Is the flood of AI-generated material on You Tube an addition or hindrance to creativity?
You Tube was once celebrated as a platform where anyone with a camera, an idea, and a little creativity could build an audience.
Over the years, You Tube has evolved into a vast ecosystem of educators, entertainers, journalists and musicians sharing genuinely original work. Today, however, a new wave of content is sweeping onto the platform: AI-generated and synthetic media.
Artificial Intelligence has made it easier than ever to produce videos at astonishing scale. Scripts can be written in seconds, voices cloned or synthesised, images and animations created, and even editing can be handled by software with minimal human input. While these tools have opened exciting creative opportunities, they have also contributed to a growing problem; a flood of low-effort, mass produced content that can make the platform feel artificial.
AI is a great tool. Used responsibly, it can assist with production and it can be used as a checking tool. I often use AI software to check how my lyrics might sound, but I would never use it on a final demo production, and I would never use it to create content from scratch, preferring instead to rely on my knowledge of music theory and my skill in writing music and lyrics.
The surge in synthetic content online presents several challenges for viewers. Firstly, it has become increasingly difficult to distinguish between the ‘real thing’ and AI-generated content. But another challenge is the creativity process itself. Can the production of AI content be referred to as being ‘creative’, or is it simply a case of inputting instructions into the AI front-end for the AI library to analyse information in its database and then to spill out the result.
I came across a fantastic piece of work recently. I say fantastic because the interpretation of one of my favourite songs has been given the AI treatment, and I do think the AI version has, at least, some merit.
The Boxer, written by Paul Simon, was released in 1969 as a single and then included in the Simon and Garfunkel album Bridge Over Troubled Water in 1970. Largely about loneliness and poverty in New York City, the song has been covered by numerous artists over the years. The lyrics, melody and harmonies are truly a masterclass in songwriting. And, it has had the AI treatment too.
I have included You Tube links to two versions of the song, one that shows a terrific performance of the song by Paul Simon and the other created by AI as an interpretation of the song. Truth be told, I love both versions! The interpretation that AI has produced, and presumably instructed by a human, is very different and very interesting. I guess this will be hugely controversial, but I do like the AI version, not as much as the real thing, but it is nevertheless an interesting take on a massive song.
But, can the AI version of The Boxer be classed as ‘creative’ or is it just AI doing its thing? Presumably, the person(s) who instructed AI to create this version did so with the intent of creating something different. Presumably, creativity was at the core of how they interpreted the song and how they went on to create something unique from it. Or, has AI been allowed to run wild and do its own thing with minimal human instruction? If the former, then possibly this version could be accredited as creative interpretation. However, if the AI version has been created with minimal human instruction, relying on its database and library of content, I am not convinced that I would refer to that as being creative.
AI is here to stay. As I mentioned above, it can be a really useful tool in the creativity process. AI works best when it enhances rather than replaces human creativity and many professional musicians already use AI as an assistant rather than as a replacement. Used thoughtfully, it can complement human intelligence without diminishing the role of the artist.
Ultimately, the difference lies not in whether AI is involved, but in who is driving the process. Music created through genuine artistic intent carries the fingerprints of its creator; their experiences, influences, imperfections and imaginations. AI can imitate but authenticity remains rooted in the human stories that inspire music. Great music is about more than technical perfection, it is about expression, vulnerability and shared experience.
The performance and expression of emotion by Paul Simon in the video is unquestionable, which is what music performance is all about. In contrast, online AI music is rather flat and emotionless. Side by side, I know exactly which version of The Boxer I prefer, although I do find the AI version an intriguing alternative.
One final point that is important to make here is the subject of legality. Using AI to create a new version of a song is perhaps a fun thing to do. However, there is a huge question mark about potential breaches of copyright. Purely generated AI music does not necessarily breach copyright, but the underlying AI databases that are used to input and train AI models is highly contested.
Sony Music Entertainment and Warner Music Group have filed lawsuits against some AI companies, alleging they have infringed on copyright of more than 60,000 recordings, because their systems use these recordings when creating new songs or new versions of songs. AI relies of vast libraries of stored music so that it can create new pieces, and the question is whether storing songs in these libraries has breached copyright.
Regardless of the legal cases that are being worked in the background, AI looks to be here to stay. Just as musicians learn from previous work, allowing them to create new songs and new instrumental pieces, it could be argued that AI is doing exactly the same; learning from experience so that it can follow human instructions and produce something new.
The debate on whether AI is contributing to creativity or is hindering recognition of ‘real’ artists is a controversial subject and no doubt the debate will go on and on. Meanwhile, we probably need to live with and enjoy, if we can, the alternative synthetic versions of songs that are now commonplace.
At least, for now, we can rest easy that actual performances by great musicians will continue to bring out emotions and experiences when playing their music, something that perhaps AI can’t do.
Please let me know your thoughts on the use of AI.
By Norman Thomson
NGT Music 