Striking portraits, claimed to be original, were generated on Midjourney
news.artnet.com
news.artnet.com
... almost...
But! we're here, so let's dive in.
| Oftentimes, Avery provided the subject’s name and a cutesy anecdote about their life, such as Lucy, “a strong woman from the Bronx,”
sure, also a bio that is obviously made up, even if he hadn't added this:
| (Fictional story. Let us know in the comments if you like a tale in the caption.)
which the post includes, right after making that statement
... then there's this blatant gem:
| among the myriad hashtags that accompanied each image, not one mentioned A.I. art or generative art.
.. so okay, there's no hashtag, but a number of his posts have
| Photo? AI-generated? Both?
So, to believe he's trying to hoodwink you into thinking it's just photography, when the completely obvious point is to create a discussion, you'd have to be lazy, not paying attention, or trying to sell me on a piece of ragebait. Or all of the above.
This just in: fake generated rage posted under the guise of journalism.
You know, like when you present a link as if it means something knowing practically nobody is going to click on it. like a lie.
1 https://news.artnet.com/art-world/fake-artists-collectors-in...
I think you missed this part:
> A further issue stemmed from Avery having previously denied that the images were A.I.-generated and were, in fact, taken on a Nikon D810.
People might have been "hoodwinked", to use your words, that it was photography if in fact he told them "I made this with a camera". Those people might then also be reasonably upset to learn they were lied to.
"Up until very recently, when asked, Avery was either vague about how he created the images or told people his works were actual photographs, even going so far as to describe which kind of camera he used to create them ("a Nikon D810 with 24-70mm lens"). "
https://arstechnica.com/information-technology/2023/02/viral...
Really? This is the internet, skepticism should be the default!
> I don't understand why you are so skeptical.
I am gladful some people are more skeptical than others, so I can be less skeptical most of them time.
My skepticism started growing when, as mentioned in another branch to this conversation, I clicked through the "dishonesty" link in this quote:
| The online response was varied with some criticizing Avery’s dishonesty and others acknowledging the quality of the work.
..and was presented with an article about a completely unrelated event that had nothing to do with Joe Avery. That's about the moment that I adopted a (perhaps slightly excessive) adversarial posture.
I get that there's a lot of FUD about this stuff, but news, this ain't.
https://petapixel.com/assets/uploads/2023/02/Screenshot-2023...
I'm not sure where your skepticism is coming from given that Jos Avery themselves has admitted they were lying and said they wanted to come clean.
Are you under the impression that the only time he ever communicates is through Instagram?
So is your comment. Maybe do not react so emotionally, and instead look at the facts?
Compelling stuff.
I'm sorry to report, that's apparently just someone else who likes Harvey Birdman and porn (both of which I will admit to, but I never turned my interest into a productive hobby, I'm afraid). Note the age of the account - it's a touch older than stable diffusion.
That is indeed, a helluva quote.
I'm not an "Ai 'artist'", scare quotes or no. I have an interest in the tech, and I've used it, but I wouldn't call myself an artist. Not that I think using the tool makes you not an artist, it's just not me.
My reaction is strong because of the dishonesty in the piece itself. I really hate being lied to. The article presents links that appear to be relevant to the story, but are completely unrelated. They talk about things from the Ars Technica interview without mentioning the Ars Techinca interview (from which it seems apparent that while he does use terms like come clean and most of my viewers don't know, he came to Ars of his own accord, and states his intent to start a conversation).
What I can see in this Insta account, is an attempt to start a conversation. I like conversation. I think it's important to discuss the finer points of stuff and really look at what we value and why critically. What this news piece seems to say is: "Hey! Look at this dishonest asshole using AI as if he were A real artist! Screw him! What a no-good lying sack of shit!".
It feels like it's designed to drum up sentiment, not talk about anything, but just to sell the reader on some narrative that "the evil AI army is coming for your art lunch, go get a torch and / or pitchfork".
So, perhaps a little immature of me, but I decided to respond in kind. But like better, by using references and context to elaborate rather than pretend those things don't exist. I may need to grow up a little, but that's what it is.
Where's the rage in the article?
