The Human Signal — with Laura Sheeran

In this episode, I'm thinking through what it means to share work online at a time when artists' images, music, ideas, and creative processes are being absorbed into AI systems, most often without permission.
I talk about finding my own music listed in AI training datasets, seeing another artist's costume work replicated through AI-generated images, and the pressure artists are under to keep posting, keep sharing, and keep themselves visible, regardless of these damaging consequences. I also reflect on my own uncertainty around releasing new music, especially when the platforms that help people discover creative work are often the same systems that make that work feel unsafe to share.

If you enjoyed this podcast, please share it with a friend or consider leaving a comment or review—it really helps. You can also support me and my work by becoming a paid member for €5 per month on Substack or Patreon.

You can also find me on YouTube and Instagram at @the_persona_project__ and @laurasheeran_ie.

Thanks for being here, and remember: Put humans first. Don't feed the machines.

---

Disclaimer: The Human Signal is an independent opinion podcast. The views expressed are my own and reflect my understanding at the time of recording. They are shared to encourage discussion and are intended for entertainment purposes. This podcast does not constitute journalism, nor does it provide legal, financial, medical, or other professional advice. I encourage you to verify information independently, think critically, and come to your own conclusions.
Hosted on Acast. See acast.com/privacy for more information.

Hello and welcome to The Human Signal with me, Laura Sheeran.

After an unexpected two-month gap in the podcast, I’m back! The first half of this episode is me explaining what happened, why I ended up going dark for so long. I also talk a bit about the AI watchdog and how I found ten of my songs in the published datasets used to train Suno and Udio… the AI music saga continues…
Then in the second half I pick up where I left off with the 11 part series I was in the middle of posting before I paused, returning with Part 7 of #AIWTF which is an episode I recorded back in April.

Part 7 here looks at the extent to which AI music has been entering the charts, what that says about the collapse of chart integrity, and what happens when music becomes less about human choice and preference, and more about automated algorithmic success.
We then move from there onto a wider reflection on social fragmentation, the erosion of shared culture, and the need for a lot more real spaces where people can gather, play, experiment, and feel connected again through their art.

If you enjoyed this podcast please share it with a friend or consider leaving a comment or review, it really helps me out. You can also support me and my work by becoming a paid member for €5 per month on Substack and Patreon. And lastly, you can still find me on YouTube and Instagram as @the_persona_project__ & @laurasheeran_ie
That's all for now.

Thanks for being here, and remember: Put humans first. Don't feed the machines.
Hosted on Acast. See acast.com/privacy for more information.

Hello! After an unexpected two-month gap in the podcast, I’m back! The first half of this episode is me explaining what happened, why I ended up going dark for so long. I also talk a bit about the AI watchdog and how I found ten of my songs in the published datasets used to train Suno and Udio… the AI music saga continues…
Then in the second half I pick up where I left off with the 11 part series I was in the middle of posting before I paused, returning with Part 7 of #AIWTF which is an episode I recorded back in April.
Part 7 here looks at the extent to which AI music has been entering the charts, what that says about the collapse of chart integrity, and what happens when music becomes less about human choice and preference, and more about automated algorithmic success.
We then move from there onto a wider reflection on social fragmentation, the erosion of shared culture, and the need for a lot more real spaces where people can gather, play, experiment, and feel connected again through their art.
If you enjoyed this podcast please share it with a friend or consider leaving a comment or review, it really helps me out. You can also support me and my work by becoming a paid member, it’s €5 per month on Substack and Patreon. And lastly, you can still find me on YouTube / Instagram as @the_persona_project__ & @laurasheeran_ie
That's all for now. Thanks for being here, and remember: Put humans first. Don't feed the machines.
Long-form edited transcript:
Hello and welcome to The Human Signal with me, Laura Sheeran. I’m back, I’m actually back. And I am about to pick up where I left off with episode 7 of my 11-part series, AI What the Fuck. But first, I’m going to explain where I’ve been, why I was gone for so long, and what I’m doing right now.

