=Paper=
{{Paper
|id=Vol-2398/Paper6
|storemode=property
|title=Quasi-Social: Software as the ‘Social’ in Socio-Technical Design
|pdfUrl=https://ceur-ws.org/Vol-2398/Paper6.pdf
|volume=Vol-2398
|authors=Mariusz Nowostawski,Christopher Frantz
|dblpUrl=https://dblp.org/rec/conf/ecis/NowostawskiF19
}}
==Quasi-Social: Software as the ‘Social’ in Socio-Technical Design==
https://ceur-ws.org/Vol-2398/Paper6.pdf
Proceedings of STPIS'19
Quasi-Social
Software as the ‘Social’ in Socio-Technical Design
Mariusz Nowostawski[0000−0002−2809−8615]
Christopher Frantz[0000−0002−6105−8738]
Norwegian University of Science and Technology, Norway
{mariusz.nowostawski,christopher.frantz}@ntnu.no
Abstract. In traditional socio-technical system design, we typically discuss three
core layers: the social, the technical, and the socio-technical layer. The social
layer represents human aspects, the technology represents the advancements in
software and technology development, and the socio-technical layer captures the
interplay between the social systems and the technology-enabled or technology-
mediated interactions. The socio-technical research programme responds to this
pattern by integrating these layers and focusing on the interplay between social
and technical. However, modern peer-to-peer technology, cryptography and en-
cryption protocols together with decentralised technology enable mixing and the
interplay of various layers, and the emergence of other, novel intermediate lay-
ers that are not well-captured, or at least not systematically identified, by the
traditional methods of socio-technical design. There is a shift of software, with
code and automation penetrating places that have, historically, been moderated
by social aspects. This shift reflects a continuous pattern that reflects the abil-
ity of software to take over increasingly sophisticated tasks previous occupied
by human actors. The resulting secondary order of complexity is insufficiently
captured by traditional methods and blueprints for socio-technical systems. As a
result, modern technology-mediated social systems form complex dependencies,
in which the social gradually shifts towards technical, and software is replac-
ing other components. When software starts crowding out traditionally human-
mediated (social) tasks, what can be used to model and design these next genera-
tion systems? The purpose of this paper is to highlight the need for new models
and metaphors that could be used to help in design of complex contemporary
socio-technical systems.
Keywords: quasi-social · decentralised · blockchain · governance · sovereign ·
anonymity · autonomy
1 Introduction
A socio-technical system can be thought of as a social system operating on a technical
base. It is a system that is both: social and technical. Traditional socio-technical system
design [2] focuses on the social aspects, and how the social constructs can be designed,
or built-in into the technology fabric. The traditional 4-layer model, explained by Whit-
worth and Ahmad [15], depicts mechanical/technological aspects as the base, followed
©Copyright held by the author(s) 42
Proceedings of STPIS'19
by the information layer, on top of which an HCI/Personal layer is constructed, to culmi-
nate in the social layer on top. The approach claims: “Whether electronically or phys-
ically mediated, a social system is always people interacting with people. Electronic
communication may be virtual, but the people involved are real.” [15], section 24.1.7.
This might have been true for a broad range of systems, however, there is a growing
category of systems for which this is no longer the case. The interaction patterns are
no longer as clear cut: people are interacting directly with people, other autonomous
systems and institutions. With institutions we refer to any organisation enacting cen-
tralised regulatory functions. Those can be also done in software [6]. Institutions are
interacting with other institutions, too. With the advent of AI and autonomous systems
is gets increasingly complex.
