Music speaks to our souls in mysterious ways. But scientists in South Korea may have discovered why those musical instincts exist — and they unlocked the answers with artificial intelligence.
The team, led by physicist Hawoong Jung, trained a deep learning model to categorize sounds from the AudioSet database without any music examples. Surprisingly, some neurons became selectively tuned to musical features over time.
Jung noted it was intriguing that the neural network appeared to develop an ability to detect music patterns, even though it had not been directly trained on any musical sounds. Instead, the network seemed to pick up on musical patterns instinctively while learning to distinguish natural environmental noises.
Further tests showed the music-focused neurons encoded temporal song structures like rhythm and melody. Just as importantly, blocking their activity hurt the model’s performance on other tasks.
Professor Jung, who advised the research, said, “The results of our study imply that evolutionary pressure has contributed to forming the universal basis for processing musical information in various cultures.”
The researchers believe musical instincts arose from evolutionary pressures as an adaptive strategy. By developing a better capacity to recognize patterns in natural sounds like animal calls or storms, early humans gained benefits for survival.
Over generations, this conferred advantages for detecting threats, opportunities, or simply staying together as a group through song. Musicality became hardwired without conscious training — and continues influencing us today.
Jung sees potential applications in generating personalized music therapy. By describing desired qualities, patients could obtain specially tailored compositions. AI may also inspire new creative frontiers when musical curiosities merge with machine learning frontiers.
He also commented on the limitations of the study, adding, “This research however does not take into consideration the developmental process that follows the learning of music, and it must be noted that this is a study on the foundation of processing musical information in early development.”
With more research, Jung hopes to better model the developmental learning involved after our innate foundations. But this study brings us closer to understanding music’s deeper impact — and our inbuilt need to find rhythms in life.