Today, there are probably no corners of the world where artificial intelligence (AI) has not been heard of. But despite its wide distribution, not everyone understands its essence, its importance for modern mechanisms of civilization development and the consequences that can await humanity with the development of AI.
After all, artificial intelligence is a self-learning technology that transforms every sphere of life and a universal tool that allows people to rethink the way we integrate information, analyze data and use the resulting knowledge to improve the decision-making process.
Most people are not very familiar with the concept of artificial intelligence. For example, in a 2017 survey of top business leaders in the US, 1,500 respondents were surveyed, only 17% answered that they were familiar with AI. Many of them did not know what it was and how it would affect their companies. They understood that there was significant potential to change business processes, but they did not know how they could implement artificial intelligence in their organizations.
It’s only been 6 years and today AI will be applied in almost every field of activity – in finance, national security, health care, criminal justice, transportation and smart cities, and we also consider issues such as data access problems, algorithmic bias, AI ethics and transparency, as well as legal responsibility for AI-based decisions.
Artificial intelligence is already changing the world and poses important questions for society, economy and management.
The working principle of AI
Artificial intelligence algorithms are designed to make decisions, often using real-time data. They differ from passive machines, which are only capable of mechanical or predetermined responses. Using sensors, digital data, or remote inputs, they combine information from multiple sources, instantly analyze the material, and act on the conclusions drawn from that data. Thanks to significant improvements in data storage systems, processing speed and analytical methods, they are capable of extremely complex analysis and decision making.
Of course, the development of AI could not fail to touch such a huge field of human activity as art and, in particular, the music industry. In 2023, the hype and hysteria surrounding the use of AI in music has been mixed with a flurry of AI-generated soundlikes that have demonstrated the potential to transform artistry – and fandom – as we know it, while many companies evaluate how best to protect their artists, copyright rights and revenue streams from the growing threat.
But not all AI in music are “fake artists”. Already today, it is possible to single out several ways in which AI can influence the development of the music industry.
1) Revolution in production
The portability and availability of technology allows you to create professional-sounding music without special knowledge and professional skills – a beginner artist who can afford Apple products can start messing with production in GarageBand or buy a “set of beats” on the Internet and record vocals using a phone. Users of the Boomy app, for example, can select a few relevant options and generate an instrumental track in seconds, which can then be rearranged, rebuilt, or recorded with vocals over it. BandLab’s SongStarter can generate instrumentals based on specific text and emojis.
2) Obtaining stems
AI technology has the ability to break songs into component parts known as stems. These are a kind of sonic building blocks, highlighting which can be important if, for example, a film director wants to use an instrumental version of a track in a trailer, or a brand wants to include a vocal a cappella performance in a commercial. Some musicians have lost their stems over time; other artists may have cut up albums before recording technology came along to isolate all the different parts, and now those albums may end up in the hands of catalog owners looking for new revenue opportunities. For example, producer Rodney Jerkins used AI technology to extract the audio of Wu-Tang Clan’s “Ol’ Dirty Bastard” from a VHS tape and sample it for SZA’s track.
3) Flow
The modern music industry was created in a world where the supply of professional music was quite limited and mostly controlled by a few large companies. But with the development of artificial intelligence technologies, it became possible to create a music stream very quickly. This, of course, caused considerable concern among the major labels. Because if AI is able to increase the amount of music being created outside of the purview of major labels, it could hurt their payouts under the streaming business model, which involves paying pro rata for services.
4) Personalized soundtracks
A number of startups are creating flexible music that changes in real time to highlight action in video games, virtual reality, workouts and Snapchat filters using cutting-edge technology. Companies like Reactional Music, Life Score, Minibeats and others, often called “dynamic” or “personalized” music, use artificial intelligence not to create music at the push of a button, but to take human-made music and shuffle it its individual elements (called “stems”), creating new compositions that best emphasize the user’s needs and actions, just like the soundtrack to your favorite scene.
5) Pitch Records
Some songwriters and publishers are now experimenting with artificial intelligence-based voice synthesis technology that helps them place their compositions among top artists. Today, “pitch records” – songs that only professional songwriters write and then pitch to recording artists – can be a particularly difficult sell as more artists want to play a bigger role in the songwriting process, so artificial voice technology is helping advanced publishers and writers show to the team of artists how the singer might sound in the track before they even record it.
AI in the Ukrainian music industry
In 2019, Warner Music became the first label in the world to contract with artificial intelligence to create music. The music was generated by the Endel program, making it personalized, taking into account the user’s heart rate, geolocation and weather conditions. The label planned to record 20 albums with the Endel music mood app and 5 albums have already been released under the titles: Clear Night, Rainy Night, Cloudy Day, Cloudy Night and Foggy Morning. Investors of the albums were Amazon Alexa Fund and Jillionaire Major Lazer.
In Ukraine, the first music album completely generated by artificial intelligence was an album called RoboWave, the author of the idea is rapper Krechet. The cover was created by Midjourney, the lyrics by ChatGPT, the music and clip were also written by AI.
The album features four tracks: Minecraft, Wednesday Addams, Huggy Wuggy and Garten of Banban. The most successful Minecraft clip has collected more than 25,000 views in 12 days. Krechet’s previous music video for the song “To Parents” with singer Kola received more than 53,000 views in three months. The album’s most popular track on Spotify is Garten of Banban with nearly 16,000 streams.
Singer Kola released an AI clip for the song “Other We”. The video was shot on the streets of Kyiv using single-frame video. So far, the AI is a work in progress – the tool adds various effects and objects to the clip, such as flowers and houses.
As the singer herself says: “We used AI only in some parts of the video and to create the cover.” The artist “collaborated” with the Stable Diffusion AI model, which adds new objects or changes existing ones in the finished images. It is enough to write down what should change in the picture and what should not.
But working with AI has its drawbacks – it takes a lot of time. Generating an image for 15 seconds of video takes about three hours, says cameraman Oleksandr Bagaziy, who works on the Kola clip. It usually takes about two weeks from the end of filming to the creation of a clip. In the case of using AI, the work can last one to two months. But if time is spent, production will compensate for the decrease in price. Specialists are expensive and it is not known how much a clip would cost if such effects were created by a person.
Neural networks can make the organization of concerts cheaper, Yevhen Lysenko, the co-founder of Concert.ua, is convinced. In his opinion, AI can reduce the number of hours spent on organization, provide more accurate cost planning and optimize the use of resources.
MOZGI Entertainment turns to neural networks for solving operational tasks, optimizing processes and gathering information.
The soloist of the Go_A group, Kateryna Pavlenko, likes the possibility of having an AI assistant, but she is convinced that artists who try to put in their soul, and not stamp tracks, will not have a significant impact on AI, because nothing can replace the artist’s energy, his unique style.
But to all the artists mentioned above, in 2021, the leader of the world-famous guru “Ocean Elsa” Sviatoslav Vakarchuk said that the song “Without you, I don’t exist” was written by artificial intelligence. How was it? As the artist himself says, “The program listened to all OE’s songs and wrote its own… well, it turned out like that in some places.”
So, not every case of using new technology involves instant creation of computer songs or voices. While some applications of AI certainly raise pressing legal and ethical issues, there are also many applications that provide new creative opportunities for musicians and rights holders, from music creation to release and beyond.
artificial intelligence / copyright / intellectual property / music