The relevant components of modern socio-technical design are: humans that are the
actual social (human) element of the system; the technology, that provides the means of
interactions and the infrastructural support; and finally the software artefacts (code [7])
that act autonomously on behalf of users, individually, or on behalf of a group. With the
growth of AI and autonomous systems, the software artefact can also, potentially, act
on behalf of itself. We refer to those layers respectively as socio-, technical-, and quasi-
social layers. Those layers interact with one another, influence one another, and co-
evolve. The modern information systems that provide means for value transfer, social
organisation and engagement utilise decentralised democratic governance mechanisms,
run without a single point of control, provide consistency under various distributed trust
models and act autonomously, not requiring human control or external interventions,
beyond the initial design of the system. For this reason, such systems are sometimes re-
ferred to as allegal as they operate in the zone of unregulated social systems and those
traditionally regulated by legal systems. The systems that mediate human interactions,
organisation and value transfer allow human to interact and engage in contractual agree-
ments anonymously or pseudo-anonymously, both, with other humans as well as with
software artefacts and computational systems acting on behalf of other humans, other
institutions, or on behalf of itself. The existing socio-technical design methodologies are
not well-equipped to model and represent the new challenges related to this technologi-
cal progression. We lack proper terminology, abstractions, and foremost, methodologies
to research and analyse those new social systems that are intertwined within this quasi-
social layer. Some of the early experiments with the autonomous anonymous systems
have highlighted the importance of the systematic approach to addressing governance
challenges, such as decision-making, information signalling, and consensus protocols
as well as the actual evolution and maintenance of the system itself. Those experiments
have also highlighted the need to conduct experiments in the real world, since the be-
haviour patterns observable in the real world are a response to the interaction with the
system – driving the need for action research [11]. In addition, there are a number
of other challenges such as the contextualisation of legal jurisdictions, the concept of
sovereignty, identity and enforcement, both, within the system itself as well as in the
wider, social or legal contexts. In this work, we focus on the new developments in the
anonymous peer-to-peer space and discuss their influence on the designs and modelling
of socio-technical and quasi-social systems. For illustrative purposes, we will apply
those ideas to advanced decentralised systems in the financial sector.
Edited by S. Kowalski, P. Bednar and I. Bider 43
Proceedings of STPIS'19
2 Trust
Traditional financial, insurance and governance institutions work through a trusted, cen-
tralised, controlled and monitored organisations established within particular national
boundaries and bound by the national and international rule of law. These centrally-
managed institutions consist of increasingly complex, monolithic computational sys-
tems that are difficult to adapt to new, growing and changing requirements. In fact, the
implicit lack of transparency regarding operational details relies on sustained trust into
their functioning – trust in humans that constitute the fundamental elements of those in-
stitutions. This gets increasingly complex in the context of multi-national jurisdictions,
often conflicting legal requirements, cross-border taxation, political instability, privacy-
preserving and data protection directives such as the new European Union GDPR pro-
visions. Similar problems are also present in the context of large-scale cross-national
social systems in which software algorithms that are primarily used to determine and
enforce operations and constraints on the actual workings of the system itself. The com-
plexity of the social systems mediated by technology becomes impossible to be handled
by traditional means and the detailed workings of those complex systems become ex-
tremely difficult, or impossible to trace. This is particularly true for situations in which
decisions within the system are mediated through machine learning black-boxes.
Good examples of that complexity are modern financial systems. On the one hand,
such systems are regulated and monitored by the complex legal frameworks that, in
large, is agreed upon internationally. The regulations safeguard, to some extent, the ac-
tual workings of the system, but they also limit the innovation and customer choice. On
the other hand, the complexity of financial instruments offered by the incumbents makes
it actually impossible for full control and predictability of the resulting emerging prop-
erties of the system itself. The history of financial crashes provides us with a plethora
of examples. The next financial crises cannot be predicted or prevented, partially due
to the human elements that influence it, but partially, due to inherent self-dependencies,
built-in feedback loops and internal mechanisms that give raise to untraceable emer-
gent properties. Therefore, traditional systems lack transparency and exhibit strong re-
sistance to analysis and predictability. The regulations have limited control over the
potential misuse/abuse. Nevertheless, the systems require high levels of data disclo-
sure from the participating individuals, an aspect that can be in conflict with personal
freedoms and private data protection directives.
