Artificial intelligence as digital agency

Ågerfalk, Pär J.

doi:10.1080/0960085x.2020.1721947

Cited by 105 publications

(57 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While we agree that it is essential to differentiate between different kinds of AI and not simply assume that it ought to entail machine learning, which in itself is a multifaceted concept (Ågerfalk, 2020 ), a detailed elaboration on the specific origins and different angles around this debate is beyond the scope of this study. However, a brief exploration around treatments of AI reveals that the topic has been mainly considered in the context of machine learning applications.…”

Section: Introductionmentioning

confidence: 90%

“…Understood this way, we agree with Fonseka ( 2017 ) that academics are in arrears of holding AI accountable. In particular, as Ågerfalk ( 2020 , p. 5) suggests, “there are good reasons to worry about misuses of AI,” given its potential to perpetuate society’s inequalities and injustices through implicit biases due to race, gender, and sexual orientation (Manyika et al, 2019 ). Insofar as for instance, recently, researchers found that COMPAS, an AI-based recommender software used to assign recidivism scores (and help predict which convicted criminals were likely to re-offend), labeled Blacks who did not actually re-offend as a higher risk at nearly twice the rate as Whites.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Questioning Racial and Gender Bias in AI-based Recommendations: Do Espoused National Cultural Values Matter?

2021

View full text Add to dashboard Cite

One realm of AI, recommender systems have attracted significant research attention due to concerns about its devastating effects to society’s most vulnerable and marginalised communities. Both media press and academic literature provide compelling evidence that AI-based recommendations help to perpetuate and exacerbate racial and gender biases. Yet, there is limited knowledge about the extent to which individuals might question AI-based recommendations when perceived as biased. To address this gap in knowledge, we investigate the effects of espoused national cultural values on AI questionability, by examining how individuals might question AI-based recommendations due to perceived racial or gender bias. Data collected from 387 survey respondents in the United States indicate that individuals with espoused national cultural values associated to collectivism, masculinity and uncertainty avoidance are more likely to question biased AI-based recommendations. This study advances understanding of how cultural values affect AI questionability due to perceived bias and it contributes to current academic discourse about the need to hold AI accountable.

show abstract

Section: Introductionmentioning

confidence: 90%

Section: Introductionmentioning

confidence: 99%

Questioning Racial and Gender Bias in AI-based Recommendations: Do Espoused National Cultural Values Matter?

2021

View full text Add to dashboard Cite

show abstract

“…Similarly, Faraj et al [23] refer to learning algorithms as "machine learning, computation, and statistical techniques .. [that] rely on large data sets to generate responses, classifications, or dynamic predictions that resemble those of a knowledge worker". Prediction, here understood as using information you do have to produce information you do not have [15], then, constitutes a fundamental task of algorithmic decision-making systems and the learning algorithms on which these are based [24].…”

Section: Automating Augmenting and Assembling Algorithmic Decision-mentioning

confidence: 99%

Assembling Algorithmic Decision-Making under Uncertainty: The Case of 'Edge Cases' in an Open Data Environment

Grønsund¹

2021

Proceedings of the Annual Hawaii International Conference on System Sciences

View full text Add to dashboard Cite

Algorithmic decision-making is rapidly evolving as a source of data-driven competitive advantage with important implications for analytical practices in multiple settings. Despite the ambitions for algorithmic and intelligent technologies, however, the requirement for quality data input to the algorithm poses a significant challenge for its actual adoption. The trend towards open data might bring additional challenges such as strategic gaming and distortion of meaning. To address this problem, we draw on a twoyear long qualitative case study of a firm in international maritime trade to understand the role of uncertainty associated with open data upon the uptake of a novel algorithm. We combine an uncertainty and assemblage perspective to unpack the arrangements by which the organization configures relations of humans and machine to mitigate this problem. We highlight the phenomenon of edge cases as a key challenge for automation and propose that an assemblage of augmentation and automation allows a dynamic arrangement that support the introduction and organization of algorithmic decision-making under uncertainty.

show abstract

“…It basically captures the desire to be causal agents of one's own life and act in harmony with one's integrated self. This is an area that is seriously threatened by new technologies (e.g.,robots, autonomous cars, and service bots), that can replace people and decide for them (Ågerfalk, 2020). Thus, there are opportunities to examine important research questions regarding: When and how does technology used by digitised individuals replace people?…”

Section: A Framework For Studying the Interactions Between Individualmentioning

confidence: 99%

An intertwined perspective on technology and digitised individuals: Linkages, needs and outcomes

Turel

Matt

Trenz

et al. 2020

Information Systems Journal

View full text Add to dashboard Cite

Information technology (IT) has changed dramatically over the last several decades. Although, in its early days IT has been mostly used as a tool for conducting business or running complex governmental and organisational operations, it has shifted to also become a productivity and hedonic tool for individual users (Matt, Trenz, Cheung, & Turel, 2019). Readers can reflect on how many technologies surround them now, as they read this article. These can include a range of mixed-use technologies that cater to both hedonic and utilitarian objectives. Examples include, but are not limited to, the device through which this article is read (desktop, laptop, tablet, or smartphone), a smart TV, smart kitchen appliances, smart watches, wearable fitness trackers, and connected or autonomous cars that drive the readers whilst they listen to a text-to-voice generated version of this article. These changes in the technological landscape have presumably been supported by technological advancements that have made technology more connected and affordable than before, smaller, yet broader in its capabilities, beyond merely being job-or leisure-oriented (Turel et al., 2019). Since many of these technologies are used exclusively in leisure or non-work settings, or in both work and non-work (including leisure) settings, they have created what we call digitised individuals, defined as users who use at least one digital technology in their non-work life domains. 1 Note that digitised individuals have existed since the dawn of personal computers, but we see major growth in the last decade with the vast penetration of smartphones, social media and personal lifestyle and health technologies. We view the collective of digitised individuals as contributing to the phenomenon of the digitisation of individuals, defined as the proliferation of digital technologies in the lives of individual users (Matt, Trenz, et al., 2019). Although the bare minimum to qualify as digitised individuals according to this definition is using one technology for non-work including leisure purposes, nowadays, many people use multiple technologies to different extents, being integrated into their lives in many different ways. This creates a large variability in the extent of their digitisation. The combination of the significant diffusion of digital technologies used by individuals with the variability in their (partly joint) usage leads to the necessity to develop an intertwined perspective that considers technology and the individual at the same time, that is, a socio-technical perspective. Several studies have argued for a need to understand digitised individuals, the drivers of the digitisation of individuals and the consequences of this digitisation, because technologies aimed at the digitisation of individuals have unique features that distinguish them from commonly examined business technologies, or that are insufficiently highlighted and understood in studies of non-work technologies (Matt, Trenz, et al., 2019; Turel et al., 2019). These characteristics inclu...

show abstract

Artificial intelligence as digital agency

Cited by 105 publications

References 52 publications

Questioning Racial and Gender Bias in AI-based Recommendations: Do Espoused National Cultural Values Matter?

Questioning Racial and Gender Bias in AI-based Recommendations: Do Espoused National Cultural Values Matter?

Assembling Algorithmic Decision-Making under Uncertainty: The Case of 'Edge Cases' in an Open Data Environment

An intertwined perspective on technology and digitised individuals: Linkages, needs and outcomes

Contact Info

Product

Resources

About