“The musicians see that the sky is very cloudy and we are happy that it is not raining yet,” says Berto Rojo, composer and bassist of Nina Polaca. The spectacular disruption of applications created with artificial intelligence in artistic or cultural fields such as illustration, photography, video, writing or education creates a strange uneasiness in the world of music: will the same thing happen to us? Think about the cast and the producers.
Rojo is interested in innovation and already uses non-productive software in his artistic work, for example to improve models. But he believes the great music disruptor has not yet erupted: “It’s early. It’s very different for an AI to understand an image than a song, a song carries a lot more information. We’re still are a long way from being able to tell AI ‘make me a Madonna song like Chopin played it on the piano,'” he says.
Google published a system in January called MusicLM that accurately generates music from text. For example: “a fusion of reggaeton and electronic dance music, with an expansive, ethereal sound. It inspires the feel of being lost in space, and the music is designed to evoke a sense of wonder and awe while still being danceable.” It’s a request for one of the 30-second tunes Google has created. The application isn’t open yet. Rojo has heard it and believes that “if you don’t Lets say the AI did it, so this is music they can enjoy”, without knowing it’s not the work of a human. This is the tune Google got from that request:
Google’s innovation isn’t just one of the last. Famous Disc jockey david guetta did Sing To Eminem in a concert, however the voice was not the rapper’s real voice but a similar but artificially generated voice. “I’m sure the future of music is in AI. Sure. No doubt about it. But as an instrument,” Guetta told the BBC.
What if the music was (somewhat) different?
These are just two examples of AI chasing music from various fronts: melodies, arrangements, voices, lyrics. It’s hard to think of an immediate future where AI doesn’t have a role in artistic creation, at least as a “tool,” as Guetta puts it. The level of seismicity probably depends on the intrinsic characteristics of each artefact. Music has two perceived advantages over other subjects. The first is that the composer and his audience have a more direct relationship than in other disciplines; It’s rare not to want to face a singer or see him live. The second is that music is already tampered with by a lot of software and has been digitally constrained for years. Some styles have naturally internalized tune-making from others, such as the use of samplesOr generate sounds with a computer.
None of the experts consulted by EL PAÍS see these two exceptions as definitive or impregnable. The pace of change in the sector is unprecedented and anything is possible: “We have no idea what the world will be like in five years,” says Berto Rojo. “Either you learn or other people are going to do it and they’re going to tear you apart,” he says. That speed has creative implications: “We believe that the more complex a song is, the more it takes a different type of creative process or the more difficult it is to get an excellent end result,” explains Chiara Hellquist, director of Vevo Spain, a digital platform. . for music videos. “But we were saying the same thing when the MP3 came out and people started listening to it worst quality Will we have more volume with a vinyl AI, and the speed at which the songs come out will be more intense, but we’ll see.
If new songs from famous artists are constantly coming out now, how many more can be created with the facilities provided by AI? “We will adapt. Right now I find it difficult to increase the pace of creation, but who knows, maybe 20 years ago it would have been impossible for an artist to release a song every month. Now it is almost essential to keep you in the limelight,” Hellquist Says. Newer, more personalized platforms may also appear. Five years ago, few people thought that there was going to be an explosion of melody snippets on TikTok. Now there are themes in which 15-second specials are ideal for the platform.
The other obstacle to the explosion of AI in music is a more direct relationship with the creator, something less common, for example, in illustration or video. But why can’t there be a lot more “musicians”, artists with little musical knowledge who for some reason know how to generate hits with AI? The composer, Berto Rojo, sees the conquest of this musical space as possible by people who know how to write requests to a machine: “This type of person can generate a community of fans around their music and AI Can make a living by using it. The artist is the one who decides what is left and what is not given. Another thing is to what extent are they going to be able to use AI for free and to what extent are they going to be able to monetize that music freely”, he says. Flood isn’t completely ruled out, but will a new platform emerge where we can consume more and more personalized music?
