Beethoven’s 10th Symphony and the jarring notes of AI

By Biju Dominic
The Beethoven Bonn Orchestra premiered the 10th symphony on 9 October 2021. (REUTERS)

Over 250 years after Ludwig van Beethoven died, his 10th symphony has been composed. The symphony was completed by a team of music historians, musicologists, composers and computer scientists who combined efforts to transcribe Beethoven’s sketches and process his entire body of work, so that Artificial Intelligence (AI ) could be properly trained on it. No doubt ‘Beethoven X: The AI Project’ is a significant milestone in the development of AI, a major field today. Is the composition of a new symphony a loud and clear demonstration that AI has no limits on what it can achieve?

According to Hans Moravec, adjunct faculty member at the Robotics Institute of Carnegie Mellon University, a representation “landscape of human competence" has often been used to gauge the potential of AI. He represented AI capability as a rising sea level, increasingly covering the landscape of human competence. Human functions like rote memorization and arithmetic were the lowlands to be first flooded by AI. Playing the Go game, speech recognition and language translation were considered mountains. But recently, AI has clearly outdone humans in these areas too. Tasks like driving are on the verge of being mastered by AI. According to Moravec, the highest peaks that AI might never conquer were creative skills, like book writing and art composition. But with the creation of the 10th symphony of Beethoven, it is quite clear that AI can perform almost every human activity to perfection, or even do it better than humans.

The Beethoven Bonn Orchestra premiered the 10th symphony on 9 October 2021. A casual glance at the viewership of YouTube videos of Beethoven’s 10th showed that 3 months after that debut performance, not more than 75,000 people have viewed those recordings. When you consider that the viewership of other Beethoven symphony videos runs into several million, the views achieved by the 10th symphony look paltry. Technically, the 10th symphony composed by AI could be considered a functionally superior product. It used sounds from electric organs that actual Beethoven symphonies did not have, for example, because electric organs did not exist during the composer’s time. But despite its functional superiority, the 10th symphony has not caught the fancy of music lovers. Why is that?

Beethoven 10th symphony is a reminder that creating an AI product that is functionally superior to human capabilities is one thing, but getting it adopted and loved by human audiences is a different ball game altogether. When music lovers are listening to a symphony, they are not just listening to the music alone. They are emotionally connecting with the several stories around that music. For example, there is a story that when Ludwig van Beethoven’s magisterial 9th Symphony premiered in 1824, the composer had to be turned around to see the audience cheering—he could not hear its rapturous applause. This knowledge that Beethoven had grown deaf by the time he composed his last symphonies adds a very different dimension to the appreciation of his music that goes far beyond its functional quality.

Many in the AI industry are driven by the belief that once a product that can do a job better than what a human could achieve is developed, it is home and dry. This is similar to the ‘product oriented’ belief that drove manufacturers at the peak of the industrial era: Build a better mousetrap and the world will beat a path to your door. But soon, there will be many more manufacturers of good- quality mousetraps. This is when ‘marketing orientation’ comes in and reminds manufacturers of the need to build brands that hold special appeal.

Building a brand is quite different from making a product. With some effort, sweetened water with a little fizz can be made into a cola drink. This is just a product. Making it into a brand, something with which consumers develop an emotional connection, is an entirely different challenge. When we are only focused on the engineering of a driverless car, one is acting only from a product point of view. To think from a brand point of view, one has to keep the end user at the centre of it all. So, while developing a driverless car, one has to think what the end user is going to do now that s/he does not need to drive. Read more, listen to music more, sleep more? If so, how can a driverless car help facilitate these activities better than existing cars? Also, what does the end user lose by using a driverless car? Today, there is so much prestige attached to being driven around by a chauffeur. How will autonomous vehicles create the same levels of prestige?

The AI industry does not have to wait for competing products to emerge in each category before it starts developing a branding orientation. Realistically, AI product developers cannot say that the likely future consequences of their technology developments, be it job losses or ethical dilemmas, are not their bother and these are someone else’s responsibility to deal with. They have to start thinking of potential outcomes ahead of time, with emotional factors duly taken into account, just as brand builders do. To acquire a wider perspective, AI teams should be composed not just of technology experts, but behavioural science and design experts too.

At this nascent stage of the AI industry, the functional superiority of an AI product might be big news. But Beethoven’s 10th symphony reminds us that an emotional connection with the end user is critical for success.

No doubt, the future belongs to AI. Not AI products, but AI brands.

Biju Dominic is the chief evangelist, Fractal Analytics and chairman, FinalMile Consulting


What is inkl?

Important stories

See news based on value, not advertising potential. Get the latest news from around the world.

Trusted newsrooms

We bring you reliable news from the world’s most experienced journalists in the most trusted newsrooms.

Ad-free reading

Read without interruptions, distractions or intrusions of privacy.