[Paper Review] eXtended Artificial Intelligence: New Prospects of Human-AI Interaction Research

Carolin Wienrich, Marc Erich Latoschik|arXiv (Cornell University)|Mar 27, 2021

Virtual Reality Applications and ImpactsComputer Science70 references70 citations

TL;DR

This paper introduces an XR-AI continuum framework to systematically study human-AI interaction through immersive extended reality (XR) environments. It demonstrates that XR enables controlled, valid experimentation on how AI embodiment affects user perception and behavior, revealing gender effects in human-robot interaction and an Eliza effect in recommender systems.

ABSTRACT

Artificial Intelligence (AI) covers a broad spectrum of computational problems and use cases. Many of those implicate profound and sometimes intricate questions of how humans interact or should interact with AIs. Moreover, many users or future users do have abstract ideas of what AI is, significantly depending on the specific embodiment of AI applications. Human-centered-design approaches would suggest evaluating the impact of different embodiments on human perception of and interaction with AI. An approach that is difficult to realize due to the sheer complexity of application fields and embodiments in reality. However, here XR opens new possibilities to research human-AI interactions. The article's contribution is twofold: First, it provides a theoretical treatment and model of human-AI interaction based on an XR-AI continuum as a framework for and a perspective of different approaches of XR-AI combinations. It motivates XR-AI combinations as a method to learn about the effects of prospective human-AI interfaces and shows why the combination of XR and AI fruitfully contributes to a valid and systematic investigation of human-AI interactions and interfaces. Second, the article provides two exemplary experiments investigating the aforementioned approach for two distinct AI-systems. The first experiment reveals an interesting gender effect in human-robot interaction, while the second experiment reveals an Eliza effect of a recommender system. Here the article introduces two paradigmatic implementations of the proposed XR testbed for human-AI interactions and interfaces and shows how a valid and systematic investigation can be conducted. In sum, the article opens new perspectives on how XR benefits human-centered AI design and development.

Motivation & Objective

• To address the challenge of studying how different AI embodiments influence human perception and interaction in real-world complex environments.
• To propose an XR-AI continuum as a scalable, controlled, and systematic research framework for human-AI interaction.
• To validate the framework through two empirical experiments on human-robot interaction and AI recommender systems.
• To advance human-centered AI design by enabling systematic investigation of interface embodiment effects using XR.
• To bridge the gap between abstract AI concepts and tangible user experiences through immersive simulation.

Proposed method

• Proposes an XR-AI continuum as a theoretical and practical framework to model and study the spectrum of human-AI interface embodiments.
• Uses immersive XR environments to simulate diverse AI embodiments (e.g., humanoid robots, virtual agents) in controlled experimental settings.
• Employs standardized human-centered design principles to evaluate user perception, behavior, and interaction quality across different AI interface forms.
• Conducts two controlled experiments: one on human-robot interaction with gendered robot avatars, and another on a recommender system with Eliza-like responses.
• Applies established psychological and HCI theories (e.g., media equation, Proteus effect, uncanny valley) to interpret user responses in XR.
• Uses quantitative metrics such as behavioral conformity, social presence, and interaction duration to assess user responses.

Experimental results

Research questions

• RQ1 How does the gendered embodiment of a robot influence human interaction behavior and perception?
• RQ2 To what extent does a text-based AI with Eliza-like responses trigger anthropomorphic perception and behavioral conformity?
• RQ3 How does the XR-AI continuum enable systematic and valid investigation of human-AI interface design choices?
• RQ4 What role does embodiment play in shaping user expectations and trust in AI systems?
• RQ5 Can immersive XR environments effectively simulate and test future AI interface concepts before real-world deployment?

Key findings

• A significant gender effect was observed in human-robot interaction, where participants exhibited different behavioral patterns when interacting with male- vs. female-voiced robots.
• The Eliza effect was confirmed in a recommender system experiment, where users perceived a simple AI as more empathetic and responsive than intended.
• Participants demonstrated behavioral conformity in response to avatar appearance, supporting the Proteus effect in AI interaction contexts.
• The XR-AI continuum framework enabled valid, repeatable, and scalable experimentation on human-AI interface design.
• Immersive XR environments effectively simulated real-world AI interaction dynamics, allowing controlled study of embodiment effects.
• The study confirms that interface embodiment significantly shapes user perception, trust, and interaction quality in AI systems.