|
egg_dog posted:Sorry for this Appropriate username though, Bing users hate the eggdog
|
# ¿ Dec 16, 2023 11:27 |
|
|
# ¿ May 10, 2024 21:19 |
|
Nice to see that most people don't want this to end up as PYF lovely AI Art: a picture tells a thousand slurs I'm trying to find one of the AI blogspam sites that was hilariously bad uncurated GPT articles with automated AI art illustration/captioning, it's terrible but also innocent enough for the thread, I'm sure there's threadworthy material on DeviantArt but I don't really wanna see that kind of thread either. Having said that I had to see this the other day so now the rest of you can too :
|
# ¿ Dec 28, 2023 10:56 |
|
Found this while trying to find the other site: NFT bro publishes an AI book! The Sausages has a new favorite as of 11:17 on Dec 28, 2023 |
# ¿ Dec 28, 2023 11:14 |
|
like freakin clockwork
|
# ¿ Jan 14, 2024 01:01 |
|
They didn't even remove the Bing watermark The Sausages has a new favorite as of 06:32 on Jan 18, 2024 |
# ¿ Jan 18, 2024 06:20 |
|
Most of the elaborate prompt-fu is cargo culting - people are copying phrases and words that more often than not don't affect the output, because they saw it in someone elses prompting and it seems logical to put stuff like "masterpiece" as a prompt or "bad anatomy" in the negative. But the models/loras/embeddings used affect the style and composition far more than "masterpiece, bad anatomy" do, and some things that seem sensible to AI creators are provably meaningless to the AI e.g, iirc "extra limbs" has zero effect on the number of limbs when included in a SD1.5 prompt, no matter how many (()) you include, The A1111 interface makes it trivial to drop in hundreds of style tokens in every prompt, which is great for lazy AI creators. But they'd get better results starting with a clean slate and incremental tweaking of prompts instead of that overwrought word salad.
|
# ¿ Feb 23, 2024 11:26 |
|
Apologies for nitpicking but that's how GAN image generators work. Diffusion Models such as SD, MJ or Dall-e are trained by adding noise to an image in a way that allows the AI model to learn to reverse the process. "Plagiarism Machines" is an excellent and accurate description in both cases. The other nitpick is the text used for the SD dataset was the alt text i.e. what gets displayed if the image doesn't load, and not the text from the page or comments. Apparently you can search the dataset here to get an idea of the GI part of AI's GIGO, or to see if your pics are included in the dataset. There's also another anime/hentai trained dataset ripped from danbooru or something that had more accurate captioning and some of that has spilled into people's prompts e.g. 1girl, or pretty much anything with underscores. So "extra_limbs" might have an effect as a prompt (positive or negative) with the Waifu models, but both "extra_limbs" and "extra limbs" won't be effective with SD models. For Dall-e 3 the dataset was re-captioned by AI image analyzers; MJ and SD are probably following suit as fast as their budgets allow. Harvey Mantaco posted:AI is too worthless to do anything of value but just functional enough to destroy any remaining semblance of trust or credibility for anything across the entire internet cool love that for it
|
# ¿ Feb 28, 2024 14:47 |
|
I totally understand artists feeling betrayed by other artists using AI but I hate this witchhunting BS more than I hate crappy AI art. "I've kept refining the AI prompt to make it look more and more like a picture and now it looks like the picture therefore the original is also AI" is some AI paranoia brainworms. The supposed resemblance is clearly not a coincidence - because it's something this dipshit is doing on purpose. Bonus points for dipshittery if they included the artist's name in the prompt.
|
# ¿ Mar 12, 2024 13:05 |
|
Maybe it's both for all I care but "I replicated it with AI therefore the original is AI" is an idiotic non-sequitur that shouldn't be accepted as evidence of anything.
|
# ¿ Mar 12, 2024 15:59 |
|
Bottom Liner posted:
The charitable interpretation is they're pointing out that claude.ai lacks a "no medical diagnosis" clause which most publicly accessible AI services have. Charitably it's all bullshit. why would anyone send their medical information to a blue tick? otoh, tech bros have terminal brain worms, moreso if participating in the attention economy. I don't care to dig to find out how true this bullshit is.
|
# ¿ Mar 24, 2024 10:51 |
|
does anyone have any idea how much money this bullshit saves vs using stock art? I'm guessing the "audiobook" using this cover is also minimum effort AI-generated audio slop as the narrators don't exist.
|
# ¿ Mar 30, 2024 04:03 |
|
Surely Peter Gabriel could do something interesting with AI watches video of same janky rear end deforum effect as every other AI video
|
# ¿ Apr 7, 2024 00:21 |
|
Splicer posted:Eh... beautifully animated but extremely generic in its own way. Anyway the only Pink Floyd/Animation fan combo I'd recommend is this one: https://www.youtube.com/watch?v=zBWbpFz3wac Splicer posted:The difference between doing something interesting with AI vs doing something interesting with stop motion animation or whatever is that you look at stop motion animation and think "OK, what are the strengths of this, what are the weaknesses, what are the quirks, and how do I emphasise the strengths, leverage the quirks, and either avoid, compensate for, or do something weird with the weaknesses." The Sausages has a new favorite as of 11:08 on Apr 8, 2024 |
# ¿ Apr 8, 2024 10:53 |
|
|
# ¿ May 10, 2024 21:19 |
|
credburn posted:God dammit that "eyes almost closed" thing actually works. old news
|
# ¿ Apr 15, 2024 23:44 |