The AI Unbundling
181–190 of 207 posts
Re: The AI Unbundling
#182What about the negative side of this? All these generators will be used for evil. Looks like we need a way to figure out if an image or video is real or not and we need it desperately. Not sure if we are working on that at all. The implications of not having that are huge.
Re: The AI Unbundling
#183Earlier quoted context omitted.
You don't get it. > A never-ending stream of novel music that tailors itself to you. That's worthless. A firehose of novelty is almost immediately no longer novel. Why would anyone want this? It's the same problem already with streaming services and their own series. They're all already joyless and immediately forgettable. They have less and less value. Your breathlessly described "amazing transformation" sounds incr…
> You don't get it. It's easy to say that as someone on the outside without skin in the game, not working on this, not in the know. > That's worthless. A firehose of novelty is almost immediately no longer novel. And yet here you are on HN. I'll bet you have a Twitter and phone and consume social media, too. That's all just a poor man's version of what we'll be building. > Why would anyone want this? I have literally…
Re: The AI Unbundling
#184I really wish I knew what DFW had in mind when he said irony wasn't the point.
Re: The AI Unbundling
#185Earlier quoted context omitted.
Trash, but trash minted through an entrenched network of art curators, directors and connoisseurs - this is what makes trashy modern art scarce and valuable, not to mention the tax breaks! What I and some insiders consider the real art - that is, applying rare well-developed talent to production of unapologetically high-quality sensual artifacts of art - is mostly commercialized by now, with best people employed by t…
> I have nothing but respect for the talent it takes to create something that is valuable on its own (contrast with modern art, not valuable without of the context of its social network) All art is only valuable within the context of its social network. If you leave the Mona Lisa in the woods, the only real "value" it will hold is as a shelter for bugs. I think what really bothers you here is about which social netwo…
Let's notice the elephant in the room: HN commentariat is a biased sample with an obvious overrepresentation of autistic people. I'm likely not too much ahead of the curve here, as you can see.
I hope this pretext will add to your charitable interpretation of what I have to say. As they say "autism speaks, it's time to listen". Consider this a hypothesis.
I cannot say this condition is overwhelmingly beneficial - honestly I can interpret it as a bit of a handicap, something that, among other effects, makes the child disinterested in people around him in his early formative years (yes it's most commonly "him"). Instead the child is made to focus on the world and things in it on his own terms, as a one-of-a-kind being starting from the almost pure blank slate - with the blank slate here, perhaps, being an overgrown PFC full of fresh synapses. This makes learning one's world and one's inner depth harder compared to normal individuals - which have the right set of biases and heuristics to start imitating other people early on, standing on the shoulders of giants right away, learning the behaviors that passed through many aeons from parents to children and quickly arriving to a socially-bootstrapped form of self-awareness. And for the autistic One the world (and, apparently, the self) is an alien world which is being conceptualized for the first time ever, and the society is an alien configuration of beings which feel and behave quite differently from how one would expect them to, if they were kin to the autistic. Even when one grows up to like it, it's still like being an eternal foreign student everywhere you go - yet your relationship to the world you found yourself in becomes deeply personal.
What I'm trying to say is this predicament makes one much more likely to learn the world on the world's terms, same with oneself and one's feelings - beauty sense included. Having developed in this way, the social concept of beauty you describe is a vastly alien thing to me - I can understand it, but it's like understanding physics of some abstract phenomena. Myself, I just feel beauty, as it developed inside of me as a palette of feelings resonating with certain structural patterns in the world's percepts I come across - something having to do with information-theoretic regularization in primate neocortex, as I may guess. No social reinforcement was necessary to arrive to this feeling, it has a sense of finality and self-referentiality to it - something basic and indispensable, something valuable.
Does it sound scary, cold and alien to you, or simply incoherent? That's how your idea of socially mandated value sounds to me - shallow repressiveness of one's social graph pushing supervised learning examples into your very soul, re-flashing it with the weights of the collective simulacra, of the lovecraftian entity - a disembodied communal sense of value (can't even call it "beauty" as in the limit it's completely arbitrary, a value-language defined by social custom) - to make you one of them, a part of the eusocial organism? This is 1984-level scary, reeducation camp-level, basically.
... And with these new neural networks I immediately felt some sense of commonality, as if they were from the same metaphorical planet I came from - tiny blank slates learning compressed sense of the world as it passes through them, their beauty and my beauty approximating the same information-theoretical Eidos.
