And I can tell the difference between tube amps, profilers and modellers in both feel and tone, but it's good enough for the average Joe that they won't care if they like the image or sound. It only serves to illustrate the point that trying to evaluate the best x work is futile because tastes vary.
Personally, I don't like the image, nor do I like any AI-generated "art" thus far. I'm a big camera wonk and a tone chaser, so I appreciate detail and that elusive quality that makes something more than just competent.
Camera wonking is great, but if you don’t like any AI art, that suggests you are rejecting the concept out of hand.
Being so low-barrier, there is naturally a flood of images from beginners who are just futzing around. But, I hear a common refrain that sounds to me like “DeviantART.com has millions of sketches from bored 13 year olds. Therefore I don’t consider pencil to be an artistic medium.”
It's not so much a rejection of concept, I just find that it does nothing for me emotionally or intellectually.
I had a scroll through the Twitter profile, and whilst at least unusual in terms of content, I close the app and remember none of it. None of it is bold enough to be memorable compositionally for me. It's competent, sure, it's just unremarkable and lacks humanity. Does that mean it's "bad"? It's just not for me, in the same way I don't care for tuna even though I acknowledge that plenty of people enjoy eating tuna.
That's giving far too much credit to judges in the first place. Consider art critics as a whole, they have been proven to be cork sniffing blowhards many times over with them believing that paintings done by elephants, monkeys and toddlers are somehow masterpieces and that the artist had fine motor control and clearly put a lot of thought and emotion into their work.
Art is subjective. If you can accept that art is subjective then quantification of the quality and value of art is absurd.
Those who can, do, those who can't—judge photo contests?
All it tells you about is the taste of the judging panel.
I agree, but contests within the context of art are absurd. They're as surreal as the music charts where because boy band x sells y units, that somehow makes them "better" than z obscure artist making microtonal music that's never been heard before. The charts are a measure of commercial success.
A contest is a measure of what? The taste of the judges?
As much as AI has trouble with hands and we can assume that it’s AI because we know what to look for, those hands are not outside the realm of human deformities and/or photography tricks. Pair that with the absolute PR nightmare possible in calling a real human fake just because they don’t fit the expected norms and it’s totally reasonable for judges to accept it as a real photo.
The hands in the photograph are immediately noticed (e.g. the smudged, mushy look of the hands). I wonder: why did the judges not noticed the hands in the photo?
Yeah thats a 5 second exercise for anybody who ever came to contact with ie Midjourney output, fingers here have various usual bugs.
Honestly I believe author did include this on purpose, since its very well known shortcoming at this point. He could have done image without fingers displayed and then nobody would have a clue. But clearly judges are oblivious to recent development in some too-old-for-this-shit fashion, so they didnt pick it up and then embarassed themselves.
Could be that they assumed it must be a photo and the smudge was done intentionally as an artistic expression? The art world is so out of touch now I wouldn't be surprised if that was the case.
My guess is that they focused on the focal point of the image - the face, and especially the eyes - and didn't try to pick apart the whole image.
The white smudges in the center/top, and the yellow glow on the top of the left side seem like artifacts of a manual development process. That's likely something that would interest a photography judge - maybe the judge(s) here were just biased toward experimental analog processing and was therefore blind to the issues we immediately see?
His photo has been removed from the gallery, but here are the other finalists in the category he won. The category tends towards the abstract, so it's not super surprising that they'd confuse AI images with real ones.
It's the same thing as the Loebner Prize 'victory'. You take a highly subpar AI output, and spin it as some sort of deliberately degraded output (such as a ESL child supposedly taking the Turing Test); humans charitably overlook the flaws and assume it's the style or medium; then you ding them on supposedly not noticing the flaws.
My initial reaction was "great, so this is a thing now" but then I opened the article:
> "I applied as a cheeky monkey, to find out, if the comeptitions (sic) are prepared for AI images to enter. They are not."
I withdraw my heavy sigh.
---
This did get me thinking though -- could camera manufacturers add a new tag to the EXIF data that is cryptographically signed by the camera? This could then be carried over from the RAW format to whichever format the photographer chooses to export as after their touch-ups. This could be verified by contest officiants.
> could camera manufacturers add a new tag to the EXIF data that is cryptographically signed by the camera
They could, and some already do, but this doesn't fully solve the problem. You'd still be left with the analog loophole[1]. A sufficiently advanced attacker could feed a generated image directly to the camera sensor.
Also, I think every one of these has been broken. There have been instances of images being entirely replaced by historical images, without breaking the signature.
For a long time, polaroids have been held as the "this image can't be faked."
