- cross-posted to:
- hackernews@lemmy.bestiver.se
- cross-posted to:
- hackernews@lemmy.bestiver.se
You must log in or # to comment.
From Deezer’s website, the detection system tags songs that are either fully AI generated rather than produced or mastered with the help of AI tools. You can also appeal if you believe your music was falsely flagged.
I strongly oppose the use of generative AI in art but if it has to be done, it should at least be labeled as AI (ideally by the “creator” themselves).
I wonder how accurate the AI detection tools are though, considering how common are posts where AI detection tools used in schools falsely flagged student assignments.
There was a song I quite liked which had several million views on YouTube which I was surprised to see was flagged as AI generated. No one I showed it to it could hear any obvious signs of AI. The main red flags were that the artists released several albums in a short time span and had no online presence on any platform you would expect to see musicians on (Bandcamp, Discogs, etc) besides YouTube and the streaming ones.
The main red flags were that the artists released several albums in a short time span and had no online presence on any platform you would expect to see musicians on (Bandcamp, Discogs, etc) besides YouTube and the streaming ones.
Honestly, those seem like pretty big red flags since that is how actual bands manage to actually get paid.
I strongly oppose the use of generative AI in art but if it has to be done, it should at least be labeled as AI.
I know I’m mostly preaching to the choir here, but I don’t think there’s any situation in which AI ‘has’ to be used in art.
I’m no artist. If I ever had the inspiration to make a song it would have to be AI generated. I’m sure I’m not the only one. Of course that would be a one off with a small audience.
Tunes generated by LLM bots should never considered as music.
I have no idea what Deezer is, and I’m afraid if I ask, somebody is just going to say “DEEZER NUTS!!!” and I will realize it was a big prank.
It is a less shitty alternative to Spotify, while costing less. They are also paying artists considerably more.
The last sentence is a little scary to me, not because it’s a bad thing, but because it’s probably catnip to scammers/AI generators. I hope they can do a good job of detecting it and keeping those scammers at bay, and not paying them for unaware listeners’ mistakes
Not necessarily, if they are more hostile towards that kind of “content” than in this case Spotify, it isn’t necessarily more attractive to AI scammers.
As far as I have read they do a lot to prevent that. AI “artists” (shartists?) don’t show in the all tab when searching, don’t get added to radio mixes, and dont get any payments from Deezer. Their AI generated tagging seems pretty accurate, I just wish it was exposed in the API so other projects could use it
Self-gottem
You couldn’t infer from the headline?
You can infer Deez Nuts.
Lmfao gottem
From the headline? No. But I could have just searched for it, or read the article. But it’s more amusing to make a slightly amusing comment.
Music streaming like Spotify or Napster.
Spot deez nuts!
Nutster
Maybe they should stop allowing it
And here I am struggling and fighting with my distributor ever time I upload a new instrumental album because they can’t confirm that it’s all original work.
Use someone else. If the AI royalty farmers can get thousands of AI generated tracks through without issue, then your real albums should be okay too.
Who’s your distributor? I used SongTradr for a while and I’m looking for a new one for my next release
I’m using OneRPM at the moment. Other than the sometimes frustrating approval process, they have been ok. It’s free with the 30% cut. I don’t get enough streams to earn a significant amount so it made the most sense to me vs using a paid distributor.
I personally started to use Qobuz. Their algorithm isn’t great, their target group is more the more distinguished music listener but their library is pretty much as big as any others plus they do have the largest library of hi-res music too and they actually sell also hi-res and CD lossless music if that is of interest to you. Most importantly though, they have a “ban-AI-music” stance on their platform. Soon enough, one will have to rely on platforms like that if one does not want to wade through a sea of AI slop.
The downside is that Qobuz is a bit more expensive than others (while paying the most to artists however, as far as I know).
Relevant, Qobuz AI policy: https://community.qobuz.com/ai-charter
I switched to Qobuz recently and started to actual discover new music again after years on Spotify.
Also Qobuz is not globally available to all.
Indeed. Its service is available in only a limited number of countries. Interestingly though it is one of the oldest streaming platforms around.
I wanted to get into Qobuz but couldn’t ultimately because the higher-res streaming killed my cellular. The app would freeze constantly when and music wouldn’t load properly when I’d try to stream it instead of playback downloaded music. I like the idea and it’s awesome that it’s higher-res music, but my phone couldn’t handle it
This (moreso for youtube music, since Deezer seems to not have a lot of East Asian labels signed) is a huge part of why I’ve been building out a selfhosted Navidrome.
Obviously there is the old school way of getting music. But Bandcamp is WAY more beneficial to the artists and ebay and Half Price Books are also awesome for grabbing music.
And combine all that with musicbrainz for scrobbling and discoverability of new bands.
I rrecently built a home server and tried getting into Navidrome, but I kinda disliked the UI - didn’t feel as intuitive to me and kinda clunky. How do you listen to your music primarily? If on a phone, do you have an app to stream the music to you?
Symfonium
It’s not FOSS and has the worst payment system possible (donate to the dev then send him an email, and he will give you a license for a single device), so don’t feel bad pirating it.
