Why use an AI image for the folding phone when they’re are literally tens of thousands of images of foldables elsewhere?
There is honestly no real justification for creating slop images for a meme or really anything. Last I checked, stock image companies or image copyrighters don’t file lawsuits against creators of memes for badly cropping a stock photo and putting text on top of it.
Fair enough honestly, it’s unfortunate how large of a pool of AI images aren’t labelled as such, and that the only feasible way to get them labelled after the fact is [at least in the case of duckduckgo,] to gather feedback data given by people who search and know that there is the “flag AI generated image” button.
I suppose indentifying AI images is easier said than done, and requires at training and good enough pattern recognition, so it doesn’t mean everyone has to become an expert at recognising them, it really means that these images are flooding everywhere and are a great reason to fervently oppose generative AI.
At this point, I’d say if you want to avoid AI images, you need to see enough of them to find what makes them look off, and with enough time it becomes second nature to discriminate images based on how real they appear.
Since we’re giving out advice: Your original comment was very quick to put me in the slop corner. I just want to shitpost, I’m on your side, and as you correctly assumed I don’t want to become an AI expert. Hanlon’s razor wouldn’t have hurt here.
Why use an AI image for the folding phone when they’re are literally tens of thousands of images of foldables elsewhere?
There is honestly no real justification for creating slop images for a meme or really anything. Last I checked, stock image companies or image copyrighters don’t file lawsuits against creators of memes for badly cropping a stock photo and putting text on top of it.
I took the first image from duckduckgo where the “no AI” filter is on. All the images I post are edited in GIMP alone, by my own untalented hands.
Fair enough honestly, it’s unfortunate how large of a pool of AI images aren’t labelled as such, and that the only feasible way to get them labelled after the fact is [at least in the case of duckduckgo,] to gather feedback data given by people who search and know that there is the “flag AI generated image” button.
I suppose indentifying AI images is easier said than done, and requires at training and good enough pattern recognition, so it doesn’t mean everyone has to become an expert at recognising them, it really means that these images are flooding everywhere and are a great reason to fervently oppose generative AI.
At this point, I’d say if you want to avoid AI images, you need to see enough of them to find what makes them look off, and with enough time it becomes second nature to discriminate images based on how real they appear.
Since we’re giving out advice: Your original comment was very quick to put me in the slop corner. I just want to shitpost, I’m on your side, and as you correctly assumed I don’t want to become an AI expert. Hanlon’s razor wouldn’t have hurt here.