Scientists Use AI to Predict Popular Songs, Potentially Improving Music Apps
A team of scientists has made a groundbreaking discovery in music prediction by utilizing artificial intelligence (AI) to analyze brain activity. Their findings could potentially offer a more accurate method for predicting popular songs compared to popular streaming services like Spotify. The AI model developed by the scientists may soon enhance our favorite music apps.
The versatility of AI development is an intriguing aspect of artificial intelligence. It has the ability to be applied to fields that seemingly have little connection to technology. In this case, scientists have found a way to discover new songs online, bringing great news for those seeking to refresh their playlists.
According to psychological research news outlet PsyPost, the research team, led by Paul J. Zak, aimed to predict popular hits using AI due to the low accuracy of methods used by streaming services like Pandora and Spotify. These platforms rely on proprietary digital systems to recommend songs, but less than 4% of new songs become hits. Analyzing lyrics, social media mentions, and other methods have also proven to be unreliable.
To address this issue, the researchers explored whether measuring neurophysiologic responses to music could help predict hit songs. Neurophysiologic responses include heart signals associated with emotional resonance and attention. In other words, the brain’s response to specific lyrics and melodies affects the heart.
Claremont Graduate University Professor Zak and his team conducted a study involving 33 participants who listened to 24 songs from a streaming service. The researchers measured the participants’ brain responses and derived three variables from the data: average immersion, peak immersion (the highest moments of immersion during a song), and retreat (the lowest 20% of immersion moments).
The researchers also asked the participants about their preferences for each song and whether they would replay and recommend it to others. By analyzing this data, the researchers were able to predict market outcomes, such as the number of song streams, based on a small amount of data. They referred to this method as “neuroforecasting.”
Statistical approaches achieved only 69% accuracy, but machine learning, a form of AI, raised the accuracy to 97%. The researchers acknowledge the limitations of their study, such as the small sample size, and plan to conduct further research to validate their findings. However, their study has provided valuable insight into the unconscious mind through the use of AI.
The AI music study mentioned above is not the only research conducted on the human brain using artificial intelligence. Elon Musk’s Neuralink has been developing brain implants for various purposes, including the ability to surf social media using only thoughts. Osaka University researchers have also created an AI that can read minds, training it to link specific brain activity to images with 80% accuracy.
In conclusion, Professor Paul J. Zak and his team have conducted a groundbreaking AI music study, discovering a method for predicting hit songs using artificial intelligence. While further research is needed to confirm their findings, their study offers new insights into the human mind and the potential to enhance music apps like Spotify. To learn more about the study, it is featured in Frontiers In Artificial Intelligence.
