In my latest series of tweets [1] I tried to argue that this is not the case in a less technical way, not sure if I was successful though. For anyone not fluent in Korean, my argument was as follows:
Stable Diffusion was trained with a data set composed of about 2.32 billion text-image pairs. The model itself by comparison weighs about 4 GB. If the model somehow stored those images (completely disregarding texts) in the compressed form, it would be less than 2 bytes (EDITED) per each image. Whoever downloaded an image from the internet can see this is absurd. In reality 4 GB of data is the juice extracted from the data set, and if you could actually recover an original image from it then the juice must have had a recognizable residue, which would be pretty undesirable.
It’s a very very fancy compression algorithm, but it still is one.
It’s easy for these models to produce very close images to the source material with the right short input.
The fact that people are writing prompts with named artists, and phrases such as "Trending on Artstation"- yeah, I'm not sure if you can just handwave that away.
I don’t see the moral significance of the former genres versus the latter
Also, even if the model didn’t come with knowledge of an existing artist someone could fine-tune it in, and it’s possible the model can learn about your art style without seeing any of your images anyway.
Others have alluded to this, but I'll make a larger claim: human artists actually are doing very little that is original. The styles, composition, and subjects are all derivative works. That means their "essence" can be recreated without them.
Another fascinating consideration, if artists names were excised from future models, either by law or by choice, the "mirror" artists (fake) could be the ones which become famous and the humans forgotten.
AI could become more generalized or there may be millions of models, doing specific things, strung together like with APIs. Either way, if an artist or trademark owner chooses some kind of explicit blocking, from the input side, it is possible that it would be the equivalent of your keyword censored from Google's index, Google Maps, Amazon, and so on. Disney? I don't recognize that word.
A few weeks before OpenAI made DALL-E 2 open, I went through a series of prompts using well known artists as a basis. Vermeer was one of them. It generated some pretty amazing works that were 'inspired' but not direct copies of Vermeer's works.
I started feeding the same prompts into StableDiffiusion (the HuggingFace online demo) and it output direct copies of Vermeer's works with minimal modifications.
Two examples: https://imgur.com/a/KAxmZCl
The original: https://en.wikipedia.org/wiki/Girl_with_a_Pearl_Earring#/med...
It's reproducing the folds in the headband, riffles in the clothing, etc. It's not just a "similar" reproduced image, it's clearly some minor variation on the original.
https://drop.qoid.us/vermeer.png
And, for a slightly different scenario, here's what I got for 'pikachu':
https://drop.qoid.us/pikachu.png
But I think this sort of thing is the exception rather than the rule (similar to Copilot regurgitating the text of software licenses). Girl with a Pearl Earring is so famous that there are countless copies of it on the internet. I don't know what kind of deduplication process was used for Stable Diffusion, but if you consider multiple photos of the same painting, plus images distorted by compression, or cropped, or embedded in other images, plus attempts to redraw it or even parody it, I can easily imagine that the training set contained thousands of images or more that are essentially the same work. And so the model memorized it, albeit at low resolution.
Similarly, it's seen countless different images of Pikachu.
However, this seems to only apply to the very peak of popularity. I tested some other Pokemon: it does a decent job with Charizard, but with Bulbasaur and Squirtle and Jigglypuff its output only somewhat resembles the character design, and for less popular Pokemon there's no resemblance at all.
I think that's part of the point, though.
Even if it's working like some kind of feature detection - "this is what 'nose' looks like [nose1,nose2,...,nose9999]" and "this is what 'girl with a pearl earring vermeer' looks like [vermeer1,vermeer2,...,]"
Clearly it is storing copies of those images internally, and constant re-enforcement of it finding noses and vermeer paintings in the sources pushes the priority of those features up so that it stores them.
We can't trust these models, as they exist right now, to reliably generate new works, rather than spit out something that already exists.
There’s a difference between an artist and a machine (for now), but it’s not unreasonable to assume that some internalisation of others’ work is fair.
I didn't claim that it was storing all of the images.