Somewhere in my collection of lesser used equipment, I've got a Daylab. https://www.instantphoto.eu/pola/pola_daylab_300.htm and I used it to transfer some 35mm slides to 4x5" to practice the various manipulations you can do there (emulsion lifts and transfers).
The thing was I had a 35mm slide that was projected onto the film. If I could make a 35mm slide of some sort, I could then make a polaroid of a digital image. Even back then there were a few services that could do it.
The idea that you can ensure that the image produced is what was recorded, even two decades ago when polaroids were still a reasonable standard for authenticity of an image shows how easily it can be compromised. Well, maybe not easily but it isn't difficult for a serious hobbyist to create the appropriately constructed image in the desired media.
And so what if you sign the EXIF data. That only really guarantees that version of the image. If you adjust the saturation or curves or exposure... or correct for some lens aberration does that EXIF information still match the resulting photo? For that matter, what if you crop it to a square from a rectangle or stitch two images into a panorama?
> And so what if you sign the EXIF data. That only really guarantees that version of the image. If you adjust the saturation or curves or exposure... or correct for some lens aberration does that EXIF information still match the resulting photo? For that matter, what if you crop it to a square from a rectangle or stitch two images into a panorama?
Oh, that's actually possible with cryptographic techniques. Very exciting!
Those are interesting, but there's a difference in the manipulations that are done for image for journalism and image for art. That has crop, resize, and grayscale... but not things like "selectively dodge and burn", "adjust the white balance", "skew or rotate to level horizon", or "correct for chromatic and coma aberration in a particular lens."
While that is good and useful (and I would be willing to even go so far as saying "needed" for journalistic uses), the transformations available are remarkably limited for artistic use that even represent what can be done in a traditional darkroom.
> what if you sign the EXIF data. That only really guarantees that version of the image. If you adjust the saturation or curves or exposure... or correct for some lens aberration does that EXIF information still match the resulting photo? For that matter, what if you crop it to a square from a rectangle or stitch two images into a panorama?
While you won't be able to verify the final work mechanically by just verifying the signature, at least, if you bundle the original raw file with the final work, an human (or an AI!) should be able to tell if the final work is at least derived from a real picture.
You just have to dive deeper. You need two cameras that have cryptographic signatures automatically attached.
In order to submit picture 1, you need to have picture 2 that was taken at the exact same time (and time is in the signature too) by a different camera, of picture 1 being taken, as a verification picture.
That does no good. You can present any image to a camera you want. You can't solve that problem by requiring another camera also being fed an arbitrary image.
The problem is that you want to ensure that an image corresponds to some real object. This problem is unsolvable, because a camera does not sense "real objects". It senses photons, and we already have the technology to shine any arbitrary combination of photons that a camera sensor can detect onto a camera sensor. There is no possible "signature" to apply to the output of a camera sensor that can validate that it is the result of "real objects", a definition that would get fuzzy if you tried to really nail it down in the presence of hostile attackers deliberately gaming your definition anyhow.
Besides, the entire idea that a photo is a concrete, specific thing is a very Hacker News, computer-programmer idea anyhow. Photos aren't just files. Long before we had the tech to shine arbitrary photons onto CCD camera sensors, photos were a process, not a single file. Even if we stipulate that the image on film can be reasonably called a "file", itself a rather large stipulation, the final photo would be the result of substantial decisions in how the image was developed. The program "Photoshop" is called that precisely because "photoshops" had numerous tools to affect the image at that point in the process, many of the tools in Photoshop are still named after these processes. Photo competitions absolutely include these elements; if there have ever been any photo competitions that accepted film as the input and rigidly developed all contestant's films the exact same way, they are the vanishing minority. Photos have always encompassed post-processing as part of their identity. A signature on a file is no good. You'd have to fundamentally rewrite the entire photo stack to include all the transforms, in a completely official and 100% specified manner, all signed, so the final signed photo has the complete record of everything done to it from the source... it's theoretically possible but absolutely not going to happen, especially in light of the fact the source is still meaningless for the above reasons. The whole idea is comprehensively unworkable.
Even with this elaborate setup, you still can't prove that the person taking the photo is the one that set up the composition or camera settings. I think you'll need a third camera recording a video of the setup process. Also cryptographically signed, of course.
When Canon did this, the signing key was the same for every camera of the same model, and you could extract it by running code on the camera. I don't have faith in camera companies implementing this properly.
Hence the “if”. The technology exists, and is for example in widespread use in smart cards, smartphones, and laptops. The article you link to is also from 2010, so the Canon development is probably 15 years old or more. I would expect a better level of security awareness by now.
