the debate about The role you will play the artificial Intelligence (AI) in the future of music is served. A few days ago that the canadian singer Grimes claimed that the live music is doomed to disappear, because it is consolidating the technology in the day-to-day human beings. The assertion caught fire networks. The artist Russian Zola Jesus answered him: “Your words come from privilege, fascist, Silicon Valley”, since Grimes is the pair of Elon Musk. To the discussion have been joined by several personalities of the music as the composer Holly Herndon, who after completing a phd on the relationship between music and technology bet for the coexistence of IA and live music. “A machine that simulates a battery you can create simple rhythms but never will replace a big battery,” he says.
The experts look for the analogy with chess. “For years the best chess player is a computer, but that does not mean that there is still world champions of chess, or continue to championships,” says Nuria Oliver, a phd in artificial intelligence from the MIT (Massachusetts Institute of Technology). Today the majority of artists use algorithms of AI in their songs through the years and the contents of some of them is 100% synthetic. However, the concert experience is a lot more of what q are the songs. “It’s a feeling shared. The human being is a social being, and by this, we are going to continue to need social experiences with other people. In addition, as used all kinds of tools in the music with AI, many times we do not know where to start what is artificial and where just the biological”, adds Oliver.
in Addition, we have more than 10 years living with the live music and the AI, although sometimes we don’t realize it. The systems of lights and sounds make up for years systems of IA. But not only that, we also have artists entirely artificial. Hatsune Miku, became the first virtual artist in the world in 2007. The shows of the creation of japanese go fans of the union of art and technology and lovers of the projections of three-dimensional surround sound because Miku is not a singer real, but a hologram. In 2020 it will be the first time you step on Spain. In January will give a concert in Barcelona. The incubator of Hatsune Miku been Vocaloid, an application with a software voice synthesis developed by Yamaha, inspired by the work of the Music Technology Group of the Pompeu Fabra University of Barcelona. This tool is able to synthesize songs. When you insert the melody and the lyrics, the software may change the emphasis, add effects and even change the tone of the voice.
There are algorithms that allow you to recreate musicians and simulate it as it touches a cellist or a violinist, and play automatically
There are algorithms that allow you to recreate musicians and simulate it as it touches a cellist or a violinist and play it automatically. “This AI is very useful to analyze in real time the movements and physiological state of the musicians and orchestra conductors,” says Oliver.
in Addition, there are prostheses that have made it possible for professional musicians, who have lost a limb can continue to dedicate itself to compose. The team of Georgia Tech Center for Music Technology (GTCMT) has created a prosthesis for a battery with a software of artificial intelligence. “This robotic arm includes the activation is controlled by the muscles or the mind, that is to say, the user do not completely control the actions of the prosthesis, but a mechanism does that for him,” explains Gabriele Bunkheila, responsible for audio product of MathWorks. The area that has been developed with the advent of the digital services of music is the modeling of the tastes of the person or the recommendation of music. “Platforms like Spotify are designed with an algorithm that analyzes tastes and suggests issues similar to those desired by the user,” points Bunkheila.
An algorithm can predict the success of a song
But in a short time we will be in another screen. An algorithm may predict the success of a song. “You can analyze historical data as the songs have been successful, and using AI techniques to identify patterns of success and apply them to new songs,” says Nuria Oliver, a phd in artificial intelligence from the MIT (Massachusetts Institute of Technology). There is another way to do it. “You can train a model of AI with all the songs of an artist who has succeeded as Justin Bieber or Rosalia. You can generate the style of the artist from the song previews that have been a success,” he adds.
According to experts, the artificial intelligence exists as an area of research since there are computers, but during this decade have produced changes appropriate for the technology available to the society. “The availability of large amounts of data and digital music, the resources to perform analysis at a low cost, and the development of architectures or algorithms of deep learning have made possible the rise of these new systems to generate music,” says Oliver.
When he left the television the concern was q going to finish the theatre. Each time there has been new media, is concerned that it will disappear all of the above and what tends to occur is the coexistence of both systems. “The human being is a social being with a great capacity of interaction. To have an experience shared with thousands of people is something that regardless of who the artists are humans, robots, or mixed we are going to continue wanting to experience”, concludes Oliver.
The keyboard who composed melodies
This month of December has been put up for sale Amazon DeepComposer, a keyboard with 32 keys and the ability to connect to a computer. There are many keyboards similar but the novelty of DeepComposer is that, connected to the cloud, and learn for yourself. “Is the world’s first keyboard with machine learning integrated”, they cite sources from Amazon.
The AI using this instrument is capable of generating different musical styles (from classic to jazz, passing by the rock and pop). For this there is a neural network that generates and analyzes the created to finally produce the piece. The previous work is the evaluation of large amounts of data to be able to distinguish musical genres. Experts believe that this kind of instruments are allies for the musicians to create compositions without much effort. Although it is not all negative, it may be an extra motivation for the public at large want to learn to play an instrument.