Due to the problems above and with the help of technological innovations a new
class of systems is emerging, which exhibit qualitatively novel properties. The ideas
behind distributed consensus and ledger technology have gained significant traction
since the launch of the first public experiment known as Bitcoin in 2008. Multiple ex-
amples of decentralised technology have been deployed and successfully used for value
transfer, digital assets, document validation, and crypto-currencies. Despite these ex-
periments computer and social scientists are in the early stages of understanding the
impact this technology has, and is bound to have, on social institutions and organ-
isations more generally. Emerging distributed institutions that utilise the blockchain
technology are able to facilitate international interactions, contracts, and value transfer,
all of which can be achieved without the need for the human-based third-party trust,
central authority, or externally managed audits. Moreover, those interactions can be au-
©Copyright held by the author(s) 44
Proceedings of STPIS'19
tomated and conducted autonomously and anonymously by the distributed peer-to-peer
networks. In other words, on the highest level, layer 7, or application layer, it is not
communities interacting with communities. It is software interacting with software, on
behalf of users that provide only implicit regulatory mechanisms and policies within the
self-governing computational peer-to-peer system. Those innovations have a significant
impact on the future structure of our social and economical environment. We argue,
that due to the properties and the design, anonymous, privacy-conscious decentralised
technologies represent a qualitatively new and disruptive change in the construction and
structure of traditional socio-technical systems. It also provides novel ways to manage-
ment, governance, conflict resolution and the societal organisation in general, which
are directly relevant to the class of systems traditionally investigated by socio-technical
research methodologies. To argue our point, we will explore both the technological and
organisational foundations of blockchain technology. We will do so by first highlight-
ing historical examples for information and communication systems in Section 3, be-
fore discussing the corresponding characteristics in blockchain technology in Section
4. This is followed by a discussion of specific applications of blockchain technology
that are dissected along specific layers in Section 5.1 to provide a basis for a compara-
tive analysis of different blockchain-based applications in Section 5.2. We conclude the
discussion on the refined perspective in Section 8.
3 Peer-to-peer and decentralised systems
The Internet is one of the largest multi-national projects that we, as humanity, have
been engaged with. It has enabled new and innovative ways for the social organisa-
tion as well as communication. It has provided an unmatched repository of knowledge,
training, and various information-centric or knowledge-centric business models. The In-
ternet itself represents “technology” in traditional socio-technical design, yet, one feels
there is something missing. A simple Word Processor can be thought of as a technology,
as a tool, that facilitates a class of interactions that are build on top of it. Yet, systems
such as Facebook, which are part of the Internet fabric are in themselves socio-technical
systems, in fact require a recursive application of socio-technical design principles as
part of the system analysis, let alone such systems’ further constituents. This exempli-
fies a break of the idealised hierarchical design of 4-layered socio-technical systems as
argued by Whitworth and Ahmad [15]. We need something new, to capture both, the
recursive nature of the inter-dependencies of the system itself, as well, as the funda-
mental inter-dependencies of the social, technological, and software artefacts that are
the fundamental building blocks of the complex systems we deal with today.
The fundamental design of the Internet protocols is inherently distributed and peer-
to-peer, with centralisation occurring only in places where it is aligned with the geopo-
litical organisation of the actual physical world. For example, Domain Name Services
are hierarchically organised with country root domains managed by the central services
of a given country. Nevertheless, ownership and rights are distributed among multiple
entities, even on the top-level of this hierarchical structure. Similarly, all core protocols
such as SMTP (for mail delivery), or HTTP (for web content delivery) are inherently
peer-to-peer-like, due to the nature of the underlying TCP/IP layer. This peer-to-peer
Edited by S. Kowalski, P. Bednar and I. Bider 45
Proceedings of STPIS'19
environment and design gave rise to a large number of innovation and a plethora of new
services, that were not possible before. The standardised way of communicating en-
abled innovation and exploration of possible interaction patterns and organisations that
human groups could form freely. The early days of the Internet could also be referred to
as alegall, as the groups organised spontaneously and followed their own codes of con-
duct, often across borders (e.g., topical forums, social networks). This dynamic process
continues today too, even though many of the core services are dominated by few large
stakeholders, such as Amazon, Apple, Facebook, Google and Microsoft.
