Beethoven’s Unfinished Tenth Symphony Gets Completed by Artificial Intelligence: Hear How It Sounds




Few symphonies are as well-known as Beethoven’s Ninth, an assertion supported by the fact that it’s no doubt playing in your head even as you read this. Few symphonies are less well-known — at least by Beethoven’s standards — than his Tenth, primarily because he never actually got the thing finished. He did make a start on it, however, and at his death in 1827 left behind notes and drafts composed alongside the Ninth, which had also been commissioned by the Royal Philharmonic Society. Such is Beethoven’s stature that his enthusiasts have been speculating ever since on what his incomplete symphony would sound like if completed, employing any techniques to do so that their time put at hand.

“In 1988, musicologist Barry Cooper ventured to complete the first and second movements,” writes Rutgers University Art & AI Lab director Ahmed Elgammal at The Conversation. “He wove together 250 bars of music from the sketches to create what was, in his view, a production of the first movement that was faithful to Beethoven’s vision. Yet the sparseness of Beethoven’s sketches made it impossible for symphony experts to go beyond that first movement.”


When Beethoven’s milestone 250th year approached, however, the age of artificial intelligence was well underway. To Matthias Röder, the director of Salzburg’s Karajan Institute, uniting this towering composer and this promising technology had become an irresistible proposition.

Elgammal and Röder were just two of the team that came together to take on the formidable task of engineering a form of machine learning capable of helping to complete Beethoven’s Tenth. The others included composer Walter Werzowa (“famous for writing Intel’s signature bong jingle”), computational music expert Mark Gotham, and musicologist-pianist Robert Levin, who “had previously finished a number of incomplete 18th-century works by Mozart and Johann Sebastian Bach.” Deutsche Telekom provided funding for the project, and also produced the short documentary video on its result above. However conceptually intriguing, this A.I.-driven musical endeavor could finally be put to the test in only one way: hearing it performed by a 100-percent human orchestra. As Werzowa puts it, looking skyward, “We hope when he hears it now that he smiles.”

Related Content:

Stream the Complete Works of Bach & Beethoven: 250 Free Hours of Music

Watch Animated Scores of Beethoven’s 16 String Quartets: An Early Celebration of the 250th Anniversary of His Birth

Did Beethoven Use a Broken Metronome When Composing His String Quartets? Scientists & Musicians Try to Solve the Centuries-Old Mystery

The Story of How Beethoven Helped Make It So That CDs Could Play 74 Minutes of Music

Artificial Intelligence Writes a Piece in the Style of Bach: Can You Tell the Difference Between JS Bach and AI Bach?

Based in Seoul, Colin Marshall writes and broadcasts on cities, language, and culture. His projects include the Substack newsletter Books on Cities, the book The Stateless City: a Walk through 21st-Century Los Angeles and the video series The City in Cinema. Follow him on Twitter at @colinmarshall or on Facebook.


by | Permalink | Comments (2) |

Support Open Culture

We’re hoping to rely on our loyal readers rather than erratic ads. To support Open Culture’s educational mission, please consider making a donation. We accept PayPal, Venmo (@openculture), Patreon and Crypto! Please find all options here. We thank you!

Comments (2)
You can skip to the end and leave a response. Pinging is currently not allowed.
  • Sam says:

    Trying to recompose beethoven. Wheres,the respect??? Go read a book. He was devine.no sensitivity.

  • Isaac Malitz says:

    It’s about a 30% success: It imitates the Beethoven middle-period sound fairly well; many of the individual episodes have familiar compositional character. Does not have what I’ll call “Schenker aspect” much either at macro or micro level. Doesn’t have LVB inventiveness, integration. Doesn’t have hooks, explosive effects, deep contrast.. but it is progress for AI. If this were chess, I would say that this AI has emulated a player in the 800 – 1000 range (“competent beginner”) but not at all in the 2200+ range (“master level”). Which is a great accomplishment, a big step forward. Looking forward to future iterations …

    HOWEVER, two criticisms

    [a] The AI team does not seem to realize that it is only at the 30% level. This is the “Florence Foster Jenkins” phenomenon, where the creator is not aware of the limits of their current scope and capability. I know that Levin was involved, and he is a superb musician, but I do not detect his influence in the current incarnation of the project.

    If the AI team needs a gentle critique of their musical work so far, please feel free to contact me. I have background and relationships that I think could provide useful input, in a respectful non-disruptive way. To help you to move forward …

    [b] The hype of this project is nauseating. It transforms a really-good technical effort into a public fiasco. Again, I think of Florence Foster Jenkins, her 1944 Carnegie Hall performance, where she was highly promoted AND totally beyond her capabilities; she accomplished one of the greatest fiascos in music history. https://en.wikipedia.org/wiki/Florence_Foster_Jenkins#Vocal_career

    I would encourage the AI/music creative team to put a leash on their hype-meisters. Tell ’em to go away and do some hype on Nora the Cat (“CAT PLAYS PIANO, SIGNS A RECORDING CONTRACT !”)

Leave a Reply

Quantcast
Open Culture was founded by Dan Colman.