One of the Big-4 tech companies, well-known to everybody, seems to have developed an artificial intelligence piano. Their initial and bold question was how they could use machine learning to create compelling art and music!
At this point, you are probably expecting a seasoned pianist and piano pedagogue to freak-out and start blaming technology for all the evils in our societies. But actually, it is not so simple… You see, I had no doubt there would be a robot (or computer program) at some point in history which would be able to give a full-fledged recital and even improvise, perhaps producing results of actual artistic merit. I just wasn’t sure if it was already happening in a lab somewhere or not. Now we know, and in my opinion, this is great news because it showcases the incredible advancement of technology as much as it illustrates the faults and wrong assumptions within the classical-music industry.
Artistic creation used to be a product not only of intelligence, skill, and practice but also a result of Inspiration, marked by Unpredictability, serving as a vehicle of breaking preconceived Limitations. That was Art. After years of evolution, such notions are mostly related to the Avant-garde, while established artistic work of perceived value seems to be governed merely by structure, logic and/or eccentricities for their own sake. Michelangeli used to call the latter, ‘experiments’…
Parallel to this developments, we have (for a good few decades now) international piano competitions, music academies, university departments of music, and millions of exams and auditions every year that confer Prizes, Degrees and Scholarships based on evaluations, as they also have sponsors, donors, state committees, and brainy critics to please. This leads to the Quantification of artistic result: a wrong note and you are disqualified, a slight deviation from the printed text and you are eliminated as an ‘amateur’, a liberty taken at the spur of the moment during a ‘live’ performance and the critics will assume you are ‘taking risks’. It is not unreasonable to believe that the past greats like Schnabel, Cortot and perhaps even Rubinstein would be neither accepted as 19-year-old freshmen at Juilliard in New York nor as major prize-winners in an international competition such as the Tchaikovsky in Moscow.
Classical music management, the recording industry (or what is left of it) and the international circuit of famous piano-mentors have set rules that promote artistic objectivity, clarity, and professionalism above anything else. As a result, the world had never seen before as many highly Qualified practitioners as we enjoy today, and at the same time, never had fewer Artist, even during periods of World Wars…
If all this sounds too grim and pessimistic, one needs to look no further than the shrinking and bored audiences throughout the USA and Europe to get really depressed: decreased audience attendance, fewer subscriptions, empty opera houses… Let us be fair, if it wasn’t for China, Korea, and Japan, countries that continue to supply the market with eager, talented and passionate youngsters who love what They perceive classical music to be, classical music would have been a museum subject by now.
Good news then that things will finally reach a true level of objectivity pretty soon, at the more capable hands of Artificial Intelligence! In fact, I have been told that the new programs not only create excellent performances but they also improvise with incredible skill! It seems that this was accomplished by downloading billions of performances into the ‘brain’ and then developing some sort of algorithmic software which is capable both for finding the next note you would ‘expect’ and also deviate from it by still doing something reasonable!
Some will find the idea of a robot playing Scriabin in a compelling and sensual way Very Very scary… I find it impressive and a perfectly logical evolutionary step towards the wrong direction. A direction we modern pianists and teachers fostered for decades by discouraging probing and exploratory performances instead of predictable and homogeneous results. The great pianist K. Zimerman calls it the ‘transatlantic playing’, I think. You turn on the radio to check your favorite classical music station, and if the recording is a performance after 1990, you cannot tell who the pianist is, possibly with the exception of M. Argerich (my opinion). In the past, a Horowitz, a Gilels or a Gould would be instantly recognizable. Now, it can be Anybody. Why not a robot then?
The next evolutionary step? International Competitions for robots! Since making a mistake will simply be impossible, we may finally focus on the music and award prizes based on artistry… We just have to make sure that real people will still be sitting in the jury as well, because counting and subtracting points, negotiating internal politics, and writing reports (as most current jury-members do during a competition) are also things computers do better…