AI in music: implications for artists

Generally, artificial intelligence (AI) refers to the simulation or approximation of human intelligence by software-coded machines. In other words, it is a machine’s ability to perform functions that we associate with human minds and which traditionally require human intelligence.

An especially important term related to AI is so-called ‘machine learning’ which refers to a field in AI, which allows systems to learn and improve from experience without being explicitly programmed. Real-world examples of machine learning include face or speech recognition, social media optimization, spam email filtering, product recommendations, etc.

It’s not a secret that, particularly in the past, when most people heard the term artificial intelligence, the first thing they would think of were robots. That’s been mainly caused by Hollywood blockbusters and novels creating stories about human-like machines and their (not very successful) integration into the lives of people.

As we know, most of these narratives don’t have a really happy ending. Instead, they portray robots as the ultimate destroyers of humankind. Even DALL-E 2 by OpenAI, one of today’s most prominent AI image generators, responds with images of robots when asked to generate illustrations representing AI. And not all of them have a very positive vibe about them.

However, the reality is slightly different. Sure, there are indeed robotics companies that design and engineer AI-powered robots which are then used in a variety of fields. For instance, Boston Dynamics develops humanoid robots to make the work of people easier, more effective, and safer, while Hanson Robotics seeks to utilize their inventions in areas such as medicine, education, or entertainment. Although, as you can see in the video below, the Boston Dynamics’ robots are to a certain degree meant for entertainment, too.

Additionally, It was only last year that the generative pre-trained transformers, of which by far the most popular is ChatGTP, brought artificial intelligence into the mainstream. As a result, while some of us still associate the term AI with robots, others perceive it as synonymous with ChatGPT. However, robotics and GPT models represent only a small fraction of ways and fields where AI is being used.

In the music industry, AI technology has been slowly revolutionizing all of its areas, from music production to data analytics to recommendation engine algorithms, and even to songwriting and music composition. Cherie Fu, the founder of the research and intelligence network for the new music business Water & Music, has identified AI as perhaps ‘the most disruptive technology for the music business since the Napster era of piracy’.

That’s because technology comes with a number of challenges, ultimately forcing us to start rethinking our traditional understanding and perception of creativity, ownership, attribution, and ethics.

Only in the last year and a half, tens of different music AI tools and platforms have been released by both independent and big tech companies, such as Google, OpenAI, or ByteDance. Their tools allow users to generate original, custom tracks in just a few seconds using only a text prompt.

Suddenly, people with no prior experience and/or knowledge and skills can make their own music, submit it to streaming services, and generate revenue for it. There are reportedly hundreds of thousands of such AI-generated songs that can be listened to on streaming services, including Spotify, Apple Music, Amazon Music, and others.

This is not necessarily a bad thing, particularly for individuals who aspire to become musicians but lack opportunities, support, or financial resources to start their music journey in any other way. However, there are also such creators that use AI tools to generate music and also clone other artist’s voices. This was, most prominently, the case of a song called ‘Heart On My Sleeve’ featuring voices of musicians Drake and The Weeknd.

The track, which went viral on TikTok and accumulated more than 8 million views, was created by the platform’s user Ghostwriter997, using AI and previous soundbites of the artist’s voices. The outcome is a song that is completely AI-generated and aims to mimic not only the musicians’ voices but also their songwriting and music styles.

Now, for those familiar with Drake’s ghostwriting accusations, you may find the name of the Tikok user slightly amusing. However, it’s exactly the people like ghostwriters and others feeling overshadowed by major labels, who believe that AI-generated music may ultimately lead to music democratization.

The ‘Heart On My Sleeve’ song was eventually taken down from all streaming services, including YouTube, due to copyright infringement claims. However, the question remains whether such musical creation is ethical and what regulations need to be put in place to secure fair conditions for musicians who create their own music without using AI or any other supportive tools.

A key advantage of AI is its ability to analyze a vast amount of data to identify patterns, predict emerging trends within the industry, and discover correlations in music consumption. While such tools are still rather used by bigger companies, they can eventually help musicians and producers create music that is more likely to resonate with their target audience.

AI is provided with data mostly from various streaming and social media platforms such as Spotify, Amazon Music, Instagram, or YouTube. This data includes a variety of metrics reflecting a musician’s performance, such as the number of views and streams, as well as engagement, including data on likes, comments, shares, etc. Additionally, AI models can analyze audio signals, lyrics, and other metadata.

Once all data is collected, machine learning algorithms are able to analyze it and identify patterns and trends that can show whether an artist is gaining popularity and is up to great commercial success with their current art. Such information may also help musicians better understand their audience preferences and thus make data-driven decisions.

AI algorithm models are already being employed by streaming platforms as recommendation systems. The purpose of such systems is to analyze users’ listening habits, patterns, preferences, and behaviors to provide them with personalized music recommendations. This, if working well, can benefit musicians by recommending and playlisting their music regardless of how big and popular they currently are.

