ARTIFICIAL intelligence has had another major breakthrough this month, with US researchers saying they have been able to utilise machine learning to predict whether a song will become a hit.
The research published in Frontiers in Artificial Intelligence detailed experiments in which their AI platform observed and learned from brain scans of people listening to music, and could predict which songs would be a banger with 97% accuracy.
Applying the system to "high frequency neurophysiologic data" significantly improved song prediction accuracy, with the scientists from Claremont Graduate University saying "this can open doors to providing consumers with the entertainment they are looking for, rather than flooding them with options".
With tens of thousands of songs released every day, radio stations and streaming services find it difficult to select which ones to add to playlists.
Analysing the brain responses to the music is a new approach, and involved capturing the neural activity of just 33 test subjects to undertake the accurate "neuroforecasting".
Despite some limitations in the study, including a lack of diversity among the demographics of participants, the researchers believe their approach is likely to be applicable to providing accurate popularity predictions for other forms of entertainment including movies and TV shows.
The above article was sent to subscribers in Pharmacy Daily's issue from 22 Jun 23
To see the full newsletter, see the embedded issue below or CLICK HERE to download Pharmacy Daily from 22 Jun 23
