What is Ardour's Generative AI policy?

#21

Your point about accountability is fair. I’m not here to argue for AI, obviously. The medical side of it isn’t something for me to figure out I just hope that whatever genuine use it does have, the “AI” they use isn’t the “AI” we have now, and is actually ethical.

I digress.

I don’t see how AI being forced or used at will makes a difference in this case. My points about gen AI being unethical still apply across the board, which I say not to be rude, I just don’t see how any of us should accept “gen” AI into Ardour in any form.
The whole concept has only gotten this far because people roll over, saying we must “get used to it” or be left behind. I apologize for the extreme example but… you know what they say about a Nazi bar? Software works similarly, and we best keep tech bro “AI” nonsense far away from open source, as it goes against the whole concept of FOSS in the first place. FOSS is about lifting each other up and trying to create things by people, for people, with the help of other people. “Generative” AI is the antithesis of that, and I don’t really want to ramble more because I’ll end up repeating myself.

If you are concerned about AI, please, please keep Ardour completely free of it, in code and functionality. It would give us much more faith in the project for this action to be taken, and honestly, it’d simply make a lot of people happy to see from their favorite DAW. <3

#22

As someone who loves punk music and all it stands for, it should be obvious that “gen” AI spits in its face. If you think a community that welcomes inexperienced musicians to learn and grow, not mocking them for their lack of form or knowledge, is comparable to asking maliciously crafted software for audio built from or assisted by metric tons of stolen art, I don’t know what to say.

"If you remember when punk - music was born (if you are old enough :slight_smile: ). It lowered the threshold for common people to enter music, since it kind of gave you the permission to make songs even when you weren’t great at playing an instrument or singing. The same happened with rap, it was easier to enter music because one didn’t need to go through the long process of learning to play and understand music. Now AI is doing the same and lowers the threshold entering music.

AI is just a tool to help anyone achieve a goal without a long learning process."

Where is the joy in learning or creating anything? I’ve had a lot of trouble with the education system over the years, and I’m sure I’m making no headlines by saying it needs copious reform (and should be free to all), but if I had decided to farm it out to a plagiarism machines even part of the time, not only would all my “work” have been invalidated, but I would have learned nothing, and I would have wasted hours upon hours of my professor’s lives with slop I didn’t even create.

I’m trying to remain calm, but the idea that the learning process for music, visual art, code, whatever, is something to be skipped over rather then relished in is an insult to all creatives.

2 Likes
#23

I’m so sorry you’ve been forced to use AI in the past, for commercial things or otherwise.

#24

Rather the opposite. Free as in freedom software does not discriminate against any use. One of the fundamental rights is to allow anyone to study and modify it. Banning AI as tool would void that freedom.

Similarly someone may make a song with Ardour to promote message that I disagree with and they’re still welcome to do that.

All that being said, it’s unlikely that Ardour will have an LLM generated code anytime soon, and not just because of (C) reasons. There are also no plans to directly integrate any genAI, though interfaces for 3rd parties to use at their digression are welcome.

2 Likes
#25

Robin, with all due respect, why single out one phrase of my message and not respond to anything else? Yes, the concept of FLOSS is that anyone can study or modify it. However, just as somebody might release a video on YouTube for free consumption or I might write a post on social media, such things being public does not make them okay to throw into a plagiarism machine that was literally invented to devalue human creations more then they already are. The person having their work or likeness regurgitated is right to be upset, even if they did everything for free.
No, you can’t control what others do with your work, but that’s not exclusive to FLOSS software, it goes for nearly everything. Regardless, if my tools were used to create something awful, I would call it so. I wouldn’t shrug it off because the F stands for freedom, in the same way you wouldn’t ignore someone hurling baseless insults around because “free speech” allows them to. You’re right, we should all be free to use Ardour and its code how we want, just as we can use our voices. That does not make it any better to abuse either, and doing so should net such a person rightful pushback.

Why are you dedicated to open source tech? Is it privacy? AI’s a nightmare for that. The ability to learn from others’ code? AI spreads only misinformation and (to be generous) calculated guesses at the truth. Is it the ability to leave big tech companies behind, using community driven software for your endeavors that you can actually trust? AI is big tech companies. Any output from a “generative” LLM is nothing more then stolen material run through a planet-killing black box which, considering you can’t see what its even trained on, makes it arguably worse then most pre-AI closed source software. I don’t see how these things align with anything FLOSS.

someone may make a song with Ardour to promote message that I disagree with and they’re still welcome to do that.

I effectively responded to this over on GitHub. If a song of mine got used in a fear-mongering propaganda ad or something, I’d be right to call it out, even if copyright was somehow against me. Likewise, if my software was scraped or used in some slop project, I’d make my stance on such things known, if nothing else. I certainly wouldn’t consider myself open to adding AI or even “interfaces for 3rd parties to use” because other forks will add them regardless. They are not Ardour, they are a modified version of Ardour controlled by other people. Like I said in that comment, its about what you can control, and Ardour is controlled by Paul and yourself. We use Ardour as opposed to a fork of it because we respect and trust you two.

