Can AI Beat Humans in Brand Music? Study Says Not Yet

Sonic branding, a strategic use of sound in brand identity, has become a vital element alongside logos and slogans. This approach aims to enhance brand awareness and strengthen brand associations through unique music and sounds in marketing efforts.

With the increasing demand for content and advertisements, there is a growing interest in using artificial intelligence (AI) to create jingles, music, and other sonic elements associated with brands. However, the question of whether AI-generated music and sounds can outperform human creativity remains a significant topic of debate.

A recent study conducted by Stephen Arnold Music (SAM) in collaboration with SoundOut, a leading sonic testing agency, sought to shed light on this question. The study examined the appeal and emotional response of brand music composed by generative AI compared to music composed by human composers. The research involved over 200 consumers and analyzed four categories of compositions: AI-generated, AI-supplemented by human modification, human-generated, and human-supplemented by AI modification.

The study utilized a popular AI platform and directed both the AI platform and a human composer from SAM to create music with a somber and emotive tone, featuring core instrumentation of piano, strings, guitar, and vocals. The results of a blind test revealed that all compositions scored highly in appeal, surpassing the 70% “excellent” baseline. The human composition scored the highest at 78%, indicating a strong preference for human-created music among consumers.

Despite the positive reception of both AI-generated and human-composed music, the study found that AI involvement led to a decrease in overall appeal. For instance, the appeal rating for the human composition decreased to 74% when supplemented by AI modification. Additionally, in terms of emotional accuracy, human-created tracks were found to be more emotionally in-tune with the brief, exhibiting attributes such as loving, discerning, sincere, easy-going, organic, and dreamy. In contrast, AI-generated tracks were rated strongly for attributes like precise, corporate, scientific, focused, and visionary.

David Courtier-Dutton, Founder of SoundOut, noted that while humans still outperform AI in terms of emotional expression, AI’s ability to compose music is not far behind. He emphasized that AI can compose music to evoke emotional responses, but it requires further technical development to match human creativity and empathy.

The study’s key finding was that human compositions outperformed AI-generated material in terms of overall appeal and emotional accuracy. However, the research also indicated that collaboration between humans and AI could lead to highly appealing music. When SAM’s composer modified an AI-generated track, both emotional accuracy and overall appeal improved.

Chad Cook, President of Creative & Marketing for SAM, emphasized that while AI struggles to accurately express human emotion, it excels at creative inspiration. He suggested that AI is best utilized as a tool to inspire and speed up the ideation phase of music composition, rather than delivering the final piece.

Looking ahead, the study suggests that AI is best utilized as a creative tool to assist with ideation and composition, allowing human composers to focus on emotion. While AI can enhance the creative process, human creativity remains invaluable for creating emotional connections with audiences.