It is, however, quite clearly storing at least some of the original training images. What that definition is, I don't know.
To go back to the original point of this thread:
> The thing is, people have wild misconceptions about what this technology does. Many clearly think it's a photocopier. As if StableDiffusion is hiding copies of all these works inside.
I don't think it's a wild misconception, or even an unreasonable concern to be worried that it could output some of it's training images. As an end-user of the technology, you have no idea.
There's different sorts of art for different purposes (commercial, fine, etc.), and for many of these sorts of art having an AI make it would be acceptable.
But when I think "art" without a qualifier, I think "fine art". And the purpose of fine art, for me, is human communication of ephemeral things that aren't possible to communicate in language.
I don't want AI generated fine art because it turns the art into just a pretty picture. It's no longer about another person communicating something significant and human to me. It removes its purpose for existing.
In that view, I totally understand why some art communities would ban AI-generated art. I don't see a problem with it. I'm sure other communities will not ban such art. That's fine, too.
But I know which community I would be more interested in paying attention to.
There is a person thinking about the prompt, typing in the prompt, painstakingly selecting the image that matches what they imagined in their head.
It just less elitist, as now even a child can create a complex, meaningful and beautiful images by leveraging the essence of human art (compressed into a single ML model).
I, for one, never had the time & chance to learn painting. Even if I did it as a hobby, it would take me up to a decade to become proficient at it.
And now I can ask AI to paint what I dreamt about last night (!!!), or create an impressionist painting almost exactly like I see it (or even better).
Fun factor is also there: I can see how an anime character could look in real life, or put a movie character in a new context , or just generate cute corgis wearing funny hats and driving cars
Which does make this a type of art. It just isn't the same form of art as digital painting, the same way painting a tree is not the same type of art as growing and 100 trees, choosing one of them and taking a photograph of it. Photography is definitely an art-form, but it's a different art-form from digital painting, even if the end result of the both of them can be a digital image. Not better or worse, but definitely different.
True, which makes it more akin to photography. I'm not saying it's "not art" (I'm saying the exact opposite of that), but that it's not art that is of interest to me. Much like my interest in art photography is very limited. Different people have different tastes.
> It just less elitist
I disagree. I think this whole subject is orthogonal to the concept of elitism. Sure, there are certainly art snobs who take an elitist perspective, but they're a very small segment of the art world.
And there we should just ask our new creative oracle machines the obvious: "devise an understandable minimally invasive method of uplifting our own cognition and extending our biological lifespan". Surely at least some gene/cell/whatever therapy which does what we need is theoretically possible - the human beings aren't special enough for this not to be the case.
At least that's the zeitgeist I perceive among others who understand the significance of the unfolding progress.
Why should we pay taxes, if the government won't even train AI for us, given such opportunity?
Large language models trained on cross-entropy loss for next token prediction aren't going to be dangerous per se, even in the limit - simply due to the nature of the objective and the distribution they are set to approximate.
We should make sure these powerful tools don't fall into a small exclusive set of power-hungry interests, though. Transformative tools such as these should benefit all of humanity and let it prosper.
An AGI, by virtue of being at least as general as a human, would be able to effectively form a plan for how to act in the world to achieve fairly general goals, in fairly general conditions.
Even if it isn’t agentic, if a program can formulate these plans in a way people can use, then the combined program+user still implements the optimization process.
The democratic solution here looks better than the expected rule of the (very) few.
Image generating AIs are trained with massive amounts of images and text, however image generating humans train with much broader spectrum of experience.
Also, feeding a model tons of images created by humans (directly or by proxy) and claiming that AI is generating something completely new is a bit naive IMHO. Humans mix a much more broader and deeper experience pool to create things without prompts.
An AI model blurts out something derived from a corpus of images and text created by humans, that's all.
The technology is impressive for sure, and it marks a new era in terms of possibilities, but it doesn't take my breath away, sorry.
The AI emulates. Humans create. - The significance here is not trivial.
I've seen lots of human-made art (especially on Artstation) that failed to elicit any emotional response in me.
