Nikon photography contest winner loses title after scientists spot signs of AI | Hacker News Reader

Nikon photography contest winner loses title after scientists spot signs of AI

techspot.com

145 pointsthunderbong88 comments

Why everything has to be touched by AI? I really don't get it. Not long ago a canadian writer was also in the center of a storm accused of using AI in a novel that won a prize.

Why do people fall for it so easily? Is it just laziness? Rush to finish things? We are so much better at creating new things that using the tech feels like cheating to ourselves.

Human body, esp. brain (or more broadly even nature) has a motivation to do more with less energy expenditure.

Drugs allows you to cope with hard things or allows you to escape the toll you have to pay.

Similarly, AI allows brain to do more with less energy expenditure. This is why it looks so tasty and is very addictive.

Your brain doesn't work at that hard, AI allows this, however there's no free lunch on that planet. The effort you give into something returns as quality and finesse, however it's tiring as well.

Addiction is a complex is a hard problem regardless what you are addicted to, it's not a light or easy matter to discuss, though.

OK, but not only that. People learn quickly that in most contexts, expending effort to create a high quality work is not noticed much less rewarded.

It's like anything goes in most human endeavors. Society is winging it; duct-taped solutions abound. Most people lack taste or a keen eye for quality, and only notice when there is something seriously wrong. Even then, politics and all sorts of other concerns skews how people respond to things. This is why we complain about AI slop in the fields we know about, while the public at large think AI is the best thing after sliced bread.

Faced with this, unless someone is really passionate about their work, they eventually kind of tune out, and then look for how to expend the least effort possible to get whatever (uncertain or politicized) reward they can in their endeavors.

Oh yeah, I'm not going to lie, once I started using ComfyUI with FIBO2 and Qwen Image 2.1 to restore my photographs, I was hook! I have 25+ years of digital and film photos that I've been restoring due to my older camera/lens quality. The addiction of restoring these older photos and then processing them once again in either Lightroom or Photoshop. I caught myself spending countless hours revisiting the past and restoring it to what I remember. I love making sure it is perfect to what I last recall it to be. There is no going back for me, these diffusion models are bring live to forgotten images and for that I'm thankful.

Why do we use a shovel instead of our hands? Why do we use cars instead of walking? Why do we use libraries instead of writing all the functions ourselves?

The issue here is you're trying to make something a binary that is a continuum. Humans are tool users and the vast majority of animals depend wholly on their bodies to accomplish everything. It's our greatest strength and weakness.

>We are so much better at creating new things

Are we? If this were really true we'd have had rockets at least 10 thousand years ago if not more. Actual 'new' creations that are not mixes of previous things are incredibly rare and most of those are found by shear luck with interacting with the natural world. We love telling ourselves a particular story about how creative and wonderful and fucking great we are but the vast majority of human behavior, including my own is just an iterative copy of other humans behavior.

In the do stuff with you hand competition, using a shovel is cheating and in a walking contest, using a car is cheating. Very simple concept.

Cheating is an emergent high order effect, not some ground rule of the universe. When 'cheating' becomes good enough it's just the cost of entry. Just look at the Olympics these days. A huge number of the people involved in different fields of it are doing all kinds of questionable medical procedures that are difficult to near impossible to detect at the time of competition.

Wouldn't be surprised if in a few decades that we have to X-ray the participants in hand competitions for robotic enhancements.

You can rationalize anything if you ignore your responsibilities to other people. Ultimately that’s just a long way around to being a crappy person.

just a meta comment about the comment you replied to.

I am consistently surprised by the levels of self-righteous hubris in the techbro community. People rationalising horrible behaviour and perpetuating the tragedy of the commons, just with computers.

I'm old enough for nostalgia so i regularly find myself pining for the days when the software industry was smaller and the nerds arena was not taken over by the MBA types and temporarily embarrassed millionaires.

It's worth remembering that for all of its strengths, this place is also first and foremost a forum for people chasing VC money and support. Name aside that is the essential draw, and that's why it's so much bigger and more active than alternative sites or more specialized technical communities. In my experience those smaller, more... well places with higher standards for actual technical knowledge tend to be a lot less tolerant of the MBA poseurs and the LessWrong/SlateStar/EA/etc cults.

