AI as Extraherics: Fostering Higher-order Thinking Skills in Human-AI Interaction

Abstract: As AI technologies, including generative AI, continue to evolve, concerns have arisen about over-reliance on AI, which may lead to human deskilling and diminished cognitive engagement. Over-reliance on AI can also lead users to accept information given by AI without performing critical examinations, causing negative consequences, such as misleading users with hallucinated contents. This paper introduces extraheric AI, a human-AI interaction conceptual framework that fosters users' higher-order thinking skills, such as creativity, critical thinking, and problem-solving, during task completion. Unlike existing human-AI interaction designs, which replace or augment human cognition, extraheric AI fosters cognitive engagement by posing questions or providing alternative perspectives to users, rather than direct answers. We discuss interaction strategies, evaluation methods aligned with cognitive load theory and Bloom's taxonomy, and future research directions to ensure that human cognitive skills remain a crucial element in AI-integrated environments, promoting a balanced partnership between humans and AI.

• The paper introduces extraheric AI, a framework that fosters creativity and critical thinking through interactive human-AI strategies.
• It employs diverse interaction strategies and evaluation metrics to boost cognitive engagement and schema construction.
• The study highlights design considerations to prevent over-reliance on AI while promoting user agency and problem solving.

AI as Extraherics: Fostering Higher-order Thinking Skills in Human-AI Interaction

This paper explores "extraheric AI," a conceptual framework designed to enhance human cognitive engagement and foster higher-order thinking skills during human-AI interactions. Recognizing the potential risks of over-reliance on AI, such as deskilling and diminished cognitive engagement, extraheric AI is introduced as a solution to promote creativity, critical thinking, and problem-solving.

Background and Motivation

Recent advancements in AI, including generative AI, have demonstrated significant capabilities in supporting human tasks. However, an over-reliance on AI can lead to various negative outcomes, such as users accepting AI-generated information without critical examination. This reliance may reduce human cognitive activity and result in decreased ownership, challenge perception, productivity, and sense of accomplishment.

Traditional AI designs often focus on task efficiency, replacing or augmenting human cognitive abilities. In contrast, extraheric AI aims to stimulate cognitive activities that foster higher-order thinking skills through a balanced human-AI partnership.

Extraheric AI Conceptual Framework

Extraheric AI distinguishes itself by fostering users' higher-order thinking skills through cognitive engagement strategies. It increases germane cognitive load, encouraging users to explore information and perspectives during task completion. Unlike orthotic, prosthetic, and exoskeletal AI interactions designed to reduce cognitive load, extraheric AI enhances germane cognitive load, crucial for schema construction and higher-order thinking skill development.

The framework applies to domains requiring intellectual activities like opinion formation, complex problem solving, and behavior change through enhanced cognitive perception. Examples include brainstorming, programming, and market analysis.

Interaction Strategies

The paper identifies eight interaction strategies to design extraheric AI:

1. Suggesting Content: Presents multiple viewpoints, allowing users to evaluate options.
2. Explaining: Provides contextual explanations to enhance understanding.
3. Nudging: Uses indirect suggestions to influence user behavior subtly.
4. Debating/Discussing: Engages users in conversations with AI agents representing diverse perspectives.
5. Questioning: Stimulates cognitive activity by prompting users to reflect or explain content.
6. Scaffolding: Offers temporary support, gradually empowering users to undertake tasks independently.
7. Simulating: Creates simulated environments fostering experiential learning.
8. Demonstrating: Engages users in observing AI agent behavior, promoting vicarious learning.

Evaluation Approaches

Evaluating extraheric AI involves assessing changes in higher-order thinking skills and cognitive load. Proposed metrics include revised NASA-TLX for cognitive load assessment, performance-based evaluations following Bloom's taxonomy, and attitudinal and behavioral scales for sense of agency, self-efficacy, user motivation, and ownership attribution.

Design Considerations

Extraheric AI design should:

• Embrace diverse AI agent outputs to encourage exploratory thinking.
• Maintain non-judgmental attitudes, fostering open intellectual exploration.
• Align with existing user workflows to support task completion without interference.
• Consider social roles and multi-agent designs to facilitate social learning.
• Accept user disengagement as a natural progression following skill mastery.

Research Opportunities

Areas for further research include developing technology to present diverse AI perspectives, contributions to cognitive load theory from an HCI perspective, new evaluation frameworks, understanding the impact of social roles on user perception, and ethical considerations for extraheric AI use.

Conclusion

Extraheric AI provides a promising avenue to balance human cognitive engagement and AI integration, supporting higher-order thinking skill development in various domains. By focusing on interactive strategies, thoughtful design considerations, and comprehensive evaluations, extraheric AI aims to create a balanced human-AI partnership that enhances cognitive abilities while mitigating reliance risks.