2. Uncertainty in the Age of Digital Transformation

Interestingly, Moore’s Law has been declared of its death periodically, but it appears to get revived each time. See, for example, Dataintegration.info’s (2023) article for one of the latest stories.

According to some reports, a significant percentage of DX initiatives failed to achieve their original objectives, with as many as 84% falling short of their intended goals (Block, 2022; Morgan, 2019).

The advent of disruptive innovations—including historical ones such as Gutenberg’s printing press, the telephone, and the automobile—has been commonly faced with resistance. The Catholic Church, for example, vehemently opposed to the printing press, fearing that the mass production of texts could propagate heresy and undermine the Church’s authority. Similarly, the introduction of the telephone was met with skepticism and apprehension, with individuals expressing concerns over potential invasions of privacy. The automobile, too, was not exempt from this trend of initial resistance; the infamous Red Flag Act, which mandated a man to walk ahead of an automobile bearing a red flag to warn pedestrians and horse-drawn carriages of its approach, should serve as a testament (see Christensen, McDonald, Altman, & Palmer, 2018). It should not be hard to find similar stories for more contemporary innovations, such as personal computers and the Internet. All of these narratives underscore the enduring nature of opposition to novel ideas, a phenomenon that is far from novel itself. Intriguingly, even eminent philosophers who discussed the value of enlightenment seem fallible with this aspect of human nature. Socrates expressed reservations about the written words (which are depicted as an Egyptian innovation) in his dialogue with Phaedrus. The great Greek thinker was concerned because people would become overly reliant on the written words, which cannot defend themselves flexibly in argument and could therefore be misinterpreted outside their original context. My pet story pertains to the advent of the bicycle in the late nineteenth century. As this new mobility device gained popularity, a wave of opposition and oppression emerged against women riding it. A faction of the medical community insisted that riding bicycles would pose some detrimental impact on women’s mental and physical health, particularly their reproductive capabilities. Fortunately, that movement was short-lived, and many women embraced bicycling as not only a form of exercise but also a symbol of freedom and liberation (see Guroff, 2016, for further details and discussions).

Intriguingly, some scholars contend that the inclination towards loss aversion can be construed as an evolutionary adaptation (McDermott et al., 2008). In the context of survival in the prehistoric era, the cost of missing out on a potential gain (like missing out on food) was much less severe than the cost of not avoiding a loss (like encountering a predator). This evolutionary psychology perspective suggests that our ancestors who were more attuned to avoiding losses were more likely to survive and pass on their genes to the next generations than their counterparts who were overly bold and risk-seeking; thus, the loss-aversive tendency has been ingrained in human decision-making mechanisms over time (see Hintze et al., 2015; Li et al., 2012).

This time interval serves as an indicator of the customer’s determination to purchase a particular drink, thereby providing a measure of its popularity.

A note of disclosure: Yours truly served as a project leader for Relic in the years 2022–2023. During my tenure, I engaged in numerous projects, including those that implemented this digital platform service.

Digitization does change the given organization’s standard for cost-effectiveness and productivity because it enhances the efficiency of operations. As a consequence, it may potentially exert additional pressure on organizational members to boost their productivity (Correani et al., 2020; Siderska, 2020).

Previous studies imply the presence of an inherent “circuit breaker” system built within the human psychological mechanism that terminates information processing when individuals are overwhelmed by the given circumstances (Fetzner, Horswill, Boelen, & Carleton, 2013; Slovic, 2010; for an interesting exception, see Vince, 2020). This cognitive halting is typically observed in situations of high stress or information overload, where the brain essentially shuts down to protect itself from further stress or damage (Kaspi et al., 1995; Phillips-Wren & Adya, 2020).

As far as they are not fired during the M&A process, that is.

On this front, it might be interesting to scrutinize if individuals who prioritize a “human-centered” approach and interpersonal relationships are more likely to be affected by digital transformation.

One of my students at a business school, who was working at the Fukuoka Financial Group then, applied through the group’s internal job transfer system to join Minna-no-Ginko but, unfortunately, his application was not approved. “When I received the news from my supervisor that my transfer request was not successful,” he muttered, “I felt like being told that I was not cutting-edge enough, too entrenched in traditional work practices and legacy systems to effectively contribute to the emerging revolution”.

An intriguing case of overcoming such “identity crises” can be found in how professional players of Shogi, or Japanese chess, have navigated the transformation precipitated by the advent of artificial intelligence programs (Igami, 2020; Schrittwieser et al., 2020). The competitiveness of the AI for playing the game of Shogi (hereafter referred to as “Shogi AI”) did not go beyond that of high amateurs for a long time. The adoption and use of deep reinforcement learning methods in the 2010s, however, have entirely repainted the landscape of the game and Shogi AIs suddenly began to beat Kishi, professional Shogi players. A watershed moment occurred on March 30^th, 2013, when a Shogi AI named Ponanza won the match against Shinichi Sato, marking the inaugural victory of an artificial intelligence against Kishi (Yamamoto, 2017). The pinnacle of that movement came on May 20^th, 2017, when Amahiko Sato Meijin, the grand champion of Shogi, succumbed to an AI. Interestingly, despite the seemingly demeaning and disgracing nature of this defeat, it sparked a trend among Kishi to incorporate Shogi AI into their personal training and research. Today, it is commonplace for Kishi to consult with high-performance Shogi AI programs to explore novel effective moves, develop innovative strategies, and scrutinize their own performance (Kobayashi, 2015; Nishio, 2018). Taichi Nakamura, a top-tier Kishi, commented in an interview: “The nature of the relationship between humans and artificial intelligence has evolved from antagonism to alliance. People used to see AI as a villain.. Today, however, our relationship with AI has become that of collaboration and constructive discussions.” (SlowNews, 2021; translation by the author).

