RuPaul under fire for using AI special effects in new movie

[SimonBaetens 9,366]

he literally fracks

[Sepsami 20,536]

10 hours ago, gagzus said:

As a creative I can tell you it is, AI is making people lose jobs. Especially in tech and creative industries. And we won’t stop making a fuss until this is properly regulated.

Not to mention all the unethical ways AI can and will be implemented. We think these social media algorhythms are an issue, which to be fair they are, but AI is going to be much much worse.. 

[chathonnete 983]

I promise the way Rupaul would be the biggest queer icon if they had real left wings convictions (about trans people, ecology, anti AI, remembering the queens' names lmao)

[Helxig 45,001]

Look Rupaul may have been a trailblazer who paved the way for gay people of today. But he is also a capitalist piece of ****. His mottos have always been around making money no matter the cost. After the fracking scandal the he was totally unapologetic about, you're really surprised he's using gen AI?

[MountainMonster 2,370]

7 hours ago, nATAH said:

a paintbrush doesn't steal, genAI's only option is to steal 

A paintbrush doesn’t learn, but neither does generative AI simply copy. It analyzes patterns from existing works and generates new outputs from those patterns. Whether that’s fair use, infringement, or something in between is a legal and ethical debate, but calling every AI-generated image ‘stolen’ assumes the conclusion rather than proving it.

It uses mathematics and patterns to create novel images - and claiming that it can only steal is radically reductive. By that same logic, all art that is created mathematically and by patterns are stolen and not art - which means we have to say Escher and his tessellations aren’t art, perspective drawings and their geometric rules and projection mathematics are not art, music is not art because there are only so many notes, anything created using the golden ratio is not art, photographs aren’t art because the camera can only steal light and convert it to math, video games, CGI, 3D animation, graphics design, etc are all invalid because they can only create models from math and vectors from formulas. 
 

I just think claiming it blanket steals is too reductive. At a technical level, modern image generators don’t have a database of images that they cut up and paste together. Instead, they learn a mathematical representation of visual concepts during training and then generate images by transforming noise into a coherent picture. It’s doing exactly what humans do. If a room of 50 people were all asked to draw a cat, the vast majority would look similar - all based upon their prior knowledge of what a cat looks like. 

[nATAH 55,475]

24 minutes ago, MountainMonster said:

A paintbrush doesn’t learn, but neither does generative AI simply copy. It analyzes patterns from existing works and generates new outputs from those patterns. Whether that’s fair use, infringement, or something in between is a legal and ethical debate, but calling every AI-generated image ‘stolen’ assumes the conclusion rather than proving it.

It uses mathematics and patterns to create novel images - and claiming that it can only steal is radically reductive. By that same logic, all art that is created mathematically and by patterns are stolen and not art - which means we have to say Escher and his tessellations aren’t art, perspective drawings and their geometric rules and projection mathematics are not art, music is not art because there are only so many notes, anything created using the golden ratio is not art, photographs aren’t art because the camera can only steal light and convert it to math, video games, CGI, 3D animation, graphics design, etc are all invalid because they can only create models from math and vectors from formulas. 
 

I just think claiming it blanket steals is too reductive. At a technical level, modern image generators don’t have a database of images that they cut up and paste together. Instead, they learn a mathematical representation of visual concepts during training and then generate images by transforming noise into a coherent picture. It’s doing exactly what humans do. If a room of 50 people were all asked to draw a cat, the vast majority would look similar - all based upon their prior knowledge of what a cat looks like. 

genAI literally steals

[TheSine 6,808]

26 minutes ago, MountainMonster said:

A paintbrush doesn’t learn, but neither does generative AI simply copy. It analyzes patterns from existing works and generates new outputs from those patterns. Whether that’s fair use, infringement, or something in between is a legal and ethical debate, but calling every AI-generated image ‘stolen’ assumes the conclusion rather than proving it.

