- Ahmed Elgammal
- The Conversation*
When Ludwig van Beethoven died in 1827, it had already been three years since he completed his ninth symphony, which many consider his masterpiece.
And he had started work on his 10th symphony but, due to his deteriorating health, he could not advance much: all he left were some musical sketches.
Since then, musicologists and Beethoven lovers have wondered and lamented what it might have been. His grades hinted at a magnificent reward, albeit one that seemed eternally out of reach.
Now, thanks to the work of a team of music historians, musicologists, composers and computer scientists, Beethoven’s vision will come to life.
I chaired the artificial intelligence (AI) part of the project, leading a group of scientists in the start–up Playform IA, which He taught a machine both the complete work of Beethoven and his creative process.
A complete recording of Beethoven’s 10th Symphony is scheduled to be released on October 9, 2021, the day of the world premiere scheduled in Bonn, Germany.
It is the culmination of an effort of more than two years.
Around 1817, the Royal Philharmonic Society of London commissioned Beethoven to write his 9th and 10th symphonies.
Written for an orchestra, symphonies often contain four movements: the first one runs at a weather fast, the second to a slower, the third to a weather medium or fast and the last to a weather Quick.
Beethoven completed his ninth symphony in 1824, which concludes with the “Ode to Joy”.
But in regards to the 10th symphony, Beethoven did not leave much materialApart from a few musical notes and a handful of ideas he had jotted down.
There have been some previous attempts to reconstruct parts of Beethoven’s 10th Symphony. The most famous, in 1988, was that of the musicologist Barry Cooper, who ventured to complete the first and second movements.
He intertwined 250 bars of music from the sketches to create what, in his opinion, was a production of the first movement faithful to Beethoven’s vision.
However, the paucity of Beethoven sketches made it impossible for symphony experts to go beyond that first movement.
At the beginning of 2019, Matthias Röder, director of the Karajan Institute, an organization in Salzburg (Austria) that promotes music technology, contacted me.
He explained that he was teaming up to complete Beethoven’s 10th Symphony in celebration of the composer’s 250th birthday.
Aware of my work on art generated by artificial intelligence, I wanted to know if AI could help fill in the blanks that Beethoven left.
The challenge was dazzling. To achieve this, you would have to do something that had never been done before. But I said I would try.
Röder later assembled a team that included the Austrian composer Walter Werzowa.
Famous for writing the characteristic jingle At Intel, Werzowa set about putting together a new type of composition that would integrate what Beethoven left with what AI would generate.
Mark Gotham, an expert in computational music, he led the effort to transcribe Beethoven’s sketches and process all of his work so that the AI could be properly trained.
The team also included Robert Levin, a Harvard University musicologist who is also an amazing pianist.
Levin had previously completed a series of incomplete 18th-century works by Mozart and Johann Sebastian Bach.
In June 2019, the group met for a two-day workshop at the Harvard Music Library.
In a large room with a piano, a blackboard, and a stack of notebooks with Beethoven’s sketches encompassing most of his well-known works, we talked about how the fragments could become a complete piece of music and how AI could help solve this puzzle. , while remaining faithful. to the process and vision of Beethoven.
The music experts in the room were eager to learn more about the kind of music that AI had created in the past.
I told them how AI had successfully generated Bach-style music. However, this was just a harmonization of a melody that sounded like Bach.
It wasn’t what we needed to do: build a complete symphony from a handful of phrases.
Meanwhile, the scientists in the room, myself included, wanted to know what kinds of materials were available and how the experts envisioned using them to complete the symphony.
The task in question finally crystallized. We would need to use notes and full compositions of all Beethoven’s work, Together with the sketches available from the 10th symphony, to create something that Beethoven himself could have written.
This was a tremendous challenge. We didn’t have a machine that we could feed sketches to, push a button, and have it spit out a symphony.
Most of the AI available at the time could not continue an incomplete piece of music beyond a few additional seconds.
We would need to push the boundaries of what creative AI could do, teaching the machine the creative process from Beethoven: how he would take a few bars of music and painstakingly develop them into moving symphonies, quartets, and sonatas.
Beethoven’s creative process
As the project progressed, the human side and the technological side of collaboration evolved.
Werzowa, Gotham, Levin and Röder deciphered and transcribed the sketches for the 10th symphony, trying to understand Beethoven’s intentions.
Using their complete symphonies as a template, they tried to piece together the puzzle of where the sketch fragments should go: what movement, what part of the movement.
They had decisions to make, such as determining whether a sketch indicated the starting point of a scherzo, which is a very lively part of the symphony, typically in the third movement.
Or they could determine that a line of music was likely the basis for a fugue, which is a melody created by weaving together parts that echo a central theme.
The AI side of the project – my side – found itself grappling with a variety of difficult tasks.
First, and most fundamentally, we needed to figure out how to take a short phrase, or even just a motif, and use it to develop a longer and more complicated musical structure, just as Beethoven would have done.
For example, the machine had to learn how Beethoven built the fifth symphony from a four-note basic motif.
Then, since the continuation of a phrase must also follow a certain musical form, be it a scherzo, a trio, or a fugue, the AI needed to learn Beethoven’s process to develop these forms.
The to-do list grew: we had to teach the AI how to take a melodic line and harmonize it.
The AI needed to learn how to put two sections of music together. And we realized that he had to be able to compose a coda, which is a segment that brings a section of a piece of music to its conclusion.
Finally, once we had a complete composition, the AI would have to figure out how to orchestrate it, which involves assigning different instruments to different parts.
And he had to carry out these tasks in the way that Beethoven could have done.
In November 2019, the team met in person again, this time in Bonn, at the Beethoven House Museum, where the composer was born and raised.
This meeting was the litmus test to determine if AI could complete this project.
We printed musical scores that had been developed by AI, built from Beethoven’s tenth sketches.
A pianist performed it in a small museum concert hall before a group of journalists, music scholars and Beethoven experts.
We challenged the audience to determine where Beethoven’s sentences ended and where the AI extrapolation began. They could not.
A few days later, one of these AI-generated scores was performed by a string quartet at a press conference.
Only those who were intimately familiar with Beethoven’s sketches for the 10th symphony could determine when the AI-generated parts began.
The success of these tests showed us that we were on the right track. But these were only a couple of minutes of music. There was still a lot of work to do.
Ready for the world
At every moment, Beethoven’s genius loomed in, challenging us to do better. As the project evolved, so did the AI. Over the next 18 months, we built and orchestrated two complete movements of over 20 minutes each.
We anticipate the reaction to this work: those who will say that the arts should be off-limits to AI, and that AI doesn’t have to try to replicate the human creative process.
However, when it comes to the arts, I see AI not as a replacement, but as a tool, one that opens doors for artists to express themselves in new ways.
This project would not have been possible without the expertise of human historians and musicians. It took an immense amount of work, and yes, creative thinking, to achieve this goal.
At one point, one of the team’s music experts said the AI reminded him of an anxious music student who practices every day, learns, and gets better and better.
Now that student, having taken up Beethoven’s baton, is ready to present the tenth symphony to the world.
*Ahmed Elgammal is director of the Art and Artificial Intelligence Laboratory at Rutgers University in the United States.
Now you can receive notifications from BBC News Mundo. Download the new version of our app and activate them so you don’t miss out on our best content.
Eddie is an Australian news reporter with over 9 years in the industry and has published on Forbes and tech crunch.