Some of the most complex computer systems are designed in a peer-to-peer, evolu-
tionary fashion, through something that is referred to as the bazaar model [13]. Linux
or FreeBSD kernels, for example, are the most advanced and widespread operating sys-
tems kernels. They have not occurred within the traditional, top-down management but
rather in an open, dynamic, and constantly changing and adapting workplace, that is in-
creasingly often virtual. Peer-to-peer and decentralised systems are essential to facilitate
unconstrained innovation and exploration on a large scale. This is one of the main rea-
sons why decentralised technology is re-shaping the communication paradigm. Build-
ing on this idea and encroaching functions of the social-coordinative realm, permis-
sionless open-source blockchain developments and blockchain-based systems change
the way liability, trust and ownership are handled, aspects of which we discuss in the
following.
4 Absence of central authority
Decentralised technology relies on an agreement of a system state achieved in a situ-
ation without central authority and with potentially hostile and fraudulent actors. The
technology, through an interesting interplay of incentive system, automation and distri-
bution of power, allows achievement of an agreement (consensus) on the system state,
and system state record. In essence, a public blockchain technology is solving the con-
sistency problem, that is, ensuring a consistent indisputable representation of state and
transitions outside of the control of either single stakeholder. The consistency of the
events log is assured by aligning the incentive model with the goals of the distributed
network of peers. In this context ‘public’ implies that blockchain applications operate
in the open public sphere and coordinate interaction between unknown participants in a
permissionless fashion, i.e. anyone can participate.
In the absence of a central sanctioning authority, blockchain modifications (i.e., trans-
actions) need to be cheap enough not to discourage the system’s use, yet expensive
enough to prevent opportunistic abuse (e.g., by submitting fraudulent transactions).
Mechanisms that facilitate this trade-off include the consumption of high amounts of
processing power or per-transaction payments or mechanisms involving stake that can
be lost when abuse is detected. This balance of incentive and deterrence is the hard
socio-technical challenge. The best mechanisms to-date rely on so-called proof of work [5].
In that model, a cryptographic riddle is posed and requires a provable amount of time
spent on computation, to be resolved. Validating the riddle result is easy, solving the
riddle can be made arbitrary hard.
©Copyright held by the author(s) 46
Proceedings of STPIS'19
An alternative approach that avoids the inefficiencies associated with the proof of
work, such as wasted power and processing time, as well as to limit the computational
‘arms race’ for computing power, is the proof of stake. In the proof of stake [1] the
individual participants’ influence is constrained by their commitment to the system,
such weighing the influence of the number of resources individual participants hold.
Naturally, this introduces hierarchical characteristics into the system, but increases the
efficiency of the system without unproductive use of computing resources. Whatever the
specific protocol employed by a given blockchain implementation, the proof of work,
proof of stake, and the voting model used for validation work in unison; the stable long-
term strategy is not to cheat. Decentralised blockchain technology offers third-party
trust without any single entity taking full responsibility or having full authority.
To discuss novel artefacts of blockchain technology, we need to first understand
some of the architectural underpinnings – which we systematically highlight in the
following section.
5 Architectural Layers of Blockchain Systems
5.1 Layer 1: Base layer
The decentralised technology can be seen as consisting of two fundamental layers.
Layer 1 facilitates the consensus and transactions sub-system. They represent the core
functionality. Layer 2, ie. any protocols build on top of Layer 2, provide additional
facilities and can provide application layer logic.
One of the core Layer 1 technology today is Bitcoin. The creator of the system,
known as Satoshi Nakamoto, wrote about the system in a founding white paper [10].
The global network of miners and users is one of the largest and most powerful com-
putational resources currently in operation. Bitcoin is a good example that highlights
the main components of a blockchain-based ecosystem. It comprises of software de-
velopers, that are either paid by the users owning the virtual currency, or own the cur-
rency themselves, the miners, and the node operators. Bitcoin operates with pseudo-
anonymous identities. There is no reputation subsystem, and the transactions are safe-
guarded by proof-of-work mechanism. Participation is encouraged through mechanisms
of incentives, including e.g. mining rewords.
To address some of the shortcomings of the original Bitcoin structure, alternative
currencies have emerged. One example for this development are DashCoin, whose
structural characteristics we will compare to Bitcoin, in order to disambiguate blockchain
technology from specific applications built on its principles. DashCoin uses the original
codebase for Bitcoin node, however, it changes some of the core fundamental mech-
anisms. DashCoin introduced the notion of masternodes, that provide a different, hi-
erarchical structure to the consensus mechanism. It also provide built-in mechanism
for so called Private-Send, which is equivalent to the CoinJoin protocol in the Bitcoin
blockchain [8]. Those mechanisms obfuscate the source and destination of transactions
to form long complex chains of ownership that is difficult to analyse. The mechanism
also provides provable deniability.
Edited by S. Kowalski, P. Bednar and I. Bider 47
Proceedings of STPIS'19
5.2 Layer 2: Application Layer
Fully anonymous, atomic, and reliable peer-to-peer transfer of value is one of the most
common examples of the blockchain technology application. That can be built-in into
Layer 1, and all of the existing blockchain systems have a native built-in currency.
Bitcoin, Ether, Dash, and many others crypto-currencies operate directly on Layer 1.
However, new assets or digital currencies can be developed and provided on Layer 2. In
fact, Ethereum provides a formal specification through ERC-20, and as of April 2019,
over 180,000 ERC-20-compliant tokens are found on the Ethereum network. Most of
the stable coins (crypto-currency with the value pegged by one of the existing FIAT
currencies, such as EURO or US dollars) are also using Ethereum as Layer 1. As an-
other example of a blockchain-enabled application, consider a simple escrow service.
Typically, an escrow service is used to ensure atomicity of a transaction between two
non-trusted entities, and to have the ability to roll back a partially fulfilled transaction.
An escrow service, a trusted third party is used to work as a trusted intermediary to
facilitate the transaction. With the blockchain, such transactions are atomic by design,
without the need for a trusted third party. Escrow services, or decentralised exchanges
are now possible through the mechanisms that allow to conduct Atomic Swaps. Those
are available on many blockchains and sit somewhere between Layer 1 and Layer 2.
Atomic swap allows to automate the process of creating escrow services based on sim-
ple smart contracts.
What those examples demonstrate is that many centrally-managed services, in par-
ticular those provided by insurance companies, banks, or governments, can be made
more secure and more transparent with the use of blockchain technology. This means
that the human element can be eliminated from selected institutions or contractual
agreements, especially in areas in which the ability to maintain accountability is chal-
lenging. This has a fundamental impact on how we will perceive and deal with fraud,
data leaks or power abuse. This potential and the associated challenges become clearer
when exploring examples of blockchain technology with respect to structural and gov-
ernance characteristics. We thus use the following subsections to highlight some ex-
amples of blockchain technologies, so-called blockchain applications, in an attempt to
illustrate the sketched potential.
6 Blockchain Applications
6.1 Lightning Network
The Bitcoin Lightning network is a mechanism to overcome the inherent scalability lim-
itations, in terms of transaction throughput, of traditional Bitcoin blockchain systems.
Due to the required consensus mechanisms and decentralised nature of open blockchain
systems, new entries in the ledger (transactions) can only be recorded with a predefined
step-like mechanism, that limits the per-second transaction throughput. One of the so-
lutions to this inherent limitation is a technique based on off-chain transactions. It is
based on the payment channel concept, and it relies on the idea that not all transactions
need to be recorded in the ledger. Only those that establish payment channels between
participants and those that resolve, or close the payment channels (channel closure).
©Copyright held by the author(s) 48
Proceedings of STPIS'19
The actual payments within the channel can be done without leaving any trace in the
blockchain proper, and instead happen off-chain (hence the name, off-chain transac-
tions). For this to work, however, additional guarantees need to hold for the participants
to limit the possibility of abuse or misuse of such a system. The Lightning Network Pro-
tocol addresses this by the existence of additional services, so-called Watchtowers [4].
Watchtowers will provide additional guarantees and triggers necessary for the system
to function properly. There is, however, limited research of how they would operate,
how they would be conceptualised and implemented, and how information they reveal
will be made available to the general public as well as to the system participants – and
last but not least, how we can devise mechanisms to react to irregularities identified by
watchtowers. This collection of open issues offer a fertile ground for research, and our
study of quasi-social systems as discussed in this paper targets precisely this unexplored
niche.
However, this approach is not limited to this precise example, but also existing ap-
plications with tangible economic impact, such as prediction markets.
6.2 Prediction markets
Let us consider a system in which traders can trade virtual asset that represents the out-
come of the future event. Consider it as information or decision markets, idea futures,
or event derivatives. A market is created for the purpose of trading the outcome of an
event. The outcome is binary, therefore, the virtual option will expire at the price of 0%
or 100%. A prediction market contract trades between 0 and 100%. Prediction markets
can be treated as crowdsourcing information, with the main purposes of eliciting ag-
gregating beliefs over an unknown future outcome. Traders with different beliefs will
trade on contracts whose payoffs are related to the unknown future outcome. Then, the
market price of the contract is considered as the aggregated belief. Markets like that
can be used for risk assessment or risk hedging in order to establish a likelihood of a
future event, market value, future market value, and so on. It is stipulated that three
necessary conditions need to hold for such markets to function well: diversity of in-
formation, independence of decisions, and decentralisation of organisation [14]. It is
exactly this third property that makes blockchain-based, smart-contracts driven predic-
tion markets an appealing value proposition. There is one ongoing experiment started
in 2014, Augur [12], that provides an open prediction market. However, one of the most
severely limiting factors is that those markets can be subject to manipulation [3]. Sim-
ilar to the challenges outlined for the lightning network and its trust-based operation
on sidechains, such markets suffer the same problem of intermediaries detecting and
addressing patterns of illegitimate behaviour – without being confined to monitoring
merely technical aspects of the system, but clearly operating at the intersection of so-
cial and technical activity, thus assuming the proposed quasi-social role. Coming back
to the specific example, this real-time monitoring of the events and manipulation de-
tection is something that the aforementioned Watchtowers could provide by facilitating
the necessary feedback loop and ensuring the system’s self-regulating properties. To
the best of our knowledge, there is no active research on Watchtowers for this particular
type of systems.
Edited by S. Kowalski, P. Bednar and I. Bider 49
Proceedings of STPIS'19
6.3 Autonomous Loan Dispenser Systems
Let us consider an autonomous automated secured loan system. In this scenario, the
borrower needs to borrow money, that she promises to pay back at a certain time in the
future. The security of the loan is based on the collateral, that is a property or assets,
conveniently represented electronically, that a lender accepts as security for a loan. The
borrower must first obtain an appraisal with the estimated fair market collateral value
of the property to be considered in the loan - this process itself requires expert opin-
ion and delegating it to an open market, such as the above example of the prediction
market, provides certain benefits. While the collateral value is an important component
of the loan, the loan system should also use the information on a borrowers credit pro-
file and credit history. In a traditional system, the value of the collateral, collection of
the credit information and assessment are left to the institution offering a loan. Typi-
cally, it is a bank. The market value of the collateral needs to be checked on a regular
basis against the market value and the loan value, such that the loan never exceeds a
certain threshold (50-90%) of the current market value of the collateral. In our case,
we could consider a blockchain-based system based on Maker tokens (MakerDAO,
http://makerdao.com/en/). It is a smart-contract-driven system based on Ethereum net-
work, that provides loans based on electronic collateral. The value of the loan has been
established to be at a maximum of 50% of the collateral value.
If the value of the collateral drops below a certain safety margin, the lender would
not be secured anymore against the failure of the loan repayment. In such a case, a safety
trigger needs to force the borrower to repay the loan, or, the collateral would need to
be re-sold on an open market as to provide the necessary security for the lender. In this
simple example, the lender actually is a set of autonomous smart contracts, that execute
the predefined logic of securing (freezing) the collateral (e.g. cryptocurrency or some
forms of digital assets, like house ownership) such that the loan system does not need a
human institution to facilitate the financial service. It is self-regulating and autonomous.
However, for its operation there needs to be an autonomous, real-time monitoring entity
that checks which of the borrowers agreements trigger the safety event. This form of
monitoring ensures and replace the role of financial auditors that verify and validate that
the system operates as prescribed. In the system, the same as in the previous two exam-
ples, this is left as non-colluding 3rd party service, such as the quasi-social Watchtower
concept highlighted before, that provides meta-layer on top of the base-layer protocol.
7 Discussion & Proposal for Integration
7.1 Discussion
As we can see from the provided examples, all monitoring functionality necessary for
developing open self-organising real-world applications – as currently advocated in the
form of blockchain technology – rely on technology that is, of course, technology, but
can no longer be clearly associated with a technical perspective, but rather sits in a layer
that operates in concert with the social affordances a system provides. However, those
can often only be observed at runtime, and worse, can change over time – preventing
©Copyright held by the author(s) 50
Proceedings of STPIS'19
these intermediary entities from being considered at design time only, but rather con-
sidering their change and evolution over time.
To make a simple real-world analogy, imagine that a certain institution is formed,
with the rules of how it operates, e.g., a bank. Then, a set of meta-layer institutions (such
as auditors, regulators and law-enforcement) need to be built on top of the base layer to
facilitate the proper operation of the institution itself. This is exactly what Watchtowers
are for. However, until today it is unclear how such systems of quasi-social nature are
embedded into (or within) socio-technical systems. The nuanced characteristics are not
explicitly reflected in the traditional socio-technical design perspective.
7.2 Proposed Integration in the Socio-technical Design
To explore pathways towards integrating these perspectives, let us provide an overview
of the aspects discussed in the context of explicit examples. Figure 1 depicts the tra-
ditional stratification into the social layer at the top, capturing all the characteristics of
human participants. At the bottom, we can see the technological layer that involves all
the structural aspects discussed in Section 5.1.
Fig. 1. Three-layer Quasi-Social model
The novel characteristics with coordinative function on both the social and tech-
nical layer, we propose, needs to be conceptualised on a novel quasi-social layer that
can both involve the observation and moderation of social interaction by autonomous
and human entities, but primarily adopting functions that in traditional socio-technical
systems would clearly be associated with the social side of the system. This involves
aspects such as Decentralised Autonomous Organisations based on smart contracts, AI
bots, Watchtowers, and more generally, the regulative functions that guarantee, for ex-
ample, compliance with legal regulation and associated liability – activities traditionally
associated with human actors.
Edited by S. Kowalski, P. Bednar and I. Bider 51
Proceedings of STPIS'19
Existing systems for audit and monitoring are focused on transaction tracking and
transaction monitoring. There are existing commercial systems such as Chainalysis, El-
liptic, CipherTrace, as well as a large number of block explorers that provide real-time
or close to real-time transaction visualisation of records from a given blockchain ledger.
There has been also a number of research projects related to transaction analysis [9],
and others. Those focus on transactions, that is, individual recorded ledger entries in the
given blockchain. This is an important and fundamental source of information about the
underlying system, however, this is not sufficient. ontologies to deal with the complex
interactions in the Quasi-Social layer of the complex software systems. Given that the
field is mostly practitioner-driven, it faces unprecedented challenges in terms of social
security and legal compliance. Similarly, the challenges faced by law enforcement agen-
cies, courts and legal experts rely on auditability, forensic readiness, ability to obtain
evidence and transaction and value transfer tracking. Those technologies and require-
ments are new and rely on expertise and tools that effectively are being on the cutting
edge of research and development. Even though the core crypto technologies have been
known for over 25 years and the Bitcoin has been operating for over 10 years, the socio-
technical implications of the various decentralised autonomous system deployments are
not yet well understood. Given this it is even more important to structure and analyse
such systems in a way that faciliates the differentiation of the technological, social and
coordinative quasi-social functionality. The Watchtowers are an example of this. They
provide the necessary feedback loop and mechanisms to validate, verify and audit if the
underlying systems actually behave within the regimes that they have been designed
for.
Sticking with the illustrative examples of Watchtowers, there are significant scien-
tific implications of the higher-order feedback mechanisms for decentralised systems
that operate across all three levels – with respect to technological impact, quasi-social
and societal impact.
Technological. Watchtowers will have a significant impact on how the decentralised,
autonomous and smart-contract driven systems are designed, and how they operate. The
higher-order indicators that will be measured by Watchtowers can be re-integrated into
the fundamental lower-layers such as to offer enhanced self-regulating properties. The
necessary feedback will make the systems more robust, resilient, more autonomous and
self-regulating.
