Unless you’ve been living under a rock, you’ll be well aware that artificial intelligence has already had a significant impact (both positive and negative) on the music industry, but it’s about to turn our business upside down.
I’ve covered potential AI applications for the industry here many times, but given the resurgence of interest, not least with the open letter from Elon Musk and over a thousand leaders in the field asking AI labs to pause for at least six months, I thought it might be good to dig a little deeper into the threats AI poses, as well as the opportunities.
One major concern about AI’s impact has always been its potential to displace jobs, which is valid, as this is already happening. With the emergence of AI-powered creation tools and legal capabilities, it’s becoming increasingly feasible to produce music and evaluate contracts without human involvement. This could result in job losses across various roles and reduce revenue for people including songwriters, performers, musicians, composers, arrangers, producers, lawyers and mastering houses.
AI is only as good as what it’s been fed. Take a moment to ask ChatGPT to tell you about one of your artists, there will be holes depending on where it’s scraping its information from. Now is the time to make sure your brand or artist’s online presence is up to date!
It is, undoubtedly, easier to infringe on copyright with AI. With the availability of AI generation tools, it’s possible to create music that sounds similar to existing songs, which will lead to a rise in copyright infringement cases. The recent Human Artistry Campaign launch is an example of collective action industry groups are taking to lay responsible foundations for AI use in music.
AI also has the potential to utterly saturate the market, taking away meaningful income from some sectors. An example is Activision, who are trying to patent the ability to generate personalised AI-generated music in video games. Outside of sync income loss, with issues around copyright, this could be a smart move from the gaming industry and a warning for the music business.
AI is only as good as what it’s been fed
Another threat to the music industry is the potential loss of identity and a serious lack of diversity. The business has long suffered from diversity issues, but AI could make that far worse.
Bias in AI can arise from the fact that humans are responsible for providing the data that algorithms use, so if diverse teams are not employed, unconscious biases might enter machine learning models. As a result, AI systems will perpetuate and automate these biased models.
Music generated by AI may also start sounding similar across the board. We’ve seen this play out in a human capacity on platforms like TikTok (‘make a beat that will thrive on TikTok’), so if you think it won’t happen with AI you are, quite frankly, wrong.
With AI, there is also the potential for misuse. AI video and image generators can be misused to create deepfakes or other types of misleading content. This could damage the reputation of artists or labels and create legal and ethical issues.
However, alongside the risks, there are of course tremendous opportunities…
AI offers several new ways of creating music. With AI-powered music generation tools, it’s possible to create music that is unique and has never been heard before. This could lead to the development of new genres of music and a new era of creativity within the industry.
AI-powered music recommendation systems have the potential to improve music discovery; it’s been in play for years across the major DSPs but it could be better for sure. These systems can analyse user listening patterns and recommend music that the user is likely to enjoy. If they get it right, it will propel new and existing artists into the right ears.
And then there’s faster content creation! AI video and image generators can create content much faster than humans. This could help artists release music videos and other promotional materials quickly, keeping fans engaged with minimal outlay for resources.
AI has vast potential to enhance music production cheaply. With AI-powered tools, you can create music that sounds like it was recorded with instruments in expensive studios with legendary mixers.
As mentioned earlier, whilst on one hand they will fuck with a million copyrights, AI-powered copyright protection systems do also have the potential to improve protection. These systems can analyse music and detect copyright infringement, making it easier for rights holders to get their money, just like YouTube’s audio recognition has done for years. And we all know how lucrative that is for UGC.
There are so many more AI companies and tools than my word count will allow here, but here are some that you should be aware of within the AI space: SampleRobot, Moises, AIVA, Jukedeck, Landr, Amadeus Code, AudeoBox, Lumen5, Articulate Instruments, iZotope Ozone, Wibbitz, Vidnami, Melodrive, Animoto and IBM Watson.
The people who move with the times will thrive in this new environment, but make the wrong move, and you may be out of a job.