Re: The AI Unbundling
#186Anything this guy writes gets gobbled up and worshipped mindlessly to no end. The opening here was so crass i didn't bother reading the rest (yet) until i find a better time to do so and have put my strong feelings aside. The News aren't just about distribution, you moron. If we didn't have large companies such as twitter and Facebook misclassified deliberately through government protection and instead called out as…
They aren't publishers. Me posting a link to the NYTimes on facebook is not the same as writing a NYTimes article. That fact that NYTimes articles get more boost from the algorithm than a sketchy right wing site does not make facebook a publisher
Just because you post a link somewhere, it does not automatically waive the commercial interests of the platform you're posting to.
In another strain of the argument, you posting to Facebook and Facebook later choosing to rank and distribute that content to their liking with you having ZERO control over that is the equivalent of using individuals as proxies to steal and propagate content.
You can't just start a website with stolen content plastered all over, say its coming "From the users" and deny any culpability.
Re: The AI Unbundling
#187Earlier quoted context omitted.
> People will insist that SD art isn't real art. Artists will fight back, and lose. When talking about Stable Diffusion and art there are usually two different aspects of art. I am not going to try to define art, but sometimes we refer to art as illustration, or stock images (what SD puts in danger) and some as broad modern art. I am no trying to say that one is more valuable than the other, but want to qualify these…
> In the modern art interpretation artists will not lose to SD. SD will enable them to do different things. Indeed. There was a time in the past ages in which one had to wait for a certain dye to arrive in your local merchant, take it and mix it with another dye in order to be able to paint a color that you need painted in your painting. They also waited and scoured for brushes and other supplies. Those tools and sup…
Not so fast. To me this is the part that people are glossing over a bit too quickly in those discussions.
How does the person making this know it's actually good art? How do they pick between the 40,000,000 variations of that composition they're capable of generating?
Wouldn't they have to be familiar with principles of composition, color, dynamics, and so on?
As it becomes easier to generate the art, it becomes more important to be capable of differentiating it (except for purely "industrial" production where "good enough" will do).
And people will need to refine their taste to extreme levels to get there.
It's the same with poetry. What's the AI worth if it's supposed to generate Shakespeare-level writing, but no one is capable of assessing its quality?
Does it mean art essentially becomes a purely curation, taste-based endeavor?
Just throwing some random questions out there.
Re: The AI Unbundling
#188Earlier quoted context omitted.
If it’s a 10x force multiplier, then one person could do the work of 10. There are two possible outcomes: 1) the lower rate per piece enabled by higher productivity unlocks a much larger market for commissioned work. No one has to lose their job, people make up the price difference in volume. 2) there is not a much larger market waiting for cheaper art. The existing pool of artists is too big, and some of them need t…
AI is not a "force multiplier" in any literal or semi-literal way. Having an mediocre version of an image isn't something that gives an artist much of a leg up on producing a good version. If you want something a diffusion model can produce, you can get it for nearly free. If you want something significantly better, you have to pay an artist significant money for it.
Re: The AI Unbundling
#189Earlier quoted context omitted.
Why would I want that? Even if I like Taylor Swift, what would be the point?
Today you're a Taylor Swift fan and you want more of her music faster than she can deliver. Maybe you make it yourself and share with friends, and it goes viral. She might even get on stage and sing it. Tomorrow, you want more Taylor Swift and you tell the computer to give it to you. It delivers with no human intervention whatsoever. You love it, but want it a bit more country - so be it. Piano and acoustic. Got it.…
Re: The AI Unbundling
#190Earlier quoted context omitted.
AI is not a "force multiplier" in any literal or semi-literal way. Having an mediocre version of an image isn't something that gives an artist much of a leg up on producing a good version. If you want something a diffusion model can produce, you can get it for nearly free. If you want something significantly better, you have to pay an artist significant money for it.
Except it is actually if you've been following the workflows which have popped up since stable diffusion's release. These models are not limited to just making a image which vaguely matches some description. They can also inpaint specific portions of an image, turn a vague sketch into a finished looking image, and blend together different images in convincing ways. A workflow for creating a high quality looking piece…
I also have questions, what is the "Generate detailed scene based on that sketch" process? Can an algorithm do this? Can a person easily do this? I'd be more impressed than with the raw pictures. Similarly with your #3.