Will AI replace human artists? What would the implications be? These questions grip many in the music business and outside of it. This weekend I decided to explore some lyric generation apps and see what I could get out of them – learning a thing or two about the future of music along the way.
Below I’ve posted the most coherent lyrics I managed to get out of one AI tool. I’m dubbing the song Purple Sun.
What I imagine the song’s artwork to look like.
You can make the sun turn purple You can make the sea into a turtle
You can turn wine into water Turn sadness into laughter
Let the stars fall down Let the leaves turn brown
Let the rainwoods die Let wells run dry
I love the turtle line. I guess the algorithm struggled with rhyming purple.
Two lines down is a wine / water line. Initially I was impressed by having a western cultural reference. But hold up… turning wine into water? That’s just evil.
Read it over once more. Or twice. By reading it over more, I became convinced that obviously humans are the superior songwriters.
But you know what, I’ve been lying to you.
The origins of the above lyrics are actually human, from a 90s rave song called Love U More by DJ Paul Elstak.
And they carry meaning. A lot of meaning to a whole generation of people in The Netherlands and other parts of Europe. Myself included. The meaning comes not necessarily from what the intent of the lyrics is. It comes from the music, nostalgia, memories, associations.
This is listener-assigned meaning. As soon as you release music, you give over control of the narrative to an audience. Artistic intent may have a lot of sway, but sometimes a song that’s a diatribe against fame turns into something stadiums full of drunk people chant.
A few statements to consider:
AI has a role as a tool to be used by people to apply their creativity.
Not all successful human created art objectively requires a lot of skill.
Creativity doesn’t end with the creator. The creator sets intent, the listener assigns meaning.
Let’s pair #1 and #3. In the first statement I talk about people, rather than mention specific roles as in the thrid statement. That’s because AI allows more people to be creative, either as listener, creator, or the space in between.
It’s this space in between that will be impacted and shaped by AI. Think of the dadabots projects, such as their infinite neural network generated death metal stream, apps like JAM, Jambl, and Endlesss which allow people to express themselves musically in easy ways, or technologies that turn music into something more adaptive like Bronze and FLUENT (disclaimer: I’m an advisor to the latter). Not all of the above use AI, but all cater to this space in between listener and creator.
The reason why I added statement #2 is because AI-created music doesn’t necessarily have to be objectively good. Music is subjective. Its sucess depends on how well it can involve the listener. That’s why AI is destined to be the most important force for the future of music in a more creative world.
Credits for the lyrics above: Lucia Holm / Paul Carnell. Thank you for the wondrous energy, the memories, the music.
New to MUSIC x TECH x FUTURE? Subscribe to the free newsletter for weekly updates about innovation in music. Thousands of music professionals around the world have gone before you.
What will the next format be to usher in a new music industry, like the record did in the 20th century?
The 20th century saw the rise of consumerist culture as a response to mass production causing supply to outgrow consumer demand. An example of this phenomenon is 20th century fashion which became highly cyclical (and wasteful), marketing new clothes for every season. After World War II, it became common to use clothing to express oneself through styles and fashions which often went hand-in-hand with music subcultures, just think of hippies, skinheads and punk music, hiphop, funk, or disco.
“Our enormously productive economy demands that we make consumption our way of life, that we convert the buying and use of goods into rituals, that we seek our spiritual satisfaction and our ego satisfaction in consumption. We need things consumed, burned up, worn out, replaced and discarded at an ever-increasing rate.”
– Victor Lebow (Journal of Retailing, Spring 1955)
Consumerism helped turn the recording industry into the most powerful part of the music business ecosystem, something which had previously been dominated by publishers. It changed music. The record player moved into the living room, then every room of the house, and the walkman (now smartphone) put music into every pocket. Music gained and lost qualities along the way.
Previously, it had been common for middle class families to have a piano in the home. Music was a social activity; music was alive. If you wanted to hear your favourite song, it would sound slightly different every time. With the recording, music became static and sounded the same way every time. And the shared songs of our culture were displaced by corporate-controlled pop music. People stopped playing the piano; and creators and ‘consumers’ became more clearly distinguished culturally.
With streaming, we are reaching the final stage of this development. Have a look at the above Victor Lebow quote and tell me streaming does not contribute to music being worn out, replaced and discarded at an ever-increasing rate.
The rules of mass production don’t apply to music anymore, since it’s no longer about pressing recordings: anything can be copied & distributed infinitely on the web. The democratisation of music production has turned many ‘consumers’ into creators again. Perhaps this started with drum computers, which helped kick off two of today’s fastest growing genres in the 70s and 80s: hiphop and house music. Today, this democratisation has turned our smartphones into music studios, with producers of worldwide hits making songs on their iPhones.
We see more people producing music, our Soundcloud feeds are constantly updated, Spotify‘s algorithms send new music out to us through daily mixes, Discover Weekly, Release Radar, Fresh Finds, and we now have the global phenomenon of New Music Fridays. With this massive amount of new music, we are simply not connecting to music in the same way as we did when music was scarce. We move on faster. As a result, music services, music providers essentially, place a big emphasis on music discovery as a result. We shift from the age of mass media, and mass production, to something more complex: many-to-many, and decentralised (music) production on a massive scale.