Judging by the reviews on the app store, it’s at least a 5-buck one-time purchase and not a subscription as well as a worthy purchase. I’ll give this a try
Update: it works surprisingly well. I linked it to my Jellyfin server and the music is played very well, even on the go and even if the song quality is lossless. I haven’t quite figured out which UI design I prefer, still testing things out, but so far, I might consider a purchase
Symfonium on my phone (so also android auto and just connecting to a bluetooth speaker while I cook) and Feishin on my desktop. Still need to verify that scrobbles are propagated correctly for discoverability purposes (so far it looks like ratings in Feishin aren’t propagated to the server).
I could probably have gotten away with just mpd but figured “why not?”.
Fuck me man, I guess I’ll just never consume art again.
Even if I tell you that Hans Zimmer is cooking Dune 3 score?
Nah, there’s still all the art that was created before GenAI, from the Epic of Gilgamesh to Undertale.
Yikes
And this is not going to end even if they ban them. Mark those songs as AI and let people filter them out.
But we do need a new music service where every artist has to prove they are the ones making music on live stream and only then they are allowed to upload songs.
Mark those songs as AI and let people filter them out.
Deezer does just that. As per the article:
Songs tagged as AI-generated on Deezer are automatically removed from algorithmic recommendations and not included in editorial playlists. The company announced today that it will no longer store hi-res versions of AI tracks.
They’ve been working on systems to recognise AI songs for quite some time now.
Music is a weird art form, because something sounding familiar is very important to our ear. Many people have a really hard time liking music that is too foreign to their taste and end up sticking with only a select few genres.
Where familiarity is important, AI can deliver easily. I would think as much as we hate the idea, there is a pretty significant market for AI-generated music, specifically because it’s so predictable and follows convention to a tee.
There is indeed a market for people who don’t care what is playing or who made it, and just want to hear the same familiar generic chords, rhythms, and vocals of whatever genre(s) they’ve grown up listening to. Not to be too blunt, but some people have no taste, and yes, they can eat slop and not notice the difference. Ok, good for them.
But those people are throwing fertilizer on AI weeds that will consume all the water and sunlight that nurtures actual music. That is really a problem.
There are also good reasons for people to use AI music. If you just want some music as background in a video you want to post somewhere, that totally is a legal nightmare here where I live. If you’re some small business, that is even more nightmarish. Licensing songs is expensive and hard to do, so just generating some ok tune is the best way forward
I think people should be very careful about how dependent they become on such things, because inevitably if adoption ever does creep up the spike in prices of accessing those models is going to be astronomically more than having some jingle writer slap something together. Right now they’re desperate for adoption but those servers aren’t free to run. If they’re ever going to turn a profit the fees for accessing these tools are going to be orders of magnitude more than any small business owner can afford, and by then, there won’t be any aspiring new artists to take a cash job; they’ll have either starved to death or moved on. You’re basically Wille E. Coyote-ing yourself off an advertising cliff using AI like that, and same for other similar uses.
And the sooner this happens, the better.
Music doesn’t stay under copywrite, forever. You could use anything that’s aged out of copywrite, too. And then you, as a business won’t alienate people who choose not to consume ai for ethical reasons.
Music recorded before 1923 has a distinct vibe to it that most businesses probably dont want
Especially since thats before the microphone so its all acoustically recorded
Yeah, about that: In most jurisdictions copyright lasts until 70 years after death. So that means in 2026 that both composer and all performers must have died before 1956. Using such old songs from old recordings is simply not feasible for most companies.
There are music libraries dedicated to that like the YT Audio Library for YouTube.
And what is the fundamental difference between the background music styled songs from the YT Audio Library that were done for a small penny in a short time by some underpaid performer vs. something a computer created?
I hear that, but it really depends on the service and prompt (including services’ internal prompt that is hidden) and result, which are many times black boxes.
I personally think artists & labels will have a tough time proving infringement for non-infringing outputs based purely on training data. But there’s really no way of being sure that the “generated” and “uncopyrightable” AI track that’s distilled from unlicensed source music is not itself infringing as a pure substantial similarity (or whatever your locality’s infringement legal test is) question.
Or someone trying to eek out a living with their music can get paid to do so. There’s no shortage of music to suit people’s tastes, the problem is discovering it because Spotify sucks at recommendations, and actively promotes AI slop to pad their profit margins while stiffing real musicians. So many mixes use AI instead of recommending actual artists specifically so they don’t have to pay royalties.
There might be a market for talentless trash music generators, but it actively harms real people creating real music with their real talent, and I refuse to participate. Fuck AI music. Just because there’s a market for something, doesn’t automatically mean it’s good, or the right thing to do.
i’m sure most people using streaming platforms don’t care about it. a lot of people don’t even know what their favorite genre is, they just play whatever is popular or getting into their feed.
I quit Spotify when I found that half of the music on random jazz playlists I’d listen to were all AI. My whole family told me I was full of shit and they’ve never encountered that haha. Caused a lot of drama since we have a family plan.