The reason people are having this discussion is a fear of AI creating aesthetically pleasing and deep emotionally through statistical approach.
And the deeper fear that AI can potentially "hack" our senses by providing us with exactly what we want to see, read or hear (statistically).
i.e. if some subset of art is pushing our biological buttons more than all other art, and the effect is generalizable over human population, what would that say about us?
To better illustrate that difference: Just because anecdotally some human might produce crap doesn't mean that is the limit of all humans.
So you can imagine new colors?
That wide variety of experience changes the decision of what to create, but I think it has less influence on the actual output than one might expect. A person can have enormous life experience but if they have never seen a logo, a photo, a painting, a drawing, or any other kind of rendered image they will be no more sophisticated in their artistic output than a caveman. Modern artists stand atop a mountain that their ancestors had to climb inch by inch.
I want my photos to make the viewer feel a particular way when viewing what I took. This affects colors, style, and everything.
Arguably, as long as the medium carries the message the creator intended, it's equal in my view. It can be a photo, a drawing, a digital painting or a physical painting with oils. They can create the same feeling if the artist aims for it.
This "optimization for feelings" is a result of my own non-photographic experiences. Sometimes a song, sometimes an event I went through, sometimes an art piece I saw on a medium I don't work.
From my understanding and observation, I saw that many artists work that way. They reflect their emotions in one domain (generally personal life) to another domain (the art they create). Also, there was a video which I fail to find over and over which shows how three designers got affected by the things they saw (and deliberately planted into their minds slowly) during a 20 minute car trip.
Yes, we're building upon this great mountain of experience and knowledge, yet our output is affected by what we experience in other parts of life, unrelated to the art we create.
Consider the following experiment: The music you listen changes your mood, your working speed, and what you feel; effectively changing how you operate and what you create. Exclude work from this, which is by definition something you have to do, unless you're literally dying.
Right, and what happens when you let that loop feedback to itself?
An asphalt road on every corner, so we build more asphalt roads, we design only what has been designed before. Amplify this thousands of time through computer generated design, the feedback loop is closed and unchanging. Unless your goal is actually to create an independent organism (different moral issue), you are creating sophisiticated feedback loops, not worthwhile content generation. Unless by content generation you mean endless remixing...which is not the same.
Sort of like changing the color of a 3d model and claiming it is a new race with new attributes, as is often done on the cheap e.g. in videogames.
I don't see how AI art is less new than the vast majority of human art. Both can create unique compositions that are still deeply rooted in patterns and principles thousands of years old.
> Humans mix a much more broader and deeper experience pool to create things without prompts.
AI does not have understanding, but it's stealing the underlying patterns that are the end result of that human experience.
OTOH, if the 2 bytes is an index into a larger corpus of collected works, the question becomes more interesting.
OTOH, if the corpus is an ensemble of other 2-byte objects...
It's only "easy" because bunch of people over years perfected every step of the process.
You can see that it has memorised all its training material very clearly in its ability to perform style transfer. You can ask it to geneate an image of Pikachu in the style of Vermeer. It can do that because it has memorised images of Pikachu and images by Vermeer in its weird compressed form and it can combine them by finding a point on a gradient between their representation in its memory.
If it didn't have memorised images of Pikachu, or images by Vermeer, how do you guess it could generate images of Pikachu or in the style of Vermeer? It's not magick.
Better analogy: suppose you take 10 pieces of A4 paper and scrunch them up into 10 separate balls. Now suppose you scrunch them all up together into one ball. One manifold, taking ten times less space.
In the same way, you don't need to store each pixel of each image separately, or in its own sub-space. Yes? It's really good compression, that's all.
Over the past 150 years it became possible to compute cube roots in milliseconds, instead of minutes. Slide rules became obsolete, and the things we learn to do do arithmetic changed.
Over the past 20 years, the way we look up arbitrary facts changed. We learned to construct Web search engine queries and we learned to evaluate the quality of results. Other library skills are still relevant, but certain kinds of rote memorization become obsolete.