But a photography contest is not a “take a picture with your camera” contest. Post processing dictates a decent portion of “photography” and it seems like the rules for this processing are a bit of a grey area. I personally hate where it has gotten to, as a photographer, but just pointing out my observations in the assessment.

> But a photography contest is not a “take a picture with your camera” contest.

Of course it is! You post process a picture you took with your camera. As a photographer, can you open paint/photoshop and create a picture out of nothing and submit it to this contest without it being considered cheating?

That's not what appears to have happened. It seems that AI was used as part of the post processing step.

It made up important details. That is why it was disqualified, AI or not.

> Are we? If this were really true we'd have had rockets at least 10 thousand years ago if not more...

Yes it is my personal belief that we are definitely much much better at creating than AI models, which the only thing they do is output an empty and dull regurgitation of its training. There's nothing truly personal and original coming out of it when it comes to creating music, literature, art, etc.

Do you know why I am so against all this hype? Because it is a poison for the mind, it deprives you from thinking from understanding and from questioning things by ourselves, and I'm already witnessing AI junkies that consume and take for granted every single output from a model.

We have already proven that most social networks are toxic and detrimental to people, especially teenagers. On top of that now add no thinking for ourselves anymore.

Why think if you can now outsource it to a computer, right?

People who aren't thinking with LLM's, probably weren't doing a great deal of it pre-LLM

> Yes it is my personal belief that we are definitely much much better at creating than AI models, which the only thing they do is output an empty and dull regurgitation of its training. There's nothing truly personal and original coming out of it when it comes to creating music, literature, art, etc.

I guess those AI models were trained on solutions to Navier-Stokes and a proof of the Quasi-Riemann hypothesis.

> Do you know why I am so against all this hype? Because it is a poison for the mind, it deprives you from thinking from understanding and from questioning things by ourselves, and I'm already witnessing AI junkies that consume and take for granted every single output from a model.

This argument has been made against every human innovation, including writing. Maybe it's true this time but I doubt it.

> Why think if you can now outsource it to a computer, right?

I was already outsourcing plenty to the wikipedia, do-yourself-youtube videos, and so on. This doesn't seem that different.

I recently had ChatGPT Astra write a bunch of short stories for me in a universe I'm fond of. I shared a few of them, and for one of them I got not only wildly complimented that it was one of the best pieces of short fiction people had read in ages, people strongly encouraged me to submit it to various publications.

When I revealed it was AI written everyone immediately ate their words and said they hated it all along and it was obviously derivative.

(throwaway because when I posted this under my main account I started getting death threat emails. like, grow up.)

Do you think it’s because those people were trying to encourage YOU? Maybe, just maybe.

And when they found out you stole it and claimed it as your own, they weren’t impressed?

Jesus Christ you AI Delusion Syndrome people (I use that term loosely) have brain rot.

>And when they found out you stole it and claimed it as your own, they weren’t impressed?

This needs to be emphasized.

Basically they realized it was droppings from the theft machine when they'd been told it was original content and they felt betrayed and disgusted -reasonably so, I might add.

Well, I think a key part of it is what exactly you mean by:

> I recently had ChatGPT Astra write a bunch of short stories for me in a universe I'm fond of.

Because that way of stating it makes it pretty clear you yourself don't really see yourself as the author of the stories.

Simultaneously, you presented them to your friends as if you were the author. So basically your friends found out you BS'd them. Why wouldn't they be upset? Seems like a totally natural reaction.

Now, if you performed extensive labor to get the story into the shape it it's in, hours long conversations with the model, some manual tweaks, etc etc, maybe you have a better claim to authorship than you would if you used a one shot prompt (though your own framing makes it sound like you did very little)

Using Navier-Stokes doesnt seem like a good choice given the controversy around it.

Well, there are tons of other things newly proven any LLMs, as well as lots of other things once termed creative and the domain of humans only, that LLMs and other models have surpassed the "creative" humans, that all you have to do is find one to realize that humans are not the only creative force on earth.

Something like the forest and trees applies here.

As far as I know, the controversy involves a pair of mathematicians using LLMs to solve the problem and another group who used LLMs to solve the problem. No one seems to be disputing the solution itself.

The controversy was that OpenAI may have stolen the work of mathematicians, which IMO is not hard to believe given they've stolen quiet alot for training data.

