AI-generated jingles? Absolutely! Nothing says "buy now" like a robot with perfect pitch.
AI-generated ad copy? You bet! Soulless selling from a soulless source.
AI-generated news synopsis? Sure! So long as it’s labeled clearly and doesn’t start quoting Nietzsche.
But when it comes to music, poetry, fiction, or art-I draw the line.
I don’t ask my hammer to write me a sonnet, and I don’t need my toaster composing symphonies.
Let the machines handle the mundane. The messy, beautiful, human stuff? That’s ours.
Sohcahtoa82 · 11h ago
What about people who just want to express themselves, but don't want to spend the hundreds of hours learning to draw, paint, play an instrument, etc, but certainly don't want to pay the money it would cost to commission?
That was me when I wanted a heavy metal song about a common office annoyance and prompted Suno to create "Per My Last Email":
Can it really be called an expression of yourself when you don’t write the lyrics or have any musical input on the song? This sounds more like having an idea for a song rather than an actual expression of yourself.
Would you still call it an expression of yourself if you paid an artist to write a song for you around an idea you had? I don’t think most people would generally refer to it as “expressing themselves” when they commission another artist.
punkrox999 · 9h ago
...you express yourself still. Art isn't about a mechanistic and functional product that needs to be "good", but what's on your mind, and often those imperfections add flair and a human story to the end result. There's tons of pop music created by people who don't even know music theory and are just messing about
bagful · 8h ago
Uh, too bad for them? What if I want a hot muscular body, but don’t want to spend the hundreds of hours running, lifting weights, planning and executing a healthy diet; and I certainly don’t want to use steroids?
Proofread0592 · 11h ago
> Deezer’s AI music detection tool sets an industry standard, with the ability to detect 100 % AI-generated music from the most prolific generative models – such as Suno and Udio, with the possibility to add detection capabilities for practically any other similar tool as long as there’s access to relevant data examples.
It would be excellent if they explained this. How? Do these systems add some sort of non-audible audio "fingerprint" of sorts that allow for the 100% detection? Or do these AI tools commonly reuse audio samples or something?
"Suno is designed for creating original music, and our models don't recognize references to other artists. We are not here to make more Fake Drakes.
"To further protect against misuse, we have developed proprietary, inaudible watermarking technology that can detect whether a song was created using Suno."
lotsoweiners · 6h ago
Glad I have very little interest in new music.
otterley · 10h ago
It kind of makes sense. If you use copyrighted music without a license, Content ID and similar identification strategies will result in your upload being taken down. Using AI-generated music for your video avoids legal concerns while fulfilling your need to have, say, a musical background in it.
yorwba · 3h ago
The article is about music streaming, nothing to do with videos. Just people hoping to make a quick buck by producing large volumes of cheap content.
AI-generated jingles? Absolutely! Nothing says "buy now" like a robot with perfect pitch.
AI-generated ad copy? You bet! Soulless selling from a soulless source.
AI-generated news synopsis? Sure! So long as it’s labeled clearly and doesn’t start quoting Nietzsche.
But when it comes to music, poetry, fiction, or art-I draw the line. I don’t ask my hammer to write me a sonnet, and I don’t need my toaster composing symphonies.
Let the machines handle the mundane. The messy, beautiful, human stuff? That’s ours.
That was me when I wanted a heavy metal song about a common office annoyance and prompted Suno to create "Per My Last Email":
https://www.youtube.com/watch?v=ZVia46yAoMU
Would you still call it an expression of yourself if you paid an artist to write a song for you around an idea you had? I don’t think most people would generally refer to it as “expressing themselves” when they commission another artist.
It would be excellent if they explained this. How? Do these systems add some sort of non-audible audio "fingerprint" of sorts that allow for the 100% detection? Or do these AI tools commonly reuse audio samples or something?
"Suno is designed for creating original music, and our models don't recognize references to other artists. We are not here to make more Fake Drakes.
"To further protect against misuse, we have developed proprietary, inaudible watermarking technology that can detect whether a song was created using Suno."