Now people are rapidly learning a new "prompt language" which uses English words but combines them in a way which is for now something like computer programming (given an input string, compile it to an output image). Will "prompt languages" for text / image / video generation systems turn into a full-fledged, non-English language that you can learn?
What will be the social effects of people learning and using this language to communicate?
Coallescing on prompt languages is probably unlikely, if only because part of the thing that defines a language is that a community has rules, both opaque and explicit, about what is acceptable in the language and what is not, and there's no effort to standardize inputs.
There is a chapter in the Stanford AI Index (2021) that opens with "The rise of AI will inevitable raise questions of how much these technologies will impact business, labor and the economy in general .. "
there are winners and losers here, not just interesting questions. In the 1930s America there was mass unemployment among many skilled "workers" .. can't imagine the 1990s ex-Soviet economies.. I dont think these questions are neutral or innocent.
We tend to value scarcity. Things that resemble AI art will become passé and anything that is difficult/impossible for AI to produce will become highly regarded. I'm very curious to see how this shift in popular and/or high-brow taste will play out.
Things are really going to suck for working artists who can't adjust. That's not cool and I should have acknowledged it.
Sure, but these are temporary disruptions. Now it's true that the pace of change feels immensely faster than ever before, and I believe that's an observable fact. But we adapt. The incentives to do so are, sometimes, life or death. When faced with that kind of situation, some small % will sink into depression, perhaps never to climb out. I'm sure that some online magazine (which surely used to be known for its print version, but is no longer...) will write an extremely long piece full of anecdotes about the plight of the artist whose livelihood was destroyed by the evil AI.
But I assert that it will always be a tiny percentage of the whole, because the vast majority of people like to eat food and live in houses. Many of them will look at these new tools and make amazing new things. Many will go do other things. That's life, man.
However, there are different "lenses" through which one might examine large topics, and one lense might be that of personal challenge, adjustment and endeavor; but another lense is closer to The Economist Magazine, where factual snippets of market behavior, participants and results are traded every day, every week, every year. Any college educated person ought to be able to say, there have been real, serious and long-standing economic changes where thousands and millions of capable, good-enough people, had serious, years-long hard times up to and including starvation, war, and abundant death. Those without personal experience of that, or a close relative or similar imprinting, may not really consider this real. I had to learn it from books myself based on where I grew up. Others reading these words, know it very well.
Hand-craft preservation is a thing, I have heard.. so there is certainly a broad spectrum between "no more blacksmiths downtown" to "I send my print jobs via phone for pickup near the metro at an automated kiosk". It is said that nobody has a right to a job. However, The Economist Magazine exists for a reason, and things are not normal where I live.. Welcome to the New Not-Normal, as Jerry Brown said..
Arguably, that's basically directly saying that nobody has a right to life.
You either ensure the basic resources necessary to live are free, or you want everyone to get a job. The alternative is saying - "some people die from lack of work, get over it".
Good :)
> The Economist Magazine, where factual snippets
I have nothing special for or against the mag, but I will point out that economists are famously known for disagreeing on huge things once you get beyond the law of supply and demand. Not every discipline of study has whole alternative schools of thought the way they do. Economists are closer to philosophists than most philosophists.
I really love economics! But I have no illusions at all that they serve up one and only one version of the truth like physicists. :D
> real, serious and long-standing economic changes
Ok, 100's of counterpoints: https://www.humanprogress.org/datasets/ (but check out the articles also, lots of great stuff on there.)
> starvation, war, and abundant death
Yeah, there's lots of that. Pre-panbdemic all the trends were in the right direction. This too shall pass.
Things aren't that abnormal in my area, and it's getting even better daily.
"founded in 1977 by Ed Crane, Murray Rothbard, and Charles Koch, chairman of the board and chief executive officer of Koch Industries."
some of these people have funded climate denial on behalf of the Oil and Gas Industry Lobby.. they have the largest wealth in history to defend, at the cost of Climate Change.
> But I assert that it will always be a tiny percentage of the whole
US/Uk life expectancy is falling, and more people are in precarious work than ever before.
