Loopy Pro: Create music, your way.
What is Loopy Pro? — Loopy Pro is a powerful, flexible, and intuitive live looper, sampler, clip launcher and DAW for iPhone and iPad. At its core, it allows you to record and layer sounds in real-time to create complex musical arrangements. But it doesn’t stop there—Loopy Pro offers advanced tools to customize your workflow, build dynamic performance setups, and create a seamless connection between instruments, effects, and external gear.
Use it for live looping, sequencing, arranging, mixing, and much more. Whether you're a live performer, a producer, or just experimenting with sound, Loopy Pro helps you take control of your creative process.
Download on the App StoreLoopy Pro is your all-in-one musical toolkit. Try it for free today.
Benn Jordan's using AI to detect AI Music
Benn put a lot of work into this one:
Comments
Very interesting watch. And damning evidence of Spotify and some other streaming services. This exposes even more barriers for musicians who want to monetize their creations. Seems to be a natural consequence of the hyper-capitalized music industry.
To what end though. I value this message. As do other musicians (probably, possibly). But does the average listener care about or even understand what the issue is. If people are enjoying what they are listening to will anyone believe (or care) that there is a problem?
To put it another way. If I read a book and like it, do I care who or what wrote it? Yes, I do, because I value the arts as a creative outlet and potential income source. Does the average reader though?
I liked this video. I think he has a good presenting style and has his media chops well and truly in order. The statistics were an interesting thing as well.
One take away I get from all of it is that in the end the music market is completely saturated. Basic economics is being shown in the manner in which the amounts artists are getting paid per stream is reducing while at the same time the number of tracks being released is increasing at a phenomenal rate. So it makes sense that the streaming pay rates are what they are and they are only going to drop even further as time goes on.
I think it might really be time to walk away from this Internet thing.
I’ve heard there is a world out there with people in it and things you can actually touch and interact with.
Sharing moments in real time without an interloper…think I might take a look.
If you don’t hear from me you know it’s all going well….
I feel you. It seems obvious that the same problem is going to hit YouTube. With more AI content and more creators in general flooding the platform, a niche like iOS music making just can’t grow its viewership to match the overall explosion of content. That means the already tiny slice of the advertising revenue pie for people like me and Doug is only going to get smaller.
I think it's more about legal uses than about informing the public about the origin of a piece of music.
He talks about the idea that an algorithm capable of identifying procedurally generated music could allow the streaming service to deny payout, since the work has no copyright, as it has no author. That would lessen the financial incentive to upload the stuff, as it may be identified and result in the uploader having worked for the streaming service for free.
Another use could be countering Content ID strikes based on music that has no human author.
It looks like a proof of concept for now, and who knows what the streaming services would end up doing (they are paying out so little, that they may not even care) but the idea of using an algorithm aimed at reducing the information contained in music files to what humans can perceive, to identify when this limited amount of information has been used as the base for machine learning, sounds a little like poetic justice to me.
I love Benn Jordan rants, very insightful!
Yeah, I think there’s just way too much content now. Humans are creative creatures, and we’re so desperate to make something meaningful that we’re willing to do it for mere pennies (or even for free). And that means there’s this ocean of free or nearly free content and it’s hard to even figure out what to pay attention to.
And as if that situation weren’t dire enough for the arts, we’ve now got these AIs that can generate semi-competent AI slop in seconds, far faster than any human. It feels like we’re at war with art itself, trying to kill it by depriving it of meaning altogether.
I believe the video is great though seeing any machine learning solution with a 100% accuracy, as Benn states, makes me skeptical about the methodology. Could there be a problem with how data got assigned for validation or a bug anywhere in accuracy code? At first glance solving this problem with near 100% accuracy seems unlikely.
Normally that would be true, but lossy audio is so heavily compressed compared to most other formats that his explanation makes sense. If Suno pulled most of their data from the same source (like they scraped Spotify) then most of their training data would be compressed in the same way.
If you keep seeing compression artifacts from 128kbps MP3 using the same exact encoder, your AI is going to generate artifacts like that even though your audio isn’t compressed yet. Encoding that audio as an MP3 will double down on those compression artifacts. At a certain point, I could see this forming a certain kind of compression fingerprint.
I’m sure there could be some false positives, but it just didn’t come up in his training set. I think it’s an interesting idea at the very least.
Is he able to detect use of sampled instruments, versus real ones? Or how about other (better) songs which have entire sections lifted and used as the basis for a new song? No one cares. Seriously. If people like a song they're not going to care how many people were involved (or not involved).
I'm no expert so please consider this friendly discussion. Just dot pointed to keep my thoughts clear.
I get the idea about "an algorithm capable of identifying procedurally generated music". But aren't the streaming platforms (or maybe just Spotify) creating their own AI music? Or outsourcing it, or preferencing it somehow. IF, that's correct, I can't see a streaming service partaking of the algorithm.