#26

As @x42 noted, we have zero plans to incorporate generative AI stuff into Ardour, and zero plans to include “purely LLM generated code” into Ardour. We are not seeking to expand our “AI capabilities” or “AI integration”, and we’re not seeking to make it easier for AI tools to use Ardour than it already is (which is actually pretty easy).

We’ve stated on the development page that we cannot accept LLM-generated code into Ardour because it cannot be copyright and thus cannot be licensed. There’s significant wiggle room here because no court has yet established how much work a human has to do in order to tranform it from a machine-generated work into a human one. We will watch that space with interest.

The MCP server was added, as @x42 explained above, to potentially facilitate better interactive use of Ardour by people with visual impairments of various kinds.

I’m not really sure what it is that you want us to do.

4 Likes
#27

Oh, that’s easy…
We are not shure either :slight_smile: .

Let’s see…
For starters, do everything opposite of Suno, and simply go Paul full blast, i suspect that’ll do :slight_smile: .

If somebody comes and demands some generative AI feature, just release the tomahawk from you hand, it’ll find it’s target :slight_smile: .

2 Likes
#28

That long post was mostly in response to the statement that AI isn’t antithetical to FOSS, an idea I also addressed in a reply to @x42 on GitHub that went ignored. Perhaps you preferred to respond to me here, Robin, which is fine, and maybe that reply was left before I was done with my own aforementioned one on GitHub.
From my perspective, however, it appeared that you saw my response over there, ignored it, then reiterated the point in this thread as if I hadn’t just attempted to engage with it in another. Additionally, the only acknowledgment of what I said on GitHub was directed at the pull’s creator, urging them not to be discouraged by my words. I didn’t intend, nor see how I was rude or disrespectful, so that stung a bit. Jumping back here to a comment similar to the one I just addressed, I felt the need to elaborate further since my last reply seemed to have been ignored.
Hence, long message. Hence, I likely came off as overly upset or spiteful. I was the former, but if none of what I described was intended, I’m genuinely sorry for all of that.

The MCP server was added, as @x42 explained above, to potentially facilitate better interactive use of Ardour by people with visual impairments of various kinds.

I will never downplay the importance of accessibility, but as someone with a disability that makes it more difficult to create visual art, it would not justify me stetting ethics aside to make a computer regurgitate a mishmash of other people’s creations. I’ll admit I’m not sure how best to approach accommodations for the visually impaired, but I doubt anything AI related has to be involved.

we cannot accept LLM-generated code into Ardour because it cannot be copyright and thus cannot be licensed. There’s significant wiggle room here because no court has yet established how much work a human has to do in order to tranform it from a machine-generated work into a human one. We will watch that space with interest.

While once again, I am grateful for your apprehensiveness towards “generative” AI, you’re highlighted the point I’ve been trying to make. If you’re concerned about the effects of this harmful tech being pushed by the kind of companies FOSS is meant to oppose, I fail to understand why copyright is the only concern. I wouldn’t put money on governments making half-decent decisions regarding “gen” AI if it existed decades ago, let alone now, given the state of the world. Does this mean Ardour will embrace such things if the law lets it? I’m not trying to put words in your mouth, this is why I recommended elementary OS’ policy as a good example all the way back in my first message. Maybe those details got lost in all my yapping. I’ve been told I haven’t shut up since day 1 so I guess old habits die hard. Apologies.
I’m not trying to attack or intimidate you, I don’t intend to be an annoyance, and I don’t want you to wince when my name appears in these forums. I only want to see the most valuable project I’ve used since going full FOSS, a piece of creative software, reject machines that exist only to further disrespect creatives and our planet. That’s all.

#29

It’s not the only concern. It is, however, the simplest concern: no need to do any analysis of the pros and cons of any AI platform or process - we just can’t do it, because it violates our own license. If that issue didn’t exist, we would certainly have to have a more evolved position on the much more complex issues. But it does, and it acts as a convenient barrier.

If the law were to change to allow machine generated code to be copyrightable (and thus licensable), we would put more time into considered the broader issues. But right now, in the USA at least, that is not the case, and I’d rather work on bugs and features.

I doubt anything AI related has to be involved.

Well. maybe but maybe not. If you wanted to build an interactive front-end that allows a person to take control of Ardour from speech or text inputs, it would be a huge task if you chose to do it any other way than the LLM way. That doesn’t make the LLM way “correct”, but I wouldn’t be so confident that other approaches are sensibly realizable.

