311
This was supposed to be NoAI
(i.ibb.co)
AI images: hide. Yeah, sure.
I guess old printed encyclopedias are much MUCH better than this shit for your child's homework. Ugh, this is so disturbing.
(ノಠ益ಠ)ノ I∀
I'm not sure I understand, are you thinking they are all AI? At least some of them are not - I recognize them from the before times. Tardigrade images have always been goofy looking.
The adobe stock images are AI though. It's probably a safe bet to avoid stock image sites although Alamy seems to have some good ones. Adding the photography filter helps a bit too, but I still got some stock images used on other websites.
Of course not, but a lot of them are, which is not good enough for No AI search. Yeah, stock images got worse, they shouldn't be called "stock" anymore. We almost got to the dead internet part and it came so fast, it's mind boggling.
About the death of the internet, I have slowly stopped casually searching stuff. I almost exclusively search within known domains now :(
Well, that seems to be the only way to handle this. For now. :(
Fair, the adobe ones are labeled as AI if you ~~click through to the link so that's a pretty big miss in the search.~~
Edit: I retract my misinformed comment, I hadn't looked at how the filter works.
Stock images have always included bullshit renders of microbes so maybe that's just something I take for granted.
If it's this obvious then it's on the search engine.
It's used to be fun to look at stock images. Now it's just disturbing.
I don't agree with that take.
https://duckduckgo.com/duckduckgo-help-pages/results/how-to-filter-out-ai-images-in-duckduckgo-search-results
This is an open source block list, which is manually put together by volunteers. It's not even reading metadata, presumably because that's too resource intensive (or proprietary/unreliable).
Gemini was able to tell that was AI because it was able to read the image ID water mark and associated it with how Adobe sequences AI generated images. The image ID is in the file name, but if we wanted to catch it that way the filter would need to work on logic, not just keywords or site lists. Programming a filter to search the file names to identify if it belonged to the sequence would be more complex.
Alternatively, if you clicked though to the image you are immediately told if it is AI. Whenever I search for microbe photos I always click the source to determine if it's a render or a stylized electron microscope capture, and I figure it's always been good practice to get the image from the source to verify it's not totally bullshit either in terms of the the search pulling up something unrelated or the image itself being misleading or even a humorous misrepresentation/meme.
I think it's important to understand how the tools we rely on work.
I see. If it works on manpower that means we cannot catch up easily, which sucks. And it's possible that it will be overwhelming at some point, if not soon. I hate how internet become finding a diamond in garbage hills. And when you finally find the diamond, it can still be fake since we cannot tell at first sight. :/
Anyway, this is a real tech fight and we are in the middle of this. It seems curated internet what we'll need from now on.
There's 21 images. Cautiously I'd identify at least 10 of them as AI generated. The issue is that there are any AT ALL with DDG's options set to hide AI generated images. That setting used to work really well.
Ironically, AI can tell me the adobe images are AI and why they don't get flagged in image search: