The Interplay Between Human and Machine Agency

Abstract

Abstract. Human-machine networks affect many aspects of our lives: from sharing experiences with family and friends, knowledge creation and distance learning, and managing utility bills or providing feedback on retail items, to more specialised networks providing decision support to human operators and the delivery of health care via a network of clinicians, family, friends, and both physical and virtual social robots. Such networks rely on increasingly sophisticated machine algorithms, e.g., to recommend friends or purchases, to track our online activities in order to optimise the services available, and assessing risk to help maintain or even enhance people's health. Users are being offered ever increasing power and reach through these networks by machines which have to support and allow users to be able to achieve goals such as maintaining contact, making better decisions, and monitoring their health. As such, this comes down to a synergy between human and machine agency in which one is dependent in complex ways on the other. With that agency questions arise about trust, risk and regulation, as well as social influence and potential for computer-mediated self-efficacy. In this paper, we explore these constructs and their relationships and present a model based on review of the literature which seeks to identify the various dependencies between them. IntroductionA definition of agency based on the notion of non-deterministic behaviours [1] fails to recognise the increasing variety and complexity of human-machine networks 1 (HMNs) [2], the intention of technology designers [3], and active intervention by bots within social networks [4,5]. The concept of agency is particularly problematic in human-machine interactions [6]. Machine or material agency may be seen as automation, which originally required some tolerance from human agents [7]. But this is no longer true: technology can actively support human activity [8], and manifests increasingly complex interaction types [9]. Machine and human agency may not be the same and yet equally valid [10]; machine agency may be just "perceived autonomy" [11]; and it 1 In the following we use human-machine network and network interchangeably.certainly enables human agency [12]. Indeed, agency may well be becoming a social and group construct where both humans and machines play a part [13,14]; and used effectively, agency may even lead to innovative review of working practice [15].The enabling contribution of machine agents within a network may have an effect on self-efficacy. Bandura's original definition of self-efficacy as an individual's belief in their ability to be able to achieve a given objective [16][17][18] has also been applied to technology [19,20] and its acceptance [21]. There are, however, constraints on the support and positive contribution of technology to human self-efficacy, not least in terms of anxiety and suspicion around technology use [22,23]. This may be further exacerbated by increasing machine animism: it may not always be obvious what machines ar...

Seeing Potential Is More Important Than Usability: Revisiting Technology Acceptance

“…There is a mediating effect, therefore, not of Perceived ease-of-use as predicted by the TAM, but of a willingness to think creatively and collaboratively. Such willingness depends on the perception of affordances which through actualisation seem to increase perceived self-efficacy and agency: potential adopters begin to explore how the technology can support them satisfy the requirements of their own responsibilities [14,32,33].…”

“…The perception and actualisation of affordances is related, we maintain, to a willingness to explore technology both pre-and post-adoption [12]. In turn, this leads to increased self-efficacy and agency [13,14,32]. We propose to revisit other validation activities we have been involved with in the past to re-evaluate those results for a larger cohort and in connection with other technology in a different domain.…”

“…Their experimental cohort is once again confined to students. Pickering, Engen and Walland [14] extended that model with increased theoretical validation, based on literature review only. Elsewhere, technology adoption has been associated with human adaptation to technology defects [15,16].…”

Resolving Stakeholder Tussles in Healthcare Systems: Ethical Challenges to Data Protection

“…In an actor network therefore, trust is less about reliance on regulatory control but rather on how the trustor interacts with the trustee. Where technology is involved, trust may be mediated by self-efficacy and agency [24][25][26]. Trust becomes an "organizing principle" [27] which underlies both the activity across a socio-technical network (between patient and clinician) as well as how it is perceived (by the community) [28][29].…”

Mediated Behavioural Change in Human-Machine Networks: Exploring Network Characteristics, Trust and Motivation

“…That location may expand up to national level, for instance, like the UK government survey site https://petition.parliament.uk/. Independently, the dimensions allow individual aspects of the network to be examined; further, taken within an individual abstract layer, the interplay between the pair of dimensions would allow additional investigation of network dynamics [14,15]. How this dynamism affects the network offers the opportunity to control and manipulate the network.…”