Even if the traffic was inflated by bots, there were many users that enjoyed the content itself. What that tells you is that people don’t have an inherent problem with the music, they just want transparancy around it.
Now, I don’t use Spotify, but what I use allows me to pick specific songs, but it defaults to “shuffle”. Sometimes it’s stuff I listen to, sometimes it’s new.
I’m not aware of anything showing a breakdown of intentional listens and popping up on “shuffle”.
From a label perspective, AI is the best kind of band because it will do whatever you say, it will never refuse to do anything out of integrity. So it seems a reasonable assumption that what they’re aiming for is “elevator music” something innocuous enough that people won’t hit skip.
If it’s too good, people look into it, discover it’s AI, and stop caring about it.
Like the vast majority of AI stuff, it might work short term, but that’s only a novelty and those wear off. If people could opt out of AI music, the overwhelming amount of people would take the time to do so.
If ratings were so easy to manipulate we would dispute every single top hit no matter if it was LLM-generated or not. Even in your shuffle usage, you probably don’t actively look at each new song, but just hit like when something catches your ear. So as long as the music is appealing people will listen. And Spotify only counts listens aften 30 seconds and only non-subsequent repeats.
If people could opt out of AI music, the overwhelming amount of people would take the time to do so.
It doesn’t really matter if it’s LLM-generated or not for most people. As with other LLM usage, while loud minority screams bloody murder, LLM tools breaks usage records.
There are many issues with LLMs, but majority of people only care about the end result. And yes, some people would opt-out on principle, but saying that most people would is delusional.
Personally if LLM could replicate my favorite bands and supplement the void when they are in between album releases, I can guarantee you that I would listen to it. That doesn’t mean I would stop listening to those artists.
So there are 2 disputed (both denied inflating their numbers) examples in 19 years? Spotify invested millions into detecting fake listens as they have to pay out for them. And sure it might happen on small scale, but I haven’t seen any verifiable information on it happening on a large scale. The cost of that would be enourmous as you need residential IPs to even try which are in limited supply.
Even if the traffic was inflated by bots, there were many users that enjoyed the content itself. What that tells you is that people don’t have an inherent problem with the music, they just want transparancy around it.
We don’t know that.
Now, I don’t use Spotify, but what I use allows me to pick specific songs, but it defaults to “shuffle”. Sometimes it’s stuff I listen to, sometimes it’s new.
I’m not aware of anything showing a breakdown of intentional listens and popping up on “shuffle”.
From a label perspective, AI is the best kind of band because it will do whatever you say, it will never refuse to do anything out of integrity. So it seems a reasonable assumption that what they’re aiming for is “elevator music” something innocuous enough that people won’t hit skip.
If it’s too good, people look into it, discover it’s AI, and stop caring about it.
Like the vast majority of AI stuff, it might work short term, but that’s only a novelty and those wear off. If people could opt out of AI music, the overwhelming amount of people would take the time to do so.
If ratings were so easy to manipulate we would dispute every single top hit no matter if it was LLM-generated or not. Even in your shuffle usage, you probably don’t actively look at each new song, but just hit like when something catches your ear. So as long as the music is appealing people will listen. And Spotify only counts listens aften 30 seconds and only non-subsequent repeats.
It doesn’t really matter if it’s LLM-generated or not for most people. As with other LLM usage, while loud minority screams bloody murder, LLM tools breaks usage records. There are many issues with LLMs, but majority of people only care about the end result. And yes, some people would opt-out on principle, but saying that most people would is delusional.
Personally if LLM could replicate my favorite bands and supplement the void when they are in between album releases, I can guarantee you that I would listen to it. That doesn’t mean I would stop listening to those artists.
I mean, yeah…
Drake just told on himself that was something that was done to inflate his own hits, so he assumed Kendrick had to have done the same thing.
Like, this has always been a thing, even back when it was giving a DJ an envelope of cash to get radio plays.
With an AI band, they don’t have to pay a band. So they’ll spend on this bullshit that, yes, human artists also use.
So there are 2 disputed (both denied inflating their numbers) examples in 19 years? Spotify invested millions into detecting fake listens as they have to pay out for them. And sure it might happen on small scale, but I haven’t seen any verifiable information on it happening on a large scale. The cost of that would be enourmous as you need residential IPs to even try which are in limited supply.
…
If I said California had two NFL teams, would you take that to mean there’s only two teams in the NFL?
Because that’s what you’re doing by acting like the only mentioned examples are the examples.
I haven’t heard of any other cases.
If you know of that being the case please educate me.