I also find it hard to figure out where I stand on the "no author". If you have an authoring scale from, just for arguments sake, pure AI through to a recording of a classical flute player's original piece, where is the point you determine "no author"? I hate to say it, but there is some effort in even uploading the pure AI audio. Does that deserve payment? I don't think so. To me it is just muddying the waters of what should be a world of human creativity. I realize it's not authoring but it's work to upload it, yes/no? And what is the difference in terms of authorship between a song composed purely of loops and a song that is recorded live by musicians playing instruments with an AI vocal track? Seems like a complicated space. (For the record, I believe there is a huge difference between someone mashing a few loops together and an artist who creatively selects, edits, and uses loops to create something new).
"That would lessen the financial incentive to upload the stuff". Financial incentive? I didn't know there was any
I'm too old for this shit. I just want the music industry to be a source of enjoyment, fair income etc. depending on your role, for every HUMAN involved.
And on a final note. Fuck Spotify and their ilk. Fuck AI.
That's what my gut tells me.
Fun fact: “no one cares” isn’t the same as “it doesn’t matter”.
————————————-
There are often long-term consequences for things people don’t care about.
Consumers not caring whether a book is written by AI or an individual song is written by AI doesn’t mean there won’t be long-term consequences for replacing authors, musicians and composers with AI.
The push towards the AI’ing of everything is not driven by an attempt or desire to make you better at writing or composing…it is driven by corporations who don’t want to pay musicians, writers, etc
Sam Altman is back to pleading with the government to give him free access to copyrighted material so that his company can be profitable.
I didn't use the phrase "it doesn't matter", but to whom does it matter? Not the listener / audience. The recorded material matters. Any drama or details beyond that are up for debate.
Not so much of a "fun fact" but thanks Captain Obvious.
I'm also really not an expert, and I'm afraid I didn't make the point I was trying to make at all.
The reason I quoted your post, was because you were talking about whether the average istener cares about this, and I wanted to say that I don't believe Benn Jordan was trying to inform the average listener here. This was a video for music and data nerds, that already care. And for people that can build on what he found out.
I took your post as saying that the video misses it's mark (which you probably were not even saying), and wanted to say that the video was going for something else.
And I ended up writing only about the possible applications of the tool he developed, because I was so fascinated by the methodology and what could be done with it.
To your points (sorry, it got way too long):
I believe for now Spotify is not officially using AI music, they used that "AI podcast" in Spotify Wrapped this year, and have some AI DJ, but for now they are not creating their own music. But my hypothetical was about the algorithm turning out to be very reliable, and Spotify deciding to use it to avoid paying for some of the streams (and this having some positive side effect). But they may as well decide to do that, and fill the service with their own AI music instead.
At the moment in the US it's pretty clear that the pure output of an AI cannot be copyrighted, and if a human contributes to the work, the copyright may apply only to the parts of the work the human has contributed to. In other countries (UK, maybe India) the person writing the prompt for the machine will be the full owner of the copyright on the work the machine creates based on his inputs.
Maybe this will change in the united states sometimes soon. But even now, as long as you claim the copyright, you will be recognized as the owner of the whole work, if you are not telling how you made it. A system that allows to reliably recognize works that have been created by an AI trained on data scraped from compressed music files, would at least make the lack of a human author provable in those cases.
And yes, it is a question of degrees and of questions like if an "authorial kind of contribution" has been made by the person claiming ownership.
Loops are a different issue, where you are using pieces of works that are protected by copyright, and the question is if and how much the original copyright will extend into your work. The situation with AI created music, at least in the US, could maybe be compared to a piece of music created with loops from a work that's in the public domain, were if you put no work in, the work will be under no copyright at all (say, you are just renaming a public domain song).
It's already a very complicated field, and the introduction of all the AI questions will probably change how a lot of these questions get handled.
It could be enough for just uploading something you put almost no work into, the payout rates seem to be based on that assumption... But yeah as I was saying the payout is so little that probably Spotify wouldn't even care for the money.
I really hope that we will move in that direction again sometimes, and maybe Benn's contribution here helps a bit.
I hadn't heard about Spotify creating AI music, and found this interview with the CTO where he says that's not something they want to do (so I believe they are also not currently doing it):
https://www.bigtechnology.com/p/spotifys-plans-for-ai-generated-music
And on the questions of ownership of AI generated works, this seemed to be the most recent overview (I wasn't sure if anything had changed):
https://builtin.com/artificial-intelligence/ai-copyright
@Hippo You are on to it. I know the video is aimed at us, musicians, etc. Not sure what my point was either but I did sort of have a feeling that, there is little point to the video. We already know the industry is fucked for artists, and I don't feel that it's such an issue for the bulk of the listening public, that people that would make a conscious consumer change to support artists better.