You don't even have to carry it over, just require the original signed file in a competition. Though whenever you look at security you always have to consider what happens when it gets cracked.
Where do you draw the line between heavy editing and fully AI-generated? Some edits have little in common with the source, so at which point is the source not relevant anymore? What if you feed an original photo through some AI-enhanced tool like Topaz?
I had seen Dogme ages ago, thanks for reminding me. Agree we may see more like it, but probably with about the same cultural relevance : mostly a footnote.
And how do you know the signing keys are legitimate? Are they in a hardware root of trust for a high security, completely locked down device that can’t be rooted to sign whatever? Is the CA that endorsed them trustworthy?
Where do we draw the line? iPhone photos go through ml models should they not count either?
Just watching where the tech is heading I’m not going to be surprised if almost all creative tools in 10 years go through this tech so the whole idea that we can cordon off “pure and honest art” seems impossible on the long timeline.
I think it needs the actual photographic data which is unique to be encoded with the camera model perhaps? Tagging is not enough, it needs to be part of the image to really work (I am guessing)
Some cameras do this already, and there is some interesting research (2022) that tries to use zero-knowledge proofs to make signed images editable (e.g. cropping, simple adjustments).
First, the image sucks. Do you really get any artistic whatever from that image? At best, it looks like a badly photoshopped image that struggles to simulates 80's look and feel. It also has AI-specific dirtiness. I really wonder how many people can legitimately argue for this image. Perhaps this may mark that the competition is somewhat irrelevant to the taste of general public anyway, and this kinda relates to the small trolling impact he made.
Second, the competition already accepts photoshopped images[1], as long as the scope of modification is clearly outlined in description. On top of that, the specific category he competed is more open to heavily photoshopped images in the first place. It's likely that images modified using AI will be allowed to some extent. (i.e. insert/remove items, minor fix to the posture, etc).
I mean, if he were such a pioneer, he could've just based his work on smartphone photos, and provide fully honest description and still win. That could've been a real history. This? It's just a dead copycat. ¯\_(ツ)_/¯
Further down the article there is an update that they removed the image from the competition after he revealed it was AI generated.
Edit: actually, at the top it's mentioned:
>>>>>>>>>>>UPDATE 14.3.23 / 10 AM CET
As photolari.com reports, my image and name have now been taken off the SWPA webpage. Hope an official statement will follow.
<<<<<<<<<<<<<
> As photolari.com reports, my image and name have now been taken off the SWPA webpage. Hope an official statement will follow.
> <<<<<<<<<<<<<
I didn't see that update because I was looking at the archived page linked in the top comment, since the main site appeared to have been hugged to death.
The official contest rules are here [0]. There is actually very little in the rules restricting how the image is created; technically it doesn't even say it has to be a photograph, although that is clearly implied. The rules mostly cover usage and legal rights.
So the artist intentionally exploited a lack of clarity around AI in the rules to gain publicity. This seems clever in that I might not have heard about this artist or this contest otherwise, but it was a cheeky move to win a photography contest by playing a meta game.
The winner wouldn't have been my first pick against some of these. However, some have a "generated" quality to them, but that is probably the expectation of being fooled again, or heavy post.
>“The work… is the result of a complex interplay of prompt engineering, inpainting and outpainting that draws on my wealth of photographic knowledge. For me, working with AI image generators is a co-creation, in which I am the director. It is not about pressing a button – and done it is. It is about exploring the complexity of this process, starting with refining text prompts, then developing a complex workflow, and mixing various platforms and techniques. The more you create such a workflow and define parameters, the higher your creative part becomes.”
The image is fairly derivative, and doesn’t demonstrate any traits of a well crafted generative image, nor does it show signs of the creator’s ability to draw in a topic, emotion or theme, other than ‘oooh spooky old photo’.
It also doesn’t show off their knowledge of photographic and ai technique like they claim. The scratches and effects look forced, and the hands and faces just look like classic ai.
It’s okay graphic design. He even goes on to say it’s not a photograph.
If this isn’t an April fool’s joke, then Sony have seriously tarnished the reputation of the award.
It's not an April Fool's joke; it was an intentional effort to deceive the competition to highlight the fact that such competitions aren't well-prepared for the addition of AI to the artist's toolbelt.
The artist refused the award, and declared it was their intent to highlight this concern.
What efforts like this are missing is publishing a cryptographic hash before engaging in the deception. Otherwise it can seem like the author intentionally decided to mislead but might have changed their intention after winning.
Publishing something openly like e1c7aa9150766bd5279b54951e3828ed9c19367af21f03f669eb586bd54c13b1 which can later be verified once "revealed".