It uses mathematics and patterns to create novel images - and claiming that it can only steal is radically reductive. By that same logic, all art that is created mathematically and by patterns are stolen and not art - which means we have to say Escher and his tessellations aren’t art, perspective drawings and their geometric rules and projection mathematics are not art, music is not art because there are only so many notes, anything created using the golden ratio is not art, photographs aren’t art because the camera can only steal light and convert it to math, video games, CGI, 3D animation, graphics design, etc are all invalid because they can only create models from math and vectors from formulas. 
 

I just think claiming it blanket steals is too reductive. At a technical level, modern image generators don’t have a database of images that they cut up and paste together. Instead, they learn a mathematical representation of visual concepts during training and then generate images by transforming noise into a coherent picture. It’s doing exactly what humans do. If a room of 50 people were all asked to draw a cat, the vast majority would look similar - all based upon their prior knowledge of what a cat looks like. 

Of course it doesn't learn, it's an inanimate object. The owner learns, and controls the paintbrush as they see fit.

Also, that's a fancy way of saying it steals either way. Sure, things use math, but art is defined by the independent creativity of the owner. There's no true definition to art, but it's related to craft that AI doesn't have. There's no such thing as originality, we lived through that already. What matters is how it's re-purposed and refined.

The cat drawings may look similar, but you're forgetting about how the person drawing moves their hands. A cat has ears, nose, torso, tail and legs, but each line stroke is different, maybe they're thinking about a specific cat. What if they wanted to just draw a cat head? What if they know how to draw more realistically? What if they're thinking of a specific cat? What if they want to draw the cat in an environment? AI isn't the problem, it's the way people use AI to do simple skills for them.

Edited 16 hours ago by TheSine

[Bronco 22,494]

53 minutes ago, MountainMonster said:

A paintbrush doesn’t learn, but neither does generative AI simply copy. It analyzes patterns from existing works and generates new outputs from those patterns. Whether that’s fair use, infringement, or something in between is a legal and ethical debate, but calling every AI-generated image ‘stolen’ assumes the conclusion rather than proving it.

It uses mathematics and patterns to create novel images - and claiming that it can only steal is radically reductive. By that same logic, all art that is created mathematically and by patterns are stolen and not art - which means we have to say Escher and his tessellations aren’t art, perspective drawings and their geometric rules and projection mathematics are not art, music is not art because there are only so many notes, anything created using the golden ratio is not art, photographs aren’t art because the camera can only steal light and convert it to math, video games, CGI, 3D animation, graphics design, etc are all invalid because they can only create models from math and vectors from formulas. 
 

I just think claiming it blanket steals is too reductive. At a technical level, modern image generators don’t have a database of images that they cut up and paste together. Instead, they learn a mathematical representation of visual concepts during training and then generate images by transforming noise into a coherent picture. It’s doing exactly what humans do. If a room of 50 people were all asked to draw a cat, the vast majority would look similar - all based upon their prior knowledge of what a cat looks like. 

GenAI companies are on record as having ignored copyright protections in their datascrapes.

GenAI literally steals.

[TheSine 6,808]

In fact, draw a cat with pencil and paper with no help, then ask AI to generate a cat that looks drawn on pencil and paper and come back to me with both images. Hmm, I wonder where they got the images from to generate it. 

Edited 15 hours ago by TheSine

[Ziggy 13,662]

12 hours ago, MountainMonster said:

Agree regarding its use in the military and for anything that makes it easier for people’s lives to be taken or interfered with by governments - I don’t like that or think that should be allowed. But again - how do we stop that from happening, and what does that have to do with this example of generative AI being used in entertainment?