The Two Month Gap
The last episode I published went out on the 28th of April. Today is the 28th of June, so there is a two-month gap there. It was an extremely unfortunately timed issue I encountered, because I was halfway through publishing this eleven part series and even in the space of time between recording and publishing, which had only been one week at the time, everything about AI and the music industry that I had researched and discussed in the episodes was already evolving and changing. So you can only imagine how dated the un-posted second half of the episodes are now two months on!
However, I have gone back through them and listened through, and a good bit of it is still usable. I’m going to reinvigorate each episode with an opening section, so that I can commit to finally finishing the series now that I’m back.
This is so typical and the story of my life, but right before I went dark, one of the last things I posted on here was my April newsletter, in which I discussed my sense of achievement over having published so many episodes of my podcast. I was celebrating the publication of the 50th episode, which is a really big deal in my personal world. Through inconsistent posting, I had managed to overall post relatively consistently.
I wrote: this is something I never thought would happen or be possible for me to even achieve. When I published the most recent episode and I realized it was number 50, I had a moment of disbelief. I’m genuinely quite proud of this milestone because I have struggled my whole life with follow through. This is a moment where I can now look at something I’ve done and learn a new message about myself. Instead of the ongoing self-hate and shame that I have felt my whole life over my lack of follow through, I am actually able to keep something up.
A couple of days after I sent out that newsletter, I was flying to London where I was doing a photo shoot. I work as a photographer from time to time. The photo shoot went really well and I was only there for one night before flying home. In the period of time between getting onto the plane and getting off the plane, my laptop had significant screen damage and when I got home I realised the screen was completely fucked, to the point where the laptop was completely unusable.
It threw a spanner in the works. I have a pretty streamlined system that I follow with my work, and as much as I would love to be in the position where I can just run out to the store and replace it the next day, that is not my current economic situation. It ended up taking me about six weeks until I could replace the laptop.
In that time, I tried to transfer my workload from the administrative stuff and this podcast stuff that I had been doing exclusively on the laptop over to the desktop, and it just was not working.
The Brain Said NOPE!
This podcast is called The Human Signal for a reason, and in many ways this phase of two months was actually a very authentic expression of the fact that I am a human being. This was just the natural way that this scenario played out in my ability to be productive and keep on top of my tasks.
So many things started slipping through the cracks when I didn’t have my usual system set up that I can rely on and depend on. For many people, it might be very easy to just switch tasks from a laptop to a desktop. For me, it was genuinely not possible. I tried and I failed. It wasn’t for want of trying and I’m going to go into some detail here about why I think it was so challenging.
Practically speaking, I use my laptop for tasks that generally don’t overlap with the tasks I do on the desktop computer. The desktop computer I use almost exclusively for creative work, which is related to entering into a flow state and essentially completely unhooking yourself from the concept of time. When I’m video editing, editing a photograph, recording music, writing musical parts, producing songs, you’re really losing yourself into a flow state. That’s where you access the best work.
My laptop world is a different world, and I need to be in a completely different headspace when using it. When I’m working on my laptop, I’m generally doing admin or recording the podcast. Both of those things require me to be alert and present and engaged with the action of time. Usually in the morning I bring my laptop downstairs, make a coffee, do my emails, look at my calendar, take my notes, and go through the organizational stuff I need to do. It sets me up for the day ahead, and for the week ahead. When I’m doing my podcast, again I’m engaged with time and being fully switch ‘on’, looking at notes, thinking of the things, and being alert.
Whereas when I’m at my desktop, it’s all about de-hooking from time, letting time slide away from all perception of reality in that moment. That’s what flow state is, thats how I achieve my best work in that context.
The weird thing was that when I was at the desktop computer, none of the email scenario or calendar checking scenario or podcast recording scenario worked and I think it must because I’ve become so physically accustomed to entering flow state when I’m in that room and at that computer..
I missed meetings. I showed up at the wrong NCT test center. I went to a meeting a month and a day early because of my inability to read my calendar and my emails efficiently on my desktop computer.
It’s not like I wasn’t doing the emails. I was doing it, going through the motions of it, but it just was not going into my brain. None of the information would stay in there. It was like falling through a sieve.
I did my absolute best, but the amount of destruction on my professional engagements, my work and personal things as well, just losing track… I started to wonder, is this like early onset dementia? What’s going on with my cognitive ability? Thankfully it wasn’t and as soon as I got my old system back with a laptop replacement, everything started returning to normal.
It got me thinking a lot about flow state and our neurology, the neural pathways, the roads that get hammered into our brains over years and years and years. And then when you try to shift those, I don’t know how long it takes for change to become possible. Maybe if you’re in your early 20s there’s more plasticity there still, and maybe I’ve gone beyond that point in terms of age. I don’t know, but it was fascinating.
So here we are. There was six weeks where I had no laptop at all. Thankfully, I didn’t lose too many files. That means I can actually jump back in where I was in the series I had made.
The Evidence We’ve All Been Waiting For
We are about to jump into part seven, but before I do that, there has been a big development in the whole AI music thing. A lot of the datasets used to train the AI models for Suno and Udio were published recently by The Atlantic. They are searchable. Loads of artists have been going on there and searching their names, searching for their music, and finding that yes, in fact, their music has been scraped and used to train AI.
Before this point, it was all entirely speculative. It was extremely opaque. We couldn’t really see exactly what had gone in there. But this publication of the datasets has illuminated that and exposed the truth of the matter. Of course, I did search for my name and music and, in at least one of the datasets, I found that 10 of my songs have been absorbed into the AI music machine.
Was I pissed off? Of course. But I wasn’t surprised. In many ways, there was a sense of relief. I feel entirely justified now in giving out until the cows come home about this situation because it does have a direct impact.
It’s really outrageous, the amount of work that has gone into the system. But having that clarification is helpful. I think it’s going to help move things forward in what feels to me like a more ethical direction, which would be to center real human artists, real musicians, and to remember that we are here. We care. This is our work. We wrote this stuff. It belongs to us. It was never anybody else’s to take. And just because they can take whatever they think they want off you, it doesn’t mean that it’s okay or justified or legal.
Having this information explicitly available and accessible brings the major label artists and the independent artists together under the same umbrella. We all can identify with the same sense of injustice over what’s happened. The outrage is palpable no matter where in the ecosystem you fall.
All of that being said, finally now I’m jumping into part seven.
What Do Music Charts Represent Today
This episode centers mainly around the cultural significance of the charts in music, in the music industry. All of this was recorded in April, so my references to just last week, there was an artist, etcetera, take with a pinch of salt. I haven’t assessed the chart situation in the past eight weeks, so I don’t know how many AI artists have or have not been included in the charts in that time. But anyway, as of April, this was the situation.
Right now, AI music is consistently topping the charts. We’ve been seeing this since well back in last year. We had Xenia Monet. She was number one. She got signed for like $3 million in September by a non-major label. She’s an AI. And then just last week, there was another AI called Eddie Dalton, who began charting this month in April and secured 11 spots on the iTunes country charts, despite not being a real person.
