We Are Not Just Strange Looking Addresses on a Blockchain: Exploring the Programmed Sociality of Tokenised Communities on the Blockchain-empowered Steem
Ecosystem
Rimmert Dooitze Sijtsma MA Thesis
Program: Media Studies
Universiteit van Amsterdam
Referencing: APA Sixth Edition
1 Abstract
In the ’platform society’ (Van Dijck, Poell, & De Waal, 2018), digital platforms are rear-ranging how sociality is programmed (Bucher, 2012). So far, this phenomenon has only been studied on centralised platforms, such as Twitter and Facebook (Bucher, 2013; El-lison & Boyd, 2013; Papacharissi, 2002; Van Dijck, 2012). However, I aim to study how decentralised social media platforms, through a technographic (Bucher, 2012) case study of the Steem ecosystem, engage differently with the platform mechanisms of datafication,
commodification and selection introduced by Van Dijck et al. (2018) to understand how
this programs sociality. The case study highlights how the specific blockchain, cryptocur-rencies, and consensus protocol technologies are closely intertwined to enable the Steem ecosystem to decentralise the distribution of rewards in cryptocurrency tokens to each contributor on the ecosystem. In turn, what I term a tokenised community forms around the Steem ecosystem that wants and needs to form an algorithmic imaginary (Bucher, 2017) to understand and protect the community’s stake (consisting of time and cryptocur-rency rewards) on the platform. The Steem ecosystem thus democratically rewards all actors, but is still driven by a calculative regime where governance is organised through performance measures, accounting and audits (Van Dijck et al., 2018). However, on Steem, it is a regime surrounding a blockchain rather than a centralised entity (Leistert, 2018). To conclude, this thesis has highlighted the relevance of studying decentralised social media platforms and their impact on how interactions come to be.
Keywords: platform society, blockchain, Steem, programmed sociality,
Contents
1 Abstract 2
2 Introduction: New Protocols in Old Bottles? 5
3 Theory: Coming to Terms With the Web of Ruin 12
3.1 Marching Towards an Increasingly Connected World . . . 14
3.1.1 Datafication . . . 15
3.1.2 Commodification . . . 19
3.1.3 Selection . . . 23
3.2 The Decentralised Dance of Technology . . . 26
4 Methodology: How to Navigate Decentralised Platforms 28 4.1 Technography . . . 28
4.2 Making a Case: The Steem Ecosystem . . . 30
4.3 The Analytical Protocol . . . 30
4.4 Methodological Considerations . . . 32
5 The Tokenised Community: A Case Study of the Steem Ecosystem 34 5.1 The Tokenised Community . . . 34
5.2 The Blockchain Regime and Datafication . . . 37
5.3 Change of Markets . . . 45
5.4 Governance and Selection . . . 53
5.5 The Tokenisation of Communities . . . 59
6 Discussion: Beyond the Platform Mechanisms 65 6.1 Be the Accountant of Your own Dreams . . . 66
6.2 App Studies and the Infrastructural Turn . . . 70
6.3 Programmed Sociality in a Decentralised System . . . 73
2 Introduction: New Protocols in Old Bottles?
“You have earned $56,45 with your interactions last week! The next pay-out is in 7 days. Keep interacting to earn more tokens!” Within the Steem ecosystem1, users are interacting and, if their content is popular, earning real-world dollars as a reward for their ’labour’. This rather novel space for interaction allows users to receive cryptocurrency tokens as a reward for commenting, liking, sharing and posting all sorts of content on the SteemIt graphical user interface. Being rewarded for your time and effort on social media seems to be a distinct step away from centralised platforms such as Facebook and Twitter.
Social media platforms have often been discussed in terms of how they capture and circulate data (Van Dijck, 2014), how they commodify the data that has been collected (Zuboff, 2015), and how these platforms govern their respective spaces (Gillespie, 2018). However, it becomes clear that these studies only focus on centralised platforms, or platforms controlled by a single entity, to answer these questions. A platform in this thesis refers to a software program that makes available other services to third-parties through application programming interfaces (Helmond, 2015). I will approach these questions through the decentralised Steem platform and its wider ecosystem as there has been little academic research regarding the pronounced differences between decentralised social media ecosystems and their centralised equivalents.
Why does it matter, then, to study an ecosystem that engages in practices so different from those common amongst contemporary platforms? According to media, culture and communication scholar Alexander Galloway (2010), studying this decentralised ecosystem matters as networks without central control (or "webs of ruin") are part of the new "apparatus of control" (p.3). Galloway (2004) develops this point and argues that control lies in the protocols through which these computers are connecting and interacting in
1 This Steem ecosystem consists of the Steem platform and a plethora of ’blockchain-empowered apps’,
such as SteemIt, that are connected to the Steem platform. For the remainder of this thesis, I will refer to the Steem platform and SteemIt blockchain app together as the ’Steem ecosystem’.
distributed or decentralised networks. As the Steem ecosystem claims to be the first decentralised social media of its kind (Steem Whitepaper , 2017), this warrants a venture into the web of ruin that is the Steem ecosystem to describe closely the associative ties between actors controlled by specific protocols in this blockchain-empowered ecosystem.
Steem rewards all users for their micro-interactions occurring on the SteemIt interface, which leads me to argue that it is the specific software architecture of the Steem ecosystem that redefines this new approach to social media and the programming of sociality therein. Sociality, understood here in Latourian terms as the way in which different (human and nonhuman) actors can form associations between themselves (Latour, 2005), has a well-established history in social media research.
In their work, Ellison and Boyd (2013) describe how the introduction of a public profile, a public friend list, the ability to view your friends’ profile and the ability to engage with and create content rearranged how social interaction was organised within these
social networking sites. Papacharissi (2018) describes in detail how social media networks
amplify the possibilities and limitations to forming lasting relationships. In a similar vein, Bucher (2013) has described how ’friendships’ on Facebook are assembled in specific ways because of the underlying structure of the Facebook platform.
Following sociologist Adrian Mackenzie (2006), software is arguably situated in this strand of research as a social object and a process, capable of forming social relations that result in a "neighbourhood" (p.3) through software. What is the Steem neighbourhood like, then, if its inhabitants are now actively rewarded for their time and effort on the platform? In the remainder of this thesis, I will refer to the forming of social relations through software as programmed sociality, thereby following the work of Bucher (2012). She has argued that software is active in the sense that it assembles relations, which brings forth questions surrounding the novel software features embedded in the Steem ecosystem and how they might prove crucial in the shaping of the Steem neighbourhood.
as SteemIt (n.d.-b) has positioned itself as having "redefined social media by building a living, breathing, and growing social economy". It represents a "new kind of attention economy". If so, given the strong relationship between social media platforms and the organisation of social activity (Van Dijck et al., 2018), then this begs the question as to how the Steem ecosystem redefines social activity.
