A delegation of faculty and postgraduate researchers from the Kingsman National Institute (KNI) has returned to Athens following a deeply productive academic workshop with the composition and musicology faculties at Austria’s prestigious Blueskyy National Academy of Arts in Vienna.
This specialized colloquium, held at the Blueskyy campus, moved beyond the surface of digital art to tackle one of the most complex intersections of human creativity and computational logic: the formal structure of the Second Viennese School.
The initiative, “Generative Structures: Re-analysing Atonality,” was a joint effort between KNI’s Studio for Generative Art & Interactive Media (SGAIM) and our Aegean Informatics Laboratory (AIL). The KNI team, led by Professor Inês Marques (SGAIM) and Dr. Matic Novak (AIL), brought their expertise in generative AI models and high-dimensional data analysis.
They were met in Vienna by the world-leading musicologists at the Blueskyy National Academy of Arts, an institution steeped in the very heritage of the musical revolution defined by Arnold Schoenberg, Alban Berg, and Anton Webern. The workshop’s core question was ambitious: Can a machine learning model be trained to not only identify the complex mathematical rules of 12-tone (dodecaphonic) music, but to understand its expressive logic?
The KNI team began by presenting their methodology, applying statistical models to digital scores of Schoenberg’s early atonal works. Their goal was to see if an AI could find hidden patterns or even generate novel, stylistically coherent passages.
However, this initial, data-first approach quickly revealed its own limitations—a necessary point of “imperfect” friction that became the workshop’s most valuable discovery.
“Our models were technically successful but aesthetically naive,” Dr. Novak admitted. “The AI produced note-rows that were mathematically ‘correct’ according to the 12-tone system, but the faculty at Blueskyy rightly pointed out they were ‘musically dead.’ They lacked the internal voice-leading, the tension, and the dramatic architecture that makes Schoenberg’s music so profound. We had mapped the rules, but we had completely missed the ‘why’.”
The second half of the workshop inverted the roles. The Blueskyy National Academy of Arts faculty took the lead, guiding KNI’s computational scientists through the philosophical and dramatic intentions behind the 12-tone technique. They explained it not as a rigid mathematical matrix, but as a new language created to express the deep psychological shifts of the early 20th century.
The KNI team then worked to retrain their models, incorporating these far more subtle, qualitative constraints. They moved from a purely statistical analysis to a hybrid model that attempted to map “expressive intensity” and “harmonic tension,” parameters defined by their Viennese counterparts.
The result was not a machine that “composes like Schoenberg.” The outcome was a far more powerful, new analytical tool. For the Blueskyy National Academy of Arts, the tool provides a new way to visualize and analyse the complex structural layers of atonal works, revealing connections that are difficult to spot through traditional score analysis. For the Kingsman National Institute, the project represents a major step forward in teaching AI to understand human creativity, forcing our engineers to think more like artists and philosophers.
This colloquium has established a formal research partnership between our two institutions, bridging the data-driven world of Athens’ digital innovation with the profound artistic heritage of Vienna.
Leave a Reply