The AI music is continuing to thrive in the charts. And this is obviously only possible because there is no chart integrity anymore. The fabric of how we measure the success of music is changing. Considering in Deezer’s data, 80-something percent of the AI music streams they recorded were fraudulent, the charting of these AI characters seems highly likely to involve streams or numbers that are probably not verified or not real streams.
It’s harder and harder to know because even the Billboard chart still takes streams into consideration. Obviously the iTunes chart used to be paid downloads, but bots can still influence those numbers. So the charts no longer really mean anything. We’re moving in the direction where the charts don’t measure what humans want or what humans are into or what humans are even listening to.
The way people listen to music isn’t the same. People are listening to stuff because it’s coming up on playlists, whereas before people would listen to stuff because they were obsessed with an artist. They would download that record and listen to it all the time, or they would have a playlist of their favorite artists that they made themselves. There were loads of ways that people listened, but it was targeted with the music that they knew and loved. They would be fans of an artist who would define a whole chapter in their life.
It was all part of a bigger social picture where music was creating the soundtrack to a certain time in life. Music was this sort of social glue which helped narrate the tone of our lives over time, and we all were involved in what music was fueling that because it was based on humans picking what they wanted. It was what we were vibing off. It was what music was resonating with the people.
I know the charts have always involved gatekeeping and labels dominating the charts, and I know it’s not as simple as that, but that is definitely how things were or felt at least, fado fado, and things couldn’t be further from that now.
People don’t even know who they’re listening to. They might have a song on repeat and not know the name of the artist because it’s just been on a playlist. So the people getting into the charts are not reflective of humanity and our autonomy in choosing and seeking the music that’s truly resonating with us for a cultural moment. It’s just whatever the playlist has put you with. Yes, it’s algorithmically generated and yes, you probably like it because it is based on all the other stuff you like, but it’s not intentional. There’s no autonomy in it. So the charts themselves, what do they even mean?
It reflects the broader culture as well, where there’s so much fragmentation. The way we consume our news and entertainment is all so fragmented. Everyone has their own private news cycle that they get fed because of the algorithm. No one is plugged into the same one frame of information.
There’s a sadness in me over how much music could play a role in helping us come together, accept difference, understand different perspectives, unify, be together, experience togetherness. Right now, I don’t feel that music has the kind of influence it used to have to shape cultural things. Everything has become very multi-layered, very complex, very individualistic.
The Algorithmic Creative Complex
I am calling this the algorithmic creative complex: not only the listening experience of music changing how people interact with music and how music gets spread around, but also how people approach making their music. No more than with YouTube or Instagram, there are all these rules about how you’re supposed to have the hook at the start, use this and this, have this duration, all these tricks people are supposed to incorporate to give their music a chance to be picked up by the algorithm. All these hoops you’re supposed to jump through, which I refuse to do because I don’t have time for that kind of thing, and I’m making the music I’m making and that’s just the way it is.
I think that the algorithms, and becoming a slave to the algorithms, have definitely influenced how people are approaching making their music. Not everyone, of course, but there is a shape that gets put on things collectively through a mixture of conscious and unconscious decisions. Everyone is trying to do their best to thrive and grow and produce good work, share their work with as many people as they can. And if there’s this carrot dangling in front of you saying if you do this, more people are going to hear it, part of you is definitely itching like maybe I should just do that, maybe that would help.
But that has consequences down the line, where everyone’s tiny little decisions to placate the algorithms over time collectively morph into these shifts. This is all just me thinking. This is not based on any evidence. This is just how I have been making sense of it in my head.
A lot of this podcast, the fact that I don’t pander to an algorithm, that’s a decision. It’s a conscious decision I’m making, and it’s been a decision that has allowed me to continue posting. I would have given up long ago if I was trying to make this a regular, consistent, algorithmically friendly thing, because that’s not the way my life is structured. It’s not the way the life of an artist works. I work freelance. I never have the same day twice. I can’t contort myself into the structured shape for the algorithm that it would demand off me.
Algorithmic creative complex. I think we need to dismantle that.
I guess all I’m saying is I feel there’s a reduction or a decline in the sense of having a shared cultural experience in relation to music, and I think that is making it harder for us to feel connected. I think music could really unlock a huge amount for us in this moment in time, and I don’t know what would be involved in that, but I think getting humans into shared spaces, getting music happening again.
DUBLIN AS A GRAVEYARD
Having spoken recently with many musician friends, this is the topic that keeps coming up. No matter where I am with musicians, everyone is trying to make sense of what’s happening right now. One of the things that has been consistently coming up is access to space in order for people to meet and connect and experiment and try things live, perform things, share ideas, express the humanity that is waiting to be expressed.
I would love if there were 10 times more places in Dublin where you could just rock in and set up some wacky thing and try stuff out, and that there didn’t have to be enough ticket sales to cover a 600 euro venue rental, because people can’t afford to go out. So many people I know have completely stopped going out because it is simply too expensive.
The places where culture used to thrive and grow are becoming so much harder to access. At the same time, there’s a housing crisis going on so a lot of people have been forced to move home and people have left Dublin entirely.
I know there are a few underground venues putting on gigs and promoters still hustling, but I feel like there’s not enough support in Dublin for people who are looking to organize stuff and get something going. There needs to be way more access to space. I don’t know why there’s so little space.
When I first moved to Dublin, it was absolute chaos in the best way possible, there used to be so many places.. There was stuff running every night of the week. Every basement was full. There were so many alternative spaces in Dublin where people could just do off the wall shit. It was such an alive place to be and I really miss that.
Anyway, that’s my sad love song about the old days in Dublin. I hope times are going to change soon for the better. I think we need some kind of punk resistance. We need to get off our phones is what we need to do. We need to get out into the world. We need to get back out into spaces with our instruments and with our tunes and with our voices.
That brings us to the end of part 7 of my series AI What the Fuck, and also the end of episode 52 of The Human Signal. It feels so good to be back on the horse. After recording today and filling in the blanks of what was happening over the last couple of months, it’s made me realize how much I have missed this and how good it feels to have a place where I can express my ideas, talk freely, speak my mind, and process a lot of the stuff that’s going on.
I’ll be back in the next day or two with part eight. In the meantime, you can follow along on Substack. You can also find me on Instagram and on YouTube. This podcast is shared mainly on Apple Podcasts, but you can find it on some other streaming platforms too. I don’t share this podcast on Spotify and I also don’t share it currently on YouTube, mainly for reasons to do with AI scraping and just trying to work all that stuff out as I go. But anyway, mainly you’ll find me on Substack.
Thanks so much for tuning in. Please follow along, subscribe, all that stuff. I’m back. I’m back. I’m back. And I will see you in the next one. Until then, remember: put humans first. Don’t feed the machines.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit laurasheeran.substack.com/subscribe

