|
I misread the pitbull one as eldritch humor and now I want to know what a stand up comedian who does eldritch humor would sound like
|
# ¿ May 26, 2022 16:55 |
|
|
# ¿ May 14, 2024 12:14 |
|
precision posted:imagine if they managed to prove that thoughts occur 0.0000001 seconds after the action we think they refer to, proving forever that we literally don't have any control over our actions and only think about them after they happen Do you think that would change anything about how society is structured?
|
# ¿ May 27, 2022 15:59 |
|
Can the person who had access to the text ai try these things: 1. Cooking recipes 2. Aphorisms I'd be curious if any interesting output came from those.
|
# ¿ May 29, 2022 01:43 |
|
This would be interesting to see visually, imo.
|
# ¿ Jun 2, 2022 07:30 |
|
That one is easy to see it got hung up on the word ultimate
|
# ¿ Jun 2, 2022 15:18 |
|
Where's the beef!?
|
# ¿ Jun 3, 2022 17:04 |
|
If you were making a board game and needed art, could you just use one of these to generate it all? What is the cost structure or is it totally free at the moment to use the art for whatever you want?
|
# ¿ Jun 4, 2022 19:48 |
|
Yeah, I wasn't sure if there was some proof or watermark or something built invisibly into the output or if every output was being stored with metadata so they could make a case if something got huge commercially or something. It's just a weird thing to think about, given how striking and high quality this stuff has become.
|
# ¿ Jun 4, 2022 20:01 |
|
precision posted:ultimately i think it's like, there's no real incentive for the people behind DALL-E to go after you, so even if they say "don't use this for work" it's more of a wink wink sort of thing I mean, the people who invested millions in the design and pay for the cost of the hardware and promotion etc aren't going to want money back? I get right now it's all building awareness and stuff with people gushing on Twitter, but surely they have some kind of business model in mind, right?
|
# ¿ Jun 4, 2022 20:13 |
|
Can someone do a still from an unreleased live action Calvin and Hobbes movie by a director of your choice?
|
# ¿ Jun 9, 2022 04:50 |
|
Oh man, the EverQuest ones are interesting. Anything with a really definitive style ends up pretty cool--the underwater juggalos, the trail cam ones, etc.
|
# ¿ Jun 14, 2022 20:14 |
|
Can someone try people or monsters or animals in the style of a first edition ad&d monster manual?
|
# ¿ Jun 14, 2022 22:04 |
|
strange feelings re Daisy posted:None of the models I tried knew who David Tempier or David Sutherland were so I had to use a more generic prompt of "Dungeons and Dragons monster, simple line art" The simple line art ones are great! That makes me think there could be a fun game where people add captions and speech bubbles to random line art prompts to make weird comics.
|
# ¿ Jun 15, 2022 02:39 |
|
This Is the Zodiac posted:
Is that a piranha plant?? Crazy
|
# ¿ Jun 19, 2022 12:57 |
|
I feel like already more content gets created everyday than you could absorb in a hundred lifetimes, so I don't know that it will make that much of a difference. It'll just be disposable like much of the content that's created by humans is.
|
# ¿ Jun 29, 2022 13:11 |
|
great, now you've summoned Officer Ghost
|
# ¿ Jun 30, 2022 11:44 |
|
The paladin holding the sword that way is actually pretty interesting. It adds a level of symbolism to it, reminds me of a tarot card.
|
# ¿ Jul 1, 2022 12:43 |
|
Foot Lettuce sounds intriguing. Does it taste like feet? Or look like feet? Smell like feet? Or are feet somehow involved in its preparation, like stomping on grapes to make wine?
|
# ¿ Jul 1, 2022 14:23 |
|
Black August posted:I want to see what 2 can do after a year of iterating off of mass user inputs and refinements to better teach it what people want to see and mean by certain phrases Do you rate the pictures after they're presented to say if it matched what you were going for or not? Or is it sophisticated enough to just somehow understand based on if you tweet it out that it was probably correct?
|
# ¿ Jul 3, 2022 23:45 |
|
TIP posted:it reminds me of when I used to stare at the wood grain on my closet door until I'd see things in it apophenia
|
# ¿ Jul 9, 2022 01:44 |
|
How detailed can the prompts be? Like if I wanted to see a "demon with no head, two right arms of different lengths and no left arm, skunk face on left knee, and wearing a rainbow jumpsuit with dozens of pockets", is there some special way you'd have to organize the prompt to output what you want?
|
# ¿ Jul 10, 2022 00:27 |
|
Tunicate posted:for midjourney not really, it kinda mushes things up That's kinda what I thought would happen, thanks for running it through! I'm sure as they get more sophisticated, you could have multiple details like that and it would produce things that were closer. Most of the good ones seem to lean on stuff in the style of famous artists or use a famous person that already encapsulates those kinds of details, it seems. Like, describing Buddha's or Kermit's physical characteristics instead of using the actual proper name would probably lead to less interesting outputs.