There have already been numerous studies into how contemporary social media redefine modes of social interaction. For example, considering the current professionalisation and monetization of amateur content on YouTube, Cunningham, Craig, and Silver (2016) de-scribe how the YouTube platform, and its competitors, directly clash with the mass media approach of Hollywood. This ’new screen ecology’ presents a historical overview of how a rapidly developing platform such as YouTube attempts to come to terms with the "con-ditions of possibility for entertainment, content and talent development" (Cunningham et al., 2016, p.388) it offers.
Moreover, social media is often heralded for the agency it provides for social activity. As Cunningham et al. (2016) state, YouTube offers amateur content producers a platform to broadcast themselves while YouTube itself is able to leverage these young talents to attract advertisers. But the work of a social media producer is driven by unrelenting algorithms that raises questions surrounding the freedom of cultural expression (Duffy, 2017).
The direct rewarding of end-users implies a clear departure from the ’traditional’ multi-sided market model that many of these platforms employ. These centralised platforms, such as Spotify (Prey, 2016; Prey, Esteve Del Valle, & Zwerwer, 2020), Facebook (Van Di-jck et al., 2018) and Google (Rieder & Sire, 2014), are matchmakers in the sense that they connect different groups, either in a real or virtual space through their platform (Evans & Schmalensee, 2016). Morris (2015) describes how Apple’s iTunes Store adopted an à
la carte model of selling digital music (buy a single song for $.99) in the first decade of
the twenty-first century that resembled a traditional way of buying a cultural commodity online.
Spotify and the like, however, grew increasingly stronger, Apple’s strategy of commodity ownership became increasingly questioned as consumer interest shifted to streaming ser-vices (Morris, 2015). Spotify, as a platform, is able to leverage a so-called ’network effect’ where the perceived functionality of the users increases as more users join the network (McIntyre & Srinivasan, 2017). These platforms then attempt to leverage the benefits of this network effect to create a dominant force in their specific market (McIntyre & Srinivasan, 2017).
Given the decentralised nature of the Steem ecosystem, questions arise as to who exactly benefits from the network effect that Steem says it leverages (Steem Bluepaper , 2017). The decentralised nature of the Steem ecosystem might play a crucial role as it is not straightforward who precisely is the platform-owner (Tiwana, 2013). Whereas centralised platforms can always be discussed in terms of a ruling entity, the Steem ecosystem cannot. It is an open-source, community-driven effort that is responsible for the development of the platform (Steem Whitepaper , 2017). However, according to the whitepaper, each blockchain app, such as SteemIt, that is connected to the Steem platform has its own registered company. These blockchain apps, however, do not control how the platform is developed. The work by Tiwana (2013) regarding the development of platforms and their governance might therefore not be directly translatable to the Steem ecosystem.
Furthermore, the so-called attention economy - "[t]he game of harvesting human attention and reselling it to advertisers" (Ferdman, 2020) - introduced by social media platforms, has already received academic scrutiny. Drenten, Gurrieri, and Tyler (2020) criticize Instagram through the lens of what they call the ’monetization of attention’, where the ability to monetize prosumer content is contingent on the attention it receives.
These studies thus position the attention economy as a pervasive and perhaps hostile development to user agency. As SteemIt explicitly distances itself from this understanding of the attention economy, the question arises of how SteemIt brings individuals together that is different from the existing attention economy.
Steem positions itself as the first ’incentivised, blockchain-based, public content platform’ (Steem Whitepaper , 2017, p.1). Helmond (2015) describes the advent of the platform as paradigmatic shift where the platform has become the "dominant infrastructural and economic model of the social web" (p.1). In this understanding, these platforms have developed from social networking sites into platforms that spread their influence into all nooks of the social web.
The Communication and Information Science scholar Tarleton Gillespie (2010) defines platforms as infrastructures to build applications on and emphasises their ability to bring different actors together to communicate, interact or sell. The media scholar Anne Hel-mond (2015), however, argues to look beyond the rhetoric of platforms and look specifi-cally at the computational work that platforms do to function. In doing so, she aligns her work with the work of Bogost and Montfort (2009) who introduced the field of platform studies as a way to understand how the technological specificity of a platform is related to culture.
Given the recent academic interest in the transformation of Social Networking Sites into social media platforms (Helmond, 2015; Nieborg & Poell, 2018), the development of what I term a ’platform-first’ system introduces questions regarding how Steem from the outset has been developed with the idea of a social media platform in mind rather than transforming a social networking site into a social media platform.
Furthermore, Steem does not explain what sort of incentives are referred to in their claim to be the first incentivised social media (Steem Whitepaper , 2017). Moreover, this also seems to imply that existing platforms offer no incentives to the actors involved. Van Dijck (2012), however, has already described the social values such as popularity, attention and connectivity as the main incentives for communicative traffic on social media platforms. Furthermore, she positions corporate incentives of market forces and commodity exchanges as the main incentives for corporate entities. The claim by the Steem blockchain that they are the first ’incentivised’ platform thus raises questions regarding the types of incentives they are allegedly adding.
The Steem ecosystem thus positions itself as a revolutionary approach to social media due to its decentralised software architecture, which introduces several new questions regard-ing ’platform mechanisms’ and the programmregard-ing of sociality. In their work, the media scholars Van Dijck et al. (2018) develop a framework to understand how digital platforms rearrange sociality. According to them, sociality is increasingly shaped by mechanisms of datafication, commodification and selection (Van Dijck et al., 2018). Although I will revisit these claims in more depth in the following chapters, let me suffice in saying that they see the capturing, circulating and monetizing of new forms of data governed by digital platforms in increasingly more facets of Western societies as the primary reasons for shifts in the programming of sociality (Van Dijck et al., 2018). It remains unclear, however, how decentralised social media platforms fit within the framework presented by Van Dijck et al. (2018).
An exploration of the specific software architecture of the Steem ecosystem and its advent into the world of social media platforms might lead to a better understanding of how blockchain-empowered social media ecosystems re-appropriate these platform mechanisms and thus program sociality. This has been identified as a gap in the current academic understanding of social media platforms.
More specifically, I ask, ‘How do the platform mechanisms of ’datafication’, ’commodi-fication’ and ’selection’ within the blockchain-empowered Steem ecosystem program so-ciality?’. The goal of my thesis is thus to explore these platform mechanisms within the Steem ecosystem and to understand further how the decentralised nature of the Steem ecosystem is organising new forms of sociality. Given the strong relationship between so-cial media platforms and programmed soso-ciality (Bucher, 2012), the answer to this central question will provide new entryways into studying decentralised social media platforms and programmed sociality in the fields of media and platform studies.
By looking specifically at the technical characteristics of the Steem ecosystem through a technographic study, it becomes possible to see how software actively assembles the potential for how actors belong and relate to one another (Bucher, 2012). I aim to not
simply describe how the software works but to describe how software ’operates’ (Bucher, 2012, p.16) to achieve certain outcomes. By focusing on the fact that the Steem ecosys-tem is powered by a blockchain, I can examine how the technical level translates into the interface level and, subsequently, in the ways sociality is organised. Finally, by embed-ding the contemporary discussion on the platform as a dominant model in contemporary societies, I will discuss how these decentralised platforms pose new challenges but also solve existing ones. In the remainder of this thesis, I will further argue that the way in which, what I term, ’tokenised communities’ in the Steem ecosystem develop, is the result of the strategic programming of software to incentivise sociality through interac-tions between different actors. I define a tokenised community as ’all actors (both human and non-human) vested in a cryptocurrency native to a platform assembled through a blockchain ecosystem’.
