Averse to what: Consumer aversion to algorithmic labels, but not their outputs?

Area of Science:

• Human-Computer Interaction
• Artificial Intelligence Ethics
• Social Psychology of Technology

Background:

• Growing research examines public reactions to AI tools and algorithms.
• Existing literature often highlights negative reactions, termed algorithm aversion or human preference.
• This study proposes a nuanced perspective on human-AI interaction.

Purpose of the Study:

• To propose a third interpretation of human reactions to AI: label aversion coupled with output appreciation.
• To offer insights for refining the study of human-algorithm interactions.
• To encourage a broader examination of AI perception beyond simple aversion.

Main Methods:

• Literature review and theoretical analysis of human-AI interaction studies.
• Conceptual framework development for understanding lay beliefs about AI.
• Identification of key factors influencing user perception of AI tools.

Main Results:

• Human aversion may stem from AI tool labeling rather than the technology itself.
• Users can appreciate AI-generated output even if they are wary of the underlying algorithms.
• A distinction between label perception and output evaluation is crucial.

Conclusions:

• Future research must carefully consider the labeling of AI tools to accurately gauge public perception.
• Studies should broaden their scope to include user interactions with AI tools.
• Accounting for the technical configuration of AI systems is essential for understanding user reactions.