Creativity blocks are everything but uncommon among musicians and artists. The worst part is you never know when they strike you. You may be working on a melody or lyrics for months or days and still experience a moment when you don’t know what to do next. Creativity and inspiration are both nowhere to be found and you may start feeling a bit lost.

AI tools may be incredibly handy in such situations. If you’re struggling with the melody of your upcoming song, AI music and sound generators that we mentioned at the beginning of this article, like Soundraw, Boomy, or Soundful, can become the support you need.

The ultimate goal is not to have the tool generate a melody you like (although this may be the case, as well) but rather to kickstart your creativity, again. As a result, you may end up receiving cool improvisational suggestions, generating new ideas and concepts, and exploring unique harmonies and sounds.

The same may be applied to writing lyrics. Text generators like ChatGPT, Scribe, or Chatsonic, can help you define the angle and perspective from which you’d like to write or suggest words that may fit perfectly into your song. You can also search synonyms or antonyms for words and discover unusual phrases. In other words, text generators can generate anything and almost everything (as they are continuously evolving)!

AI can be a great assistant during the process of producing, and particularly, mastering your music. Various AI mastering tools allow for the automation of certain repetitive tasks while improving the sound quality of the recordings. AI algorithms can, for example, detect and remove background noise, enhance audio sound quality, correct timing errors or pitch errors in voice recordings, or even stimulate and augment the sound of different musical instruments.

Additionally, AI tools can analyze and optimize the mixing and mastering processes, making it easier, less time-consuming, and cheaper (no need to pay a professional to have your track mastered) for artists and producers to achieve the desired, balanced and polished sound.

At iMusician, we offer our very own AI-driven Instant Online Mastering tool that enables you to have your track polished and ready for release within minutes. The professional yet affordable tool uses machine learning and offers genre-specific mastering options to ensure that your song will always sound the best. Learn more about our Instant Mastering tool on our product page and start polishing your tracks today!

As you may already know, ChatGPT is a form of generative AI, similar to automated chatbots and services which are nowadays commonly found on customer service websites. People can ask it questions, make alterations to their requests, or ask for clarifications to its replies. In other words, it allows you to specify your topic and have a whole conversation about it.

The interesting thing about ChatGPT, in comparison to other tools, such as Google, is that it doesn’t act as a regular search engine. Instead, it takes and synthesizes information from various sources to create something ‘new’ and ‘unique’ for a specific purpose and/or with a particular perspective in mind.

This can have various uses specifically for musicians and artists. Do you need a hand with your artist bio, a PR release, or, let’s say, creating your artist brand? Not a problem! ChatGPT can create something more or less specific to your needs, taking into consideration your location, target audience, artist characteristics (such as genre, music styles, expertise, etc.), or preferred length of the copy.

With that being said, whatever ChatGPT generates, you won’t get 100% finished results. You may still need to personalize the product and add details about yourself and your music that the generator simply doesn’t have. This means that the purpose of ChatGPT is not to do the whole work for you but act as the starting point in the creative process.

You should also bear in mind that ChatGPT’s tone of voice may feel a bit ‘unnatural’, meaning it simply doesn’t say/write things the way you’d do so. You’ll get a piece of text that you can work on until it feels right.

As we’ve outlined, ChatGPT can generate any type of text and so experts claim it might be the perfect tool for content creation for everything around music. This also includes activities for audience engagement, such as social media post captions, DM responses, or blog articles.

Once again, this doesn’t mean, you’ll make a request to ChatGPT and copy and paste its response, as some sources may suggest. Instead, you’ll get a workable copy that you can further amend in accordance with your needs, or which you can get inspired by (particularly in the case of blog articles).

Just be cautious about the correctness of the information you’re provided. There’s no guarantee that the data ChatGPT generates in its texts are completely or definitely true. Naturally, in most cases, you’ll want to have created copies about you and your music but if you happen to make some fact-based posts or comment, it’s essential that all facts are correct.

In addition to generating texts for content creation, ChatGPT can also be well-used as a planning and organizational tool. This can come in handy when you’re planning a tour, creating a marketing strategy or simply doing something that requires a plan, itinerary or a list.

Planning a tour, for example, is usually a lengthy process during which a lot of tasks need to be undertaken. Compiling an itinerary is therefore recommended but making it yourself may take a lot of your time and energy that you can rather spend on making music.

ChatGPT can take the details you provide about the tour, e.g. location, time duration, budget, target audience, and turn it into an itinerary customized to your needs. The same can be applied to your promo plan.

You simply provide the information regarding your promo activities (e.g. what kind of activities and approximately when) and watch the generator transform it into an organized plan within seconds. As said before, it’s important that you thoroughly proofread the text you’ve been given to confirm all data included is correct before you start using it. If you’re unsure about some technicalities, don’t hesitate to ask a professional in the field.