How do workers (in)directly affected by AI implementation experience a change in their work and work environment? In their third policy brief, the ai:conomics research team provides initial insights into their research work on this question.
Artificial Intelligence (AI) is yet nowhere near retrieving “artificial general intelligence”: the ability to undertake a wide variety of cognitive tasks that humans can. Nevertheless, different fields of AI, such as machine learning, robotics, computer vision, and natural language processing, are already succeeding in performing complex problemsolving tasks as humans would normally do; voice assistants, image recognition, and algorithmic hiring are just a few examples. Although the effects of AI implemenation on labour yet remains unclear, it is essential to understand how workers perceive its impact on their work and related outcomes.
Based on qualitative research, this policy brief provides firm-specific insights about whether and how workers (in-)directly affected by AI implementation experience a change in their work and work environment. We interviewed 25 workers at two multinational private-sector companies that have recently implemented new AI applications
These are the main findings:
- Findings suggest that, although several workers were aware of threats such as technological replacement and highlighted human’s superiority in certain tasks, most workers acknowledged the added value of AI.
- Furthermore, most workers experienced increased productivity, while some also feared higher workloads as a consequence.
- Nevertheless, no substantial changes in tasks, skill demands, wellbeing or satisfaction were found as a response to AI implementation.
- Interestingly, perceptions of future employment security depended on the implementation context: interviewees from one company were mostly aware of potential displacement effects, while in the other the AI was perceived as a complementing tool.
While this brief allows for nuanced insights on individuals’ perceptions, more qualitative firm-specific research is necessary to fully comprehend how different implementation contexts of AI can affect various types of workers.
What do these findings reveal about innovative and intelligent technologies’ effects on perceptions of work?
Remarkably, although quite some interviewed workers were aware of threats such as technological replacement, most informants were still able to acknowledge the added value of technological progress: they mostly perceived the AI as a supporting tool and highlighted human superiority in decision-making as well as flexibility. Furthermore, while most workers noticed an increase in productivity, some of them also feared an increase of their workload based on a technology-induced increase in work- and production speed. Moreover, the researchers neither found substantial changes in tasks nor current skill demands, potentially resulting from the AI technologies being embedded within familiar tools and work processes in both use-cases. However, the absence of changes in tasks and skills may also only be valid in the short run, as the interviews were conducted shortly after AI adoption. In addition, the researchers did not detect clear patterns of a direct effect on well-being or satisfaction as a result of AI implementation. However, perceptions of future employment security depended on the implementation context: in one company the interviewees were mostly contemplating technological displacement effects more pronouncedly, while the AI was rather perceived as a supporting, complementing tool in the other company. This may partially depend on organizational communication strategies surrounding the AI implementation and its consequences for workers’ job security.
Ultimately, while this brief aimed to provide insights on how workers experience AI implementation, additional qualitative firm-specific research is essential to fully comprehend how different implementation contexts of AI can affect various types of workers. Additionally, although we did not find substantial evidence that AI adoption changed workers’ tasks, skill demands, well-being or satisfaction in the short run, long term consequences of AI adoption on these and related outcomes are yet to be understood.
Transparency can alleviate concerns and worries of employees
Even though the interviewed workers have not been able to observe substantial changes in tasks, skills, well-being, and satisfaction, they can still worry about their workload and employment security in the future. This implies that even though AI may currently not incur drastic changes in jobs yet, affected workers are not limiting their concerns to the short run. Most of them seem to anticipate these changes in the future, showing clearly that workers need to be taken along in the journey of AI implementation. Aside from monitoring AI-induced changes, mitigating any occurring fears and concerns among workers seems increasingly important. For instance, concerns can
be addressed by transparently informing workers of potential effects and offering opportunities to proactively shape their professional future.