Right, but those mathematicians whose work might have been stolen were...using LLMs to attack the problem.

Nobody expects personality from their math proofs, everyone expects it from art. Why are you conflating the two as if they're related?

An LLM can be good at math and shit at art. And in fact, they all are

>AI models, which the only thing they do is output an empty and dull regurgitation of its training.

This just is such repeated drivel, a thought terminating cliche.

Remember over a decade ago when AI trained itself to whip human players at Go using moves no human ever did. Because I do.

Again, you keep using examples that are decidedly not like the examples under discussion: artistic material (actually, the article scenario is even further constrained in that it's supposed to be footage of real physical stuff not some fabricated nonsense, so yeah, no duh using generated content is a no go).

The difference between real and artificial is called error, things like a generative adversarial network is developed for for the exact purpose of reducing error. Eventually you'll get a generator that mimics reality.

That's an ahistorical oversimplification that acts as if time and provenance are not a thing.

We operate within history whether you like it or not. This problematizes any abstraction that is conceived in terms of ahistorical static information and difference.

If I stimulate your neurons with a couple of electrical charges you will experience a "real" phenomenon which isn't real by any classical sense of the definition because the origins of that phenomenon are not "natural".

Why don’t we allow printers in a painter competition?

Why don’t we allow scooters in 100m sprints in track and field events?

Over 100 metres, a typical athlete can probably run faster than they could scoot.

true but even if we allowed scooters, if top sports athletes don't have to wreck their bodies before their 30th, what's the point

You wouldn’t use a typical scooter

https://bo.world/blogs/newsroom/insane-100-mph-e-turbo-scoot...

I fully accept that AI is a good tool.

But a damned shovel cannot create the dirt it digs.

> Why do we use a shovel instead of our hands?

AI is not a shovel. A shovel digs faster than hands, but it is still digging in the ground and cannot unearth anything that wasn't there.

Generative AI tooling creates things that weren't there. It is not deterministic and not compatible with the spirit or rules of creative competition.

And when you hire someone to do all the work for you with near-zero involvement from you, do you claim “you” did the work? Do “you” win awards for it?

Jesus Christ you AI Delusion Syndrome people are exhausting.

I think your response misses pretty important distinctions between art and practical pursuits.

People get upset when other people use AI in the arts because the point is largely human expression. Technical virtuosity isn't even necessarily the most important thing in art (just look at modern art and how common the philistine refrain of "my kid could do that" is).

Art doesn't serve a utilitarian function. Using AI for it largely misses the point, and I think that's what people get upset about. It's like showing up to play tag and saying "actually I'm going to play hide and seek instead" you aren't following the established social norms of the game.

Contests, in general, always ways have restrictions around them, because they are meant to test particular human feats. I'm sure plenty of future contests will have rules against AI just as today plenty of contests arbitrarily limit technologies for the good of the competition (e.g. no doping in sports).

To respond to this situation by emphasizing the utilitarian function and value of AI is to completely miss the point.

>distinctions between art and practical pursuits.

And at the same time many of these distinctions are artificial human constructs.

The idea of these photography contests is nearly dead, if not dead. Much like the no doping in sports, once the technology gets too good, for example replicating the exact processes of your body the error rate on detection gets so high you might as well flip a coin. When making rules for any system you can always have adversaries. In contests you have two types of cheating you have to deal with. Cheaters that want to win, and cheaters that want to show you your system is broken and can't be fixed. The second type is the kind that shows you that your rules are arbitrary little abstractions in the shadows on the wall of Plato's cave.

Id say this situation is more like watching someone else use a shovel and then claiming that I dug the hole

Counterpoint: we've been making beautiful art and prose for millennia without machines doing it for us. And no, typewriters and word processors and paintbrushes obviously aren't the same class of tool as an LLM.

LLMs are truly worse at creative processes and I'm sorry you're not exposed to enough beauty in your life to recognize that

Wait until you hear about expensive lawyers and judges... https://www.damiencharlotin.com/hallucinations/

Oh man, I love using AI to retouch my photographs. I shoot with a Canon 5D Mark I (OG), which has a faulty sensor. Since I can’t afford an upgrade, I rely on vintage glass: Pentax SMC 50mm and 35mm lenses adapted from M42 to EOS. With FIBO2 and Qwen Image 2.1, I can transform my raw shots so they look as though they were captured with a high-end digital camera. I then pass them through Lightroom for fine adjustments, add some Photoshop magic, and bam—the result is exactly what I want. I’ve been photographing as a hobby for a little over 25 years now.