$ echo "I, CapableWeb, is gonna participate in Sony World Photography Award 2023 with a AI generated image in order to see if they are ready for AI generated images. 2023-04-14" | sha256sum
Don't really have to involve blockchain here (yet again), publishing on a website and letting Internet Archive archive it is good enough. Or even posting it on Twitter/HN/$social-media-with-edit-limit
It’s okay graphic design. He even goes on to say it’s not a photograph.
Well, that's the bit that I think is most odd here. It's not a photograph.
Whether if AI-generated images are fair game in art competitions in general, I guess that's up to the folks running the competitions, but I would think an AI-generated image would pretty clearly not be a valid entry for a photography contest. No more than a painting or a charcoal drawing.
What's the difference between a photo I generated with AI vs , say, one I edited using Adobe's AI-driven toolkit?
This is a great debate and the right answer here for Sony (and everyone) is to have an open dialogue with creators and consider new sorts of categories and classifications so that we may revel in the spice of life – instead of letting it undermine/destroy the indefinable nature of art!
Sony always finds itself at the forefront of these kinds of discussions because they want to eat their cake while keeping their cake: their business model is to create and sell technology to enhance and simplify the audiovisual artistic processes, but if they enhance and simplify it too much they lower the price floor and pull money out of the market... Money they want.
As the MC Lars lyric goes, "Epic's up in my face like, 'Don't steal our songs, Lars' / While Sony sells the burners that are burning CD-R's."
I am of the opinion that the process used to create this image (or any AI image) can be considered an art form. For the author to correctly find the language and prompts needed for an AI to generate this image is a new type of art-form, from my point of view.
It doesn't mean, of course, that we shouldn't be able to understand the origins of the image and correctly categorize it. Contests like this will definitely have to grapple with their handling of generated images, just like they've had to decide rules about photoshopped images.
"AI images and photography should not compete with eachother in an award like this. They are different entities. AI is not photography. Therefore I will not accept the award"
That is, the author of this dirty trick knew this very well from the very beginning, but nevertheless he submitted to the competition not a real shot (which is a unique stopped flat moment of reality), but a synthetic picture of which he is not the author.
And after all this, the author of this trick gracefully declined the award.
Someday such a logic of such bad fools will destroy this world.
Too many people happily go along with the status quo, despite correctly recognizing something should be changed. It takes energy, courage, and sacrifice to create change. A body at rest will remain at rest unless acted up on by an outside force.
In this case, this author correctly recognized that its better to force the conversation now, himself, than let things play themselves out naturally.
Excuse me, but what kind of "changes" are you talking about?
Professionalism is primarily built on trust, including a mutual trust in professional systems. Therefore, students are not allowed to cheat on exams, for example (not only for this reason, of course, but also for this reason).
If you think that the destruction of the system of mutual trust is some kind of "change", then I want to upset you, because such "changes" do not require software systems on the scale of modern "AI" at all.
By this act nothing but destroying the mutual trust was done. Now the organizers of professional photography competitions will not trust photographers, and photographers will not trust the organizers of competitions.
I appreciate your idealism, but the unfortunate reality is that we cannot simply hope people will do the right thing and it will be so.
Now the organizers of professional photography competitions will not trust photographers, and photographers will not trust the organizers of competitions.
The entire point of this exercise was to point out exactly that organizers SHOULD NOT trust photographers. That these systems must be built without trust in mind, or the competitions are quickly going to become worthless.
> Now the organizers of professional photography competitions will not trust photographers, and photographers will not trust the organizers of competitions.
"organizers of professional photography competitions will not trust photographers"
But this is already present. There were plagiarism or alleged plagiarism scandals [1][2][3]. There were cases of doctored photographs [4][5], or misinformation [6]
Most competitions use means to protect against this to some extent [7]
Lots of comments here saying how much they dislike the picture itself, so I'm just putting in my two cents as a counter weight because I actually do like the picture.
Can tell it's AI generated from the weird looking pinky on right hand side of the photo, looks like it tried to do a thumb there instead. Stellar faces and eyes though.
Credit to the artist: for the overall effect, the AI conceptual error artifacts enhance the effect. The mind picks up that "something's off" but may not immediately catch which details are the problem.
> Image 5 of 43
"Pseudomnesia | The Electrician" by Boris Eldagsen of Germany; open competition shortlist for creative category, Sony World Photography Awards 2023. (Boris Eldagsen, Sony World Photography Awards 2023)
Actually yeah, the only place I'm seeing this reported is the linked website I've never heard of before, you'd think it'd be a little more widely covered.