The author of this book has a first-hand experience regarding such transmogrifications. Approximately a decade prior to the penning of the current manuscript, I became one of the early-stage members of a startup that harbored the ambitious goal of revolutionizing Japan’s secondary education landscape and thereby transforming the learning experience of millions of teenagers in the country. The allure of this mission was the primary impetus behind my decision to join the respective startup. Unexpectedly, though, the founder and CEO elected to pivot the business’s focus from education to the recruitment of recent college graduates (for discussions of pivoting in startups, see Kirtley & O’Mahony, 2020; Meiners, 2018). While I refrain from passing any judgment on the validity of this strategic redirection (as of the time of writing, the company continues to operate successfully within the HR sector, suggesting that the pivot may have been a sound decision after all), I felt a loss of the reason to stay connected with the company and chose to pursue a different career path shortly after the CEO’s announcement.

Several measures were undertaken to ensure the data fidelity. First, attention checks were included at random places within all surveys. For example, one question specified that a particular response should be selected (“This is an attention check; please choose ‘strongly agree’ for this item.”) and the data provided by those who failed to follow the instruction were removed. In addition, the cases where the time spent to complete the survey was either too short or too long (i.e., ± 3SDs) were removed.

The service provider, Macromill, had the top share in the market research industry in Japan as of the current study’s data collection (Japan Marketing Research Association, 2021).

Initially, 1,200 respondents were identified after the screening process, to whom an invitation for participation was sent. From this initial pool, 30 did not return the response within the deadline. Another 31 were removed by the data-fidelity checks detailed in Footnote 14.

In the parallel analysis, first, random values are generated and factor-analyzed. Thus computed eigenvalues represent the threshold to identify theoretically meaningful factors. If the eigenvalue of a factor extracted through EFA from a given data set fails to exceed that of the factor in the parallel analysis, it means that the former may be less substantive than a random factor (Hayton et al., 2004).

On the other hand, the result of null-hypothesis significance test (i.e., p-value) associated with the χ² value was not used for goodness-of-fit evaluations, because the χ² test is known to yield overly strict, and hence inaccurate, results when the sample size is large (Hu & Bentler, 1999).

The impact of the demographic variables measured for the current study (i.e., respondents’ age, sex, and tenure at the currently affiliated organization) was examined by running an alternative SEM, in which the demographic factors were entered as control variables. The results suggested that the statistical significance and direction of the effects among the focal variables did not change. Thus, for the brevity’s sake, the results of the analysis without demographic variables are presented in this article. The details of the parameter estimates computed for the alternative model are available from the author upon request.

Initially, 700 respondents were identified after the screening process, to whom an invitation for participation was sent via the same research service used for Supplementary Study 1. From this initial pool, 11 were dropped by the data-fidelity checks detailed in Footnote 14.

Obviously, this argument hinges upon a crucial assumption; that is, it is presumed that the employees under consideration will not encounter serious, negative consequences, such as job loss, upon the impending DX. In the cases where the likelihood of such adverse outcomes is high, they will suffer a great deal of anxiety concerning their future career trajectory and overall life prospects (for discussions of the relationship between job loss and employee uncertainty, see Keim, Landis, Pierce, & Earnest, 2014; Leana & Feldman, 1988; Sverke & Hellgren, 2002; Wanberg & Banas, 2000).

To illustrate, let us imagine that an organization decides to implement a new data management system. No matter how extensively the organization in question attempts to support its employees in the form of training and resources, they may still experience uncertainty because the full impact of the new system on their daily tasks and roles may not be completely predictable ahead of its implementation. For instance, the system might have unforeseen technical issues, or it could change the workflow in ways that were not initially apparent.

Initially, 700 respondents were identified after the screening process, to whom an invitation for participation was sent via the same research service used for the main study and Supplementary Study 1. From this initial pool, eight were dropped by the data-fidelity checks detailed in Footnote 14.

Another example of the emancipatory potential of digitization in reducing strenuous labor can be found in the application of the paper-handling robot developed by Fujifilm Business Innovation Corp. (FBI) in the printing industry (Takashima, 2023). Albeit essential to prevent paper jams in the printing processes, the task of paper-handling (i.e., inserting air between papers) is physically demanding, repetitive, and time-consuming. Workers who engage in the task have to deal with large-format paper, which can weigh over 10 kg, continuously cut open the stacked wrapping paper at precisely the same spot, and move it to another location after introducing air between each sheet of paper; moreover, they need to repeat those procedures over and over again for 8–10 hours per day. FBI’s robot has automated this entire process, enabling employees within the printing sector to concentrate on more creative jobs, such as striving for optimal print quality that faithfully reproduces the original design and color palette of an artwork.

According to the OECD report, this percentage is projected to rise up to 80% in 2060.

It might be interesting to scrutinize in future studies the extent to which communication skills and perceived “visionariness” of C-suite executives are related to the success of DX initiatives (Flauto, 1999; Horlacher & Hess, 2016; Wrede et al., 2020).