You can't just wish away these problens and pretend they don't exist
Why do you posit that life expectancy will continue to fall? Not sure it's rational to expect an extended reversal of the trends which have been on the increase prior to the past 2 years. Look at the charts and the analyses.
Precarious work? You may need to explain that one.
So what? Slavery and The Plague and Cancer was part of the human condition too.
And its wrong - technological disruption only appreared recently. For ten thousand years my ancestors lived nomadic lives on the eurasian plains, their lives were disrupted when a gun and a steam engine were invented few hundred years ago.
or
'human condition sucks and the only way to fix it is with disruption'?
Are you just making it up as you go along? Do you feel the need to justify disruption at any cost?
Now, we cannot suppress technological disruption either, we can regulate and channel it, but you won't be getting your pastoral society back in which everything stays as is for a few thousand years . I think that's good, you Kay think that's bad, but it just is.
I am just pointing out that the post by halr9000 and your post both try to put disruption on the pedestal, but the arguments are mutually exclusive, and in fact they profess opposite values.
I got a confused, through both posts were made by the same person, hence 'making it up as you go along', sorry about that.
The headlines like "AI art banned!!!" just generate more clicks as this thing is fresh new
Which brings me to a topic which seems to never go out of fashion: Purism/Conservatism.
Like most discussions that are centred in conservatism/purism, there's always a thread of hypocrisy, because this type of person will use themselves as the golden benchmark to what is acceptable.
To use an example: a purist painter of earlier times might shun the modern purist painter for the simple act of purchasing pre-made paints. They might bemoan that they didn't make their own pigments and mix up their own paints.
So while this online art community might ban AI work, it merely an action that excludes themselves. AI opens the door to many who otherwise wouldn't have the capability of producing beautiful pieces, just as digital painting did, and the number of those people will be larger than current # of digital painters. They should have just made a subcategory on their website for AI and used it to bring new relevance to their platform.
> To use an example: a purist painter of earlier times might shun the modern purist painter for the simple act of purchasing pre-made paints. They might bemoan that they didn't make their own pigments and mix up their own paints.
That I think is pretty much why this stupid discussion is still going - people considering their work (or just stuff they enjoy) "art" want other things to not be it to feel better.
My art was made by a friend of mine who works in a union factory in Texas. Art is his real joy and I was able to snag some pieces from when we were in college together. It's easy to write off that the pieces are just sentimental and therefore subject to sentimental value, but really, his work is very abstract and each piece (if you can snag them in order) describes a time in his life.
I tried to explain to this person how art to me is the story not told or the story as I see it. More recently I had the opportunity to really appreciate his work when I was on an LSD trip, but that's a story for another time.
For those formally educated, schools should really focus on the economics more instead of relying on magic
The reason to me is that people that buy contemporary art (stuff from living artists) are already going out of their way to do so, knowing that it is divorced from the concept of investment or stores of value. But for other living artists not to really understand the motives or empathize with the customers, going as far to criticize them for any number of reasons (daring to question the stated price, having a taste at all) really undermines the point of bothering with living artists at all.
You make a good point that the trend of automating art is not a recent one.
At some point I think a line in the sand must be drawn where the human artist is no longer the primary conceptualizer.
Writing a text prompt is a fair place for that line to be drawn in my opinion.
I read an article[0] on the subject the other day. At some point it said that some art tools are inherently easy to use. Part of the difference between an amateur and an actual artist is in the mastery of such tool.
Think of a phone camera for example. It's extremely easy to take more than decent pictures with it. But it takes a good photographer to make art with it.
Bringing it back to the topic at hand it means that yes, describing what you want to see to this kind of software is easy on a basic level. But you get a picture that looks like any other. Instead, you need to speak the language of the software to obtain a better materialization of your vision, and that requires some sort of mastery of that tool. That would be the difference between an illustration made with stable diffusion and actual art.