Going forward, I’ll be using local diffusion models exclusively, and I couldn’t care less what others think. This workflow suits me perfectly, and I genuinely enjoy the process. I have over 100,000 digital and film photos I've slowly been processing using this method - I have restored photos I tried to restore the older ways without success. I'm so grateful for these local models.

That sounds hella interesting! Do you have any examples or pointers on how to try myself?

Here are some of my prompts. https://old.reddit.com/r/StableDiffusion/comments/1xb59yw/fi...

I use the Portable version of ComfyUI, most are done with Qwen Image 2.1 and some hard restores using FIBO2. You can find a bunch of video tutorials on YT for both models.

Yes, it's just laziness. The modal person who goes "look what I made with AI!" is a person who never actually put in the effort to make things before. There are a handful of people who actually have skills and are trying to use AI to get better, but they are very much in the minority. And even they aren't actually getting better, they are simply deceived by the hype.

I still don't think the disqualified photographer was intentionally cheating. He was pretty open about his process, and Nikon took some time to determine whether it was within the rules or not. https://arstechnica.com/science/2026/10/winning-nikon-small-...

Edit: an earlier statement from Nikon said that he was complying with the investigation and "At this time, we don't see that any rules were violated" https://www.bbc.com/news/articles/ck4gjn1yzprno

Did he ever share the raw source?

I don't think that's an honest summary of what happened. It seems like every other scientist was pointing out aspects of the submission that were clearly generated by AI. things appearing out of nowhere or behaviors and appearances that don't match reality at all. it's pretty clear AI was not used purely to post process b/w source images. Or at the very least, was not used responsibly. Whether or not he intended to cheat, the lack of discipline is pretty egregious

What you see as evidence of a weak case I see as evidence of the judges being responsibly cautious, and affording the photographer every reasonable opportunity to make their case. I would need to see additional evidence before I took a statement beginning "at this time..." as stronger evidence than something subsequently stated by the same party.

As it happens, the final decision is not that the disqualified photographer was intentionally cheating, but that the video was ineligible for the competition.

They were giving him the opportunity to withdraw the submission.

Since Qwen Image Edit 2509 camera out, I was able to restore some of my earlier photographs that my camera at the time just failed to produce what I envisioned. Due to poor sensor and quality glass. To this day I still after 25 years shoot on a Canon 20D and 5D with old Japanese Asahi Takumar SMC lenses(M42 to EOS), which is all I can afford. I have over 100,000+ photographs that I have slowly started to restore thanks to these diffusion models. And now that FIBO2 and Qwen Image 2.1 are out, OMG! I can restore images I've spent hours on to restore in seconds! I still run them through Lightroom and Photoshop. They are comparable to a high-end digital camera. I'll never stop using them and can keep using my same old glass for years to come. I don't care what anyone says, they are an expression of what I see as a photographer and artist.

I've also started using Flux.3 Klein for extra added details. Local diffusion models are so great.

The subject of the thread is a competition of skill with rules.

This is the second time that I had to read about your off-topic personal projects where it’s utterly irrelevant that you use AI or not. Add an off-topic tag to your messages at least.

Understood. No need to be rude about it. I will ensure that all my post when I try and relate are marked accordingly. VAV VGX.

> I have over 100,000+ photographs that I have slowly started to restore thanks to these diffusion models.

Restore ? Are you sure ? Watch out for people with 3 hands in your "photographs"

Off topic: At one point it was 150k, unfortunately this was at a time before I learned about backing up. Lost from 2006 to 2007 due to a spin disk failure. Roughly 50k photos of scans and digital taken.

Anyway, if you know of VAV VGX then welcome thus the three hands remark.

I love photography and I'm intrigued by what you are doing with this. I'm not annoyed that you are talking about photography on this post. I think there is room for judicious use of AI in photography. The fact that it was disclosed and still accepted means they were at least curious.

