As algorithms become more sophisticated and accurate, they become increasingly capable of giving us advice. Without thinking about it, many of us rely on the advice of algorithms daily, for example when Google Maps tells us which route we should take to work or when Spotify advices us to listen to a certain song based on some other songs we liked. In situations like these we often do not even doubt the capabilities of algorithms and choose to readily trust them. Even though it is clear that algorithms have started to outperform humans on many tasks, it seems that humans are not always likely to trust them. As the capabilities of algorithms increase, they are also introduced in areas and situations in which many people may prefer human decision making. Research showed that people in general do not trust advice coming from algorithm, a phenomenon that was named ‘algorithm aversion’ (). According to Castelo et al. (2019), this aversion comes from a belief that algorithm lack certain emotional abilities and according to Dietvorst et al. (2018), people have a greater tolerance for human error compared to a mistake from an algorithm. Multiple papers readily confirm the existence of algorithm aversion (Castelo et al., 2019; Bigman & Gray, 2018). On the other hand, a paper by Logg et al. (2019) shows mainly results that indicate the existence of a concept called algorithm appreciation, the opposite of algorithm aversion. Especially in the case of quantitative tasks, a preference for algorithmic advice seemed to exist. A reason for algorithm aversion may lie in an accountability problem. Research showed that people prefer a human making a decision in life or death situation in the medical sector, even if it was known before hand that the person would die anyway (Bigman & Gray, 2018). An algorithm itself cannot be held accountable for whatever it does or advice, and it is also difficult to hold the creator of the algorithm accountable.
For the implementation of algorithms to be successful within an organization, it is important that people are willing to rely on the advice given by the algorithm. In situations that concern ethics and morality, the issue of accountability makes it difficult for people to be willing to rely on the advice of an algorithm, and at least for now, I would say it is better that a human expert makes the decision while having received advice from an algorithm. For now, I do not believe that the accountability issue can easily be solved. For it to be solved, people will have to become more used to the idea of algorithms making decisions or giving advice in situations that concern ethics and morality, and algorithms must obtain a reputation of superiority in giving such advice and making such a decision.
______________________________________________________________________________________________________________________________
Bigman, Y. E., & Gray, K. (2018). People are averse to machines making moral decisions. Cognition, 181, 21–34. https://doi.org/10.1016/j.cognition.2018.08.003
Castelo, N., Bos, MW, & Lehmann, DR (2019). Task-Dependent Algorithm Aversion. Journal of Marketing Research , 56 (5), 809–825. https://doi.org/10.1177/0022243719851788
Dietvorst, BJ, Simmons, JP, & Massey, C. (2018). Overcoming Algorithm Aversion: People Will Use Imperfect Algorithms If They Can (Even Slightly) Modify Them. Management Science , 64 (3), 1155–1170. https://doi.org/10.1287/mnsc.2016.2643
Logg, JM, Minson, JA, & Moore, DA (2019). Algorithm appreciation: People prefer algorithmic to human judgment. Organizational Behavior and Human Decision Processes , 151 , 90–103. https://doi.org/10.1016/j.obhdp.2018.12.005