More importantly, I wonder if I would have been fooled. The thumbnails in the article look like AI (especially the AI airbrushed hair and skin), but I saw the headline before I saw the prompt images.
I had no idea. And thinking about this now, I find what he's done brilliant. It shows in a very simple and concise way how far the generative AI has been for months now. And it shows (and proves) this to people who otherwise wouldn't be playing with any diffusion models otherwise.
In a way it blows your mind. If that's not art, I don't know what is.
I don't think one failed prediction invalidates all predictions?
We have already partially replaced artists with technology. Have a look at how photography changed painting (eg painting portraits is no longer the staple breadwinner it used to be). Or see how recorded music removed the need for each bar to have their own performer.
In the past technology just meant that we re-defined art and what it means to be an artist. We will continue to do so in the future.
Has it, though? You can make a point that photorealism in painting has been technically superseded by photography, but we still find such paintings enjoyable and impressive. Sure, there's a smaller market demand from a purely practical perspective, but photorealistic paintings are still popular, and demonstrate the artist's skill and craftsmanship.
Technology can never replace art. It can just enhance it, broaden our abilities, and give us new perspectives to enjoy it.
In the past technology just meant that we re-defined art and what it means to be an artist. We will continue to do so in the future.
Similar: cars have not completely replaced horses. Horses still exist. People just have a very different relationship with them these days, and we also have much fewer horses (especially per capita) than 100 years ago.
I also find it curious there is still so much demand for photography when pretty much every photo you could ever want has already been taken. I can buy anything from Getty images.
Have you seen the level of photography on popular photography social media over the last ten years ? It absolutely blows your mind the amount of good content in there. I’m not sure “more” content will necessarily matter?
Then there’s authenticity. A photo of a real object, person or event actually matters to people for all sorts of reasons.
I’m still going to guess people will still pay artists to paint there portraits in 500 years, and just as we geeks like to argue computers are just better at everything, I’m not sure that’s what matters to people when it comes to art.
A better guess would probably be any job that requires the dexterity, strength, versatility, and sensitivity of human hands (not just one of those, but their combination). Think electricians, or plumbers.
IDK, I think that in the AI circles the idea that this would happen was clear long before it actually became possible.
I recall a university workshop probably 20 years ago with a discussion about generative art of various forms (music, visual, etc) - as some genres of computer-generated art were a thing even back the - and the main conclusion about genres of "mainstream" art (which were far beyond our reach at the time) was that the difficult part is "judging taste"; in essence, if we got a reasonably accurate way of evaluating that one song/picture/sculpture/whatever is "better" than the other, then the rest is "just" relatively straightforward engineering. So the consensus prediction already two decades ago was that it's going to happen.
And while we don't have that "magic taste function", what we have made is an "indistinguishability" function, which can fulfil a similar role, and is effectively what the "photographer" in this story has used - if a model generates items that are (almost) indistinguishable from "good portrait photography art", however you define that, then it can make "art" that is as good as the rest of it. It can't really go beyond that, because if it accidentally generated something better then the model wouldn't be able to recognize that it's better, only that it's different, and discard that option, so improving on the status quo would require curation from someone with "good taste", but it apparently can beat mediocre photographers by matching the results of good ones.
Even when we get AGI that can truly be emotional, it still won't replace or compete with human art. There will just be more of it, and we'll learn to appreciate it in different ways.
Art is not competitive. Just as new art styles and techniques made by humans didn't replace previous ones, AI won't either. It's just another tool or outlet.
In the case of art, I wouldn't label AI-generated works to be inferior, just different. Once we're able to distinguish them from traditional works done manually by humans, it will eventually settle into its own category, with its own supply and demand. The current problem is that it can pose as something that traditionally takes effort and skill to produce, but we'll find ways around that.
Can we find a way to tell apart human art from generated art? I'm not so sure in the long run. Today, we have a lot of pointers and with enough focus experts can tell them apart. But they can't in passing, and an average person probably can't without being told what exactly to look for and advanced tools to check those. We're only a few years into it though, so I find it hard to believe that we'll be able to tell whether a digital picture was produced with a digital camera or generated by an AI. You can probably make it much more expensive to fake it (e.g. film yourself with 3 different cameras which are positioned differently, so AI would need to also fake a multi-perspective video reliably), and maybe that'll be the thing (like "organic" diamonds vs industrially produced ones), but I'm not sure if you can put the cat back into the bag.