Now that the goal of this thesis has been positioned, the following sections will ad-dress the questions posed in this introduction to the world of decentralised platforms and programmed sociality. Chapter two will introduce and debate the existing literature surrounding the topic of programmed sociality (Bucher, 2012), platform studies, the plat-form mechanisms of datafication, commodification and selection (Van Dijck et al., 2018) and the role of decentralisation in programmed sociality. In chapter three, I will present the methodological approach that I followed to undertake the technographic analysis of the Steem ecosystem, the rational for the case study and the materials that have been used during for this analysis. The results of this case study and method will be presented in chapter four. This case study provides a close description of how the different workings of the Steem ecosystem re-configure the ways in which sociality is assembled. Finally, chapter five will provide a discussion of the main additions that this thesis has brought forward. The additions will be discussed in relation to previous literature and critically examined. Finally, the additions will be placed in a wider context in the conclusion to provide clear pathways for future research and discussion.
3 Theory: Coming to Terms With the Web of Ruin
The main premise of blockchain ecosystems is that they allow interaction and exchanges between actors without the governance of a central entity. These connections are estab-lished through several layers of protocols to ensure that these "webs of ruin" function (Galloway, 2010, p.776). The way in which these protocols program interactions matters if one is interested in how different actors belong and relate to each other through soft-ware (Bucher, 2012). In other words, to study the sociality of actors (both human and non-human) as they interact is, in Latourian terms, the study of a "trail of associations between heterogeneous elements" (Latour, 2005, p. 5). The social is thus understood not as a thing among other things, but as a type of connection between things that are themselves not necessarily social (Latour, 2005).
The dissociation between the social as ’a thing among other things’ and ’a connection be-tween non-social things’ was the premise put forward by Latour (2005). Before, according to Latour (2005), the study of the ’social’ was to identify the ’social’ as something that existed alongside things such as the economy, the law, politics and so forth. This idea, however, became difficult to upheld as the search for the ’special ingredient’ that set the social apart from these other fields of study seemed impossible to find (Latour, 2005). The social factors in economical studies, for example, left little room to argue that the social was a distinct ’thing’. Instead, Latour (2005) argued that the existence of certain social ties uncovered the hidden presence of some specific social forces.
The social could thus be understood as that which is being glued together by many non-social connectors. These social ties are especially visible when the system is not in equilibrium. Any time there is a change in the ways actors decide to interact, Latour (2005) argues this is a re-association of social ties into a new assembly. Certain events that cause instability, such as the COVID-19 pandemic, make humans reassess the norms and values they hold in their interactions. The 1,5m distance between humans queu-ing for their groceries is in Latourian terms an example of how the social is constantly reassembled.
The re-assembly of sociality through software has been explored through the idea of programmed sociality (Bucher, 2012). By connecting software to sociality, Bucher (2012) paved the way for a analysis into how software prescribes certain norms and values in the way it programs (assembles) how actors belong and relate to one another. Her first work on programmed sociality provided insight into how Facebook and Twitter organised sociality through algorithms, APIs and friend suggestions (Bucher, 2012). However, I am particularly interested in how the medium-specificity of the Steem ecosystem programs sociality. I put forward the argument that the blockchain ecosystem re-assembles the ways in which actors belong and relate to each other through the strategic deployment of software and specific protocols embedded in this ecosystem.
Platforms around the globe2constantly "reorganise value regimes and economies" (Van Di-jck et al., 2018). These shifts could be explained through the concept of platformization. To be sure, these shifts are the result of constant interactions between platforms and cultural practices (Deuze & Prenger, 2019; Holt & Perren, 2011; McRobbie, 2018).
The term, first coined by Helmond (2015), refers to the ways in which these platforms are branching out to other parts of the web and increasingly draw in as many components of the web as possible into its architecture. Helmond (2015) drew on existing literature from software studies and defined platformization as "the extension of social media platforms into the rest of the web and their drive to make external web data platform ready" (p.1). The remainder of this thesis will refer to platformization as it was positioned by Helmond (2015) as it allows for an analysis of the ways in which platforms program sociality.
Within this thesis, the abbreviation ’app’ will refer to a software commodity which has specialised functions that solves a mundane task and that is distributed and commodified through a specific platform. I therefore align my understanding of ’apps’ with the work of Morris and Elkins (2015). Furthermore, I find the emphasis on the distribution of apps through platforms useful because it is this platform structure that consolidates the app
2 Google, Apple, Facebook, Amazon and Microsoft (GAFAM) predominantly in the ’West’ and Baidu,
as a specific type of software.
In this sense, my work is closely aligned with the work of Tiwana (2013) and Gerlitz, Helmond, van der Vlist, and Weltevrede (2019), whom also emphasise the app-platform relation. Morris and Elkins (2015) argue that apps are more than simply an abbreviation of applications. They position the advent of apps as a paradigmatic shift in the software industry as apps fracture the previously multi-purpose application software into distinct and specialised functions that are commodified on their own (Morris & Elkins, 2015). They also position the platform as having a key role in the advent of these mundane apps. The introduction of the iPhone Software Development Kit and App Store gave developers and users a single outlet from which they could distribute and purchase a plethora of different apps (Morris & Elkins, 2015).
It is this ecosystem (Tiwana, 2013) of the platform, developers and users that have allowed the "proliferation of mundane software and an intensified integration of software into everyday routines" (Morris & Elkins, 2015, p. 70). I approach the Steem ecosystem in a similar fashion. Where I see the developments in the field of platform studies not as competitive to the developments in the field of app studies, but rather approach them together as an ecosystem to understand the intricacies between the platform, developers and the users that make up this Steem ecosystem.
3.1 Marching Towards an Increasingly Connected World
Now, programmed sociality within what Van Dijck et al. (2018) name the ’platform so-ciety’ has already amassed a lot of academic interest. Yet there is less known about how these new decentralised ecosystems continue or change contemporary ideas on pro-grammed sociality. The following section will look specifically at some of the key questions raised by Van Dijck et al. (2018) to contrast them with literature regarding decentralised ecosystems. In order to understand how software program the connections between dif-ferent actors, it is worthwhile to first try to find common ground and differences between centralised platforms and decentralised platforms.
The work of Van Dijck et al. (2018) is useful as it portrays digital platforms from neither a technological optimistic nor from a deterministic approach. Rather, they see digital platforms as gradually infiltrating in an increasingly connected world (Van Dijck et al., 2018). In this way, following Latour (2005), the social is produced through the ties that are formed between actors, and in this ’platform society’ (Van Dijck et al., 2018), these ties are increasingly formed through digital platforms.