[0] https://www.indiscreto.org/non-fa-tutto-il-computer-il-ritor... (in Italian, sorry)
Suppose I'm an artist, with something I feel I want to say, something I want to share with the world. How can I convince you to listen to me, when there's a thousand other people you could listen to? Especially about big issues like what's worthwhile in life.
Well, one way is commitment. Supposedly, Sophie Germain decided as a child to study mathematics because she read the story about Archimedes, who was so obsessed about his circles that he got killed over it. Clearly, Archimedes had found something that really mattered to him, and if Sophie studied it, maybe she would find that meaning too?
So likewise, as an artist I can convince you to take a look by being really dedicated, practicing the craftsmanship side of what I do so that no one can convincingly say they do it better. That doesn't guarantee that what I say is worthwhile to you, but it's at least an indication it was worthwhile to me.
But you know, the converse is not necessarily true. Someone who wasn't the best at the craft and who didn't sacrifice everything for the opportunity to reach you, could still have something wonderful to say.
Sometimes the skill at the craft limits what you're able to say, even if what you wanted to say was great, but I think that's rarely what matters. On a technical level, there's very little that distinguishes, say, the best guitarists in the world.
My inclination is to say the interface needs more breadth of options beyond single-line string input, and the outcome needs to have some level of predictability to enable the artist a significant enough level of understanding of how each aspect of their input alters the output.
I mean, the input parameters should be complex enough to let artist intentionally develop a unique style/idea, rather than style being dictated inside the black box.
Once I’ve selected one from the 20 or so images, had it printed, framed and placed it on my wall then weeks later notice it again when the light coming through the window catches it just right, I think that is a really great picture that transports me to another place.
Is it art? Because I was holding the camera does that make me an artist?
All I know is that my life is enhanced by having that image on my wall however it came to be there.
https://en.wikipedia.org/wiki/Buffalo_buffalo_Buffalo_buffal...
That sounds a bit negative I guess, but techniques like stable diffusion will be applied to other problem domains, so all other art forms like music and creative writing will be automated within the next 5-10 years. And if computers can do that, they can do anything. Because that's how our brains work. Our neurons are microscopic filters magnifying patterns in the long tail of quantum randomness where source consciousness probably resides. Although science will only have recipes for creating consciousness, it will never be able to explain meaning, because it has nothing to say about it. That's the domain of personal experience, faith, the study of magic, etc.
The big question for me is: if a liberal arts degree is worth even less than it is today, what will people do for work? We're looking at not just the death of creativity, but the death of all human contribution. This was predicted decades ago (see Asimov, Kurzweil, etc), but we're not prepared for what's coming. I just don't see us switching to UBI before we're subjugated beneath the yoke of corporate AI that gradually changes out human workers for machines. We're in the endgame now.
I still hold out some small hope that kids today will run with this and do creative stuff we can't even imagine. Or will we just merge with AI when the Singularity arrives around 2040? Who knows.
Prototyping games is getting so much better, this will really help indies game studies quite a bit.
Will those AI generated assets replace hand crafted ones? Heck no!
Have you ever tried doing AI art?
What gets posted on social media are the few successful experiments. Most AI generated stuff is absolute crap. Getting anything useful out of it requires lot's of trial an error and magical incantations.
Now yes, you can quickly make generic background #20233 with it or simply something weird that would fit into some bad dream. If you need something specific and coherent though? Yeah, this will be very hard.
Here is how to make concept art with stable-diffusion: https://www.youtube.com/watch?v=rS2kNvbIL5Q
Look how many iteration that requires. How it requires finding the proper real word assets for img2img techniques. How is still requires some serious Photoshop skill to put everything together.
AI helps us build more higher level tools for creating art but the idea that it is anywhere close to creating art itself is silly. We have not made any progress in creating general artificial intelligence at all.
The main progress we see is that AI-tech is that it is becoming available for the masses, which is awesome.
Yes.
> What gets posted on social media are the few successful experiments. Most AI generated stuff is absolute crap. Getting anything useful out of it requires lot's of trial an error and magical incantations.