There are already writers who are getting closer to producing more quickly today: “New artists who are emerging bedroom pop Or in an urban area they compose directly with a computer, and the production table is at home”, says Miguel Martorell, director of Southern Europe at Altafontein. “An artificial intelligence applied to these instruments could be very interesting because you have to do it yourself. You won’t need to create, but you can directly ask the artificial intelligence to do it for you and adjust it,” he added.
role of bookstores
There is already one part of the sector, perhaps the most anonymous or with little creative value, that AI can sweep: pre-recorded and royalty-free libraries for video or advertising. “Now when you upload a video to YouTube, for example, you choose the library music, ‘Quiet Piano Music,'” Martorell says. “It’s still a standard that a lot of background producers use for their videos. An AI here can be very interesting because each producer can give some parameters and have a particular music or song”.
These two examples are useful for understanding the kind of natural evolution that has already been done and what AI provides. Now there are already AI applications that tweak demos to leave them with great sound: “You have a rock song and you want the volume of the different tracks to be as close to rock as possible,” innovation manager Fran Leo They say. record company. Cheat. “Then that artificial intelligence, which puts a lot of rock songs into it, analyzes how they’ve been mastered, draws some conclusions and applies it to you.”
Tools such as Lalal, Supertone or BandLab, which are already widely used, come closest to a fully generative AI. This example is just a synthetic voice created with Text to Voiceful, with no technical requests for pitch, tempo, timing, or anything else in post-production. It took about 20 seconds to make this artificial voice.
They make what’s already out there so much better without the help of an app maker and hours of studio time. The following steps will progressively offer new options that are easier to use, such as the one explained by Berto Rojo: “I heard an AI with which you record a bass track at home with a bad sound card, you tell it to the AI.” pass. And you can say, ‘Make me sound like the Pixies’ bass. And it’s able to take your track and turn it into a professional, It opens up a lot of doors because it will make it easier to sound good”, he says.
Even some of the most distrustful novelties, like Antonio Luke of Senior Chinaro, have their musical reasons for skepticism: “Those of us who make pop rock songs know that the ingredients we use are short. We don’t use very complicated harmonic progr
essions like in jazz. We know everything sol do re, re la sol and ‘love songs’ and ‘I saw you like this’, ‘I died for you’ and ’til the end’ and the sun and the moon. What we all handle, some are successful and most are not. Even one that has made one hit is not guaranteed to make another. If it were that easy, I would challenge artificial intelligence to start making hits,” he explains.
not everything was original
In other genres it’s even more clear that AI won’t be such an innovative disruptor: “Not all songs that are successful are 100% original creations. a vast majority of rap is based samples, taking a piece of a song and remixing and repeating it and layering that melody on top of other sounds to turn it into a whole new song. It’s very common in some styles and it’s never been a problem”, explains Rojo.
Despite these details, it is likely that the novelty is revolutionary for some reason that is now difficult to foresee: “It is something that is scary. People who live on their creativity may fear that suddenly this technology can do the same thing twice as fast”, says Berto Rojo. “There are two types of artists: those who are going to hate it and those who are going to hate it. When everyone else is using it and those who do realize that it is here to stay, its growth is exponential and to be competitive all you have to do is learn to use it. It’s like Excel appears and you have a company and you say, no, I want to continue with my notebook”, he adds.
Ultimately, as with the rest of the arts, everyone involved insists on what they believe will be strictly necessary to give humans their final creative touches. Or that the machines will bore us and we will have to choose what is good. Or to something else: “That artificial intelligence already exists. It’s called culture. They are the songs of the centuries. When I need that help, I listen to the songs that have already been made and feel it.” without, refactoring what already exists. If the computer did it, I’d prefer to do it myself”, says Luke. For now, it’s time to smile: “It’s full of parody. Point is. I see it as a joke, even with threats from technology companies to tell us they can replace us whenever they want, as they did with supermarket cashiers or gas fillers. Have done”.
you can follow country technology In Facebook why Twitter or sign up here to receive our newsletter semanal,