Hello and welcome to The Human Signal with me, Laura Sheeran. Today we are looking at part six of my eleven-part series, AIWTF?, which looks at AI in the music industry right now—what’s happening in the digital music ecosystem, the good things, the bad things, the complicated things, and the bits that make no sense at all. We’re looking at all of it.
In this episode, I’m trying to make sense of what happens when the copyright system we’ve relied on for decades no longer fits the reality we’re in. I talk through the foundations of copyright—authorship, ownership, original expression—and how all of that starts to break down when music is generated by AI.
I also reflect on what this might mean structurally for the music industry. It feels like we’re moving toward a split between two systems that don’t align—one human, one AI—and I’m trying to understand what that could look like, and where it leaves artists.
Full Episode Description
Today’s episode is focusing largely on copyright law, the copyright law as it currently exists today—the law that we as creators have depended on to defend our rights, to protect our work, and to make sure that we can get paid and earn from the work that we do.
Since AI music has been flooding the industry to the degree that it has, the limitations of copyright law have been becoming more and more evident, and it is clear that the current model that exists to protect artists and musicians is no longer sufficient or fit for purpose, and radical change needs to happen.
The previous episode, I talked a lot about fair use, which I would recommend if you didn’t listen to that episode, or if you don’t know much about fair use, to go back and listen to episode #50, because it gives the full picture and the context for how AI music is able to thrive to the degree that it has. There are complications with the fair use law as well, which we go into in the last episode.
If you want to go and listen from the beginning of this series, AIWTF?, the podcast episode number to go to is #46.
This series is for people who, like me, are trying to find their way through a muddy, murky haze which is very uncertain and very unclear as to how one might want to release music in the immediate future, considering the rapid change happening all around us and how unstable the infrastructure we’ve known and depended on feels right now.
There are a lot of ethical considerations with AI. There are a lot of people who are very torn creatively in relation to AI. I know that there are artists using AI in creative ways. But there are still very complicated webs of uncertainty and potential vulnerability that you can fall into by not understanding the underlying rules, regulations, and systems that are supporting the AI music industry.
There’s a lot of potential liability, and potential loss of autonomy and rights. It’s a minefield.
Quick disclaimer: I’m just a normal person. I am an independent artist. I am not a legal expert, not an industry expert, not an AI expert. I’m just an artist trying to do the right thing by my work, by my music, and to protect myself from potential exploitation.
I’ve always released my music as ethically as I can, so that’s the position I’m coming from.

