ABSTRACTPurpose – This study experimentally evaluates the viability of AI-animated humanlike avatar technology vs actual organic human spokespersons for use in learning videos. It is guided by a convergence of AI Literacy Theory, the Technology Acceptance Model, and the Theory of Reasoned Action.Design/Methodology/Approach – Findings are determined through an experimental test of onboarding video variations. Employing a randomized 2x2 trial design, over 250 professionals assessed training videos featuring either synthetic avatars or organic human spokespersons. Variables such as training effectiveness, knowledge transfer, and brand perception were analyzed using ANOVA.Findings – Synthetic avatars matched the effectiveness of human counterparts in all scenarios where participants were unaware of the avatar’s synthetic nature. The majority of participants did not detect the avatars’ syntheticness, which was undisclosed to them. However, when there was perceived syntheticness, it triggered the negative uncanny valley phenomenon, albeit weakly statistically.Originality/Value – Synthetic avatars, as outlined here, are a viable new option in the array of available learning technologies. This research is thus highly valuable for practitioners and scholars interested in the effectiveness of cutting-edge learning tools. The foundational quantitative assessment here provides a benchmark for evaluating future improvements in the technology. This study provides actionable recommendations for integrating AI-powered avatars into professional development, enhancing our understanding of advanced technologies’ roles in our educational/training programs.Keywords: synthetic media, digital media, avatar, artificial intelligence, computer mediated communication, video, professional communication, onboarding, training, uncanny valley