It directly relates in the sense that these are the same companies' AI. Like OpenAI makes ChatGPT and is creating military AI. So it's connected and the more we embrace it, the more it gets adopted. The military industrial complex is, well, complex lol but as for what do we do about it, we don't engage with it! Already we're starting to see companies like Microsoft question the value of their AI investments in code citing it as more expensive than human labor, which is objectively true. Granted, in this case they're likely pivoting to Codex or something, but the fact that these are conversations being had is encouraging. It's really not an inevitability like they all say it is, they just say that because they're pushing it. If you were selling lemonade, you'd say, "it'll quench your thirst! it's taking over the world!" even if that wasn't true because you need people to believe and buy-in. And that's what's happening with this generative AI which, imo, hurts the development of the actually helpful applications artificial intelligence could have. what ever happened to the remarkable fantasy of the sci-fi films of even 20 years ago where tech made lives better? Now they're war profit machines and it truly doesn't have to be like that. there's great potential for this tech going unrealized bc the people in charge of it atm are genuinely evil people

[bpmMonkey 7,446]

19 hours ago, Creyk said:

This is just how it's going to be now. Amazon studios announced an AI-based children's programming studio last week where they will make cartoons for kids using AI technology. 2 shows are in the pipeline for now and no one is talking about it.

I mean, this will be the reality of Hollywood in a very near future. Ellison over payed immensely on the Paramount x Skydance merger. If he wants to make any amount of the money back he'll strip back production costs as far as he can do. Given that he's a lil dip sh!t with no morals genAI will run the studio in no time and other studios will do the same sooner or later. 

[Jill 31,562]

11 hours ago, Bronco said:

There is *no* benefit to genAI. GenAI is not necessary. 

We all know what I mean when I say "AI". we're talking about generative AI in this post, and that's what I'm referring to. It's implied. But saying there is no benefit to this type of AI is also incorrect, and it assumes that everyone using it and incorporating it into their pipeline is somehow crazy and doesn't know what they're doing. Not a single one of them. Obviously, people don't use something if it doesn't work for them, that's just a given. We aren't all half-witted, chasing after something that makes our work harder. That would be irrational. That's why this isn't a good line of argument.

At the risk of sounding like a broken record, China is a great example of AI (generative, to be clear!) being used daily, deeply embedded in technological and economic processes rather than just as simple chatbots. (I think that's the problem many Americans have when discussing AI, which is that they imagine its beginning and end is ChatGPT). Some of the things that are possible today but weren't before without this type of AI include, but are not limited to:

- faithful & fast transcriptions and translations of videos, books, movies, etc. This reduces the effort needed to expand access to more culture and materials that might have previously remained obscure due to the manual labor these processes used to require. It can also clarify scanned images (like book pages) or serve as an assistant when preserving languages or translating dead languages for the first time.

- Process optimization, as is currently happening with renewable energy in China and other countries. Things that used to take much longer can now be made quickly, allowing people to access and review information that would have been impossible to sift through before.

- Also improvements in bureucracy and the streamlining of paperwork, which is also evident in China.

- In the case of renewable energy, using this type of AI allows for a much deeper level of efficiency than before (which also ties into the next point).

- CODE: This is crucial and one of AIs greatest breakthroughs: generating source code for applications, software, robotics, etc. Prototypes (or even final products sometimes) that used to take months of trial and error can now be done in HOURS, which is a massive game-changer. Programmers have incorporated this fully already, alongside agents.

- Agents in daily life, which help automate workflows, handle customer service, & execute more complex tasks (like booking trips) in less time. THis is incorporated in some Western tools but it's also much more obvious in China, in apps like WeChat. In much the same way Google saved us the hassle of going to buy a ticket plan physically, this saves us the hassle of booking a trip directly, through several websites.

- Optimization of agriculture and "feeding more people without cultivating more acres" (the WEF is a great source for AI-related innovations)

(none of this is done without human supervision, btw)

China is also developing a very active culture of incorporating AI into daily life, but with a vision of AI that is more collective and less proprietary (open-source) than the Western one. A good example of this is the way DeepSeek operates.

And finally, to be clear: let's not confuse the death drive of the US economy (turning to AI almost like a "last-gasp effort") with AI itself and its potential. The internet was also the victim of a bubble back in the day, and many denied it could have any utility whatsoever. We can agree or disagree with AI, but PLEASE let's not say it has no benefits because every current fact proves otherwise.