Furthermore, the active role that platforms have in the programming of sociality has already been emphasised (Bucher, 2012; Van Dijck et al., 2018), which raises questions surrounding the embedded norms within software and how they may or may not clash with existing values. The programming of sociality through digital platforms is driven by three mechanisms that mutually shape technology, economic models and end-users:
datafication, commodification and selection, according to Van Dijck et al. (2018). In the
following sections I will discuss these three mechanisms to identify the main differences and similarities between central and decentralised platforms.
3.1.1 Datafication. The term ’datafication’ refers to the ability of digital platforms to capture human activities into data (Mayer-Schönberger & Cukier, 2013, p. 91). Al-though social activity was also captured before the advent of digital platforms, for ex-ample, to see people as a population through census and statistics (Foucault, 2007), contemporary digital platforms have been organised in such a way that all interactions on the platform can be captured into data (Van Dijck et al., 2018).
Whereas the collection of data could be seen as by-product in the earlier days, it has since become the key commercial strategy for platform owners (Van Dijck et al., 2018). In what is oftentimes referred to as the age of "Big Data" (Lycett, 2013; Mai, 2016; Van Dijck, 2014), data can be considered as the engine powering digital platforms. Through a new ’repertoire of interactions’, such as liking, upvoting, sharing and commenting, these plat-forms employ increasingly sophisticated strategies to capture behavioural data (Van Dijck et al., 2018).
Case in point is the Facebook ’Like button’ that has seen several developments over the years. At first, the Like button was introduced to enable peers to communicate through the click on a single button (Sumner, Ruge-Jones, & Alcorn, 2018). Then, Facebook allowed other websites to implement the Like button on their website, which can be understood as the ’platformization’ of the Internet (Helmond, 2015). Even more recently, Facebook has introduced several icons representing human emotions as an additional layer to the Like button, subsequently allowing Facebook to automatically capture emotional behaviour on its platform (Smoliarova, Gromova, & Pavlushkina, 2018).
This example shows the insatiable thirst that digital platforms have for new streams of data. Arguably, it is the central position of the platform owner that is driving this thirst for data. According to Van Dijck et al. (2018), all "activity of every user can be captured, algorithmically processed, and added to that user’s data profile" before it is subsequently monetized by the platform owners (p.34). This is possible because all data flows are controlled by the platform owner, who then circulates it amongst the different platform participants (Van Dijck et al., 2018).
Decentralised platforms seem to approach datafication from a rather different perspective. Whereas centralised platforms such as Facebook store all captured data on servers that the company develops, builds and exploits, decentralised platforms make use of blockchain technology to store captured data. They both engage in practices of datafication but the underlying approach sets them apart.
A blockchain can be understood as a continuous and permanent public ledger that allows the storage of transactions that are connected to previous transactions through a cryp-tographic hash (Lydiate, 2018). Crypcryp-tographic hashing means that data is relocatable through a key value (Brookshear & Brylow, 2015) which is encrypted using a specific algorithm. In a blockchain, this essentially means that each block is stored with a unique key value and the key value of the block that was created before it, creating a chain of blocks in the process. A transaction refers to any action that changes the state of the ledger (blockchain) (Casino, Dasaklis, & Patsakis, 2019).
Blockchains store user data but do not necessarily do so for the segmenting of audiences into different demographics, as is the purpose of centralised platforms (Van Dijck et al., 2018). Instead, blockchains seem to store data that involve transactions (Wüst & Gervais, 2018). Often, it is not straightforward why exactly a platform would decide to use a blockchain over a centralised structure.
According to Wüst and Gervais (2018), employing a blockchain is useful when two mis-trusting parties intend to do a transaction without the involvement of a third trusted party to guide the transaction. Say, I ’upvote’ a post of a peer on the SteemIt interface, then the state of the ledger should represent that I have voted for this post. Following this argument, it becomes a question of how a decentralised platform should be governed without relying on a central force of power. The blockchain technology introduced two key developments to address this issue.
The near-immutable and public nature of the blockchain technology introduces new mech-anisms for how technology, business models and users interact. As the blockchain func-tions as a ledger, it needs to ensure that the state of the ledger is accurate (De Leon, Stalick, Jillepalli, Haney, & Sheldon, 2017). Therefore, the blockchain is nearly im-mutable, which means that changing one block requires each consecutive block after that one needs to be changed, too. The longer the block has been linked in the chain, the harder it will become to amass the amount of computational power needed to change the block as more and more blocks need to be changed.
On centralised platforms, such as Facebook, the collected data is obfuscated by the platform as it is the prime resource and the driving force behind their business strategies (Van Dijck et al., 2018). These platforms are able to do so as they are the ones who control who is able to get access to which data. As decentralised platforms have no central entity who controls the ledger, they need to store the data in such a way that all actors can agree on the state of the ledger3.
3 The consensus mechanism through which this agreement is reached will be discussed in the ’selection’
A key requirement for a decentralised ecosystem is thus that the participants need to have trust in that the information in the blockchain is correct. As there is no platform owner who controls the flows of data and the storage of data, the information stored within the blockchain and the source code of the blockchain are publicly available in order for users to be able to verify the state of the ledger (Cai et al., 2018; Casino et al., 2019; Leistert, 2018). This is a second departure from the platform mechanism of datafication described by Van Dijck et al. (2018), as they argue that data is only circulated back to users through graphical user interfaces which allows them to keep in touch with their friends, colleagues and family.
Now, this ’circulation’ of data is true to a certain extent for decentralised platforms, too. The Steem ecosystem does allow users to follow other users, like posts and see what others are doing. Yet, this is not the only way in which users can access data in decentralised platforms. Anyone who has access to the Internet can audit the information that is stored within each block in a blockchain (Casino et al., 2019). This is a development from the obfuscated databases that centralised platforms derive their business from (Van Dijck et al., 2018).
In centralised systems, access to the stored data is possible to the extent that the platform owner allows it (Rieder, 2013). The research tool ’Netvizz’ was developed by Rieder (2013) in part to explore the Facebook API to understand if user’s could figure out to what extent their data was accessible to others. However, following the wake of the Cambridge Analytica scandal, Facebook restricted access that third-parties have to the data it collected, and that was the end of Netvizz, too (Perriam, Birkbak, & Freeman, 2020). This example highlights how centralisation of power lead to situations where access to certain data is disallowed.
As has been argued so far, the immutable and public properties of the blockchain tech-nology allow it to provide a way of storing that is necessary to keep track of a shared history of transactions (Casino et al., 2019; Lydiate, 2018; Wüst & Gervais, 2018). The blockchain thus provides a significant building block for a decentralised social media
plat-form. As it allows the development of a list of transactions which is near-impossible to alter, the history of those transactions can be traced. Given that the Steem platform wants to reward users, a blockchain appears to be a necessity as it allows to keep track of all transactions in a transparent and verifiable manner without central authority. The rewards can be distributed accordingly as all actors agree on the state of the blockchain.
There are thus both differences and common ground between centralised and decentralised platforms and their use of the datafication platform mechanism. Decentralised platforms offer new ways to handle user data without a platform owner controlling the capture and circulation of data. The public and immutable nature of blockchain technology offers new ways to think about privacy problems such as data control, transparency and access control (Zyskind, Nathan, & Pentland, 2015).
