Deepfake videos are one of the many trends keeping the quarantined masses entertained, and many of them involve music! Though deepfake might not go down as an invention that forever changes the music industry, they do open some doors of opportunity for musicians, they are fun, and the technology could certainly be cause for related advancements in the music industry.
AI is nothing new in music, either, and thanks to algorithms and such, many consumers are finding it easier to be paired up with the music they enjoy. Here are a few ways that AI is affecting the music industry during COVID.
Music streaming is, by far, the most common way we consume music now, especially younger generations. With radio stations customized just for fans of a given band or genre, and monthly subscription options to dodge the ads, streaming services like Spotify and Apple Music are just so easy to use, and deliver the opportunity to have a truly unique listening experience based on taste, it doesn't seem like they are going away anytime soon, even if artists are taking a bit of a hit to the wallet because of it. (Many argue that the ability to share the music with a larger audience allows them to gain fans easier and cheaper than touring, thus balancing out the lack of music sales that would have previously came from CDs and such.)
Why are they so convenient, though? Artificial Intelligence and machine learning, is your answer! When users stream music, there are pre-programmed algorithms that tell Spotify "if a user likes this band, they will most likely enjoy this one too," in simple English, anyway. This allows less popular band to jump on coattails of more popular bands on their way to consumer ears. Machine learning takes it a step further when it comes to artificial intelligence, and actual gathers information from the consumers and "learns" from it, and acts accordingly. For example, the hip hop industry, like most genres, has a slew of lyrical content ranging from that which empowers a marginalized society to grow, to thought provoking wordplay, to religious lyricism, and, indeed, songs about the tribulations of being subjected or involved in gang violence.
With such a wide breadth, it's uncommon for someone to like all hip hop, and with machine learning, consumers simply click a button when they like or dislike a song, and the AI utilizes the data to make more granular playlists that they think like-minded listeners can enjoy and share.
Where many musicians have an understandable case against live streaming services due to the relatively small payouts for streaming, many more may have concerns with deepfake technology as it continues to evolve. There are plenty of good uses, like quick re-recordings and dub abilities, but with those dub abilities comes a lot of potential for manipulation and illegal use of music. Deepfake tech has already been the motor behind scams putting celebrity faces on nude bodies and things of the like, and now as it continues to grow in the audio world, artists fear that their music may be misused or more easily stolen.
AI and Music – The Future
As with any technological breakthrough, keeping it in the right hands is paramount. If AI continues to be used for good, especially in recording and target demographic recognition, it should be viewed positively by the music community. If it ends up being manipulated, however, musicians will need to work together to combat the "bad guys," and history has shown that musicians are pretty darn good at teaming up when something tries to thwart their passions and careers.