Regarding Recent Technology Advancements

[DarkeSword]

Starting today, the sharing of music generated by feeding a prompt into AI-software is prohibited on OC ReMix, both here on the forums and on our Discord.

LLMs, AI, and other machine-learning technologies are both interesting and impressive, but the implementation of these technologies in popular usage rely largely on unethical training practices. Copyrighted works are used without artist consent to train commercial AI-models. Moreover, OverClocked ReMix has, over the course of over 20 years, established itself as a place for artists to learn about the composition and production of music and hone their craft. We ask artists to pay tribute to VGM through the art of interpretation; we're looking for the personal spin, the human touch. I, personally, don't believe audio generated by a machine-learning algorithm is doing that.

There are ethical implementations of AI-technology in the music sphere, such as sample libraries, physical modeling software, and audio production suites that use machine-learning to perform complicated tasks that an artist can use in the production process. Sharing music that uses this kind of tech is absolutely fine. But tracks generated wholly or in-part by services like Suno and Udio (among others) do not belong here.

And just to clarify, the discussion of machine-learning, LLMs, and AI here at OC ReMix is not prohibited, but should be tempered by an understanding of the goals of our community, which include artistic development and education and the appreciation of VGM through thoughtful exploration and interpretation.

Thanks.

-- DarkeSword

[Master Mi]

Just from the point of the artistic level and the joy and intensity of a creative journey, AI is even worse than using premade loops.

You might be able to get something that sounds good 'n' ready for the masses of listeners - but you'll never be able to put all the compositional details, thoughts and feelings from inside your imagination into the realization of the soundtrack.

And if you don't have the knowledge and experience in music theory, composition, mixing and sound design, you won't even have an idea about what's even possible in the soundtrack you create.

For the most part, AI draws on things that already exist, on things that are known or have been grasped by the human mind.
A fine consciousness of a vital life form in combination with a high level of creativity, on the other hand, might be able to recognize things, energies and phenomenons that are still unknown in this world, and to create some really new 'n' unique stuff.

...
 

Or to put it in some more romantic words of video game poetry:

Creating video game music or remixes with AI technology is like feeding the plastic/wax fruit to the hungry, music-loving Green Tentacle in Maniac Mansion.
Even if the Green Tentacle likes the artificial stuff and already feels stuffed after eating it, as a hungry composer fueled and inspired by true life force within and around you, you wouldn't feel vital, nourished and satisfied if you ate the stuff yourself. ))
 

Edited April 17 by Master Mi

[Nase]

very good.

i haven't kept track with how advanced the publically available music a.i. stuff is compared to the text and visual stuff...and i really don't wanna. getting back into music making is freaking hard enough as it is :D

i wanna take my stand against a.i. in my own little way...i have to believe that the value of making music lies in the process, not the results. the personal relationship you form with the notes and sounds while making music.

got a lot to learn until i can completely embody that stance; honestly a large part of me has always been very results oriented, kinda greedy that way.

we artists have to prepare for 'spiritual war' in that sense; like, fuck it, we're doing it live.

but nothing against folks who wanna use a.i. in a creative fashion. 

not for me. idk, i might cave in for deep fake vocals at some point...as long as i write the notes and lyrics myself. basically, as long as i feel like i am doing the damn thing, more than the machine.

like, give me a vocal model of freddy mercury or rob halford or celine dion, and give me all the fine grained controls to make em totally glitch out when i want to.

 

[Unknown Pseudoartist]

This is an interesting topic I've been thinking a lot of lately because the recent hype with AI generated music got me in a time when I just released my first original music album. Since I'm still learning and lacking of experience the results aren't that great so some people might prejudge that and assume it's AI, even though actually I manually wrote every single note in the pianoroll.

But it was a pleasant surprise to see you all pointed at exactly the points I'm concerned of. It's not about results but more about the process. Nowadays AI generated music might be still mediocre, to say the least, but even if it eventually got real good... there's something special about the satisfaction of simply having your work done after all the time and work spent. You kinda grew with your works in a way that it makes them more like friends you build memories and stories with, and when you finally see them ready to be published you feel like you accomplished something you can be proud of. It don't matters if the compositions or arrangements are bland, if the mix sounds horrible or the genres/styles are not everyone's cup of tea. It's your work and you love it and want others to enjoy it just as you enjoyed the process, with all its moments of both tediousness and happiness. And then you keep learning and being able to do more and better stuff as you keep trying and doing more and more of that.

If all you did to get it done is just writing some text line in a web/app form, however... where's the story? Where's that special feel that may connect both artists and listeners? That work ends being nothing but some randomly generated product, very much like any of these old Flash avatar generators.

Music making can be often too hard and frustrating, especially if you are under disfavorable conditions of some kind (like lacking proper tools, dealing with ancient hardware, being in some kind of depression or any mental craps that makes everything even harder, etc), so I can totally understand people loving the idea of having access to such a task with just asking some cold and emotionless AI. I'm all into making things as easy as possible for anyone to show their creativity to the world, actually. But making things easier is one thing; and I can see how practical AI tools can be to just ease some parts of the process; but for the end-product? That's a totally different thing that does nothing good for creativity but rather the opposite.

[timaeus222]

That's what I've always loved about OCR. Y'all care about human interpretation, performance realism, personal touch, and other aspects of music that go underappreciated by the casual listeners out there. At first I thought the perspective towards AI-generated music was because of that philosophy, but I do understand that there is also the bit about unethical training practices.

Good to see OCR still going strong, and I hope to continue seeing real music coming out from real artists, using real production efforts and/or real studio time!

Edited May 3 by timaeus222

[djpretzel]

It's great to see new policy forged in my absence.

I do expect this stance to change & evolve over time, as the tech also changes over time.

As written, the first line mentions "a prompt" (singular) but then further down it mentions "tracks generated wholly or in part" and cites a couple of those shiny new services.

I agree that music generated by a single short text prompt doesn't come remotely close to the expectations and concept of this site, or any similar community which emphasizes the creative process & human decision-making extensively. Right there with you on that one.

My mind just tends to then jump to all the future hypotheticals that will one day crop up:

• What if it was 5 prompts instead of one?
• What about.... 50 prompts? 100?
• What if was just one prompt, but instead of text it was someone singing a complete arrangement and tapping their fingers on the table and the AI took that, respected all the beats/intervals, and built out the whole track around that concept?

All will be possible, at some point...

It really comes down to a ratio of input to output & the overall amount of human discretion and time involved. What I expect is that these lines will only get blurrier & blurrier, and more difficult to assess, as AI-based tooling becomes ubiquitous within DAWs and as part of creative pipelines, not just as a soup-to-nuts prompt-based magic track generator.

I also do wonder about traceability/proof - beyond asking for project files as evidence of effort, seems like it would mostly be honor system. Even asking for project files will only work until AI is proficient enough to understand & navigate DAW interfaces and work within human-oriented tools, allowing for further human refinement.

It's all a matter of time, and just more to consider when it arises.

The ethicality argument, well... I don't think you actually need to go there, so I personally wouldn't. People learn from & mimic other people, just not at the same speed & scale as AI, and the body of work in the public domain alone is sufficient for a pretty badass model, even if that's not the approach Big Tech seems to have taken. The issue of creative input ratio is not only more critical, it also persists even if you (somehow) fix any ethical concerns.

[Nase]

so in the end, the A.I. assisted piece would be justified by an essay by the artist detailing the creative input process. (and it being a good piece.) right?

but it's still zukunftsmusik at this point as we say in german.

i think a flat out anti-AI-policy is the right step until we get a better grasp of the creative potential of future tools.

Edited 16 hours ago by Nase