It might be AI generated but there’s still a lot of human work gone into the images.
The sad news for me personally is taking photos and working with models and lighting people etc is fun, same for meeting people and taking their portrait, this guy got images, the hard way and didn’t seem to get the experience.
Maybe he enjoyed sifting through thousands of images and stitching them together, then editing with photo shop. I don’t know but it sounds much less interesting than photography but hey, to this guy it’s art, so good luck.
Just as electronic music didn't replace instrumental music, or photography didn't replace paintings, we'll also find ways to incorporate AI as another tool that creates different forms of art. They will all have their audiences, and will broaden our senses in different ways.
The current problem is that we can't distinguish AI-generated/assisted works from others, but eventually we'll find ways to do that too.
> I’m a photographer
Judging by your username, I'm not so sure... :-P Quick check: are you alive?
Kropotkin has something to say about the individual and the influence of others, but in short, you can not construct anything individually without resting on the shoulders of everyone that came before you.
The individual struggle is a struggle because of everyone around you, the way they behave towards you, and the way they don’t. Simultaneously if you were the last person on earth, your struggle would be the absence of everyone else.
Art is, by and large, an expression of the cumulative experiences of a person at that point in time, which come out of their interactions with everyone else. White implies black, and self implies other.
The impressionists, think Degas and Monet, went against the Salon and forged their own path, and that influenced their art. Their counter culture was by influenced by the very existence of mainstream culture.
To look the art in isolation and without context is to ignore their struggle of the individual against the broader society.
And in this view, AI art captures the cumulative experiences of all humanity, and with the right direction can show intent by an individual.
The problem is when images can be produced effortlessly by anyone to resemble anything else, which current AI tools allow. This makes them capable of posing as something that traditionally took a lot of skill and effort to generate. We need to be able to distinguish between these two, so that we can determine their relative value (both emotional and monetary). Art is tricky in that sense, since it's mostly based on personal enjoyment, but I think we can agree that things like NFTs are scams posing as high value art, while traditional paintings should be valued much more highly in every way.
Why?
Why should the nature of a work of art's creator have any bearing on its emotional impact? If the emotional impact is the same, why should the nature of the creator have any bearing on a piece's monetary value?
I suppose you can argue that part of what you feel when you look at artwork is empathy with the artist, and that empathizing with a human feels better than empathizing with software. I think that argument sucks.
I think if you felt something looking at these photos (fauxtos?) when you thought they were taken by a human, and now they make you feel something different because you know that's not true, that's not a judgement of the artwork, that's a reflection of you. I'd also argue that it's art on that merit alone.
Every new medium encounters resistance from the old guard that it’s “not real art”. This is nothing new.
I think art/performances are about what they mean to you. You can listen to a recording of a classical orchestra with a headset at home, but it's not the same experience as going to a live concert. And I don't see a handmade woodwork the same as an industrial one. Context matters a lot.
I'm certain artists will find a way to use AI, but it will not remove the need for the artist to make art. Just like I can take photos with my camera, but that does not make them art.
My point was about art. I don't think that cameras made it harder to make artistic paintings, did they?
Duchamp and Warhol would emphatically, EMPHATICALLY agree with you there.
All this is very Warhol. Andy would have loved this. I'm sure some of the surviving Pop artists are trying to think of something cool to do with it. In a way, the coolest Pop thing they can do with it is to sit back and watch literally all the world become artists (in a sense).
Everybody will be an artist for fifteen minutes. Warhol had it all figured out.
For instance, portraiture is to document people, who they are, their history, culture, etc. What’s so impressive about people that don’t exist? You’re correct.
What You're referring to reminds me of something like event photography. Photos that serve as a souvenir of an event / person / thing. Pictures that can be described as having "moment captured in time" as their raison d'etre. These are indeed unlikely to be replaced anytime soon.
