The rubicon has been crossed. I generally don't comment on AI models, simply because of their ephemeral nature, but Suno v5 made some cultural aspects of AI-generated art so much more obvious. As someone who has spent a lot of my life thinking about, listening to and making music, it is a very very important part of my life. Music is the basis of so many of my relationships, I experience a great deal of joy listening to music, and I find it to be one of the most reliable ways to share emotions with others. In short: music is culture, culture is music.
The first time I was knowingly confronted with a song that was AI-generated was the German language song "Verknallt in einen Talahon" (see here). It made the rounds in German-speaking media sometime in 2024, and features a musical style reminiscent of 1960s German Schlager. There has been quite some controversy around the term and the song that I will not get into, which is evidence that it made a cultural splash. AI generation was not a topic, and it probably didn't even make a difference, apart from the fact that this could be produced by a single person in what I can only assume to be a short time. What made this song draw attention was the input by the user, the expression, the sentiment, the lyrics that hit a cultural spark.
Suno v5 allows you to generate stylistically almost pixel perfect imitations of musical genres. It allows you to enter lyrics, or come up with themes that will then be filled in. And the output of it can have an emotional capacity for the listener. I sent some AI generated music to friends and unanimously, the reaction was one of awe first, shock second. After pondering it for a while, this model revealed to me how utterly worthless this technology is without a human touch. The first song is surprisingly good, you second-guess yourself, did I put on a real song accidentally? Is this just a carbon copy of some other song? After about song 5 you realize what is going on, and it starts to become somewhat boring (and fun for other reasons: look what I made! How can I break it?). And then you start to feed it with material and topics, and the songs become good in a commercial radio sense. This probably has already happened, but I think the market for music for advertisements is completely gone - this can be done with just AI going forward.
What it allows you to do as well, however, is to bond with others over specific expressions of culture. During some beers with colleagues, we discovered that when you ask it to go from opera to metal, it will create a perfect Eurovision-style song about anything. It was hilarious, and we had a good time. We will never listen to these songs again. And trying to show it to others who were not there for the creation is an exercise in cringe-resistance.
The demise of public broadcasting and the rise of social media were the beginning of a transformation from a shared cultural basis to hyper-local culture in bubbles. The problem of these cultural bubbles is that we no longer share a cultural basis to bridge disagreements with, especially because much of culture has been infused with local political expression. I fear that models like the new Suno model will be used exactly to the same end, and it's bad. This is not conjecture: Suno themselves published a blog post about using the application to generate personalized songs for people in need and distress. While this is a worthwhile initiative in general, I think it underestimates how important the human component of these endeavors is.
Human connection and culture are not factory-made, nor should they be. I recently heard about a funeral speech that was very obviously written using ChatGPT, including the hallmarks of generated language. If you cannot bring yourself to write a page about someone who is close enough to you that you are speaking at their funeral, you should have the decency to write a few words yourself. Both for your own sake, the others attending, and the deceased.
The hyperreality of AI-generated cultural artifacts makes them worthless for actual culture, because they lack what Walter Benjamin calls aura in "The Work of Art in the Age of Mechanical Reproduction". What will this do with culture? I don't know. In some pockets of Silicon Valley this is a hot take, but I don't think human-made music, or any art, or any culture for that matter, is going away. Real culture inspires you in a way that these models never will, and as a society, we need to continue investing in the arts for this exact reason. I'm not scared that it is going away, as long as we all realize the difference between true expression and hyperreal artifact.
And as such, my personal plea: use AI tools for whatever you like, but if it is about human connection and expression, please do it yourself.