Demo singers and their stakeholders are vital contributors as hidden service groups within the recording industry. Their collaborative efforts are crucial for content creation, navigating the market trends, particularly in the rapidly evolving digital music sphere. We utilized qualitative approach, which involved observing studio singers (N=10) and interviewing stakeholders (N=16), to reveal the three primary levels of co-innovation outputs: cultural, emotional, and technical output. Through analyzing the stakeholders' satisfaction within outputs, the research also delved into a comparison of the levels achieved by AI singers versus human singers. Notably, the findings suggest that while AI singers excel in emotional and technical outputs, they currently fall short in achieving the highest level of output, which is cultural. The research holds valuable implications for the design and reform of music education practices.