Can we discuss the impact of AI production here? If not delete I guess… ![]()
Hmmm, all i can say, i love when it pulls out usable audio from lousy noisy field recording with me doing almost nothing, but i hate when it “creates” music.
One can argue that artists always “stole” from older works, but when a machine does that in a blink of an eye i’m left feeling coned and robed of my entire existance
.
The ignorance of people who think that they created anything using things like Suno is something that i hope to forget as soon as i encounter it.
My advice is to just ignore AI-generation “tools” like this almost entirely. Pressing a button involves basically no effort, which leads to no long-term satisfaction with what you’re doing.
I think a lot of people forget about the importance of the process, not the product. The process gives the product meaning in many ways. So what ‘meaning’ can an AI-generated song really have if the entire process occurred in a split-second, and all the person did was literally click a mouse?
Yes, it’s all super interesting, but honestly, nothing beats writing your own song from ‘scratch’ (-as well as recording it, editing it, mixing it, etc.). Doing it ‘the old fashioned way’ will always have meaning and value, and it will certainly remain more unique than pushing a button.
I am in the camp of AI should be used to augment your work, not do it for you.
On one hand I have a lua script for Ardour that calls a python script that uses Azure’s OpenAI API to import a MIDI file (melody or chord progression) on a track with a synth based on a prompt you input. It works similar to something like Spawn or MIDI agent. The way I use it is I just build around whatever it generates. In that respect it doesn’t seem that much different from using loops in your production. The reason I did this is because my employer wants me to get Microsoft AI-102 certification and I thought it would be a cool idea to fuse my interest in Linux music production with my studies.
On the other hand, I was playing around with lyrics in ChatGPT and tried a free credit on one of the song generators out there. I input the lyrics and spit out a country song that sounded pretty decent, however I didn’t feel like sharing it with anyone because I felt no sense of accomplishment just inputting AI generated lyrics and pushing a button to get a song. Where is the skill and talent in that?
I have been looking for other ways to use lua scripting to leverage AI to augment production like for example maybe put in a lua script that manipulates the lv2 version of Surge XT based on a prompt for a synth sound or maybe something that manipulates the Guitarix plugins for a particular guitar sound.
There is no joy in any of that, and I expect humans will keep making music without AI, play live with other humans, and record themselves.
Take chess as an example, AI solved that game 25 years ago. It takes years to become good at it, and yet today the chess community is larger than ever.
…but yes, everything music you hear at the hairdresser, in a fitness studio or most background music is already AI slop. That’s done; few humans will be paid for that.
Yes! This!
This is by far my biggest misgiving about the whole thing, it is innately human to learn from processes and trial and error. There is a reason for the old adages like “give a man a fish he’ll eat for a day, teach a man to fish he’ll eat for life”… To develop technologies that ignore such a vital part of not only how we learn but also support and nourish our mental acuity and supplanting that with an effortless outcome in seconds really underscores how ignorant and inhuman these tech Bros are. This is exactly the kind of shit you get when knowledge becomes detached from wisdom…
Very interesting discussion here. Thanks for starting this thread ![]()
Personally, “AI” in music (and in general, I suppose) stirs up negative emotions in me but I struggle to articulate these feelings, especially because I feel like I contradict myself in my own opinions.
Take mastering, for example. I’ve tried a lot of those AI Mastering services and universally hated them all (Aria produced some nice results though, actually…). Nothing beats a human mastering a piece of music that was created by a human and that’s just how it should be done, in my view.
However I do think there are valid use cases for those AI assistant mastering services and I’m glad they exist as an option for those who want it.
In a similar vein, I had to listen to a stereo recording of a fire back when I was studying noise at university. It was an amazing piece (think it was by Cage…?) and I loved listening to it. However, the artist didn’t actively participate in the ‘music’ nor did they actively compose the piece. And that’s not a million miles away from an AI-generated piece, is it not…?
So why would I reject AI music but not a dude recording fire? Or a sound walk, for that matter? Or a field recording of a motorway in Northern England? All of these are real examples of pieces that I enjoy listening to and are valid forms of art.