So, this is part six: human copyright law. The law as we know it was not set up to deal with any of these AI complications. The original copyright was built on human authorship, original expression, and clear ownership. But now, generative AI has exploded that whole concept wide open.
AI music - who is the author? Is it the AI model? Is it the prompter? Or is it the (potentially) hundreds or thousands of writers who’s stolen work makes up the training data which spits out some kind of Frankenstein-esque amalgam on the other side of a generate button.. With AI music specifically, who counts as an artist? What does being an artist mean?
These are questions that must be asked and they are things we haven’t worked out or decided on collectively yet. At the moment, everyone is just going with what makes sense to them and what suits their own interests. There’s no culturally agreed upon set of rules.
If copyright depends on original expression, then how can that apply if something like a song has been generated from potentially millions of other works? Where does ownership begin and end?
If I own my music and it gets trained into an AI, and then I hear something in an AI-generated track that clearly resembles something I created, can I claim that? Or does my ownership end at the boundary of my original song?
How does this tie in with sampling? Sampling has been around for decades, with its own rules. I don’t know how this compares. That’s something I want to look into further (I really wish I was educated in law. It’s kind of a fantasy of mine to go and get a law degree!)
Back to the point. I think what needs to be worked out first is whether AI music counts as fair use. That will dictate everything. Because since AI outputs are being commercially distributed, charting, generating revenue, and organisations such as AIMPRO positioning themselves to start collecting royalties for AI music, then fair use becomes very hard to argue. If outputs are directly competing in the market, the fair use argument starts to fall apart.
What’s happening right now is that outputs are trying to be legitimised before the inputs have been legally resolved. I can’t see a logical pathway forward because there’s no agreed ethical code, and the law hasn’t adapted.
One thing feels certain: the music industry as we’ve known it is gone. Not that it will disappear entirely, but it will have to bend with this tide, or it will break. Maybe this is a breaking point, where the split starts to happened.
What we’re seeing is a traditional rights system, built over a century, now colliding with a new AI-native system emerging alongside it. I can’t see how these two systems can merge—they are fundamentally misaligned. It feels more likely they will split and move in different directions. We could be witnessing the great music industry split of 2026!
Ideally, they would coexist in a balanced way. But more realistically, they may compete for dominance. Based on what platforms are prioritising and how AI content is being pushed, it’s not hard to see which direction things are heading in the short term.
Looking at platforms like Deezer, if AI uploads continue growing at the current rate, they will exceed 50% of new uploads very soon. That would make AI the dominant presence in new music being uploaded.
From my own conversations with artists, many are choosing to step away from these platforms and return to Bandcamp, vinyl, or other physical formats. Of course that’s all anecdotal, but from where I’m standing, it’s clear that platforms are becoming less hospitable to human artists, while AI presence continues to rise. We see this reflected in the bilboard charts where AI’s are frequently landing number one spots and taking these places away from genuine human artists who have committed their lives to their craft.
If there really is a split beginning to develop between the human system and the AI system, it seems clear which one is positioned to dominate in the short term.