You are exaggerating. Easily over half the output I generate is on par with the standard being posted on social media.
The demos look amazing because there is no acceptance criteria. Anything gets hailed as "groundbreaking" because is process is novel. Which is a completely different scenario from what you expect from someone actually being paid for making art.
I am still figuring that out myself. My current project is a fantasy-themed visual novel for which creating backgrounds is super easy. You just pick a certain style, for example "by Greg Rutkowski" and put that in any prompt.
Normally you want to reuse your backgrounds as much as possible to not blow the budget but thanks to AI I can be much freer with my writing. Of course the important backgrounds should be remade by hand at some point but for prototyping they are great. Character art does not really work beyond concept art as you normally need different poses for a character and you are very specific about what you want, again not a strength of AI.
My advise would be to play to its strength, if you can: Horror works really well. Outdoor scenes work better than indoor.
There's nothing to say that creativity will die. We invented typewriters, but calligraphy is still an art whose pens and inks are beloved interests. We invented the digital screen and the paper machines, but handmade paper and hand-stitched book binding is still an artist's pursuit.
True, but fast food has seriously damaged and reduced both real restaurants and home cooking.
https://twitter.com/hollyherndon/status/1567694502382518274?...
It's like saying that computers killed artists. AI is just a tool, it still need a human in control and guiding it, can't generate anything on his own (yet). If anything, this can help creativity.
Admittedly, my acquired lack of creativity after a lifetime of struggle is the very thing preventing me from seeing beyond this horizon. Too old to begin the training and all that! Ironically, I find the not-knowing to be somewhat comforting.
Exactly, think of it as a symbiosis, a man is slow, but a man in a bycicle is fast. Now a man in a rocket, can go to other planets. But it is still a man, a rocket is just an inert pile of metal and fuel.
Same with AI, AI is the rocket, it propels and accelerate creativity to inhuman levels, but there is still a man controlling it. AI is just an extension of our minds like a rocket is an extension of our feet.
gibberish
> Although science will only have recipes for creating consciousness, it will never be able to explain meaning, because it has nothing to say about it. That's the domain of personal experience, faith, the study of magic, etc.
misguided conjecture
> We're looking at not just the death of creativity, but the death of all human contribution
hilariously misinformed extrapolation
AI is nowhere closer to cracking the meaning problem. Until we have meaningful AI generated art there's no point for this baseless speculation.
There's analogues in music. We have algorithms that can write jazz standards with all the same essential characteristics of Blue in Green, or So What, but no one cares. People just want to hear Miles Davis and Bill Evans do it.
That's definitely not true. In the areas that make money with some semblance of reliability for the artists (low budget film / game scores esp. asset flippers, shitty TV shows, music for ads, etc), people are going to be more than happy with a magic "make music" button without having to pay for a composer. Like, if you thing that anything in this video is out of reach for today's AIs... https://youtu.be/QgU9H9dECt8
I also think it's very interesting that these tools enable a near absence of technique. I tend to be one of those who doesn't value technique very much. For instance, I believe I've commented on this site before that while I certainly appreciate the time, effort, and technique that goes into photorealistic painting, I don't think it brings a whole lot to the table beyond what a photograph would. This is strictly a personal assessment, of course -- others are free to, and do disagree with me. I just feel as though photorealistic art leaves less room in it for the viewer than other forms. (Incidentally, I think this also applies to AI art, and may be a part of the reaction to it.)
The displacement problem isn't quite there yet. While it's a good replacement for stock photo aggregators, it's not a great replacement for commissioning an actual artist. E.g. if you want a consistent look and feel for each of your images, you'll struggle. It's only the C-listers, who create heaps of bland stock photos in hopes of earning royalties, who have something to worry about.
As the technology improves, you'll see a whole new tranche of creators who are able to harness it to create amazing new things.
[1] https://superbowl.substack.com/p/the-future-of-art-is-artifi...