Perhaps it’s something to do with meaning and/or intent…? We’re humans that perceive things as humans and relate/respond/react to stuff outputted by human beings and other organic, living things (cat meows, bird songs, etc). This overlaps a bit with the notion that the composition, construction, recording and mixing process is important when enjoying music in its final form.
In music, the input plays a significant role in deciphering the output and shapes our perception and, in turn, affects what it all means to us. The input affects, and forms a sort of dichotomy with, the output.
And AI-generated shit lacks a meaningful input, literally.
Maybe…?
It’s why some of us love listening to artists talk about their work and why some read liner notes and why some really really enjoy live performances.
Listen to “Pain is forever and this is the end” by the incredible Mantar. Brilliant, isn’t it? Just two dudes making a racket. Fuck yeah.
Then listen to Hanno and Erinc talk about the pained process of writing and recording the album. How appalling and terrible the whole ordeal was and what it meant to them before, during and after. And then listen to the album again and contemplate your thoughts and feelings afterwards. And ain’t that just something.
What you hear is affected by the process. And being aware of the process affects how you perceive what you hear. It’s all connected.
AI shit removes an enormous part of that whole thing we call music. Which is why AI is shit. It just lacks humanity.
Reject AI. Ignore and be ignorant of AI. Maintain humanity.
Whenever I see that adage I always remember the reformulated version of it, which might be especially relevant to AI and the future of humanity: “Build a man a fire and he’ll be warm for the day. Set a man on fire and he’ll be warm for the rest of his life”…
Weep ? No, it’s just funny, especially when you think about the long history of music throughout the ages. This AI tech brings nothing new, it just made the good ol’ practice of plagiarism more time efficient. And because of that, it’s a dead-end, unless it can truly be creative but for the latter to happen, I tend to think that it needs spiritual elements that are somewhat beyond logic (and therefore algorithms). So the way I see it, it is just the latest trendy toy that will become obsolete fast enough. In fact, the whole AI thing here is quite reminiscent of a bubble dynamic, it will probably pop sooner or later (when all the GPUs in the world fry ahead of their advertised lifespan).
By the way, I find the video content way too emotional and irrational, probably an attention seeker for his other youtube content.
In general he is amped up in every Video, it’s just his personality… That said, if I went to live in another country for a few years for an expensive education and was coming to grips with a technology replacing me that handily I’d be amped up too…
The video has a deep point. AI truly is a threat to all creative work. Just yesterday I was talking about this topic with a graphic designer. Their field is really starting to struggle. Why? Because with AI tools, almost anyone can produce decent results. This will destroy many, many creative jobs — even entire industries. It’s sad.
Here in Turku, Finland, we have a handicrafts museum where you can explore the methods and workshops of artisans from the past. Perhaps they’ll have to radically expand their facilities to include all creative professions in the exhibition — because soon, they might no longer exist.
What can we do? We can keep making art — for example, using Ardour and similar open-source tools. Maybe those will become the last strongholds when everything else turns AI-driven. The same will apply to the sciences. Humanity will probably start to regress intellectually if we slowly let AI make all the decisions. And we could also discuss what happens after that.
One possible scenario is what happened when photography was invented. Before that it required long studies and training to draw / paint realistic pictures. All of a sudden you could do the same with one click. Painting and photography still exist side by side filling a different niche and perhaps a different need.
I write my own music and find it very satisfying to finalize a song. It is always a struggle to create something and the pain of it makes the result more precious to me. It really is the frustration and countless dead ends that you need to overcome that makes you feel like you accomplished something worth while. No pain no gain.
I don’t find making AI compositions satisfying in any way, but if there was a way to make it do what I imagine in my head then it might be worthwhile. If I could prompt AI to go to the right direction then instruct it to reiterate and reiterate until the result reaches my vision then it would be satisfying. This is basically the same thing an art director does when making other people handy in their craft to realize his vision (movies for example).
Perhaps this is the future of AI ? Enabling one to be his own art director and make arts accessible to more people. AI is not there yet, but one can hope. Anyway there is no substitute to achieving ones goal through struggle so I think people will still find the old fashioned way to be more satisfying.
AI is already hurting people massively. One friend of mine who does illustrations, 3D modeling and animation recently lost almost all of his clients. They straight up told him that AI gave them good enough results lately.
