Resonoo and the Music Industry: could this be a gateway to standardized music?

It is no secret that behind any successful artist is not only a team of production professionals but also a flock of editors, publicists, marketers and sound engineers along with others. The new AI-driven start-up Resonoo has understood that complex ecosystem, thus offering a solution, for all parties involved, for a better return on investment. Despite their claim to identify hit songs, this new music AI has broader implications for the creative industries…

Resonoo at a glance.

Resonoo, a Toronto based company launched in August 2019, enriches the blossoming tapestry of AI-driven music start-ups. Fifteen months ago, Resonoo was better known as OptimiseLab and focused on the analysis, improvement and composition of music (Dredge). The company is now rebranded under Resonoo and comes back with a stronger focus on the prediction of music popularity (Dredge). 

The company is at the crossroad of musicianship and AI technology and claims that based on the data processing of “tens of millions of songs” (Dredge), their AI allows predicting with 90% accuracy the potentiality of a hit song. This product is aimed at editors, marketers and music publishers to either scout new talents or indicates the direction in which the producing companies should head.

Beyond the potential gain this AI could bring to the music industry, its innovative approach to music evaluation poses essential questions with regards to creativity and technology as well as music standardisation. 

Creativity, Expertise, Artificial Intelligence and the music Industry.

Can Artificial Intelligence be creative? This is not a new debate, and academic interest for the question can be traced back to Terry Dartnal’s seminal book Artificial Intelligence and Creativity. If based on the idea that creativity is “ a process which leads to the production of something that is both new and useful”(Runco and Pritzker 1) then there are multiple examples proving that creativity is not solely a “feature of human intelligence”(Boden 347). Indeed computational creativity can be found in AI written poetry, and even AI produced music as below (Cafolla).

However, the creation of Resonoo’s AI goes beyond the debate between creativity and technology but enters the realm of Artificial Intelligence as a creativity expert. 

While AI might increase creativity or produce creative outputs, it is also starting to “ inform practitioners of the kind of creative output society favours.” (Marks 1). In the creative industries, understanding a market segment’s preference is a precious asset. Today, this type of expertise is often linked, in the music industry, but the creative industries in general, to experience and knowledge acquired through experience. Thus the ability to predict the next hit song was the trade of music editors and professionals. However, by its nature, AI can analyse and process vast datasets much quicker and efficiently than any music producer could (Marks). Thus the AI transcends the expertise of human creativity to assert itself as a more cost-effective and efficient way to define trends. 

From data processing to lack of transformational creativity. 

According to Boden, two types of creativity find their roots within cognitive, psychological and sociocultural abilities. Those are exploratory creativity and transformational creativity (Boden). While the former conceives creativity as “exploring a search space of partial and complete possibilities” (Wiggins 1), the latter challenges the rules which create this conceptual space. While AI’s have explored both, they tend to succeed a lot better through exploratory creativity (Dartnall). The AI algorithm which generates music in the style of Charlie Parker, for example, can produce music that is very similar to the one of Parker, yet it will never leave the realm of the artist. It is not a simple process as it requires “considerable domain expertise and analytical power” (Boden 354); however, the AI is incapable of creating beyond its initial knowledge. Nevertheless, AI tends to not excel beyond their own conceptual space. 

It is within that realm of transformational creativity that Resonoo might encounter some limitations. Indeed, even though their AI by nature will be able to learn from the pattern in the data it is processing to identify songs that will be popular, the data it is processing is still within the same conceptual space. A conceptual space reinforced as more hit songs that have been validated by the AI are later on processed by the AI. This can, in time, create a loop, or confirmation bias of some sort that will prevent or make it impossible for transformational creativity to occur without being discarded by the learning software.

Of course, an AI’s decision is not finite, and the editors still have the final call on which songs they decide to promote. Utilising such technologies can be beneficial, cost-effective and lead to accurate predictions. Nonetheless, relying on this platform might increase that chances of missing out on groundbreaking musicians. Overlooking new forms of creativity has occurred through time, and without an AI. Indeed Vincent Van Gogh was not recognised as a successful artist until very late in his life, only selling one painting before his death (Rüger ). Nevertheless, Van Gogh played a crucial role in the evolution of modern art. So even though this occurred without the AI, this might enhance a cost motivated attitude toward music discovery and disregard transformational musicians that decided to question their conceptual space.  

In conclusion, start-ups like Resonoo have incredible potential and can help monetise and facilitate the ecosystem of music production. However, the implications of such a technology are not entirely benign and can potentially have consequences on the industry. Many musicians are stating that the industry is becoming more permeable to differences and newcomers, and an Artificial intelligence like Resonoo’s could be a gateway to transform that claim into a reality

References

Boden, Margaret. A., “Creativity and artificial intelligence.” Artificial Intelligence, 103, 1998, 347-356

Cafolla, Anna, “Artificial Intelligence is writing poetry, but is it any good?” Dazed, 14 Aug 2018, https://www.dazeddigital.com/science-tech/article/40985/1/artificial-intelligence-ai-poetry-sonnet-shakespeare

Dartnall, Terry, Artifical Intelligence and Creativity. Dordrecht, Springer Science and Business Media First Name, 1994.

Dredge, Stuart “OptimiseLab uses AI to predict ‘music popularity scores’.” Musically,  22 May 2018,  https://musically.com/2019/09/02/startup-resonoo-is-putting-ai-to-work-on-spotting-hit-songs/

Dredge, Stuart “Startup Resonoo is putting AI to work on spotting hit songs” Musically, 2 Sept 2019, https://musically.com/2018/05/22/optimiselab-uses-ai-to-predict-music-popularity-scores/

Marks, Anna, “How AI is radically changing our definition of human creativity” Wired, 15 June 2019,  https://www.wired.co.uk/article/artificial-intelligence-creativity

Rüger, Axel, “From Unrecognized Genius to Global Icon: Vincent van Gogh Then and Now”, Van Gogh Museum, Amsterdam, 25 Feb 2017, https://www.nortonsimon.org/learn/multimedia/videos-podcasts-and-lectures/lecture-from-unrecognized-genius-to-global-icon-vincent-van-gogh-then-and-now/

Runco, Marc A,  Pritzker, Steven R, Encyclopedia of Creativity. San Diego, Academic Press, 1999.

Wiggins, Geraint A. “Towards a more precise characterisation of creativity in AI.”, 2001.