Has consumerism broken music culture? I don’t think so. As a matter of fact, consumerism is also what producers of music creation software and hardware depend on, which contributes to the democratisation of music and returning musical participation to the days of the piano as the default music playback device.
If streaming is the final stage of the age of the recording, then what’s next?
Embedded deep in the cultures of hiphop and house music, we can see what cultural values are important to the age of democratised music creation. Both genres heavily sampled disco and funk early on in their lifecycles. One of the most famous samples in hiphop and electronic music culture is the Amen Break. With the advent of the sampler, the drum break of the Winston‘s Amen Brother became widespread and instrumental to the birth and development of subgenres of electronic music in the 90s.
Not so long ago, ‘remix culture’ was still a notion one could discuss in abstract terms, for instance in the open-source documentary RiP!: A Remix Manifestowhich discussed the topic at length. Things have changed fast however, turning the formerly abstract into a daily reality for many.
Since the documentary’s release in 2008, social networks have boomed. Back then, only 24% of the US population was active on social media, but now that’s ~80%. With the increasing socialisation of the web, as well as it being easier to manipulate images, we saw an explosion of internet memes, typically in the form of image macros which can be adjusted to fit new contexts or messages.
The same is happening to music through ‘Soundcloud culture’. Genres are born fast through remix, and people iterate on new ideas rapidly. A recent example of such a genre is moombahton which is now one of the driving sounds behind today’s pop music.
Snapchat filters and apps like Musically let users playing around with music and placing ourselves in the context of the song. Teens nowadays are not discovering music by some big budget music video broadcasted to them on MTV, they are discovering it by seeing their friend dance to it on Musically.
Music is becoming interactive, and adaptable to context.
Matching consumer trends and expectations with technology
Perhaps music is one of the first fields in which consumerist culture has hit a dead end, making it necessary for it to evolve to something beyond itself. People increasingly expect interactivity, since expressing yourself just by the music you listen to is not enough anymore to express identity.
Music production is getting easier. If combined with internet meme culture, it makes sense for people to use music for jokes or to make connections by making pop culture references through sampling. Vaporwave is a great example. But also internet rave things like this:
Instead of subcultures uniting behind bands and icons, they can now participate in setting the sound of its genre, creating a more customised type of sound that is more personally relevant to the listener and creator.
That’s where the future of music lies. Not in the static recording, but in the adaptive. The recording industry that rose from the record looked nothing like the publishing industry. It latched on to the trend of consumerism and created a music industry of a scale never seen before. Now that we’ve reached peak-consumerism, and are at the final phase of the cycle for the static recording, there’s room for something new and adaptive. And like with the recording business before, the music business that will rise from adaptive media will look nothing like the current music industry.
Then last week, James Lynden shared his research into how Spotify affects mood and found out that people are mood-aware when they make choices on the service (emphasis mine):
Overall, mood is a vital aspect of participants’ behaviour on Spotify, and it seems that participants listen to music through the platform to manage or at least react to their moods. Yet the role of mood is normally implicit and unconscious in the participants’ listening.
Having developed music streaming products myself, like Fonoteka, when I was at Zvooq, I’m obviously very interested in this topic and what it means for the way we structure music experiences.
Another topic I love to think about is artificial intelligence, generative music, as well as adaptive and interactive music experiences. Particularly, I’m interested at how non-static music experiences can be brought to a mass market. So when I saw the following finding (emphasis mine), things instantly clicked:
In the same way as we outsource some of our cognitive load to the computer (e.g. notes and reminders, calculators etc.) perhaps some of our emotional state could also be seen as being outsourced to the machine.
For the music industry, I think explicitly mood-based listening is an interesting, emerging consumption dynamic.
Mood augmentation is the best way for non-static music to reach a mass market
James is spot-on when he says mood-based listening is an emerging consumption dynamic. Taking a wider view: the way services construct music experiences also changes the way music is made.
The playlist economy is leading to longer albums, but also optimization of tracks to have lower skip rates in the first 30 seconds. This is nothing compared to the change music went through in the 20th century:
The proliferation of the record as the default way to listen to music meant that music became a consumer product. Something you could collect, like comic books, and something that could be manufactured at a steady flow. This reality gave music new characteristics:
Music became static by default: a song sounding exactly the same as all the times you’ve heard it before is a relatively new quality.
Music became a receiving experience: music lost its default participative quality. If you wanted to hear your favourite song, you better be able to play it, or a friend or family member better have a nice voice.
Music became increasingly individual: while communal experiences, like concerts, raves and festivals flourished, music also went through individualization. People listen to music from their own devices, often through their headphones.
Personalized music is the next step
I like my favourite artist for different reasons than my friend does. I connect to it differently. I listen to it at different moments. Our experience is already different, so why should the music not be more personalized?
The gaming industry has figured out a different model: give people experience to the base game for free, and then charge them to unlock certain features. Examples of music apps that do this are Bjork’s Biophilia as well as mixing app Pacemaker.
But it’s early days. And the real challenge in creating these experiences is that listeners don’t know they’re interested in them. As quoted earlier from James Lynden:
The role of mood is normally implicit and unconscious in the participants’ listening.
The most successful apps for generative music and soundscapes so far, have been apps that generate sound to help you meditate or focus.
But as we seek to augment our human experience through nootropics and the implementation of technology to improve our senses, it’s clear that music as a static format no longer has to be default.