If you enjoyed this podcast please share it with a friend or consider leaving a comment or review, it really helps me out. You can also support me and my work by becoming a €5 member on Substack and Patreon. And lastly, you can still find me on YouTube / Instagram as @the_persona_project__ & @laurasheeran_ieThat’s all for now. Thanks for being here, and remember: Put humans first. Don’t feed the machines.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit laurasheeran.substack.com/subscribe Hosted on Acast. See acast.com/privacy for more information.

This is part five of an eleven part series called AIWTF? where I am discussing all things to do with AI and music and the digital music ecosystem as it stands today.
In this episode, I’m working through the concept of fair use and how it’s being used to justify the current direction of AI in music. It’s something that underpins most of how content is shared online, but when applied to AI training and generated outputs, it starts to break down in uncomfortable ways.
I walk through the key principles of fair use—what it allows, where its boundaries sit, and why creative work like music is much harder to defend within it. From there, I look at the ongoing lawsuits between artists, labels, and AI companies, and the contradictions that are starting to emerge.
What becomes clear is that the system is trying to hold multiple positions at once: claiming transformation under fair use, while also moving toward monetisation and licensing of AI-generated outputs. Those positions don’t sit easily together, and it raises deeper questions about ownership, value, and what happens when extracted work begins to compete with its source.
The episode addresses the tension of trying to make sense of a legal and cultural framework that no longer seems equipped for what’s happening.
If you enjoyed this podcast please share it with a friend or consider leaving a comment or review, it really helps me out. You can also support me and my work by becoming a paid member, it’s €5 per month on Substack and Patreon. And lastly, you can still find me on YouTube / Instagram as @the_persona_project__ & @laurasheeran_ieThat’s all for now. Thanks for being here, and remember: Put humans first. Don’t feed the machines.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit laurasheeran.substack.com/subscribe Hosted on Acast. See acast.com/privacy for more information.