Regarding “directed at the pull’s creator” … the only people who set “Ardour’s Generative AI policy” are @x42 and myself, not PR creators or forum commenters, so while it may be valuable to persuade others of the deep problems with LLM technologies, doing so doesn’t have much to do with Ardour’s own policy.

#30

I see, I understand your focus remaining on the software itself itself for now. If LLMs never existed, I’m sure we’d be able to find a way to create these accessibility functions otherwise. I know similar things have been done before actually, but regardless. Thank you for responding, I appreciate it.

#31

Highly unlikely and not for a lack of trying. We failed to do that for 20+ years. There was a very early prototype ages ago, but that not viable to maintain with evolving features.

I also like the fact that LLMs can be used as studio-assistant. Your new intern can set up routing, and do basic session tasks…

#32

Personally I am looking forward to seeing how well it can handle certain aspects of session analysis for my own purpose, again nothing to do with generating work at all, everything to do with analyzing existing work. Could I do it without an LLM? Sure, in fact I already do, but it takes hours I don’t have to spare to do so and I probably still don’t go as in depth as I could.

 Seablade
#33

So can lua scripts :wink:

But you’re right, an LLM is way more convenient and accessible than the LUA API. Not everybody has the time and technical understanding to write a script that does the right thing in Ardour. Even LLMs fail at that task so far, I read a few times here.

Pretty sure that many fear that the AI will overstep in its intern role and starts either interupting or taking over the creative process. A script runs and shuts up afterwards, or is only triggered on clearly defined signals from the session.

In my opinion, the assistant would need some very hard and easily definable borders, like some checkboxes in the ardour settings what it can and can’t do, not some file in plain english asking it not to overstep

#34

… or deleting the complete project data base - all wav and midi files - because it “thinks” it’s crap.
I read about a case where an AI bot deleted the complete company data base including backups on the network. They only could recover from backups stored externally.
Those are the issues we need to address :slight_smile:

BTW: Can also happen to human “bots”. I remember a case - loooong time ago - where a collegue of mine accidentally deleted the work of a whole week with the customer engineer sitting next to him, because a mount and cd command had failed, so his working directory was still the project directory instead of the backup device …

#35

IF you know what the setup is going to be before you write the script.

I haven’t done this yet, but it has been interesting to watch videos of people get tech riders from bands, and have agents set up their console for them, even if it isn’t perfect as long as it is ‘good enough’ it can save a not insignificant amount of time on short days to get you that much more prepared for a live gig, and thus that much more successful. Mind you don’t let it do all this and at least check a bit behind it, but it is still kind of impressive what it could potentially do. In at least one of those cases they had it triggered based off emails to a specific address, again doable with scripting, but you are talking about a lot of work to get to that point, and then someone throws something like a hurdy gurdy at you one week and all your scripts fail because they don’t know how to handle it.

I can see similar potential use cases in the studio as well, again not generating audio in any way shape or form, but as Robin mentioned, an assistant to take some of the tedium out of things.

IN terms of deleting entire collections of sounds etc. yes this is definitely a concern and should be for anyone that allows this. There is a reason I haven’t trusted agents to manage much data on my drive, or given them ability to do so. The more power you give them the more destruction a wayward agent can cause, as others have said this is not to different from wayward scripts/programs or commands by human users. There is definitely a learning curve to happen here, and yes maybe the answer is to make sure that there is a similar permission structure built into Ardour’s MCP access.

 Seablade
2 Likes
#36

Has anybody actually demonstrated that the MCP server in Ardour (an opt-in feature, btw) allows one to “regurgitate a mishmash of other people’s creations”?

#37

I’m not sure if it’s currently even possible. So far there is no audio import. Though FreeSound import would be amazing feature to expose via MCP.

So far only MIDI drums and MIDI melodies are possible.

As for actual real world examples, have a look here:

#38

If my memory serves me well, using FreeSound requires setting up access to the FreeSound server for each new launch of Ardour. The service really wasn’t built for this kind of integration (I don’t blame them, it’s just a fact). So it might come in handy, but the UX would still not be great.

My own experience with FreeSound is that I need to listen to 10-20 different records to pick the right one, I’d have to write a very detailed prompt for the LLM to make a semi-right choice for me, and then the LLM would have to be good enough to translate that prompt into the right action.

Yeah, I saw that. That’s precisely what I would expect from this kind of functionality: being a sound engineer’s assistant.

#39

I am referring to LLMs in general. I do not think the server can or is supposed to do this, but any use of AI enables such things. Hence I’m against it no matter what supposed benefits it brings. Basically this post.

No amount of functionality is worth being complicit in this ever-present problem that’s only getting worse the more people try to justify its use.

#40

I see what you’re saying and understand the goals, but use of AI is use of AI. It enables and furthers the same things even if you aren’t using it for more explicitly “generative” purposes. I’ll just be blunt and say there is no use case that justifies being complacent with fundamentally fascist technology.