This has happened in a different (non-AI, but "technology enables") way already with other forms of creativity. While it is true that more people become enabled, its not clear that this is overall a good thing. The benefits of more people empowered and enabled to engage in self-expression has to be traded off against the problems associated with discovery caused by their work becoming a part of the cultural milieu. People making art in private is almost certainly a good thing without limit; people making art and then making it public is not so clear - most of the work people create (even the best) is just not worth being public, but the self-limiting discrimination that this implies is in short supply in most of us. Yes, it's good you made it. No, it shouldn't be online or public, almost all the time.
I see exactly this backlash in most art schools until eveything is so mingled a new generation just doesn't care.
Photoshopping or making an AI make art still takes skill, so it's still art - just in a different category. Putting a robot in a boxing match or an AI in a human car race wouldn't make much sense, but robot vs robot would.
https://twitter.com/matthen2/status/1566424523125276672?s=20...
(And thanks :)
for example, in the last two weeks, promotional flyers for parties in my area have been all AI generated, because Stable Diffusion is free and DALL-E 2 completely fumbled with their limited access and data credit system
its kind of game-over for the promo art trade, the transactions that were happening, are not happening
Of course people practising it would say that it is an art.
And of course they would discount stuff that is beneath them as not art. "I'm creative, I don't just copy stuff I see!", just to feel better about themselves.
So someone might say code is not an art, it's same as any other engineering. But code can clearly create art pieces, whether directly (say a shader generating pretty picture) or indirectly (AI training and generating pictures based on that).
So, what, do we just say "art is any product that you can experience and have feelings about?". Nah, can't discriminate this and that based on that vague expression! We might even accidentally call blacksmith an artist if we go too far! /s
I find this is true of a lot of topics, too bad there's not some way to shorten the time between shallow to more useful/informative debate. Your post above does that for this topic, thanks for that.
> Like a music synthesizer times a million.
Even now, every once in a while I find people arguing that synths are not "real music". Also DJs, etc. generating songs using pre-made samples. I think it's a very apt comparison, where tools become more and more available and barrier of entry gets lower, I don't believe art will die but rather just change into new forms, and what's now revolutionary will become just a very commonplace tool in the future, just like synths, and "real" artists will keep on existing alongside.
I haven't found the HN discussion; I have found a 2017 paper [1] that had relevant examples [2].
[1]: https://junyanz.github.io/CycleGAN/
[2]: https://taesung.me/cyclegan/2017/03/25/monet-to-photo-summar...
Duchamp was already one of the great painters in the world when he showed the readymades - and there's a lot of evidence that he made them, what a great joke if so.
They can afford to pay off the press, change the narrative and overpay the artist.
Be it the de Medici in the Renaissance or CIA and state department backing abstract expressionism.
Duchamp poked the art establishment in the eye with his 'readymades' (urinals on the gallery wall, etc.) Warhol with his 'factory' making silkscreens.
The 'reverse Monet' (too lazy to link) that caused such an uproar when it was shared last week was ironic – apparently people aren't aware that the impressionists themselves were considered 'not artists' by a lot of people at the time.
The camera and photography made many painters sit up and say, 'hey, it can't just be about "image" anymore' – enter abstract expressionism, fauvism, cubism, etc. Work that shifted the focus from the image represented in the painting ('a window into a world') onto the surface of the canvas itself: Rothko, Albers, etc.
The thing is, people have wild misconceptions about what this technology does. Many clearly think it's a photocopier. As if StableDiffusion is hiding copies of all these works inside. When in fact what's been done is to take all of this online knowledge and create an incredible new tool with it. Like a music synthesizer times a million.
It's understandable why this technology makes people uncomfortable. There are legit fears about displacement, economic uncertainty, etc. But we are fundamentally tool makers – it's arguably intrinsically part of what makes us human (borrowing from Carl Sagan's "Dragons of Eden").
If I were back in the art department last century, I'm pretty sure we'd be really excitedly discussing how these new tools will enable entirely new forms of expression. And sure, there'd be the old guard saying it was awful, and 'not Art'.
This is a really interesting debate – I just wish it scratched deeper than the surface more often than it does, with a bit more historical context.