Quasi-social. The Watchtowers will make the decentralised systems more transpar-
ent and robust. Harder to misuse. The Watchtowers will enable better, richer and more
complex services that can be designed in the future with the use of the automated feed-
back mechanisms. This is similar to how Lightning requires and builds on top of the
required Lightning watchtowers.
Societal. Regulators, auditors, traditional law enforcement agencies as well as the
general population can use the results of this project, indirectly, through the feedback
mechanisms that will be established between the underlying technological layer and the
indicators and metrics that are provided by the Watchtower services. This will provide
better and more transparent information spread, ease the operation of the system as well
as ensures that the systems are not misused or abused.
©Copyright held by the author(s) 52
Proceedings of STPIS'19
8 Summary
We have argued that traditional characterisations of socio-technical design are not suf-
ficient to capture the full spectrum of complexity arising in modern systems. We have
argued, that a three-level modelling approach, based on social-, technical- and quasi-
social conceptualisations will lead to a way forward in understanding and subsequently
analysing the new category of emerging systems that involve anonymous, peer-to-peer
networks and complex social-machine interactions that span individual, collective and
institutional layers. We discussed this using the concept of watchtowers as an example
and explored their potential impact across all three layers of the socio-technical design,
as well as all layers from individual to institutional. For both the socio-technical re-
search community, it will be important to consider the aspects thrown into this debate
by the emergence of such novel socio-technical system components, since, one way or
another (i.e., explicitly or implicitly), decentralised coordination technology, such as
evidenced in blockchain technology will be pervasive in any future information sys-
tems. This leaves us with two options: observe the space and risk being sidelined by
the complexity such systems pose, or to actively engage in such discussion in order to
influence how such systems will look like in the future.
References
1. Proof of stake instead of proof of work. https://bitcointalk.org/index.php?topic=27787.0, ac-
cessed on: 1st May 2016
2. Baxter, G., Sommerville, I.: Socio-technical systems: From design methods to systems engi-
neering. Interacting with computers 23(1), 4–17 (2011)
3. Chakraborty, M., Das, S.: Trading on a rigged game: Outcome manipulation in prediction
markets. In: IJCAI. pp. 158–164 (2016)
4. Decker, C., Russell, R., Osuntokun, O.: eltoo: A simple layer2 protocol for bitcoin. White
paper: https://blockstream. com/eltoo. pdf (2018)
5. Dwork, C., Naor, M.: Pricing via processing or combatting junk mail. In: Proceedings of
the 12th Annual International Cryptology Conference on Advances in Cryptology. pp. 139–
147. CRYPTO ’92, Springer-Verlag, London, UK, UK (1993), http://dl.acm.org/
citation.cfm?id=646757.705669
6. Frantz, C.K., Nowostawski, M.: From institutions to code: Towards automated generation
of smart contracts. In: Foundations and Applications of Self* Systems, IEEE International
Workshops on. pp. 210–215. IEEE (2016)
7. Lessig, L.: Code is law. The Industry Standard 18 (1999)
8. Maxwell, G.: Coinjoin: Bitcoin privacy for the real world (2013)
9. Moser, M.: Anonymity of bitcoin transactions (2013)
10. Nakamoto, S.: Bitcoin: A peer-to-peer electronic cash system (2008)
11. Pasmore, W.: Action research in the workplace: The socio-technical perspective. Handbook
of action research 2, 38–48 (2006)
12. Peterson, J., Krug, J.: Augur: a decentralized, open-source platform for prediction markets.
arXiv preprint arXiv:1501.01042 (2015)
13. Raymond, E.: The cathedral and the bazaar. Knowledge, Technology & Policy 12(3), 23–49
(1999)
14. Surowiecki, J.: The wisdom of crowds. Anchor (2005)
15. Whitworth, B., Ahmad, A.: Socio-technical system design. In: Stephanidis, C. (ed.) The en-
cyclopedia of human-computer interaction, chap. 24. Interaction Design Foundation (2012)
Edited by S. Kowalski, P. Bednar and I. Bider 53