Intro
Music and technology have always been my passions; I’ve been a songwriter/recording artist for fifteen years and a software developer for close to twenty. Even though I haven’t locked in my thesis project topic as yet, it will likely involve some intersection of these two areas. Artificial Intelligence, or AI, is the hot new wave sweeping the software industry (and many others, if we’re being honest). Its ripple effects are so far-reaching that I wouldn’t be surprised if grandmothers use AI for recipes. This ubiquity of AI inspired me to explore this topic, specifically about how it affects the music I love to make, and whether the industry can even survive its onslaught, as some doomsday pundits fear. Investigating the challenges and opportunities for musicians in this brave new world is essential. There is a lot of misinformation and fearmongering in this space, most of it due to a need for more research.
Is AI in music a new thing?
The first thing to know about AI and its relationship to music production is that, depending on your definition of AI, it isn’t exactly new. According to IBM, “Artificial intelligence (AI) is technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy“. If we focus on “simulate human learning”, Auto-Tune, a plugin that “fixes” notes by adjusting the pitch, is a form of AI. It stimulates the vocals of a singer who understands the technique of vocal modulation to achieve a pitch-perfect note that sounds “in key”. It’s a very simplistic form of AI, but its algorithms are providing intelligent tweaks on vocals.
Auto-Tune was developed by Antares and released in 1997. It was popularized by the billboard-topping hit song by Cher titled “Believe”. At the time, there was much controversy surrounding its usage. In the current era, almost every song on the radio has a dash of Auto-Tune, even for talented singers who have no problem singing in key.
There are plugins to fix the rhythm of drums on a track to either make it precise to an inhuman degree or just off-kilter to mimic a “live feel”. Melodyne is a plugin that goes beyond fixing pitch and promises to put your singing “perfectly in tune“. Landr, a mastering plugin that has been marketed as AI and promises to do as good a mastering job as a human sound engineer, has been available for a decade. In the last two years, the music industry has acted like AI has now arrived when we’ve been living with it for years.
However, it would be foolhardy to pretend that the recent wave of AI is a small step up from the tools of yesteryear. In the past, “AI tools” were limited to automating and streamlining the technical, more tedious parts of music production; the creative side was kept in the human domain. This is no longer the case. Not too long after ChatGPT launched in November 2022, a “new release” by Drake featuring the Weekend emerged on the YouTube platform. This was common, except that the performers on the track were neither Drake nor The Weekend. It was an AI-generated sound-alike performing over original production. Many were fooled.
This raised serious questions regarding copyright, intellectual property and the eroding protections that artists could expect even for their works. Music labels rushed to take down similar copycat entries, and months of whack-a-mole releases and takedowns continued. A cloud of doom and gloom hung over the industry. All pundits and analysts could speak about was whether we were looking at the end of days.
As an artist myself and owner of a recording studio, I initially felt threatened by AI. If anyone can type a prompt into a textbox and receive creativity wholesale, then what is the value of creativity at all? I’m not a pessimist, but I did see AI as the enemy. I didn’t even want to understand my enemy. I prayed it was a fad and hoped it would disappear like Napster.
Luckily, my curiosity got the better of me, and I started researching. I began to use ChatGPT and experiment with voice generation technology. I used it to assist me in writing a novel I was working on — not to generate the story itself but for research and idea generation. Suddenly, a light bulb went off. This isn’t a doomsday device. It’s a tool like any other. And it isn’t even infallible, just good enough to be useful. Very useful. It wasn’t as sinister as I made it to be. And it was coming, whether I liked it or not. So I better be prepared.
Why is this important?
Creatives must understand, like I eventually did, that AI is not the enemy. Traditionally, our local music industry has a pretty quick adoption rate of new technology once it’s affordable. We quickly jumped from analogue production using physical tapes to digital production largely due to it being low-cost and favourable to small studios. Similarly, we need to quickly investigate AI. If we want to continue contributing globally and pushing our unique culture, we need to discover where we fit in this new AI world and carve out our position.
This study is very important to our country because of how much creativity (carnival, soca) is related to our national identity. Creative practitioners need to clearly understand the challenges and opportunities with AI. One main challenge is intellectual property protection. The fact the AI algorithm trains on original works is a huge issue that doesn’t have an easy solution. The best-case scenario is the one that leads to the owners of those works getting paid fairly. We need reliable systems that haven’t yet been developed to track what content these algorithms use to learn. If an algorithm uses 1% of my song to create an original work, I should receive 1% royalties. Maybe this can be set via a smart contract on a blockchain to automatically bless my bank account. Until then, we need to keep our eyes on this issue to ensure that we are in line to benefit when things are implemented.
Another reason this study is important is how it links to the Sustainable Development Goals or SDGs. One key SDG that it addresses is “Reduced Inequalities”. Music has often been an avenue out of poverty and challenging situations. Conversely, technology, or lack of access, has often increased the chasm between rich and poor. Without studies like this to shed light on such a nascent field and its effect on music, it can remove this avenue for creatives and cause even more inequality. We must ensure that after 2024 and beyond, we preserve an environment where a future Ellie Mannette still has the agency to create a national instrument out of a discarded oil barrel.
Trinidad and Tobago is one of 193 countries that adopted the 2030 Agenda for Sustainable Development (taken from a post on the United Nations’ website about the 2030 Agenda). This means we have a mandate to push forward the seventeen outlined goals. In a nutshell, the vision for these goals is transformation to a better world for those to come.
Conclusion
It’s been exciting researching AI because of how new it is. As fast as new discoveries emerge, others are building even newer ones on top of it. I hope that through my study, key information can be disseminated to those who need it most. Hopefully, it can at least reach the decision-makers who can set adequate policies to protect our copyright and promote further learning. Culture is very important to Trinidad and Tobago, and I would like to see it continue to thrive as we enter this new paradigm.
References
Kaur, R. (2024). HUMAN RESOURCE MANAGEMENT AND ARTIFICIAL INTELLIGENCE: TRANSFORMATIVE EFFECTS AND FUTURE PROSPECTS. International Journal of Education Modern Management Applied Science & Social Science. https://doi.org/10.62823/ijemmasss/6.3(ii).6911
Alben, A. (2020). When Artificial Intelligence and Big Data Collide—How Data Aggregation and Predictive Machines Threaten our Privacy and Autonomy. AI Ethics Journal, 1(1). https://doi.org/10.47289/aiej20201106
Elliott “Ellie” Mannette. (n.d.). National Endowment for the Arts. https://www.arts.gov/honors/heritage/elliott-ellie-mannette
United Nations SDG Website, https://sdgs.un.org/
What can Melodyne do? (n.d.). https://www.celemony.com/en/melodyne/what-can-melodyne-do
CBC News: The National. (2023, April 21). AI-generated Drake/Weeknd deep fake song rattles music industry [Video]. YouTube. https://www.youtube.com/watch?v=COok3YO1L5w
AI in the Music Industry: Transforming Music Production, Discovery, and Data. (n.d.). https://www.dataart.com/blog/ai-in-the-music-industry-transforming-music-production-discovery-and-data-by-sergey-bludov
What can Melodyne do? (n.d.-b). https://www.celemony.com/en/melodyne/what-can-melodyne-do
Lamont, T. (2020, August 25). Napster: the day the music was set free. The Guardian. https://www.theguardian.com/music/2013/feb/24/napster-music-free-file-sharing
Leave a Reply