Artificial intelligence (AI) is rapidly transforming higher education, with its influence now extending into specialized artistic disciplines. A recent study, published in Frontiers in Psychology, has provided empirical evidence on how AI-generated "second opinions" impact crucial psychological and cognitive outcomes for postgraduate students in vocal accompaniment. The research highlights that integrating AI as a supplementary feedback tool can significantly boost self-efficacy and diminish performance anxiety, while its short-term effects on decision-making styles remain inconclusive. AI as a Confidence Booster in Advanced Music Training The study, conducted by Tong Zhu and Lu Xia from Henan University, focused on postgraduate students specializing in vocal accompaniment, a field demanding high levels of interpersonal collaboration, intuition, and nuanced judgment. The researchers investigated whether AI could serve as a valuable tool to support these developing artists, particularly during their critical stage of professional identity formation. Vocal coaching is a complex art form that extends beyond mere instrumental accompaniment. It involves deep musical analysis, stylistic interpretation, psychological support, and real-time artistic co-creation between the singer and the accompanist. This collaborative process is inherently subjective and dependent on mutual trust and understanding. For postgraduate students, navigating this uncertainty can be a significant challenge, often leading to self-doubt and hesitation in their professional interactions. Traditional pedagogical methods, relying heavily on the instructor’s "first opinion," can inadvertently limit students’ independent critical thinking and artistic voice. In response to these challenges, the researchers explored the concept of an AI "second opinion," drawing parallels from its successful application in fields like medical diagnosis and financial investment. The core idea is to introduce an independent, data-driven perspective to complement human judgment, thereby reducing cognitive biases and broadening decision-making horizons. Contemporary AI, particularly in music information retrieval and generative modeling, is capable of sophisticated analysis and offering stylistic suggestions based on vast datasets of historical performances. This capability positions AI as a potential augmentative tool, providing a verifiable and discussable alternative viewpoint for students. Methodology and Findings: A Controlled Study The study employed a pre-test and post-test experimental design involving 150 postgraduate students from three leading universities in China. Participants were randomly assigned to either an experimental group, which received AI-generated second opinions, or a control group, which worked without this additional AI input. The AI intervention utilized the intelligent arrangement module of the NetEase Tianyin platform. Students created a MIDI-based accompaniment for a designated art song, and the AI system then generated alternative harmonic, rhythmic, and expressive suggestions based on its learned patterns. These AI outputs were presented to the experimental group as a "second opinion" to inform their revisions. Key psychological and cognitive outcomes were measured using established scales: the General Self-Efficacy Scale (GSES) for collaborative confidence, the Music Performance Anxiety Inventory (MPAI) for anxiety levels, and the Decision Style Inventory (DSI) for decision-making approaches. The results revealed significant positive effects of the AI intervention on confidence and anxiety. The experimental group demonstrated a statistically significant increase in self-efficacy compared to the control group (p = 0.026). Furthermore, within the experimental group, there was a significant improvement in self-efficacy from pre-test to post-test (p = 0.007). Conversely, music performance anxiety was significantly lower in the experimental group than in the control group (p = 0.013). A notable within-group reduction in performance anxiety was also observed in the experimental group (p = 0.007). These findings suggest that the AI second opinion effectively mitigated apprehension and bolstered confidence in the students’ abilities. However, the study found no significant differences in decision-making style between or within the groups (all p > 0.05). This indicates that, within the short timeframe of the experiment, the AI intervention did not alter students’ fundamental tendencies towards intuitive versus analytical decision-making. Analysis of Implications: AI as an Augmentative Tool The research team interprets these findings as strong support for the role of AI as an augmentative tool that enhances human interpretative agency, rather than a force that reshapes fundamental cognitive patterns in the short term. For postgraduate students grappling with the inherent uncertainties of artistic collaboration, the AI second opinion appears to serve as a valuable, non-evaluative reference point. By providing alternative perspectives, AI can reinforce students’ initial decisions, encouraging them to explore and justify their artistic choices more thoroughly. This process, in turn, can reduce the anxiety associated with self-doubt and foster a greater sense of confidence in their collaborative capabilities. The stability of decision-making styles suggests that AI’s influence may be more about how students approach their existing decision-making frameworks rather than fundamentally changing them. Students might use AI suggestions to validate intuitive hunches or to gather more data for analytical processing, without necessarily shifting their preferred style. This aligns with the study’s theoretical framework of Human-AI Collaborative Augmentation (HACA), which posits that AI should scaffold human judgment and support interpretative agency, thereby expanding cognitive resources rather than substituting human capability. Broader Context and Future Directions The integration of AI in higher education, particularly in creative fields, is a rapidly evolving area. While AI is increasingly used for tasks such as pitch correction or compositional assistance, its impact on higher-order cognitive and psychological processes in advanced collaborative settings remains less explored. This study addresses that gap by focusing on the specific context of postgraduate vocal accompaniment. The findings have significant implications for pedagogical approaches in music conservatories and arts institutions. As AI technologies become more sophisticated and accessible, educators and curriculum designers will need to consider how to integrate them responsibly. The study suggests that framing AI as a supplementary resource for reflection and exploration, rather than as a definitive source of truth, is crucial. This approach can empower students to develop their unique artistic voices while benefiting from the analytical capabilities of AI. While the study provides valuable insights, it also acknowledges certain limitations. The sample was drawn from a specific demographic and geographical context, which may affect the generalizability of the findings. The intervention’s duration was limited, and its long-term effects on actual performance outcomes remain to be investigated. Furthermore, the absence of an active control group means that the observed benefits might be partly attributable to the general effect of receiving additional feedback, rather than solely to the AI’s unique contribution. Future research could benefit from longitudinal designs, qualitative methodologies to explore student experiences in greater depth, and the development of adaptive AI systems that cater to diverse decision-making styles. In conclusion, this research offers compelling evidence that AI second opinions can be a powerful tool for enhancing collaborative confidence and reducing performance anxiety among postgraduate vocal accompaniment students. By acting as an augmentative force, AI can support students in navigating the complexities of artistic collaboration, ultimately contributing to the development of a new generation of artists who are technologically adept and artistically confident. Post navigation Self-regulated learning constructs as predictors of academic performance among Saudi EFL students in virtual learning environments: Insights from self-regulated learning theory