But that's only a small subset of what we use photographs for. Lots of photography is about capturing an emmotion, rather than reality. Think of all the marketing (to name the biggest economic factor). If You're making an ad for watches, You don't give a damn if that cool guy wearing them is real or not. If You're a hotel or some travel agency, You are probably going to need "smiling happy couples / families around an infinity pool with a sunny beach as a background". Oh, hej, looks like that's actually a Stable Diffusion prompt...
> Every morning she wakes up at the crack of dawn and takes the subway to her job downtown. She works for a big law firm, and she knows the ins and outs of the courtroom like the back of her hand.
> But on the weekends, she's a different person. She loves to explore the streets of the Bronx with her girlfriends, and they always find something new and exciting to do.
> Today, they've decided to take a trip to the Bronx Zoo. Lucy loves animals and can't wait to see all the different species. She's especially excited to see the giraffes - she's heard they have the most beautiful eyes.
> The day passes quickly and they spend hours exploring the zoo. They laugh, they talk, and they take lots of pictures.
> When it's time to go, Lucy feels an overwhelming sense of happiness. She loves the Bronx, and she loves the people in it. She's proud to call it her home and she wouldn't want it any other way.
> Lucy takes the subway back to her apartment and smiles as she remembers the day.
Surely these descriptions are also AI generated?? I actually got angry reading this at how bland and "non-offensive" this story is, lol. Is there a word for that?
Also, which part of this made up story should Lucy hypothetically feel apologetic for (as suggested in the opening sentence)? Loving animals and the Bronx?
Milquetoast, perhaps.
"The most creative people simply to the best job of hiding the source of their creativity".
AI or not, they're good images, "deception" in art has been going on since forever.
> Lucy's Love Lucy is a strong woman from the Bronx - she is unapologetically herself and she loves it. > Every morning she wakes up at the crack of dawn and takes the subway to her job downtown. She works for a big law firm, and she knows the ins and outs of the courtroom like the back of her hand.
> But on the weekends, she's a different person. She loves to explore the streets of the Bronx with her girlfriends, and they always find something new and exciting to do.
> Today, they've decided to take a trip to the Bronx Zoo. Lucy loves animals and can't wait to see all the different species. She's especially excited to see the giraffes - she's heard they have the most beautiful eyes.
> The day passes quickly and they spend hours exploring the zoo. They laugh, they talk, and they take lots of pictures.
> When it's time to go, Lucy feels an overwhelming sense of happiness. She loves the Bronx, and she loves the people in it. She's proud to call it her home and she wouldn't want it any other way.
> Lucy takes the subway back to her apartment and smiles as she remembers the day.
It's like someone was trying to come up with the blandest, stereotypical piece of garbage of a background story for a character. Sounds like something ChatGPT could have written.
Cave painters said "You can't capture the beauty of fingerpaint as you drag our fingers against a chunk of rock. The virgin rock was porous, your paint smooths it and creating impression that lasts for centuries. You leave your mark, your identity, your fingerprint with the art. That is what makes it authentic." He added, "Painting with brushes creates a barrier with the art and the human. It can never be true art. It will never capture the painters mark and your soul in it."
---
That was a poor attempt at making a joke. I think, over the years, most people just stopped caring about the origin and process of scrollable things. You just don't have the time or energy to validate everything's source or process. You take nano-dopamine hit by seeing something, and move past that. Trying to label things to be "pure" and "authenticate" and asking the merit to be determined based on this "purity" is a thing of the past.
What I'm seeing is, he guided the construction of an evocative face, and then told an effective story about it, and THEN put it forth. It looks like he habitually adds 'fictional story' to the blurb.
The guy is a writer. If his stuff is taking off (or was), it's on the strength of his characterization, and that's perfectly legitimate. It might also be taking advantage of a hidden strength of AI: it's only us, it's only a mass-appeal take that much like ChatGPT goes for common knowledge and isn't interested in truth or novelty, and that is a way to hit an artistic mass audience with something they respond to.
