Google just made its AI music tool a lot harder to ignore, and the new iOS app is the piece that changes who can actually use it. I have mixed feelings about AI-generated music. On one side, the technology is simply amazing. Over a year and a half ago, I played around with some sort of new service and created a bunch of really short songs based solely on prompts. Back then, it was fun and mind-blowing.
Now, “artists” are launching on social media with complete albums, websites, personalities, and more… but they are completely AI-created, from graphics to lyrics to the music. I worry about actual physical musicians, the ones you see at bars and nightclubs with other real people.
And there is the “training” of models. How is this being completed without copyright infringement? If you want to create a song in the style of a particular artist, how was the AI trained to know that artist’s style? While supposedly Google Lyria is trained on licensed data, where do the guardrails start?
Still, it’s intriguing, and this same question goes across images, writing, design, and even code. And where and when does it all start becoming incredibly bland? I guess I will go listen to an artist from my childhood whom I have seen in concert many times. That’s real.