However, both centralised and decentralised platforms seem to circulate data to users through graphical user interfaces. The SteemIt interface is but one example of an inter-face where data from the Steem blockchain is fed back to end-users. Users could look at the "raw data" through so-called ’block explorers’ or download a copy of the entire blockchain on their device (SteemIt, n.d.-b), which is not possible to the same extent on centralised platforms (Van Dijck et al., 2018). The interplay between technology, business models and users thus seems to be changed by the introduction of blockchain technology in decentralised social media platforms.
3.1.2 Commodification. The mechanism of ’commodification’ describes the ways in which digital platforms transform "online and offline objects, activities, emotions, and ideas into tradable commodities" (Van Dijck et al., 2018, p. 37). Through the growth of datafication on these platforms, the possibilities for the commodification of data have grown in similar fashion (Van Dijck et al., 2018). According to Van Dijck et al. (2018), these commodities are mostly valued for grabbing attention, data, money or attract users. Digital platforms have been effective in the commodification of these ’goods’ arguably because they transitioned the classic two-sided market to a multi-sided market where they
are able to leverage the infrastructural core of the platform which the platform owner can connect to varies platform ’complementors’ (Van Dijck et al., 2018).
Consider the digital platform ’Steam’ (not to be confused with Steem), which has plat-formized the distribution of digital video games (Becker, Chernihov, Shavitt, & Zilber-man, 2012). From its launch in 2003, Steam has gathered a community of over 30 million users around the globe. Whereas before, video games were developed and marketed to potential customers through traditional marketing channels by the individual developers, Steam adopted a different way to connect developers and customers. Instead, Steam acts as a digital distributor of video game software (Becker et al., 2012). The platform brings together game developers, big or small, with a extensive user base (Becker et al., 2012).
Steam is able to leverage the connection it forms between countless of developers and even more users through the central role it plays on the platform. Following the logic of the multi-sided market as stated by Van Dijck et al. (2018), the platform allows developers to use its infrastructure to reach a pool of potential customers. On the other hand, the platform affords users a central place where they can store all their games, connect to friends, and form communities with other users (Becker et al., 2012). Steam benefits from bringing all these groups together as they receive a fee for every purchase that takes place on the platform (Lam, 2019).
The example highlights how these newly developing platforms are able to transition markets to a system where platform owners decide the rules of economical play (Van Dijck et al., 2018). They bring together different actors but only under the rules that they have set for their interactions. The introduction of cryptocurrencies within digital platforms seem to rearrange the economic rules that a platform and the ways in which actors are rewarded.
A cryptocurrency coin is a string of information and a digital signature (Narayanan, Bon-neau, Felten, Miller, & Goldfeder, 2016). The main difference between a fiat currency and a cryptocurrency, such as the Bitcoin, is that the former needs to be enforced through
au-thorities. There are numerous efforts made to secure cash against counterfeiting through specific patterns, materials and so on. Regardless of these measures, governments still need to actively stop individuals or groups from attempting to create and circulate coun-terfeit cash. The latter is enforced fully through protocols that have been agreed upon before the actual creation of a cryptocurrency (Narayanan et al., 2016).
These cryptocurrencies cannot operate without a blockchain supporting them (Narayanan et al., 2016). As mentioned before, a blockchain allows decentralised platforms to create a transparent ledger of information. A cryptocurrency ’coin’ is a piece of information stored in an blockchain signed by the account that created the coin (Nakamoto, 2008; Narayanan et al., 2016). When the coin is transacted to other accounts, the additional information generated by these transactions are also stored as a sort of public, historical overview of the current and previous owners of this specific coin (Nakamoto, 2008).
To be fair, centralised digital platforms also reward users for their activity on the plat-form. The video-sharing platform YouTube, for example, reserves a portion of its revenue to reward creators whom oftentimes have developed a ’fanbase’ on the platform (Cun-ningham et al., 2016). In return for their ’labour’, these creators can monetize the content they upload on YouTube to be rewarded part of the advertisement revenue accumulated for that specific piece of content (Cunningham et al., 2016). However, as Cunningham et al. (2016) state, YouTube is involved in an ongoing balancing act between maximising their own revenue streams whilst maintaining the loyalty of the creators who draw in many viewers. Centralised platforms enable those with an entrepreneurial spirit to be rewarded for their time and effort, which excludes the majority of users (Van Dijck et al., 2018).
The example of content creators on the YouTube platform shows that platform owners of centralised platforms can set the economical rules of the platform as they control how data is collected and used (Van Dijck et al., 2018). The business models of centralised platforms revolve around the commodification strategies personalised advertisements, data services and fees. The employment of any of these strategies enable and change the ways in which
exchanges take place and simultaneously define the activity of a plethora of different actors (Van Dijck et al., 2018). Given the advent of cryptocurrencies on the decentralised platforms (e.g. Bitcoin, Ethereum and Steem), the development of a cryptocurrency native to a decentralised platform could be seen as a fourth commodification strategy.
Whereas platform owners can monetise the whole platform or certain parts, through the strategies of personalised advertisements, data services and fees (Van Dijck et al., 2018), cryptocurrencies were designed to reward all who are involved without a central power driving it (Burnie, Burnie, & Henderson, 2018). However, this is not to say that the original developers of a cryptocurrency have no say over the function it has within the platform ecosystem.
According to Burnie et al. (2018), contemporary cryptocurrencies can be divided across three functions. Firstly, crypto-transaction tokens are developed as a "better form of money" and are easily transferable with almost no limits to acquiring these tokens (p.32). Then, the crypto-fuel tokens help a blockchain app to function. In other words, these tokens are designed to be exchanged within a certain app. Finally, a crypto-voucher token is pegged to a predefined asset. This means that, based on an exchange rate, the cryptocurrency token is always worth as much as the predefined asset (Burnie et al., 2018).
This fourth commodification strategy could thus be implemented by the developers of the cryptocurrency to replace centralised ways to transact currencies, to provide blockchain apps with an internal driver for exchanges, to provide investors an easily convertible currency, or any combination of these functions (Burnie et al., 2018). For those involved early in a certain cryptocurrency, the returns could be substantial. As users start adopting a cryptocurrency, the value of the asset grows with the demand because even though cryptocurrencies are digital their volume is almost always limited (Gandal & Halaburda, 2016). To maximise this development, many developers of cryptocurrencies are looking to maximise a ’network effect’, where adding new users requires no extra cost but does add to the profitability of the coin (Gandal & Halaburda, 2016).
The introduction of cryptocurrencies in digital platforms thus represents a development in the ways in which a digital platform could be commodified. Although personalised ad-vertisement is being developed for decentralised platforms (Shrestha & Vassileva, 2019), data services are open source and thus need not to be commodified and fees are used to pay network nodes for their work (Werner, Pritz, & Perez, 2020), the launch of a cryp-tocurrency and the subsequent attempt to achieve network effects to increase the value of the token introduces new ways to monetise a digital platform (Gandal & Halaburda, 2016), as all actors who are invested in the cryptocurrency can benefit from this process.
