Is It Human? The Role of Anthropomorphism as a Driver for the Successful Acceptance of Digital Voice Assistants

Abstract

The market of digital voice has grown significantly over the recent years. Big players like Amazon, Google, Apple, Microsoft and Samsung are focusing on the development and expansion of their assistants. Especially smart speakers are on the rise but also in smartphone integrated voice applications are getting more popular. The main characteristics of this new technology are both elements of human-computerinteraction and especially the attribution of human characteristics. Although there is an increase of the number of current users as well as of consumers intending to use digital voice assistants in the future, drivers and barriers of digital voice assistants have not yet been sufficiently empirically investigated, especially concerning the phenomenon of anthropomorphism. This study points to additional key factors that are important to foster broader acceptance. Our empirical study is based on the UTAUT2 and highlights the importance of anthropomorphism in relation to other determinants known from the literature.

Machinelike or Humanlike? A Literature Review of Anthropomorphism in AI-Enabled Technology

“…In the AIET context, the phenomenon of attributing humanlike traits to these technologies is called anthropomorphism [9,19]. Recently, anthropomorphism has been identified as an important factor influencing the perception, adoption, and continued use of AIET [12,15,16,21]. For instance, Waytz et al [21] found that anthropomorphism promoted trust in autonomous vehicles.…”

“…Moussawi et al [12] showed that anthropomorphism enhanced individuals' perceived enjoyment when they interacted with smart speakers, which led to an increase in adoption intention. Likewise, Wagner et al [16] reported that anthropomorphism increased smart speakers' likeability, thereby promoting the intention to adopt them.…”

“…These capabilities make autonomous vehicles operate autonomously, chatbots communicate with humans naturally, and social robots exhibit emotional responses [9,11], thereby allowing AIET to interact with humans in humanlike ways [9,12]. Researchers thus argue that the humanlike characteristics of AIET are unique technological features that can influence users' perception, adoption, and continued use of this technology [12][13][14][15][16].…”

Designing for Positive Emotional Responses in Users of Interactive Digital Technologies: A Systematic Literature Review

“…Qiu and Benbasat [34] studied the effects of anthropomorphic embodiments in the interfaces of product recommendation agents and found that it can increase the user's perceived enjoyment and trust in the recommendation agent. A study on anthropomorphism in digital voice assistants found that a more human-like interaction between the voice assistant and the user positively influenced the likeability of the digital voice assistant, which in turn also influenced the user's intention to use the assistant again [35]. Based on the above research, practitioners can, therefore, influence the affective responses of users through implementing virtual characters, or other anthropomorphic means/techniques or tools in their interfaces.…”

“…Practitioners may also implement affective computing techniques, as there is also evidence to support their influence on the affective responses of users [31][32][33][34][35][36].…”