Intrinsic motivation in virtual assistant interaction for fostering spontaneous interactions

Area of Science:

• Human-Computer Interaction
• Artificial Intelligence
• Psychology

Background:

• User motivation in human-AI interactions, particularly intrinsic motivation, is understudied.
• Previous research often conflated different types of motivation or focused solely on utilitarian purposes.

Purpose of the Study:

• To investigate intrinsic motivation in human-virtual assistant interactions using an affective engineering approach.
• To propose and test a novel motivation model where intrinsic motivation is influenced by expectation and uncertainty.

Main Methods:

• Experiments were conducted manipulating expectation of capability and uncertainty in participants interacting with a smart speaker (simulated Amazon Echo).
• Intrinsic motivation was assessed via questionnaires and a covert free-choice interaction period.

Main Results:

• Higher expectation of capability led to increased intrinsically motivated interaction.
• Reducing uncertainty encouraged more interactions and shifted motivation towards intrinsic.

Conclusions:

• Expectation and uncertainty are significant factors influencing intrinsic motivation in human-AI interactions.
• The findings suggest practical implications for designing more engaging and intrinsically motivating virtual assistants.