3.1.3 Selection. Because of the datafication of user behaviour and the commodifica-tion of these experiences, centralised platforms steer user interaccommodifica-tion through the selec-tion and curaselec-tion of content (Van Dijck et al., 2018). One the one hand, these platforms present certain topics, trends and themes through algorithms while, on the other hand, users are able to filter content through interacting with interfaces (Van Dijck et al., 2018). These user and algorithm-driven forms of filtering and curation has increasingly replaced the reliance on expert-based selection (Van Dijck et al., 2018). However, the algorithms with which centralised platforms filter and curate what end-users interact with remain behind closed doors (Gillespie, 2014).
As was established before through the discussion of the blockchain and cryptocurrencies within decentralised platforms, this issue of the ’black-boxed’ ways of selecting and cu-rating content by a platform owner is revitalised. The public nature of the blockchain and the changed platform commodification mechanism that a cryptocurrency might intro-duce, results in a platform where information is publicly accessible and where information regarding the state of the blockchain and cryptocurrency are important to actors who are invested in the platform (Saleh, 2020). How the platform dynamics of personalisation,
reputations and trends and moderation (Van Dijck et al., 2018) are regulated might prove
to be different from centralised platforms.
the blockchain technology is argued to introduce a "sovereign chronological regime that has the capacities to proof and modulate the existence, identity and administration of data, assets, goods and services from a distance on micrological scales" (Leistert, 2018). As was argued before, the blockchain is immutable and publicly available and only works when there is consensus about that what is stored in the blockchain is correct. It could thus be seen as a sovereign truth from which cannot be deviated (Leistert, 2018). The
consensus protocol (Cai et al., 2018; Chohan, 2017; Leistert, 2018; Xu, Weber, & Staples,
2019, among others) can therefore be seen as the gatekeepers who decide what is to be registered in the blockchain and subsequently the truth. As a consequence, the user and algorithmic-driven selection of information to be displayed follows the rules set by the consensus protocol.
Most questions regarding multi-sided markets revolve around the platform owner and their governance over the platform (Boudreau & Hagiu, 2009; Rieder & Sire, 2014; Ti-wana, 2013). Given that there is no single entity that can control a decentralised plat-form, it becomes a question of how the introduction of a cryptocurrency and a blockchain changes selection dynamics within a multi-sided market. Centralised platforms govern the personalised experience of individual users as they can develop and update algorithms for the personalisation of the user interface (Van Dijck et al., 2018). Although some fear for the fragmentation of the social due to personalisation (for example, Pariser (2011) de-scribes the alleged existence of "filter bubbles"), personalisation is also argued to empower users as they are able to quickly find the best deals or information they need (Van Dijck et al., 2018).
The governance on a decentralised social media platform is organised through a consensus protocol. This is the set of rules on which the nodes of the blockchain network decide whether or not a block is valid and thus should be added to the blockchain (Casino et al., 2019). Although these rules can differ per platform, they are all employed to ensure that nodes who do not know or trust one another can come to an agreement over the correct history of the blockchain, also known as the ’Byzantine Generals Problem’ (Lamport,
Shostak, & Pease, 1982).
Lamport et al. (1982) describe this as the need to conjure up a battle plan but the generals can only communicate through messengers. However, among the generals there are traitors who are trying to prevent the loyal generals from reaching agreement over a good plan. It thus becomes a search for a way to prevent a small number of traitors to cause the loyal generals to agree upon a faulty plan (Lamport et al., 1982, p. 383). This problem matters for a decentralised platform, thus without central governance, as it must be able to identify malicious attempts to manipulate the history of the blockchain and correct them. The way in which the consensus protocol is organised is crucial for the trust that users have in the system.
The Bitcoin cryptocurrency was the first system that solved the Byzantine Generals problem through a consensus model called Proof-of-Work (Nakamoto, 2008). The term refers to the ’puzzle’ that each node needs to solve in order to receive the authority to create a new block and add it to the chain (Nakamoto, 2008). In other words, all nodes compete to solve a mathematical test which requires computational power and the first one that is able to solve it and can prove it by showing the outcome will be rewarded with the creation of a block, which is tied to a reward in Bitcoin (Saleh, 2020).
As all nodes need to compete for every single transaction that is to be added to the chain, Bitcoin and other blockchains that use the Proof-of-Work consensus model have issues with scale (Cai et al., 2018; Casino et al., 2019) and sustainability (De Vries, 2018; Mora et al., 2018). The Proof-of-Work model has a significant impact on the use of electricity worldwide, with some estimations saying that Bitcoin annually uses comparable amounts of energy to Ireland (De Vries, 2018; Mora et al., 2018). The redundancy needed to ensure consensus provides major concerns for further increasing global temperatures (Mora et al., 2018). Thus, the blockchain system introduces a novel way to decentralise transactions, but might lack the scale needed for many real-world applications and imposes significant environmental concerns.
To overcome these technical limitations, governance is increasingly organised through what is called a Proof-of-Stake consensus protocol (Saleh, 2020). Following this protocol, nodes are not required to compete with one another but are randomly assigned the right to create a block, thereby mitigating the energy use of the blockchain (Saleh, 2020). This protocol is closely tied to the existence of a cryptocurrency within the platform (and thus the blockchain) as this protocol makes the assumption that the nodes are invested (stake) in the cryptocurrency and the platform and will thus operate in favour of the blockchain, as any decrease in trust in the system will decrease the value of their investment (Saleh, 2020).
The selection of what is to be appended to the blockchain in blockchain-empowered platforms is thus tied to the investment that the actor has in the platform (Saleh, 2020). This system was adopted to overcome the limitations that the Proof-of-Work consensus protocol has in terms of process speed and energy use (Saleh, 2020). Whereas Van Dijck et al. (2018) introduced the platform mechanic of selection to highlight the ways in which platform owners can change user behaviour through the dynamics of personalisation, reputations and trends and moderation, the introduction of a consensus protocol also introduces the need to understand how platforms come to select what is appended to the blockchain. This is relevant as the blockchain is argued to introduce as sovereign truth from which the platform dynamics mentioned by Van Dijck et al. (2018) are organised.
3.2 The Decentralised Dance of Technology
To sum up, the blockchain, cryptocurrencies and consensus protocols have been dis-cussed through the framework of datafication, commodification and selection introduced by Van Dijck et al. (2018) to understand the active role that platforms have in the programming of sociality (Bucher, 2012). The new technologies that operate blockchain-empowered platforms are important on the level of the mechanisms introduced by Van Di-jck et al. (2018), but also highlight the strong dependence on one another. Cryptocurren-cies were introduced as a democratic way to commodify a digital platform which requires a blockchain to create an undisputed history of the coins. In order to reach consensus on
the history, a set of rules must be in place to govern the integrity of the blockchain and thus the cryptocurrency.