The guy's a writer with a reasonably good artist's eye and the ability to see a resonant character, perhaps not a particularly deep one. There's gonna be lots more of this and one thing you'll eventually see is that these folks are as interchangeable as the AI works they produce.
If you don't like him there will be another.
There will also be people using the technology to carve out a more interesting spot for themselves that will NOT hit the mainstream, because they're not pandering half as much. But for every one of them there will be 99 folks with computers and a decent eye, who hit much bigger because their pandering is perfectly sincere. What they like is typical and common, and they'll go for the low hanging tropes and hit it really big by sincerely pursuing unambitious goals with some degree of skill and tenacity.
Worked for JK Rowling.
First, it's amazing how many people on this site have no bloody notion of what the artistic, compositional and qualitative side of real photography means. Their idea of this AI-rendered, prompt-generated stuff being the same as going out and taking real portraits while collecting their human stories is so absurdly, stupidly, blandly utilitarian that it just leaves a bad taste about the general emotional vacuity of tech types.
Secondly, the creator of this IG feed was blatantly dishonest and then later fessed up about it. If there's anything creative about that it's in the area of creative mendacity, a thing that any common grifter or crypto-bro can pull off. It hardly demonstrates much artistic talent or nobility.
With that said, it's interesting how many here defend this kind of dishonesty in this context, though they'd happily throw all kinds of shit on someone being equally dishonest in another tech context, such as in crypto or NFTs. emotional bias very visible.
Some are missing skin pores (some models are good at this now) the tiny details on the face are very very vague and blurry (see lines or pores) details blend into each other (try following some strands of hair, their origins are very vague). Eyes are usually bad with AI, this person has clearly copy pasted same eye in both eyes (exact same reflection in both).
There is a picture of old lady which is so bad (melted plastic face) that I don't know how anyone would take it for real. May be because I consciously look for details. To untrained eye (who don't know what SD or MJ is capable of) they may not look further.
He's pretty haphazard with his hashtagging - changing the style and contents over time - but his very first post has the #ai #aiart hashtags (as well as #naturephotography) and it seems to pop up every now and then
https://www.instagram.com/p/CjqQF9zuvkW/
If I were a real photographer I might get upset that my photos were being used by the tool he used. But there's so much use of filters and staged lifestyle photos on Instagram I'm not sure he really deserves the hate he's getting
https://www.vice.com/en/article/dy7b5y/artists-are-suing-ove...
That's some DnD level of character sheet...
> "[My Instagram account] has blown up to nearly 12K followers since October, more than I expected," wrote Avery when he first reached out to Ars Technica in January. "Because it is where I post AI-generated, human-finished portraits. Probably 95%+ of the followers don't realize. I'd like to come clean." (https://arstechnica.com/information-technology/2023/02/viral...)
The article goes own to say: "With Midjourney, anyone can pay a subscription fee for the privilege of generating art from text-based descriptions"
Google says 1k followers costs between $12-$30. Getting bots to follow you for three months seems pretty cheap... could this have been some kind of weird self-promotion or an ad for the AI? He felt so guilty that his first thought was to run to the media? This whole thing just seems off to me somehow...
Maybe I'm being too skeptical, but it seems like instagram is worse than other social media sites when it comes to shameless self-promotion, "influencers" who are basically undisclosed paid advertisers, fake accounts scamming people etc.
Cue sunshine telling me how all photographic images have post-shot interpretative processing(?).
Grayscale is not very realistic, is it? Yet we adapt our perception and agree that it captures some reality.
One can use the tools to create a representation of the scene so that observers would agree this is what they saw. And one can use the tools to create a new scene from pieces.
I like to take pictures of scenery with the moon in it. I have no problem using different contrast settings for the moon. But I wouldn't allow myself to move the moon around in the image. Why not?
When we use computers to create faces combining millions of shots, it can be argued that this is photography too. (Though I do struggle to use the term here.) The deception comes in only when we present the scene as having existed on its own.
LOL...
- a photographer needs to find the right model, it just can't 3D print it.
- the model needs the right clothes, you just can't 3D print that
- the model needs the right make up
- the model needs the right haircut.
- the photographer needs to make the model take the pose he envisioned.
