Intrinsic motivation in virtual assistant interaction for fostering spontaneous interactions

• DOI: 10.1371/journal.pone.0250326
• Publication Date: 2021-04-23T18:39:56Z
• Authors: Chang Li; Hideyoshi Yanagisawa
• Keywords: Goal theory

With the growing utility of today's conversational virtual assistants, importance user motivation in human-AI interaction is becoming more obvious. However, previous studies this and related fields, such as human-computer human-robot interaction, scarcely discussed intrinsic its affecting factors. Those either treated an inseparable concept or focused on non-intrinsic motivation. The current study aims to cover by taking affective-engineering approach. A novel model proposed, which affected two factors that derive from interactions with assistants: expectation capability uncertainty. Experiments are conducted where these manipulated making participants believe they interacting smart speaker "Amazon Echo". Intrinsic measured both using questionnaires covertly monitoring a five-minute free-choice period experimenter's absence, during could decide for themselves whether interact assistants. Results first experiment showed high engenders intrinsically motivated compared low expectation. results also suggested suppressive effects uncertainty motivation, though we had not hypothesized before experiments. We then revised our hypothetical action selection accordingly verification uncertainty's effects. reducing encourages causes behind shift intrinsic.