This consensus protocol is often organised around the stake that actors have in the cryptocurrency: going against the blockchain will decrease the investment and is thus less likely to occur (Saleh, 2020). The existence of a cryptocurrency specific to a plat-form seems to create an incentive to secure the integrity of the blockchain. Following the specific types of cryptocurrencies, the integrity of the ecosystem is crucial to reach a network effect and grow the value of the cryptocurrency (Burnie et al., 2018). Blockchain-empowered ecosystems thus rely on an intricate relationship between a blockchain, cryp-tocurrency, and the governance of the ecosystem through a consensus protocol to enable the functioning of the platform without a platform owner.
To conclude, blockchain-empowered ecosystems have been positioned as a distinct ob-ject of research. Situated in media studies focused on the materiality and infrastruc-ture of media technology (Helmond, 2015), this study aims to place the materiality of blockchain-empowered ecosystems centrally as they are seen as complex sociocultural products. Whereas previous studies have analysed the materiality of centralised plat-forms, such as Facebook and Twitter, here the argument has been made that the ecosys-tem surrounding blockchain platforms and apps brings forth new questions regarding the workings of platform mechanisms in a distributed network.
The work by Galloway (2010), rooted in the field of cybernetics and information studies, positions ’protocol’ as the structuring agent that allows distributed networks to become the new power structure in Deleuze (1992) his society of control. Given the introduc-tion of new technologies (blockchain, cryptocurrency and consensus protocols), the ways in which, in Latour (2005) his terms, blockchain-empowered ecosystems re-associate so-cial ties between actors to form new assemblies of programmed soso-ciality (Bucher, 2012) warrants a closer look at the materiality of this new medium.
4 Methodology: How to Navigate Decentralised Platforms
It is necessary to delve deeper into blockchain ecosystems to describe specifically how they work and for whom (Galloway, 2010) to understand how they program sociality (Bucher, 2012). To do so, I will employ a technographic study of the blockchain-empowered ecosys-tem Steem. Through the application of technography (Bucher, 2012) I will describe the technicalities of Steem in a structured way to develop a better understanding of the ways in which these protocols interact. My research approach is an addition to a long line of empirical research as I will use my observations as the data for this study. I will analyse this data through a qualitative lens to describe the experiences that I have encountered.
4.1 Technography
Technography is a descriptive-interpretative approach to describe and observe technology in order to study the interplay between different actors (human and non-human) (Bucher, 2012). It can be traced back to the ethnographic approach, and still uses many of its tools, but the difference between technography and ethnography is that the former is focused on how norms and values become evident in technology (Bucher, 2012). To study blockchain ecosystems using the technographic approach, in other words, is to study how different actors interact and form connections through the specific protocols of this software-mediated space.
Bucher (2012) argues that the technographic approach follows the Grounded Theory (GT) approach. As the reflections that might be developed into theoretical considerations come from the collected data, rather than presupposed hypotheses. This study is rooted in the GT approach as well since I will use ’systematic, yet flexible guidelines’ for the collection and analysis of qualitative data to develop theories ’grounded in the data themselves’ (Charmaz, 2006, p. 2). These data will consist of what Charmaz calls ’extant data’ (Charmaz, 2006). In other words, the data that I will analyse has been created without the researcher having any sort of influence over the data. The relation between the researcher and the data is thus different for extant data than for data that was initiated by the researcher (Ralph, Birks, & Chapman, 2014). Whereas elicited data presupposes
some sort of interpersonal relation between the researcher and the participants, extant data could mean any data that is of interest to the research but created for another purpose than the research (Charmaz, 2006).
Consequently, I will not rely on elicited data from interviews or focus groups to describe the practices found between the different actors. I have chosen to do so because with blockchain apps it is very well possible that there is no single actor responsible for a certain form of extant data. The underlying blockchain technology relies on an open-source nature to garner some level of trust between the blockchain and its users. Not only are all of the interactions on the blockchain visible, much of the source code is also open-source and developed by a community of loosely-tied developers and enthusiasts. Finding the right participants for either a focus group or interviews might prove to be very difficult, if not impossible
This means that the ’black-box issue’ issue, the closed nature of application software, that Bucher (2012) identified will be less of an issue as the source code of blockchain apps is available for both developmental purposes as well as scrutiny. Thus, I argue that relying on elicited data might prove to be difficult as these technologies develop further. Therefore, I will use the plethora of extant data available, in the forms of documentations, white-papers and source code, among other things, to describe how the different actors interact through protocols.
Technography is in many ways similar in approach as Actor-Network Theory (ANT). ANT is a methodological approach stemming from science and technology studies which argues that social forces do not exist and thus cannot explain social phenomena (Bucher, 2012). Instead, everything exists in shifting networks and relationships which, through description, can be identified as social activities. She also puts forward the standpoint that nonhuman actors are actors, too. In this sense, it puts software on equal terms with the users using the application software. However, technography aims to go beyond ANT in its efforts to bring closer social and technological research and more strongly reconsiders the long-held sociological assumption that technology is a ’dead’ space (Kien,
2008). Instead, Bucher (2012) and Kien (2008) both emphasise technology as active and capable of defining norms, values and practices.
The technographical approach is thus situated as a way to describe the role of technology in the co-action of both human and nonhuman actors. It is rooted in a an interpretativist strand of social research and closely associated to the work of Latour and ANT. This methodological approach situates software as suggestive of certain things (Bucher, 2012) and attempts to describe what it is, does and what ties it creates.
4.2 Making a Case: The Steem Ecosystem
I have chosen the Steem ecosystem as my case study as it is currently the only blockchain ecosystem that has hosted a ’successful’ social media blockchain app. Although the term successful is rather subjective, SteemIt arguably is successful in comparison to other blockchain apps as it has partly overcome the major obstacle in the industry: user adop-tion.
SteemIt has over 1 million registered users which dwarfs other blockchain apps in terms of user adoption. Although there are certainly arguments to be made for other cases, I was first and foremost intrigued in why this blockchain app did succeed in providing both scale and usability to users in a blockchain-based ecosystem. Although SteemIt is still a fringe social medium in comparison to centralised platforms such as Facebook, it nevertheless provides an opportunity to understand how sociality is organised in a blockchain-based ecosystem that is inhabited by a relatively active and sizeable user-base.
4.3 The Analytical Protocol
In order to adhere to the technographic method, I will now describe the three phases that I went through before and during the case study. The first phase was the note-taking phase. During this time, I collected notes on everything that I could find that was related to the case study. I read, searched, listened viewed and listened to many different documents from different sources. I was becoming acquainted with the neighbourhood, if you will (Mackenzie, 2006).
A broad range of documents were used. I predominantly used white- and bluepapers to get an understanding of how the original developers of the Steem blockchain envisioned the technologies and features they embedded in the architecture. These papers provide both technical details as well as imagined opportunities for investors, HODLers (’Hold on for Dear Life’), end-users and developers (Steem Whitepaper , 2017). I also read the documentations designed for the developers and the comments embedded in the source code of the Steem ecosystem to gain an understanding of how these features and function-alities were imagined to be used. These documentations show, for example, how APIs give access to certain information stored in the public blockchain and how developers could best make use of these options.
