this post was submitted on 08 Jul 2023
66 points (100.0% liked)
Technology
31 readers
1 users here now
This magazine is dedicated to discussions on the latest developments, trends, and innovations in the world of technology. Whether you are a tech enthusiast, a developer, or simply curious about the latest gadgets and software, this is the place for you. Here you can share your knowledge, ask questions, and engage in discussions on topics such as artificial intelligence, robotics, cloud computing, cybersecurity, and more. From the impact of technology on society to the ethical considerations of new technologies, this category covers a wide range of topics related to technology. Join the conversation and let's explore the ever-evolving world of technology together!
founded 2 years ago
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Everyone has takes like this like we can’t just draw the line at making it illegal for a computer or algorithm to do this.
We can, but why should we? The end result is the same.
It’s not though. That’s you naively ignoring the aspects that are different. Computer generated imitation is easier to create, can be created at a scale far eclipsing human action, and can be finely tuned to make it harder to discern.
You can look up impressionists and you’ll find it’s a rather small club when looking for the true greats. Computers remove this barrier and allow any asshole with an internet connection to create a video of you screaming racial epithets if your voice is easy enough to access.
The vast difference in scale can’t be ignored.
Talking up the capabilities of AI voice acting is not really helping the case against it. If it's really so good, and laws are enacted that forbid mixing human and AI voice acting, then I expect the straightforward optimal solution would be to entirely eliminate the human voice actors going forward.
You write laws for the future. Unless you think ai generated content has plateaued. Which is again, naive. Just because social media wasn’t popular at first doesn’t mean we should’ve waited on passing data privacy legislation like we have. It’s good to identify potential issues and attempt to mitigate them early. So we don’t get situations like our current climate status.
I don't think it's plateaued, I think it's going to get significantly better from here onward.
I'm not sure what laws you're proposing at this point. Are you suggesting that AI should be forbidden from "mimicking" a human voice actor? That's what I'm suggesting will lead even more quickly to AI-only projects that get rid of the human voice actors entirely, since having a human voice actor under laws like that would end up as a huge hindrance.
I think you are on the right track with this. Reminds me of the tales of people smashing automatic looming devices with the early luddites.
Ai voice impressions should be rendered illegal without explicit consent from the entity being imitated. Simple.
Also your extrapolation of potential events feels ridiculous. Tech is banned so it’s used more in commercial projects heavily subject to such legislation?
If impressions are banned, ie voices that sound like existing human voice actors, then yes, I expect it would be used a lot more. Because the only safe way to use AI voices in that case would be to never have a human actor to begin with. Create a novel AI voice from scratch and use that for your character, and then you can freely generate new lines with no further legal or practical concerns.
Whereas if you were to use a human voice actor for a character, you're stuck with that human voice actor. You can't do a quick virtual re-shoot without hauling him in for it, there's royalties for everything, and if the human voice actor dies or spouts off some unfortunate racist rants on social media or simply quits then you're screwed.
Unless you're proposing banning all AI voices completely, including novel ones that were never imitating a specific human to begin with? That's rather the more ridiculous scenario.
It’s like you just skipped past the “with consent” caveat to go on your diatribe.
So now that I’ve reiterated that it wouldn’t be entirely illegal, can you explain how that requirement will cause there to be no more VAs? In this world you’re imagining, how are these computational models creating voices? You talk about a “safe way to use” it in these circumstances but again, these companies still need data to generate voices. And this data would be protected through active seeking of consent.
You aren’t “[creating] a novel voice from scratch, that’s just not how the technology works. It needs a human to extract data from and compile something intelligible. Unless you want every animated feature to use the robotic assistant voices. Another aspect to why your perspective makes no sense and seemingly shows a complete lack of understanding of how these computational models work and how that comes together with my proposition.
Then your last paragraph is just a confirmation that no, you haven’t fully read what I wrote. So one more time with gusto:
Using computer generation to imitate a person using their own biometric data should be illegal unless explicit consent is given.
Much like with art AIs, the outputs don't necessarily have to slavishly mimic the style of any of the inputs. Train an AI with a bunch of different voices and then you can get it to generate a novel voice that isn't a copy of any specific one that it was trained on.
This doesn't affect what I've said. If imitating a specific human comes with a bunch of annoying legal and economic hassles, then don't imitate a specific human. Create a novel voice and you're free of all of that.
And yes, the technology lets you create a novel voice different from any of the ones it was trained on. I do know how these things work.
You should not be able to use someone else’s biometrics (voice is one) to generate content without their consent. Your example would be similarly illegal because it is unethically using a persons personal data without their consent for commercial or other purposes.
It’s “novel” in that it’s an approximation of all its input data, tweaked to match the specifics of the request given. It still needs to use the data of real people or it can’t create anything. You have a surface level understanding if you don’t understand the importance of that seeding data.
You don’t seem to value consent when it comes to the systematic harvesting of personal data for another persons benefit. I’ve been very clear that the issue is the lack of consent combined with current and future capabilities of the technology.
The end result is going to be basically the same regardless. Plenty of people (such as myself) who believe in the huge potential of AI to give creative power to regular people will volunteer our voices. Giving that creative power to everyone is worth far more, in my opinion, than gatekeeping the creation of art.
Unless they're planning on making it illegal for a computer to imitate any human voice, I don't see where making a law against using a voice without consent would make a big substantive difference. Just re-voice the existing lines in Skyrim with new voices to maintain consistency and you're good (there's a Serana mod that already does this, for instance).
If you’re using new voices then congrats. That’s not the issue. The issue is people using existing voices to create audio the person didn’t consent to record. You gave an example of it being done right. Not to mention I’m pretty sure the Serena dialogue isn’t ai generated so it’s not even close to relevant in actuality.
But since you seem to love the potential of AI would you be willing to send me an audio file of you pronouncing every possible phonetic sound the human mouth can make? I promise there won’t be audio of you talking about eating babies afterward because as you say, there’s no practical reason to require consent for these things. No one could possibly abuse technology to hurt other people. It’s never happened in history.
AI is indeed a powerful tool that can be used to let more people explore their creativity. Your assumption that I felt otherwise is because you’re on the opposite end of the spectrum. So self assured of it’s value that you’re blind to real shortcomings and abusable points.
I would describe my position more like: AI, like any new technology, is neutral. It’s usable for good and for bad. Thus it’s important to watch for ethical pitfalls that we may not have had to consider before due to the drastic way the new technology impacts society.
In theory, absolutely.
In practice, I'm not going to go through that much work just to make a point for a single fediverse comment. I'll be honest, though -- I'm not particularly worried about somebody using my voice to do a bad (or do a racism or whatever). It may happen, and I can live with it; I think the benefits far outweigh the cost, and in my experience, far more people use those sorts of things to do awesome stuff than to be shitty. Earlier today I was considering trying to put together an Open Voice project and collect volunteers to do exactly what you said.
I've already released open source code over the years; people could potentially use that to do things I don't agree with as well, but frankly, as someone who has had work out in the wild available for use by everyone, the panic is vastly overblown.
Just because I feel that the potential benefits far outweigh the costs (as well as the draconian technical restrictions that would be required in order to prevent people from expressing themselves in a bad way), it doesn't follow that I'm somehow blind to the real shortcomings and abusable points of AI. I would appreciate if you not make silly strawman assumptions about whether I've given something due consideration just because you don't like my conclusions.
If you have a solution that wouldn't absolutely kill it (or put a horribly filtered version in the hands of a few massive corporations who charge the rest of us for the privilege of using it while using it themselves however they want), I'm all ears.