Saying that shows very little interest in seeing the "other side" of the situation, which is highly necessary even to "confront" it, if that's what you want to do (which is also why a Marxist, for example, has to read libertarians, anarchists, and other authors).

 

Edited 12 hours ago by Jill

[Bronco 22,494]

16 minutes ago, Jill said:

We all know what I mean when I say "AI". we're talking about generative AI in this post, and that's what I'm referring to. It's implied. But saying there is no benefit to this type of AI is also incorrect, and it assumes that everyone using it and incorporating it into their pipeline is somehow crazy and doesn't know what they're doing. Not a single one of them. Obviously, people don't use something if it doesn't work for them, that's just a given. We aren't all half-witted, chasing after something that makes our work harder. That would be irrational. That's why this isn't a good line of argument.

At the risk of sounding like a broken record, China is a great example of AI (generative, to be clear!) being used daily, deeply embedded in technological and economic processes rather than just as simple chatbots. (I think that's the problem many Americans have when discussing AI, which is that they imagine its beginning and end is ChatGPT). Some of the things that are possible today but weren't before without this type of AI include, but are not limited to:

- faithful & fast transcriptions and translations of videos, books, movies, etc. This reduces the effort needed to expand access to more culture and materials that might have previously remained obscure due to the manual labor these processes used to require. It can also clarify scanned images (like book pages) or serve as an assistant when preserving languages or translating dead languages for the first time.

- Process optimization, as is currently happening with renewable energy in China and other countries. Things that used to take much longer can now be made quickly, allowing people to access and review information that would have been impossible to sift through before.

- Also improvements in bureucracy and the streamlining of paperwork, which is also evident in China.

- In the case of renewable energy, using this type of AI allows for a much deeper level of efficiency than before (which also ties into the next point).

- CODE: This is crucial and one of AIs greatest breakthroughs: generating source code for applications, software, robotics, etc. Prototypes (or even final products sometimes) that used to take months of trial and error can now be done in HOURS, which is a massive game-changer. Programmers have incorporated this fully already, alongside agents.

- Agents in daily life, which help automate workflows, handle customer service, & execute more complex tasks (like booking trips) in less time. THis is incorporated in some Western tools but it's also much more obvious in China, in apps like WeChat. In much the same way Google saved us the hassle of going to buy a ticket plan physically, this saves us the hassle of booking a trip directly, through several websites.

- Optimization of agriculture and "feeding more people without cultivating more acres" (the WEF is a great source for AI-related innovations)

(none of this is done without human supervision, btw)

China is also developing a very active culture of incorporating AI into daily life, but with a vision of AI that is more collective and less proprietary (open-source) than the Western one. A good example of this is the way DeepSeek operates.

And finally, to be clear: let's not confuse the death drive of the US economy (turning to AI almost like a "last-gasp effort") with AI itself and its potential. The internet was also the victim of a bubble back in the day, and many denied it could have any utility whatsoever. We can agree or disagree with AI, but PLEASE let's not say it has no benefits because every current fact proves otherwise.

Saying that shows very little interest in seeing the "other side" of the situation, which is highly necessary even to "confront" it, if that's what you want to do (which is also why a Marxist, for example, has to read libertarians, anarchists, and other authors).

 

You've said you're talking about GenAI and then used multiple examples of how it can work that aren't GenAI. 

Hell you even repeatedly use the correct terminology for the AI models in question so you know they aren't GenAI.

As already stated, my opposition is GenAI not all AI. 

Should have kept you on the ignore list. 

[Jill 31,562]

Just now, Bronco said:

You've said you're talking about GenAI and then used multiple examples of how it can work that aren't GenAI. 

Hell you even repeatedly use the correct terminology for the AI models in question so you know they aren't GenAI.

As already stated, my opposition is GenAI not all AI. 

Should have kept you on the ignore list. 

Why are you attacking me? lol. My examples were about generative AI. What are you talking about? Are you talking exclusively about AI art? That's a portion of generative AI.