political economy of this entire platform ecosystem is becoming increasingly complex, which also makes the ecosystem more and more difficult to regulate for legislators.

The governing institutions of past sectors also used to regulate their professional norms; with the decline in the power of these institutions comes a decline in these norms, and so the question becomes how we might build trust in the emerging platform society. This requires a comprehensive knowledge of the ecosystem, and an understanding of how it is governed, but such knowledge is difficult for ordinary users to obtain and maintain. As users our power to impact on this global system is very limited, but we must require transparency in platforms; resist trading public value for the sake of convenience; and be vigilant and informed. Governments, too, must stand up for the rights of citizens (not just of consumers), and ensure that their rights are protected.

For owners and developers, the advice is to put public trust over short-term gain; be transparent about data flows, business models, and governance structure; and help encode public values in platform architecture (e.g. by implementing privacy by design). airbnb could easily inscribe a rule into its platform that hosts in specific areas may not rent out their homes for more than sixty days per year, for instance, in order to avoid quasi-commercial airbnb hosting that puts pressure on the ordinary housing market. By pushing for such rules, governments could defend public values and the common good, and negotiate public interests with platforms.

There is also a need to update regulatory institutions and processes, however, in order to bring them up to date with the digital age. Eventually it may even be necessary for governments to develop an international blueprint for the platform society, however difficult this may be. Overall, the platform society is therefore in the first place a field of struggle, with a wide variety of stakeholders having an interest in its further development – but governments are especially called to task to make the platform society governable. This may be a particularly pressing issue in Europe, where the leading platforms are largely US-based and therefore often evade existing regulation.