By Dr. Maya Ackerman
Only a few years ago, aspiring artists who wanted to break into the music industry had the same narrow pathway to success: Recording songs in a studio, sending their demos to A&R reps at record labels around the country, and hoping that they get discovered.
But technology has disrupted the music industry in a big way, and aspiring musicians no longer need to wait around to get discovered by a record label in order to find success. Whether it’s through creating a strong YouTube following or by releasing original music through SoundCloud, young artists are finding new ways to break out on their own, turning the traditional music industry on its head.
But thanks to new technology, it’s not just music platforms that are being democratized—new artificial intelligence apps and programs that have been developed in recent years are transforming the music-making process, as well. Emerging artists are starting to use AI-based systems to create their albums and release original music without the help of big producers or studios.
Discover how artificial intelligence is used within the music creation space to empower aspiring artists in a new age of music: Computational creativity: The next generation of AI
The rise of deep learning led to radical improvements across a wide array of applications, from facial recognition and fraud detection, to language translators and home assistants like Google Assistant and Amazon’s Alexa.
Machine learning typically requires the collection of large amounts of data, which is subsequently analyzed in order to extract reliable pattern. All of those patterns then act as predictors for future data use.
Despite the wide net cast by machine learning, until recently it was believed that creative applications are far beyond the reach of AI-based systems.
But for decades now, another subfield of AI was brewing: computational creativity. This is a subfield of artificial intelligence that explores the creative capabilities of AI-based systems, from autonomous creative machines to those that help us reach our creative potential.
Computational creativity addresses challenging questions such as “What is creativity?” and studies the differences and similarities in the creative capabilities of humans and AI. Humans are exceptional at evaluation—we can determine whether we like a piece of visual art, a story, or a melody, even if we’ve had no training in these spaces. On the other hand, AI-based systems have incredible regenerative capacities, but they struggle to determine which of their generations to select.
The combination of novel machine learning methodology—most notably deep learning, with techniques for computational creativity, is taking computer creativity to new heights.
By using AI and computational creativity to help create music, artists can retain creative control over their music while using AI as a creative supplement to inspire their lyrics and melodies. Furthermore, AI-based systems can open up the world of song creation to aspiring musicians who may not otherwise possess the skills required to engage in this creative art form.
How AI and Computational Creativity Are Democratizing Songwriting
Singers and songwriters are starting to take notice of the musical capabilities of artificial intelligence systems. Earlier this year, Taryn Southern, a singer and former American Idol contestant, released her debut album co-produced by AI. A system build upon novel methods from both computational creativity and AI was also recently used to create the world’s first EP made completely with an AI app called ALYSIA. Several songs created by the app were recorded by Gwyndolyn, an aspiring singer-songwriter who has leveraged the app’s AI technology to help her write and produce her own original songs.
So how does computational creativity actually work within the music creation process? As an example, the ALYSIA smartphone app uses several machine learning models for the creation of original lyrics and melodies. It then incorporates methodology deriving from computational creativity research to create unique vocal melodies which integrate music with natural language. This marks a significant step forward in musical AI, which was previously considered too difficult in both industry and academic circles.
Singer-songwriters can use AI technology to generate lyrics and melodies one line at a time, while relying on their own personal taste in music to choose amongst them and piece together original songs. In addition to machine learning models that generate unique lyrics and melodies, using computational creativity offers genre flexibility. For example, ALYSIA’s generations fit any musical genre, allowing the app to span across a diversity of styles from pop, rock and R&B to hip-hop, jazz and country.
This allows everyone to create original songs in their own taste, even if they have no musical training or experience, bridging the gap between ability (by leveraging the capabilities of AI technology) and taste (by leveraging our human capacity for evaluation).
AI Isn’t Replacing Songwriters, but Rather Helping Them Write
In the next decade, it’s estimated that between 20% and 30% of the Top 40 singles will be written partially or totally with machine-learning software. While some songwriters and musicians may be worried about getting replaced by this technology, many simply see it as a helpful tool.
“I wasn’t sure I could use artificial intelligence to compose songs,” says Gwyndolyn. “It’s a very personal thing, writing lyrics and coming up with melodies, to capture the feelings I want in my songs. But when I started to use ALYSIA, I saw how it was not a replacement for my creativity, it was a means of inspiring my work and using it as a tool to get to the songs I wanted quicker. Now, I see AI as an effective tool in my songwriting efforts.”
“It’ll take the sentence, and I’ll divide it up between the columns, and then when I’ve got three or four or five—sometimes I’ll go as much as 20, 25 different sentences going across here, then I’ll set it to randomize,” said Bowie, while demonstrating his program in 1995. “So what you end up with is a real kaleidoscope of meanings and topics and nouns and verbs, all sort of slamming into each other.”
Using artificial intelligence as a tool is especially helpful for songwriters who may not have a technical musical background or any experience in producing songs. Taryn Southern said in an interview that the technology particularly helped her in this aspect.
“I’d find a beautiful chord on the piano and I’d write an entire song around that, but then I couldn’t get to the next few chords because I just didn’t know how to play what I was hearing in my head,” said Southern. “Now I’m able to iterate with the music and give it feedback and parameters and edit as many times as I need. It still feels like it’s mine in a sense.”
A New Era in Music Creation
Thanks to computational creativity and deep learning being used in these innovative ways, aspiring musicians and songwriters with little musical background can now start creating their own music—without the help of expensive producers and recording studios.
A new era of AI-generated music is blossoming as we speak. From smartphone apps like WaveAI’s ALYSIA, to IBM’s Watson Beat program and Spotify’s Creator Technology Research Lab, music creators are starting to take note of how this technology can be used as a helpful tool, rather than replacement of musicians and human creativity.
As more emerging artists begin using these artificial intelligence programs to create music and break into the industry on their own, it will be interesting to see how this new democratized process of making music impacts the traditional music industry. No matter how AI changes the industry in the next few years, aspiring artists can look forward to accessing the process and creating music easier than ever before.