Earlier quoted context omitted.
So let’s say in 10 years they have the processing power to generate full movies on demand. OpenAI trains its model on all movies from IMDB. Then I ask ChatGPT to produce a movie « like » avengers but with the faces of all my friends. And I just pay the 20$ fee (or 200$ by then) to OpenAI and watch my movie. Then I can ask to tweak some parts of the story and update the movie in real time. And this for any movie. Give…
And then obviously no-one will spend $220 million on a movie anymore because, hey, we can just use generative AI. So I guess subsequent AIs will be based on the outputs of a previous AI? Or will all movies from a certain point onwards be wholly based on a corpus of existing movies used for training? Maybe AI companies will start shooting small film segments in meatspace just for the purpose of training or providing s…
ChatGPT turned generative AI into an “anything tool”
201–210 of 245 posts
Re: ChatGPT turned generative AI into an “anything tool”
#202Earlier quoted context omitted.
I'm actually worried of the inevitable enshitification of Open AI. Right now, GPT 4 is great and I use it every day. I rely less and less on Google and save a ton of time. Google has gone downhill hard for a few years, thanks to all these silly blog engines like medium, substack and the like. I'm really worried and would like to 'freeze' GPT 4 as is, at least with the data up to Sep 2021, and for them to not change o…
I'm almost hoping for some enshittification. It's so depressing seeing how many skills and jobs have been instantly devalued. Take a look at the artist communities and it's total depression that peoples life passion and source of income has just been deleted over the course of a few months. We aren't talking about burger flipping and checkout scanners. This is a creative pursuit which takes years of grinding to becom…
Even the street vendor spraying away at stencils is selling an experience. The artwork itself isn't what's valuable.
And I can say the same about nearly every supposedly devalued profession.
Re: ChatGPT turned generative AI into an “anything tool”
#203Earlier quoted context omitted.
A human doesn't exist in servers, work for free (well, minus utilities) 24 hours per day and work as long as you want on a project you've given it with zero input. If you can't appreciate the difference between an AI model and a person I think that says more about you than AI.
None of those differences really says anything about morality of having an AI do things, unless you want to argue that the AI suffer from their (mis)treatment. People sometimes state that The Simpsons is now drawn in sweatshops somewhere in Asia. Assuming that's true, or at least that it's happened at least once for at least some cartoon, does it make any difference at all to the question of where the people directin…
I mean, if it was actually AI and not machine learning, that would certainly be a question wouldn't it? That being said, I'm not saying the machine is suffering. I'm saying the people it's replacing cannot possibly hold their own. There is simply no way in hell a person can compete with generative AI, assuming of course you don't need perfection just "good enough." And given how massive components of the modern economy get by on far less than "good enough" I'd say that's a solid reason to be concerned.
> People sometimes state that The Simpsons is now drawn in sweatshops somewhere in Asia. Assuming that's true, or at least that it's happened at least once for at least some cartoon, does it make any difference at all to the question of where the people directing them get their ideas from or how those workers learned to do their thing?
Collaborative creative work among a team of individuals regardless of their location is not the same thing as generative art trained on a dataset it's designed to mimic, come on now. You're grasping at some pretty flimsy straws here.
Re: ChatGPT turned generative AI into an “anything tool”
#204Earlier quoted context omitted.
AI is a tool and is not going to replace artists any time soon. Proper 3D art generators don't even exist yet. The ones that do are useless toys. We aren't even close to generators that can output anything even close to what a beginner 3D artist can produce. And yes - it will make sense to physically draw stuff, because it's a much better way to tell the AI what to do. You can massage prompts all day and you will sti…
> AI is a tool and is not going to replace artists any time soon. It has for my personal projects. I used to hire freelance designers often. I see no need anymore. > Proper 3D art generators don't even exist yet. The ones that do are useless toys. We aren't even close to generators that can output anything even close to what a beginner 3D artist can produce. Disagree, unless you believe we weren't that close to good…
You need to generate models, materials and textures separately. Need to split the model in separate parts that make sense. Need to know where to use spline meshes, where to use instancing, how to group meshes and so on.
Also which format to pick? Gltf is a good choice, I guess. But it's pretty limited compared to formats like .blend or .max.
One of my favourite examples is a clockwork mechanism. Current 2D AIs always generate complete nonsense, because they have no idea how a clockwork mechanism works.
A much more advanced model is required for something like that and it currently doesn't exist.
A model like that is the bare minimum for 3D.
Re: ChatGPT turned generative AI into an “anything tool”
#205Earlier quoted context omitted.
To use a fishing rod and drive a car ?
To have the social and physical capabilities to create and pass on the concepts of "fishing rod" and "car" as well as how to use them. We did start from scratch, after all. But I think the learning of simple, complex, abstract, and nuanced concepts throughout a person's development as their brain grows is more than enough to compare to an artificial general intelligence without taking into account development of a sp…
Re: ChatGPT turned generative AI into an “anything tool”
#206Earlier quoted context omitted.
None of those differences really says anything about morality of having an AI do things, unless you want to argue that the AI suffer from their (mis)treatment. People sometimes state that The Simpsons is now drawn in sweatshops somewhere in Asia. Assuming that's true, or at least that it's happened at least once for at least some cartoon, does it make any difference at all to the question of where the people directin…
> None of those differences really says anything about morality of having an AI do things, unless you want to argue that the AI suffer from their (mis)treatment. I mean, if it was actually AI and not machine learning, that would certainly be a question wouldn't it? That being said, I'm not saying the machine is suffering. I'm saying the people it's replacing cannot possibly hold their own. There is simply no way in h…
ML is a subset of AI, not a disjoint set.
> I'm saying the people it's replacing cannot possibly hold their own.
