Image: © ginettigino/Stock.adobe.com
Old Town Road by Lil Nas X is breaking all sorts of chart records, but now AI is trying to figure out what genre the song fits into.
A new AI tool developed by University of Southern California PhD student Timothy Greer automatically predicts music genres based on how lyrics and chords interact with one another throughout the song.
Set to present his AI at an upcoming conference, Greer – a lifelong musician – tasked his creation with trying to figure out what genre the hit song Old Town Road by Lil Nas X falls into. Earlier in the year, the song was controversially removed from the Billboard Hot Country Chart because, according to its curator, it did “not embrace enough elements of today’s country music to chart in its current version”.
Now, using his algorithm, Greer has shown that its lyrics definitely define it as a country track, but its chords – based on a music sample from Nine Inch Nails – see it becoming infused with the rock genre. When combined, it becomes a straight-up pop song.
“Old Town Road is an interesting song,” said Greer. “The lyrics are steeped in the country genre, but the chords and the instrumentation don’t sound like country at all.
“The algorithm highlights the complexity of music, both in terms of how the music is constructed and how it is perceived – in other words, how people process it.”
A total of three models were tested on Old Town Road using only chord embeddings, only lyric embeddings and using chord-and-lyric embeddings combined. He trained the system on a dataset with 190,165 musical segments from 5,304 pop songs with lyrics and corresponding chords.
Previous genre prediction tools have used a song’s entire audio file requiring the processing of high-quality recordings. However, Greer said the benefit of his tool is that it can search through Google results of chords and lyrics substantially quicker.
“This interplay between chord sequences and lyric sequences may give us a better glimpse into how we perceive genre than using either alone, although both of these modalities contains useful information alone as well,” said Greer.
In terms of applications, he said that it could be used by the music industry to better market and tag content, as well as influence neuropsychology and the mechanisms of human thought.