14 days ago 38 points, 64 comments (though some threads were... questionable.)

https://news.ycombinator.com/item?id=50038640

Photography in general is in a weird place right now. AI-assisted autofocus, subject tracking, zoom/upscaling, exposure, white balance, noise reduction, color correction, HDR and lots more that all cameras have now is considered fine. Heavy AI-assisted editing in tools like Lightroom is considered fine. But the second an LLM gets involved for even the tiniest detail suddenly everyone is in agreement that "this is AI!!"

I feel like the distinction comes down to "edited" vs "generated", but even here the line can be blurry. For example, is AI denoise really just removing noise from an image, or is it generating a new image of what it thinks the original signal was?

> is AI denoise really just removing noise from an image...

This is why I like Darktable's profiled denoiser. It works by using a noise profile measured with a deterministic process using the camera you're targeting. In short, it's tailored to your sensor by profiling your sensor.

ComfyUI with either FIBO2 or Qwen Image 2.1 can do wonders for denoising. You can refine sharpness as well as make the image RAW quality for editing. My 3060 can do a 2k image in 40 seconds. Then process through Lightroom and Photoshop.

These 2 diffusion models are amazing.

Thanks for the tip. I'd look into them, but for processing photos, I'm not sure I'm going to use a tool which might add information was not there to begin with.

Think of those Samsung moons? That was really just creating new information on demand for a while. I think it looks pretty tacky too.

> AI-assisted autofocus, subject tracking...

This is already present for the last decade. It's the foundation which makes "sports DSLRs" possible. AI here is not what we know as today's "AI". These are specialized models baked into DSPs to run at inhuman speeds. In Sony's case it's phase data, image itself, information from the lens, and sometimes the accelerometers on camera. This is not making something up from thin air.

AI assisted autofocus is smarter phase detect autofocus. Subject tracking is an educated guess for where the focus shall move, based on movements in the image.

My ~20 year old Nikon D70s has a 1004 pixel "scene detection and metering" sensor "trained" on thousands of images to make educated guesses in full-frame metering mode to exposure correctly. This is in no way helping or altering the image in any way. Same for faster autofocus and object tracking.

Exposure, white balance, noise reduction, color correction, HDR et al can be all achieved with deterministic and traditional algorithms as well. Some of these processes have history in film photography and carried over to digital from analog. Some photographers are famous for their darkroom treatment of their shots during development.

Processing an image with deterministic algorithms using the data in the raw photo is fine. This is how darkrooms work as well. However, guessing things, altering images by adding something which was not there and claiming otherwise is not acceptable. This can be a new element inside the scene or pixel details which was not resolvable by the lens/sensor/film to begin with. Both are "photomanipulation", not photography.

If the point of a competition is to show a photographer's skill, it sounds like providing deblurring or similar by educated guess would help quite a bit to win?

I guess the key point is that use of features needs to be adopted slowly in competition so that it doesn't turn into a materials sport where it's just the most expensive equipment winning.

AI photo editing is surely a competition-worthy art of its own.

Edit: moot point; it's object recognition to find the auto-focus target, not applying digital refocusing.

It's not deblurring, there's no post processing. It just selects where to autofocus, and commands the lens to focus to that point, that's all.

AI assisted autofocus and object tracking doesn't alter the image after capture. It just moves your focus point and tries to keep the object optically in focus during the half-press phase.

The most expensive equipment doesn't give you any advantage in composing, exposure, etc. If you're shooting birds, you need some tracking capabilities and fast focusing lenses. These are not expensive because they post process the image, but they have expensive glass, fast motors and powerful DSPs to track that bird in that frame. If you can't frame the bird and press the shutter the right moment, you're toast.

Nobody is using relighting tricks like Nikon D-Lighting in competitions, it'll disqualified anyway. I'm not sure my Sony has similar post processing options, even.

These are shot with a D70s: https://www.flickr.com/photos/zerocoder/39910477633/, https://www.flickr.com/photos/zerocoder/41901384135/in/photo..., https://www.flickr.com/photos/zerocoder/53363865806/

This is shot with an iPhone 5, at 4AM: https://www.flickr.com/photos/zerocoder/29340231528/in/photo...