- the photographer needs to chose the right spot to take a picture
- the photographer needs to take that picture at the right time of a day
- the photographer needs to chose the right lenses
- the photographer needs to get the right lighting equipment, place the lights correctly and so on and so forth.
and I'm sparing you the details of all the steps in post production...
in the meantime, a magical prompt did generate 10,000 images as your alleged hard working "prompt engineer" simply had to enter a few sentences in a text area or just copy paste another prompt with slight changes, zero photograph experience needed.
So no, being a photographer is much harder (and expensive) than using midjourney.
Like in development, what's hard in portraiting is not pressing the button; it's spending weeks to find the model, dress the model, groom the model, do the model make-up, find a good spot, find a good lighting, find a good angle, chose the best compromise in all the settings, and so on.
It takes a lot of time to find the right model and settings with AI.
1. Use a python script to iterate over model,seeds and prompt (pos/neg) variations 2. Use the resulting image in img2img or ControlNet for further processing 3. Use photoshop for finalizations.
that's the _entirety_ of your analysis on the difficulty of taking photos? the duration of light hitting the sensor?
> write a reply to the following comment, in a style of a stuck up hacker news commenter. Ridicule the idea that photography is easier than creating AI art. Point out how much effort a photographer needs to put in: (your comment)
Result:
Oh dear, it seems like we have another individual who's completely missed the point. Let me break it down for you, dear commenter. Photography isn't simply about "spending milliseconds taking photos." It requires years of practice and honing one's skills to capture the perfect shot, understanding of lighting, composition, and an eye for detail that can only be developed over time.
Meanwhile, creating AI-generated art with Midjourney or any other software is just a matter of feeding the right data and parameters to the machine. It may require some technical knowledge, but it certainly doesn't require the same level of artistic skill and intuition that photographers need to possess.
So, if you really think that copying text from PromptHero or Lexica and pasting it into Discord is the equivalent of the amount of effort and artistic vision that photographers put into their work, then I'm afraid you've got another thing coming. Perhaps it's time to put down your keyboard and pick up a camera, so you can truly appreciate the art and effort that goes into photography.
---
I can't wait for the top story next week where one of us comes clean as not actually being an expert programmer and a successful entrepreneur!
It took me 3 tries to "engineer" a prompt that'd produce a sensible reply to this, so one could argue that it is indeed harder than, say, understanding sarcasm on the first try.
Had he mentioned that they were AI generated, he might not have got the exposure (since anyone could do it).
This would have been fine if it was a researcher (not a photographer) who was doing this.
However it is not soul it is egoism. The artist spent thousands of hours developing a style to make it seem like they are special. To someone with real intelligence and creativity this level of toil is unbearable and there is no motivation to do it simply to satisfy their ego because a real artist has no ego, they hack around for the fun of it.
In reality , the end product is always just paint on a canvas that has specific patterns and style applied to it and this can be replicated and improved upon by an AI Artist. The machine simply automates away the middle man, pointless egoism and the toil.
You do you, by all means fight for your right to be accepted and loved as the creative person who just randomly does whatever and it's just as good. Be natural! By all means. At the end of the day you've gotta be what you are and the integrity of that matters.
Without the toil, you ain't me.
Ego is a malignant master. As a rule the folks who rise so high accomplishing such elaborate, auteurist things are MISERABLE, just awful humans. They're driven in a way that doesn't exempt them from the consequence: they are as awful to themselves as they are to others and the prizes they produce are some compensation for what they've been through. Of course they'll flip out when you come around, mister natural, and say that you deserve equal cred—more!—for just hacking around and having fun.
They needn't flip out. Neither you nor the AI are going to be them. It's you who are the middle man, in pointless egolessness.
Take heart in this: you're closer to the mainstream. If you add enough ego and toil you can get to where NOBODY 'gets you' and die in weird obscurity. Which is kind of fun too :) but there will always be more people like you, not very special, able to reach the mass appeal more sincerely and easily.
Except now, AI is better at that than you will ever be ;)
Deckard : Is this a real snake?
Zhora : Of course it's not real. You think I would be working in a place like this if I could afford a real snake?