Suno, Alexa, and the rise of ‘create to consume’

This blog was created with collaborative inputs from Ben Woods, Mark Mulligan, Tatiana Cirisano, and Kriss Thakrar.    

It is hard to think about generative AI without coming back to copyright. There is a pressing need to find IP safeguards before even more copyrighted content is used to train AI models. Without such agreements, IP holders will find it harder to maintain control and secure a fair financial return as AI tools grow in sophistication.

However, copyright is only one part of the challenge. Despite being in their infancy, AI music and video generators are posing questions about how consumers will come to watch, listen and play in the future. This includes how AI generators will be used to create content for others, and how they will be used to create content for ourselves (create for self-consumption).

That future still felt some years away in May 2024 when Fable AI announced Showrunner, an AI generator and distributor of TV shows and movies dubbed ‘the Netflix of AI’ by CEO Edward Saatchi (per Forbes). While the test cases looked promising, the studio behind AI South Park episodes has kept the servicein early access. Like many AI tools, Showrunner created a major media splash without the necessary polish needed for mainstream use.

But there are some exceptions. Amazon’s new AI assistant Alexa+ includes a partnership with Suno, the controversial AI music generator (as outlined by The Verge). This tie-up enables Alexa+ users to voice prompt Suno to create ‘personalised music on the fly’. While Amazon suggests this could be used for a personalised birthday message or a rap by a child’s favourite cartoon character, the long-term implication is to create for personal consumption. The most obvious examples are:

• User has a creative idea for a song based on their mood or feeling

• User is not satisfied with their music recommendation and opts to create instead

These chatbot-empowered AI generators will enable creation to more readily compete with traditional consumption when it comes to time-spent. Not only will AI creation be able to fill the gaps left by recorded music, but it may come to offer an entirely different music consumption experience.

For example, will generative AI music provide the option of being output driven or continuous? Will the consumer be able to interact with the music in real-time? Such as “Alexa, play me some smooth jazz. Now play it faster, with more saxophone”. The next step of this evolution could even see music adapting to one’s heartbeat through their connected device as they work out in the gym. For example, the music gets faster as the work out gets harder.

Both examples point to generative AI music having impact with the type of background music consumption that made the Lofi Girl YouTube channel so popular (14.8 million subscribers). Such innovations could threaten recorded music if generative AI music listening hours increase at the expense of recorded music listening hours. However, this would only be the case if music rights holders choose to not become players in AI themselves.

There will of course be those who dismiss generative output as ‘AI slop’ that is incapable of competing with human creations. But brushing aside the create-to-consume potential of generative AI risks a similar trap that streaming TV fell into with social video content. Assumptions were made that consumers would always find time for TV shows and movies over creator content because the two outputs existed in separate lanes. Put simply, social video content was perceived as inferior to the premium quality of streaming TV content. However, YouTube has now supplanted Netflix for smart TV watch time in the US. And this has prompted even more rightsholders to publish TV shows and movies for free on YouTube. All in a bid to connect with younger audiences and turn them into SVOD subscribers.

However, the long-term effects of create-to-consume behaviours could be to drive consumers even deeper into their echo chambers. Social algorithms have created a world where everyone is able to indulge in their niche interests. Content generators will enable consumers to create content based on those niche interests or by how they feel in the moment. A need to breach these strengthening echo chambers is what may push rightsholders towards going all-in on AI generation, regardless of the copyright risks.

For more on AI’s impact on entertainment, look out for Ben Woods’ upcoming report set to be published later this month: ‘AI chatbots | The new conductors of entertainment’.