The Wolf Is Here! AI Defeats Humans to Win a World Photography Award!

Just a few weeks ago, in "In the Age of AI Painting and Video, Where Is the Future of Still Photography?" (《AI绘画与视频时代，平面摄影的未来在哪？》), we discussed the impact of AI-generated art on photography.

Never expected the wolf would actually come!

Two days ago, an AI-generated work fooled professional judges and defeated many human photographers to win the Creative category award of the open competition at the Sony World Photography Awards!

And in the near future, AI will be able to learn and generate RAW format files, and even fabricate videos of the "shooting process."

Will photography still exist? Will humanity still have truth and history? What do you think?

Video

▶ Watch the video tutorial on Bilibili

This is the illustrated companion to the video above. Shortly after Thomas discussed the future of photography in the age of AI image generation, the wolf really did arrive: an AI-generated image fooled professional judges and beat human photographers to win the Open Creative category at the Sony World Photography Awards. This video walks through that earthquake moment for the photography world — and the deeper questions behind it.

Start with the three questions the video ends on: can AI generate convincing fake RAW files? Convincing fake videos? If neither can be verified, does photography still exist? Does truth still exist? Keep those in mind as we replay the whole story.

01 The Wolf Is Here: AI Takes the Sony Award

The Sony World Photography Awards is one of the industry's most prestigious contests, and this year's edition had its own highlights: two Chinese photographers, Li Fan and Zhang Kechun, each took first place in a professional category with outstanding work (at 0:18 in the video). But the spotlight was stolen outright by a German photographer's stunt.

His entry was an AI image styled like an old photograph: an older woman embracing a younger one. The flaws are visible on inspection — skin detail on the faces is slightly missing, and the hand at the lower right is an outright defect. But since it entered the Creative category, the judges presumably assumed some unconventional technique was at play, and the image fooled everyone into a win (at 0:43 in the video).

02 A Piece of Performance Art: The Public Refusal

What followed was even more theatrical: the image went into the awards ceremony and the exhibition as usual, the photographer flew in for the event — and then announced on stage that the picture was AI-generated and that he was refusing the award (at 1:35 in the video).

On his website he explained the intent: to accelerate people's realization of what AI means for photography and for ethics. In other words, a carefully staged piece of performance art — using a trophy that was never rightfully his to force the whole industry to face a question that has already arrived.

03 Rule Gaps and RAW Verification

Thomas points out that comprehensive contests like the Sony awards are not documentary competitions: they impose no restrictions on post-processing, and a careful read of the entry terms shows no clause banning AI work — a huge loophole in itself (at 2:08 in the video).

Expect most contests to spell out whether AI entries are allowed — most will ban them — and expect a new verification step in both documentary and creative categories: checking your RAW files to prove the source material was truly captured by a camera and belongs to the photographer (at 2:27 in the video).

04 AI Images Are Flooding the Communities

AI's advantage over human photography is real — it fooled the judges and every viewer until the creator himself revealed the trick; had he stayed silent, nobody would ever have known. And it is not just contests: browsing the Tuchong homepage, Thomas counted multiple AI images on the very first page, portraits included (at 3:23 in the video).

05 Some Day in the Future: How Much Truth Remains

Thomas contrasts it with his own photograph from Greenland: to get such an image you must travel somewhere extremely remote, wait for waves to push the sea ice right around the little town, for the wind to drop, for a storm to have just passed and stacked spectacular burning clouds overhead — locals said they had never seen anything like it — and then grade it in post (at 3:47 in the video).

In the future AI may generate such frames every single second, thousands to tens of thousands a day — and once scarcity is gone, the photograph itself loses its meaning. At that point only on-site video evidence may let people feel the spectacle, and appreciate the photographer's waiting, reflexes and composition (at 4:24 in the video).

The scarier part: a RAW file is, in the end, just a pile of zeros and ones recorded by the camera's sensor — AI could well learn from massive RAW datasets and fabricate convincing 'originals'; and if one day even RAW files cannot be verified, or AI can produce perfectly realistic fake shooting videos, we will lose the very ability to prove a photograph was ever taken. Thomas sees the remedy not far off: AI legislation and mandatory labeling — requiring major model vendors to embed multiple visible and invisible watermarks; and since AI is open-sourced and runs locally, future laws may require every website and social platform to label AI content automatically, with penalties for deliberately hiding the tags — much as pornographic and illegal content is kept off mainstream media today. Only that can keep our news, our media, our public discourse and our history real (at 4:47 in the video).

1. An AI image fooled the Sony World Photography Awards judges and won; the photographer's public refusal was performance art forcing the industry to face AI head-on; 2. Comprehensive contest rules generally don't ban AI yet — expect explicit bans plus RAW verification soon; 3. When AI can generate thousands of great images a second, scarcity dies, and photographs will need on-site records to prove their meaning; 4. A RAW file is just zeros and ones and one day may be forged too — legislation and mandatory watermarking are the last line of defense for truth and history.

HD video:

https://www.bilibili.com/video/BV15M411L7tT/

Related tutorials:

"From Beginner to Professional: Learning and Creation Methodologies Ordinary Photographers Won't Share!"

"Want to Improve Your Aesthetic Sense? 6 Websites Kept Private by Professional Photographers!"

"How to View Imitation in Photography? How Can Imitation Truly Improve Your Shooting Skills?"

"The Winning Photo of an International Contest Was Stolen? How Artists Steal Inspiration and Style"

"How to Improve Your Aesthetic Sense and Learn Photography Through Competitions? Methodology Sharing Is Here!"

"A Farmer Leading an Ox with Artificial Morning Fog - Why Are Such Photos Soulless?"

"Can a Photoshopped Photo Still Be Called Documentary Photography?"

"In the Video Age, Does Still Photography Still Have a Future?"

"How Can Landscape Photography Avoid Being Repetitive? Some of My Thoughts"