Sure. And? This is the exact same problem with offshoring to sweatshops, and also with all automation from the pottery wheel onwards; and in the industrial revolution this automation happened in a way to cause the invention of Communism.
I'm certainly curious what the modern equivalent to Karl Marx publishes, and how much it will differ from The Communist Manifesto.
> Collaborative creative work among a team of individuals regardless of their location is not the same thing as generative art trained on a dataset it's designed to mimic, come on now.
The argument you're making doesn't appear to care about the differences — it should bite on both, or neither.
Re: ChatGPT turned generative AI into an “anything tool”
#207Earlier quoted context omitted.
> None of those differences really says anything about morality of having an AI do things, unless you want to argue that the AI suffer from their (mis)treatment. I mean, if it was actually AI and not machine learning, that would certainly be a question wouldn't it? That being said, I'm not saying the machine is suffering. I'm saying the people it's replacing cannot possibly hold their own. There is simply no way in h…
> if it was actually AI and not machine learning ML is a subset of AI, not a disjoint set. > I'm saying the people it's replacing cannot possibly hold their own. Sure. And? This is the exact same problem with offshoring to sweatshops, and also with all automation from the pottery wheel onwards; and in the industrial revolution this automation happened in a way to cause the invention of Communism. I'm certainly curiou…
Yes but crucially: the industrial revolution and automation in general has historically been targeted at things of need, in various stripes: one can argue that the demand for goods, be they food, cellphones, televisions, cars, what have you makes automation extremely helpful: yes there was a period where laborers were replaced and priced out of the market, but many eventually returned. The ability to produce 65" televisions at scale and sell them for cheap enough to let people buy them at scale has led to the prices of televisions cratering, and in the context of products consumed by large portions of the market, this is a desirable and good thing (with caveats).
This breaks down with creative output though. No matter how cheap they are, one person can only consume so many, for example, movies and television shows. Even if generative movies were a thing (I don't think they are yet?) and you could produce them just block by block at scale... there's a ceiling there. People can only consume so many movies. You will saturate that market incredibly quickly, and even that is assuming the market of moviegoers will be interested in an AI movie.
Hell, Disney has already learned this without even needing AI to do it. A big part of their ongoing failure that is their streaming service is they completely and utterly saturated the market for Marvel content. They poured billions into all manner of series and films, the quality has steadily declined, and despite being cheaper and easier to access than most, people have still managed to get sick of it. And however you want to slice it one thing AI cannot overcome is that it cannot create new, novel concepts: it can only remix and recombine existing things into new combinations. This is probably good for hobby tier stuff, but for industry? This is going to crater VERY quickly I believe.
> I'm certainly curious what the modern equivalent to Karl Marx publishes, and how much it will differ from The Communist Manifesto.
I enjoy David Graeber personally.
> The argument you're making doesn't appear to care about the differences — it should bite on both, or neither.
I mean there are definitely things to be said about outsourcing in that conversation, the incentives at play that make these animators willing/interested in learning a style of animation that is not part of their culture, the benefits involved in their education, why they're paid a fraction of what westerners would be for the same work while the product is then sold for the same if not a higher price. I just don't think it's relevant, they're still people and still work within the limitations of people. You're not talking about a widget press that forges 10 widgets with the work of a single operator vs. a blacksmith making them by hand with regard to generative art, you're talking a widget press that uses less resources to produce effectively infinite widgets at incredible paces that are simply unfathomable to the blacksmith, and also somehow the machine stole the souls of millions of blacksmiths to enable it to work.
The metaphors just break down when talking about this, because of the sheer differentials involved.
Re: ChatGPT turned generative AI into an “anything tool”
#208Earlier quoted context omitted.
I'm really stunned by the fact that, it seems, so few people get this. It's not a robot in the dense of Commander Data from Star Trek. It's a recursive software method that just has unfathomable amounts of data to base it's guess for some correct sentences on. There is no logic, no reasoning, no nothing. Yet people complain it's 'giving me wrong answers'. Well, it doesn't know what's it's giving you either way. It on…
You could say the same about your brain. You predict the best thing to say at the moment all the time. You make predictions off everything you do.
As an aside, the claims that people are willing to make about language models are quite astounding considering that they never seem to realize that most of those claims apply to humans also...
Re: ChatGPT turned generative AI into an “anything tool”
#209Earlier quoted context omitted.
I disagree with you, simply for the fact that artists have been learning from one another for thousands of years. We can see a clear timeline of art and it’s progression throughout human history, and it’s often clear how a later work took inspiration from an earlier period. Art school teaches techniques and methods pioneered by earlier artists, for the express purpose of their students to know how to incorporate them…
Alike in method is not like in output, and it's output that matters. A human takes ~4-20 years to become a good artist. They can then produce works at a single human rate. A model takes ~30 days to become a good artist. It can then produce works at an effectively infinite rate, only bounded by how many GPUs and much electricity can be acquired. These are very different economic constraints and therefore require diffe…
This is often listed as the reason why it’s ok for human to learn from a prior art, but not for a LLM. The question is why? If the act of learning is stealing, then it is still stealing, no matter how small scale, and every single human on earth has committed it.
The LLM vendor may benefit more than a mere mortal pupil because of the scale and reach. At the same time the LLM may make the prior art more visible and popular and may benefit the original creator more, even if only indirectly.
Also if content creators are entitled to some financial reward by LLM vendors, it is only appropriate that the creators should pay back to those that they learn from, and so on. I fail to see how such a scheme can be set up.
Re: ChatGPT turned generative AI into an “anything tool”
#210I'd like to have LawGPT, where it reads all laws of a certain country, and I can ask it things like, "what are my options when the neighbors tree crosses over into my property?"