His sketches when we were in 2nd grade of elementary school were already astonishing. Level of talent was unbelievable, and he worked hard, finished high school and college all related to his calling…I don’t see any logic in what is happening to him.
For many people this is not just “find another job” question, it’s more like “find another life” thing.
Part of it for me is that it is absolutely flooding the market, and also that this theft (of scraping the world’s tunes) is blatant and accepted. I unfortunately use this issue as an excuse for not putting out any music. I keep hoping someone smart like Robin will come up with a license and platform which will protect against some of this insanity.
I tried to get Ted Gioia interested, but didn’t get much traction. Regardless of what you think of him, I hope that Musk will see the problem and integrate something into X.
This week (obviously far too late) I pulled everything down from all streaming services with the exception of Bandcamp (I would imagine it is a scrapefest too). Nobody will care and my meagre number of followers will doubtless find bigger and better things to listen to so there is no ‘sticking it to the man’ victory here, in fact it was one of the most disappointing things I’ve done but as you say the blatant theft of everyone’s online data, presence and likenesses without their permission has been one of the most heinous thefts in history with barely a whimper of protest. Before most people had any idea what the Broligarchy was up to or that there was anything to be concerned about at all of their pertinent data had already been scraped. I guess we’re lucky it’s all just a simulation…lol(notlol)…
In a way I’m quite glad that I’m now nearer the end of my career than the beginning, I (made) a reasonable living from my accumulated knowledge and ability to understand and design professional audio electronics and latterly to translate that into DSP code (in the form of plug-ins amongst other things), now anyone, without prior knowledge, can quite literally ask ChatGPT to e.g. ‘write a digital model of [some circuit / effect / component] in C++’ and it ‘just works’ surprisingly well
(whereas six months ago it would have been more like ‘nice try, but…’ )
History seems to move in circles. Where the first Iron Curtain once divided the world ideologically, a new – digital – Iron Curtain now appears to divide it ontologically: into what is still human, and what has already merged with machines.
This new curtain is neither visible nor political, but invisible and technological. It does not descend across borders accompanied by tanks or armies, but through cloud services, platforms, and algorithms. It absorbs all of humanity into itself — our words, thoughts, memories, and dreams — and converts them into data to be exploited, sold, and directed.
At the same time, a countercurrent is emerging — a quiet, as yet unnamed movement of solidarity. Across the world, people are beginning to sense that we are losing something essential. This movement may not march in the streets, but instead builds open systems, shares knowledge without gatekeepers, and defends the right to remain human in a digital world.
Could this become a new wave of human solidarity — a version 2.0 of the Iron Curtain’s collapse?
Perhaps. But it won’t happen through political declarations, rather through moral and technological courage: through decisions to code differently, to share differently, to see differently.
The digital Iron Curtain may not tear apart through one great event, but through millions of small acts — lines of open-source code, conscious choices, and the refusal to surrender our worth.
Maybe the revolution of the future won’t be a protest in the streets, but a quiet chorus echoing through the networks:
*“We are still here. Not data, but living beings.”
Unfortunately, I had to translate this using a machine, because my English wasn’t quite enough to express the idea properly
The physicist Sabine Hossenfelder just released a video that is pertinent to this discussion, and I imagine would resonate with people here, including some stats on projected music market size changes.
Thanks for the link, I was recently introduced to her channel, lots of interesting topics there aside from that one. I think it will be very difficult to be Pro or Con AI in any blanket way without being somewhat hypocritical because like ‘Climate Change’ it is 1000 issues referred to as one and there are of course numerous beneficial uses for certain kinds of AI and it is obviously not going away whether we ignore it or not. My beef and the Video topic in the OP is mostly discussing the generative stuff although the so-called ‘good AI’ doing massive calculations and database stuff is also going to empty a lot of office cubicles too…
He has some bizarre takes on things. Is someone going to be getting in a car and say to an A.I. Bot “make me a bossa nova about me going to the park today”, I don’t think so. Also it’s odd to make some mediocre song with suno and then say “That does it better than I could”. I wouldn’t say that on video. He makes out like he’s a professional arranger likely to be robbed of his living but he looks like a full time influencer to me.