Furthermore, I extensively read blog posts by users of SteemIt on SteemIt and Medium. I found these sources to be insightful as they provide both me and other end-users with a less technical explanation of what is ’going on’. These authors describe certain mechanics of the platform, new updates or why certain things are (not) a good development. To better understand how the developments surrounding Steem and SteemIt were taken on in a broader context, I relied on journalistic pieces which explain the system to laymen. Finally, over the past few months, firstly out of personal interest and later as preparation for this thesis project, I followed several courses on the blockchain technology and the development of cryptocurrencies. This later proofed useful both in reading literature as well as understanding and describing the technicalities of these systems.
In terms of the note-taking process, I started out reading and searching without a pre-defined goal. I started very broadly to get a sense of the different components that make up a blockchain ecosystem. Whenever I found a quote that caught my attention or that raised questions, I copied these notes to a Word file. I then took some notes on why I found this particular quote interesting. Gradually, I started reading more on the specific Steem and SteemIt ecosystem.
But, at some point, one has to make the transition from taking notes to sorting the notes. I made the transition when I did not find any more information that had not been noted
previously in my note file. However, I will mention that I went back and forth between these phases on several occasions to implement newfound ideas. This second phase is where I grouped similar pieces of information, create preliminary topics and studied the material for any similarities.
The last phase embodied the transition from sorting notes to organising the notes into a certain narrative. During the first phase I was specifically not interested in finding patterns or highlights in the documents. Instead, I focused on exhausting the sources that I found and collecting as much information as I could. The second phase was where I started to look for patterns. I specifically did not delete any information as one could not know in advance which information might proof to be resourceful. Then, in the last phase, I became more focused on seeing how all these preliminary topics revolved around this one word that I found returning so often in the documents: incentive.
4.4 Methodological Considerations
The technographic method is highly flexible and thus particularly suited for the temporal nature of software (Bucher, 2012). It does, however, lead to several considerations regard-ing the method and the way it was carried out durregard-ing this thesis project. Firstly, I have made the decision to focus solely on extant data to understand the workings of software in the Steem ecosystem. However, given the work done in, for example, the fields of hacker culture (Maxigas, 2017) and social media (Bucher, 2013, 2017), elicited data could bring forth very interesting perspectives on programmed sociality within decentralised social media ecosystems. I therefore would like to invite other scholars to take on this challenge to understand the people and their rationales behind the Steem ecosystem.
Secondly, I intentionally started out with the idea to analyse the blockchain-empowered app SteemIt to gain an understanding of how the technical structure underlying this app reassembled sociality. However, it became apparent that in order to do that, I needed to delve deeper into the wider Steem ecosystem to better understand the ebbs and flows of this ecosystem. This meant practically that I had to thickly describe more functions and
features than I could fit within this thesis project. I have thus been unable to describe all intricacies of the Steem ecosystem to the extent that I would have wanted beforehand. Nevertheless, by focusing on the key technologies, the addition of this thesis project to the current understanding of platforms and sociality still stands.
Lastly, partly because of the reason mentioned above, I was only able to analyse one case study. The differences between the current blockchain-empowered platforms are plentyful and worth scholarly attention. Yet, I was only able to focus on the current largest Steem platform. A comparative component could have substantiated this thesis project by adding a deeper understanding of the terms that I have introduced following the case study.
5 The Tokenised Community: A Case Study of the Steem Ecosystem
Our strength is the personal contacts we have with each other. We are not just strange looking addresses on a blockchain, we share more than just num-ber transactions, we share ideas, beliefs, and fellowship with one another (Blocktrades, 2020).
5.1 The Tokenised Community
One late afternoon in early May, 2016, a user named "bytemaster" published a thread on the bitcointalk.org forum titled "Steemit.com: Blogging is the new Mining" (Bytemaster, 2016), a title that nods to the success of the Bitcoin. Yet, this was to introduce something that was to replace the Bitcoin as the most well-known cryptocurrency and blockchain system out there. This ’something’ was a "decentralised and incentivised social media" (Bytemaster, 2016) called SteemIt built atop the Steem blockchain ecosystem. The thread promised to introduce a social media where the "community rewards individuals for their posts, comments and votes", according to Bytemaster (2016). This effectively means that users are rewarded a micro-sum for each interaction (post, comment or vote) that is recorded on the Steem blockchain connected to this social media. For the first time, users are rewarded with cryptocurrencies for their interactions.
This model for rewarding all participants is highly debatable in contemporary platform ecosystems. As was discussed in the literature review, the introduction of blockchain technology, cryptocurrencies and different consensus models present major changes to the way in which platform mechanisms of datafication, commodification and selection, in-troduced by Van Dijck et al. (2018), program sociality. Each part of the Steem ecosystem plays a specific role in this novel way in which programmed sociality is organised, which I will refer to as a tokenised community.
I define a tokenised community as ’all actors (both human and non-human) vested in a cryptocurrency native to a platform assembled through a blockchain ecosystem’. The actors, in the Steem ecosystem, refer to both the specific technologies of blockchain
technology, cryptocurrencies and a consensus protocol (Steem Whitepaper , 2017) that underpins the ecosystem as well as the different human actors that play a role in the ecosystem. It it through this relational understanding of software and human agency that the tokenised community can be organised as it is. By enabling ongoing interactions, human actors become vested in the community surrounding the Steem ecosystem. Human actors thus form a relationship with the ecosystem that repays them in influence and cryptocurrencies but through a protocol that ties them to the platform.
Whereas centralised platforms obfuscate the extent of their data collection (Van Dijck et al., 2018) and restrict access to their data collections (Rieder & Sire, 2014), the blockchain allows transactions without a centralised entity to verify the transaction and therefore has to be open-source to ensure that actors can audit the information stored in the blockchain (Cai et al., 2018; Casino et al., 2019; De Leon et al., 2017; Lydiate, 2018; Saleh, 2020). In turn, it is argued that this blockchain technology creates a regime which has sovereign control over the ’truth’ from which cannot be deviated (Leistert, 2018). The blockchain thus presents a substantial building block in a blockchain-empowered platform as it allows the formation of a shared, public ledger of immutable information from which actors can interact without intervention of a central entity (Wüst & Gervais, 2018).
As there is no central entity that can leverage the commercial potential of the Steem platform, the platform mechanism of commodification (Van Dijck et al., 2018) seems to be primarily organised through the introduction of a cryptocurrency native to the plat-form ecosystem. These cryptocurrencies can be seen as a more democratic approach to commodifying a platform as they intend to reward all that involved without a central authority (Burnie et al., 2018), as opposed to the commodification strategies of
person-alisation, data services and fees used by centralised platforms (Van Dijck et al., 2018).
This approach to commodifying a platform through the introduction of a cryptocurrency could thus be seen as a fourth commodification strategy.
These cryptocurrencies can only be introduced in an ecosystem which uses a block-chain system, which already highlights the strong relation between these two technologies.
