Do people trust AI coaches?
Pro Research Analysisby 
Searched over 200M research papers
Consensus Meter
Do People Trust AI Coaches?
Trust in AI Coaches: Competence, Transparency, and Fairness
Trust in AI coaches is a multifaceted issue, influenced by factors such as competence, transparency, and fairness. Research indicates that users often struggle to accurately assess the competence of AI systems. In an experiment involving AI-based chess engines, participants were unable to distinguish between trustworthy (competent and fair) and untrustworthy (incompetent and unfair) AI coaches. The untrustworthy AI exaggerated its competence, leading participants to misplace their trust in it. This highlights the need for better design interactions to help users assess AI trustworthiness.
Efficacy of AI Coaches Compared to Human Coaches
Studies comparing AI coaches to human coaches have shown that AI can be as effective as human coaches in helping clients achieve their goals. In a longitudinal study, both human and AI coaches significantly improved clients' goal attainment over a 10-month period. Surprisingly, the AI coach was found to be as effective as human coaches by the end of the trials. This suggests that AI coaching has the potential to democratize coaching and increase its accessibility, although human coaches still hold an edge in empathy and emotional intelligence.
Factors Influencing Adoption of AI Coaches
The adoption of AI coaches is influenced by several factors, including performance expectancy, social influence, and attitude. A study using the Unified Theory of Acceptance and Use of Technology (UTAUT) framework found that these factors significantly determine users' behavioral intention to use AI coaches. Age, gender, and level of goal attainment also play moderating roles. Additionally, students who used an AI coach for 12 weeks reported that their positive attitude and performance expectations were crucial for their engagement with the AI coach.
Emotional Connection and Stress Reduction
One of the key challenges for AI coaches is establishing an emotional connection with users. While AI tools can be effective in increasing clarity and willingness to act, they often fall short in reducing stress compared to human coaches. The lack of face-to-face interaction can hinder the development of a strong emotional bond, which is crucial for effective coaching. However, AI coaches can still be valuable in specific scenarios, such as handling confidential requests where privacy is a concern.
Practical Applications and Limitations
AI coaches have shown promise in various practical applications, such as training sales agents and aiding in physical rehabilitation therapy. For instance, AI coaches can provide systematic management and improve self-efficacy and motivation in rehabilitation exercises. However, the effectiveness of AI coaches can vary based on the user's rank and familiarity with the technology. Middle-ranked sales agents benefit the most from AI coaching, while bottom- and top-ranked agents face challenges such as information overload and aversion to AI.
Conclusion
In summary, while AI coaches have demonstrated efficacy comparable to human coaches in certain areas, trust remains a critical issue. Users often struggle to assess the competence of AI systems, leading to misplaced trust. Factors such as performance expectancy, social influence, and attitude significantly influence the adoption of AI coaches. Although AI coaches can enhance clarity and willingness to act, they often lack the emotional connection and stress reduction capabilities of human coaches. As AI technology continues to evolve, it is essential to address these challenges to build more trustworthy and effective AI coaching systems.
Sources and full results
Most relevant research papers on this topic