Hacker News Reader

Comment by smallerize | Hacker News Reader

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

View on HN

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.

Did he ever share the raw source?

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.

Reply on news.ycombinator.com