|
# ¿ Jul 10, 2022 01:09 |
|
Tunicate posted:did some monster manual covers in midjourney Man, I really want to see more old school D&D art, that newer style just does nothing for me. I doubt it has enough exposure to like DCC RPG style art to make a real go at it, though. Maybe in the style of 60s fantasy paperback would get something closer?
|
# ¿ Jul 10, 2022 13:09 |
|
KakerMix posted:I'm not familiar with the specifics of this art, but I did like a ton of the line art from Shadowrun books in the 90s Now that is what I'm talking about! Very cool! That's exactly the style I was imagining and personally find really evocative. It's cool to see it able to channel that style even if you don't mention the specific artists.
|
# ¿ Jul 10, 2022 19:17 |
|
doctorfrog posted:What happens when the internet begins to fill with AI-generated images, and the AI generators begin training on AI-generated images? Broke: in the style of Dali Woke: in the style of Dall-e
|
# ¿ Jul 14, 2022 00:48 |
|
The yard sale reminds me of that one image that looks like reality but then you're hard-pressed to describe anything that's actually in the picture. Like maybe I can make out a blue garment but primarily it just has like the essence of a yard sale.
|
# ¿ Jul 20, 2022 14:40 |
|
Bloody waterslides reminds me of a really hosed up Megg, Mogg and Owl strip, so gross
|
# ¿ Jul 20, 2022 15:15 |
|
It's weird that would need teased out, you'd think part of the academic or research team documents would explain how they intended to handle that kind of bias in the sample set.
|
# ¿ Jul 21, 2022 18:22 |
|
Wheany posted:Well, one benefit of running something locally is that it's disconnected from the panopticon. You can be as horny as you want. Huh, so you think every image generated is stored along with the prompt and user info? I guess it's naive to think that wouldn't be the case, it just seems like the output generated in most cases is probably meaningless.
|
# ¿ Jul 24, 2022 14:33 |
|
Tunicate posted:Talked with one of the developers, the filter on DALLE is a combination of a wordlist ban and some sort of machine learning classifier (I'm guessing GPT3 since they have it in house). LOL, they should just give the people what they want
|
# ¿ Jul 24, 2022 22:05 |
|
Thinking about it, are there actual ways to tell if an image was ai generated or 'real'? Like is there some secret watermark that would always be present or something? Just trying to imagine what weight photographic evidence or even video will carry in the future, if ai can generate such things convincingly.
|
# ¿ Jul 24, 2022 22:43 |
|
Everything is so based on existing art (obviously), I'm curious if you combined the text ones with the image ones if you could get a pitch for a new IP with concept art and plot ideas, from a well worded prompt. And then theoretically even scripts, though a lot of the text stuff I've seen seems to forget what's happened or who is in a given scene, so there's probably more sophisticated models to come.
|
# ¿ Aug 3, 2022 13:42 |
|
I like the homebrew diy version of Monopoly
|
# ¿ Aug 3, 2022 16:12 |
|
Really, everyone should just start wearing outlandish masks everywhere they go in public and adopting fake Internet names. Then any kind of deep fake surveillance couldn't be used anyway, since anybody could have been wearing that mask.
|
# ¿ Aug 9, 2022 00:22 |
|
KakerMix posted:No it's exactly what is happening, Dall-E is altering your prompts before you get your image. The data is in the model, they are intentionally and deliberately not giving you what you ask for. So it transformed "zoidberg" into "red anthropomorphic crustacean man" or something before passing that on to the ai?
|
# ¿ Aug 10, 2022 20:31 |
|
I guess I would imagine the model doing a basic Google image search, at which point, the hundreds of thousands of images of zoidberg that look like zoidberg would inform the output. It's not like the prompt was for something really obscure, like a third string Dickens character from Bleak House.
|
# ¿ Aug 10, 2022 21:02 |
|
mobby_6kl posted:Well it works now. I nuked everything and started over with this. I think an issue was that some stuff failed to install automatically and I did it manually with the pip installer, but that hosed up the environment creation. No idea why it would fail to find the OptimizedSD module though as that has nothing to do with installed packages. loving python. You managed that stuff on your 1070?? I have a 1080 ti and just assumed getting a local instance of this was out of reach.
|
# ¿ Aug 22, 2022 13:26 |
|
Do you think someone will make an easy way to install this locally? Or will you always need to know python and stuff to do this on your own machine?
|
# ¿ Aug 22, 2022 20:36 |
|
Mozi posted:this guide worked for me for getting this running locally on Windows That's really sweet! I might look into this given the 1080ti I have sounds like it could be beefy enough, even if it's not as quick as a newer GPU.
|
# ¿ Aug 23, 2022 04:51 |
|
|
# ¿ May 14, 2024 12:14 |
|
Is Virgil Finlay known by the models enough to be used for bad rear end black and white fantasy and horror stuff?
|
# ¿ Aug 31, 2022 04:12 |