None of them are post-processed with AI tools. Heck, AI is in today was not even a dream when I processed these. Pocket Tiger has been denoised using Profiled Denoise feature of Darktable if I remember correctly. It's taken at Maximum ISO of D70s, which made it a colorful mess, recovering that image in post-processing was a miracle of its own.

> into a materials sport where it's just the most expensive equipment winning.

It's easy to take bad photos with good cameras. Also it's easy to take good photos with bad cameras, as well. It's a squarely skill sport. Equipment can't help you beyond a certain point, and you need to be able to demand that from your camera first.

> AI photo editing is surely a competition-worthy art of its own.

We have photomanipulation competitions for that.

> It's not deblurring, there's no post processing. It just selects where to autofocus, and commands the lens to focus to that point, that's all.

Ah, thanks for the clarification.

Errr… not that long ago using the zone system (during shooting, developing and printing!) to got a good negative and print. Your 20 years old D70 is doing a lot of heavy lifting you are not even aware of!

Even ~20yo lower tier digital cameras still required understanding what light meter meters and correcting for it.

I know. I cut my teeth with a Pentax MZ-50, so I have experience with understanding the camera, its light meter and how to compensate for it.

I probably shot less than 100 frames in full auto mode with that D70s, rest was apeture priority or full manual. The reason I still love my D70s is it behaves like an analog camera very much, from its light meter correction needs to ISO response of the sensor.

When compared to it, shooting with A7-III still feels like cheating sometimes. Maybe I should get the Pentax fixed or start playing with the Canonette 28 I have.

Cannonets are great! Sync ay any time due to leaf shutter if you are into strobes!

There has always been some fuzziness behind what counts as "genuine photography" and people have been gatekeeping it for decades as new camera tech became available. Generating statistically likely pixels given the context of the image and massive amounts of (untraceable, stolen) training data definitely seems like a different kind of post-processing than what you'd normally do in an editor, or even changing pixels by hand.

Nikon gets to gatekeep their competition however they like. Personally I think it would be a shame to reward photographers for the work they do outside taking the shot and light editing/compositing. What's the point of a competition if you're not going to enforce high standards and challenging constraints?

I would argue that the post processing is just as important as when the shutter was pressed. Film is nothing until developed, digital is much the same.

Taking the photo captures what is visible, editing captures what was seen.

Why would someone do this?

It makes no sense why you’d do something so foolish.

There are low-integrity folks who are willing to risk getting caught for fakery in public spaces in trade for the promise of notoriety.

Chicanery seems to scale anywhere that there are insufficient institutional barriers.

Publish or perish. The stakes are incredibly high. Cheating may be the difference between keeping and losing your job.

Especially since everyone else is aware of that. If you don't take the risk, you will lose to someone who did and got away with it.

This researcher is in China, where I gather the competition is even fiercer than in the West. I do not know that for a fact, but if it is true it makes it even clearer.

Viewed that way it's not foolish. It's actually mandatory. And it's just going to get more prominent as the AI makes fewer mistakes, and competitors get better at spotting and removing them.

Man... This is so sad, disappointing and worrying. For some reason, your explanation in particular struck a nerve with me. Who could have predicted all the nuanced social consequences that AI would bring upon us. I don't yet know what position to take in my head on all of this. It's too much change, despite believing that the world will change and adapt. In particular, that human work will become more valuable - but as a craft. As an endeavor you do because it's artful, not because it's mandatory, or the most optimal way to quality or speed. And those two categories - human and ai work - will become very separate. While the economical value or even expectations of absolute quality in the human work will plummet.

Similarly, I expect "raw"-ness to gain on value. Photos untouched by AI or filters. Physical objects. The non-digital. Who knows, maybe physical media will make a comeback.

Either way, it's not good times we're living through, and I don't particularly appreciate all of this happening after *just* becoming an adult in the world.

I was surprised at how obvious the AI was. Cilia do not, in fact, pop in and out of existence while moving around.

Anyone have a link to the video? It seems to have been taken down from YouTube.

In a news broadcast: https://www.youtube.com/watch?v=Sz3V9398LEY

Aw, they took the video down. I was curious to see how egregious the AI usage was.

Read the article and see the video there

2 week old news OP;

[dupe] https://news.ycombinator.com/item?id=50038640

Reply on news.ycombinator.com