Human-Machine Knowledge Conversion 

Human-Machine Knowledge Conversion 

Human-Machine Knowledge Conversion 

Soo Young Choi and Stefan Haefliger, Cass Business School, City University of London October 2019 

Recent advancement in AI technology and adoption of such technology have dramatically transformed organisations’ innovation processes and their outcomes. AI can be used in almost every process of product development (Klintong et al., 2012), which enables digital innovation, and it is reshaping firms’ innovation process and innovation outcome (Lyytinen et al., 2016). Technologies such as generative design, machine learning and knowledge-sharing platforms now play a powerful role for product innovation by not only enabling multiple actors with diverse capabilities to produce change in innovation processes, but also acting its own agency to make this change (Zafari and Koeszegi, 2018). AI functions as an intelligent agent that transforms the actions of individuals and organisations. AI technology and its integration into innovation practices underlying product development processes is creating a new relationship between the human and the AI, which affects innovation processes and (product) innovation outcomes. As a result, there is need for re-examining and re-conceptualising the role of digital technology in organisations. 

Knowledge management has long argued for knowledge to be treated as an asset, fundamentally extending the resource-based view of the firm (Kogut and Zander, 1992; Grant, 1996) and knowledge creation theory argues about the ways in which knowledge is mobilised by human actors in firms as integral part of any innovation process (Nonaka, 1994; Nonaka and von Krogh, 2009). With AI technology, we argue, generating new ideas is no longer limited to humans but also performed through collaborations between humans and machines. Building on the tradition of the knowledge- based view of the firm, we aim to re-conceptualise knowledge creation theory by exploring how knowledge creation modes are altered with AI agents that can lead to digital product innovation. 

First, we start with the literature on digital innovation and take a step back to explore how AI differs from other digital technologies to complement the work of the individual innovator and how an innovation process involving AI in product development affects how knowledge is mobilised, converted, articulated, and applied to problem solving and innovation. Second, we build on ideas from sociomateriality to focus on the role of material agency in knowledge creation and to theorise where and how interactions between AI and human innovators are ‘entangled’ or ‘imbricated’ to generate digital innovation (Leonardi, 2013; Orlikowski, 2009). This allows theorisation about the precise role and potential of AI in both the innovation process as well as the outcome. We propose a few conjectures in order to account for knowledge work by material agency, the socialisation of AI technology in organisation, and the knowledge conversion between human and nonhuman actors. Implications for the economics of innovations and organisation science are discussed. 

Based on Nonaka’s knowledge creation theory (Nonaka, 1994) yet radically departing from its assumptions, we aim to describe how the four modes of knowledge creation (SECI – Socialisation, Externalisation, Combination, Internalisation) operate between human and nonhuman actors during the process of new product development. Knowledge interactions are seen as social practices of human and nonhuman actors configured together to innovate. This study goes beyond just sequentially equating one knowledge creation mode to another but rather to argue that all knowledge creation modes are intertwined in generating digital product innovation. The major assumption of Nonaka’s knowledge creation theory is that new knowledge is created through knowledge conversion between and within individuals. However, we illustrate through a sociomaterial lens, how incorporating AI agency transforms this fundamental assumption and how this change affects knowledge creation modes, and ultimately digital product innovation. 

For the purpose of this paper, we address the use of AI in designing a new product, recognising that the setting provides a primary context for materialisation of the knowledge creation modes enacted by both human and nonhuman actors. This is particularly relevant, because what was previously perceived as human activity is now shifting towards activities of humans and technology. For 

example, a generic AI system can be regarded as a new digital apprentice that does not possess any knowledge but is embedded in data and equipped with unique learning abilities. Its human mentors ‘train’ and ‘share’ their experiences with the AI, such that the AI abstracts and converts existing tacit and explicit knowledge with the aim to develop new knowledge. The socialisation of knowledge, and knowledge conversion modes, operate between human and non-human actors. This knowledge interaction between human and non-human actors implies that innovation agency is more distributed among different designers, engineers, managers and AI. This raises questions about how technology and human agencies are constituted in the activity of creating new knowledge, which is key to digital product innovation (Lindberg et al., 2019). In particular, how innovation agency changes compared to other previous digital technologies and how human actors engage with AI to create new knowledge. 

A theoretical framework that describes the mechanisms of digital innovation, created by both human and AI is required to propose an organising logic of digital innovation and to understand effects of distributed innovation agency with AI. Theoretically and empirically accounting for AI agency in the knowledge creation modes and in generating digital innovation would reveal a novel aspect of AI, which is fundamental to organisational outcomes for managing digital innovation. Innovation practices can be clarified by understanding how innovation occurs through a knowledge conversion process between the (individual) humans and the non-human (learning) machine. Within our conceptualisation, we argue how using AI for product development is about socialising and combining knowledge with AI, where the focus is on 1) individuation of AI agency in context (the digital apprentice), 2) the nature of socialisation that involves both human and nonhuman actors, 3) how contextualised (or trained) AI combines acquired knowledge in meaningful ways, 4) the process of externalisation of tacit knowledge held by AI yet intelligible by humans, and 5) the limits of AI and the level of substitutability of the human designer. 

References 

Grant, R.M., 1996. Toward a knowledge‐based theory of the firm. Strategic management journal, 

17(S2), pp.109-122. Klintong, N., Vadhanasindhu, P. and Thawesaengskulthai, N., 2012, February. Artificial intelligence 

and successful factors for selecting product innovation development. In 2012 Third International Conference on Intelligent Systems Modelling and Simulation (pp. 397-402). IEEE. Kogut, B. and Zander, U., 1992. Knowledge of the firm, combinative capabilities, and the replication 

of technology. Organization science, 3(3), pp.383-397. Leonardi, P.M., 2013. Theoretical foundations for the study of sociomateriality. Information and 

organization, 23(2), pp.59-76. Lindberg, A., Lyytinen, K., Zhang, Z., Yoo, Y., 2019. Towards a theory of liminal agency. Case 

Western Reserve University Working Paper. Lyytinen, K., Yoo, Y. and Boland Jr, R.J., 2016. Digital product innovation within four classes of 

innovation networks. Information Systems Journal, 26(1), pp.47-75. Nonaka, I., 1994. A Dynamic Theory of Organizational Knowledge Creation. Organization science, 

5(1), pp.14-37. Nonaka, I. and Von Krogh, G., 2009. Perspective—Tacit knowledge and knowledge conversion: 

Controversy and advancement in organizational knowledge creation theory. Organization science, 20(3), pp.635-652. Orlikowski, W.J., 2009. The sociomateriality of organisational life: considering technology in 

management research. Cambridge journal of economics, 34(1), pp.125-141. Zafari, S. and Koeszegi, S.T., 2018, September. Machine agency in socio-technical systems: A 

typology of autonomous artificial agents. In 2018 IEEE Workshop on Advanced Robotics and its Social Impacts (ARSO) (pp. 125-130). IEEE.