|
Combining the text/audio AIs seems like an interesting approach to TV shows, at least animated ones to start with. Begin with a concept/backstory, say a comic book universe and a general plot idea for a a story from the universe. Introduce a random seed, and anyone who was interested in that concept could watch a procedurally generated version of that input. If there was a particularly good version, people could share that random seed to see the same version.
|
# ? Apr 24, 2022 05:24 |
|
|
# ? May 30, 2024 13:16 |
|
Computer. Show me inside the ladies' locker room
|
# ? Apr 24, 2022 07:05 |
|
A magic space wizard in pixel art style
|
# ? Apr 24, 2022 10:00 |
|
I want to toss so many ideas at this thing. I appreciate if you get to any. Choose whichever you think might give an interesting result from past experience. The Eiffel Tower, Big Ben and The Statue of Liberty waiting in line to enter a lively nightclub, everything is in the style of an old comic, with the exception of the bouncer, who is photorealistic. A large Ancient Greek vase with a design etched in it, depicting a bulldog riding a skateboard. A photo realistic ice sculpture depicting a giant standing ape which has a ball dancing couple standing on its palm. A chubby man eating an oversized burger depicted with the shapes of US states in various sizes and colors. A magazine cover celebrating the first koala to pilot a helicopter to the Moon. A kangaroo wearing a jacket, top hat and monocle, depicted in an old sepia toned photographic portrait. A collage of numerous elephant images, each in a different artistic style. A magician making an angry Viking disappear in the style of Roy Liechtenstein. The Last Supper except all of the characters are humanoid reptiles, and there is a briefcase with shining gold bars laid on the table. A photorealistic depiction of a forklift racing a giraffe down the empty streets of Miami A simpsons style illustration of a roast being ruined. A patent diagram of a border collie being ridden by a small monkey dressed like a cowboy A 1950’s diner staffed entirely by chimpanzees in costumes and roller skates.
|
# ? Apr 24, 2022 10:11 |
|
AreWeDrunkYet posted:Combining the text/audio AIs seems like an interesting approach to TV shows, at least animated ones to start with. How would this be different from the multitude of Spiderman + Elsa videos The Algorithm cranks out daily?
|
# ? Apr 24, 2022 10:59 |
|
Tarquinn posted:Don't mind the goats.
|
# ? Apr 24, 2022 11:10 |
|
vaporwave painting of gamer dmt elves inside vishnu's brain in the style of alex grey
|
# ? Apr 24, 2022 11:13 |
|
i must see four balls at the edge of a cliff
|
# ? Apr 24, 2022 12:03 |
|
the movie poster for honey i shrunk the kids in the style of dees illustrations. id pay money to get some time to mess with this thing, this poo poo is honestly incredible.
|
# ? Apr 24, 2022 14:10 |
|
Hopefully some more goons are able to get access soon. This stuff is amazing and just a little scary at the same time. "a dog wearing tennis shoes struggles to take the mail-bag away from a bare-footed USPS mail carrier in the style of a coloring book"
|
# ? Apr 24, 2022 14:19 |
ekuNNN posted:they are not in any way close to being even a part of a person. p sure humans are just chunks of processes glommed together into one big thing that wakes up alive one day. If you keep slamming enough of these sub units together eventually something that can have experiences will be formed. Justin Credible posted:Someone should feed into it the concept of Roko's Basilisk. well they got it to use big words at least
|
|
# ? Apr 24, 2022 14:36 |
|
An ambulance trying to leave the hospital Man plays games online while wife is giving birth Guy wearing wedding ring opens his donkey to show everyone World's largest man plays the world's smallest violin (and the opposite) Anthropomorphic corn stalks grow a field of humans Anthropomorphic salt shaker sprinkles humans onto his food The rise and fall of Lucifer Human waiters serving robots a meal at a restaurant The great pot and kettle is black debates Man proves the nonexistence of God using a PowerPoint presentation Tim is hungry for apples E-cigarette vapes a human The edge of the universe and beyond
|
# ? Apr 24, 2022 14:50 |
TIP posted:you could always try reaching out to someone from openAI and pleading your case, that worked to get me bumped up on the list for playing with GPT-3 tempted to do thhis. last night i spent like an hour trying to combine hi res images of pasta with the only hi res public domain pornographic image I could find in photoshop just using clone stamps and healing brush and whatnot. Definitely felt like a job for AI, "full penetration sexual adult pasta magazine cover, in oil paint"
|
|
# ? Apr 24, 2022 16:57 |
|
SniperWoreConverse posted:
I think I posted this screenshot in the UFO thread and it just hit me that yeah GPT-3 has probably already been fed some Nick Land, lol
|
# ? Apr 24, 2022 17:27 |
Yeah I def poached it from there, cutting edge uaptek posting
|
|
# ? Apr 24, 2022 17:30 |
|
these are amazing. if you're still doing these, could toss an image request in as well? a mechanical bokrug surveys the wastelands of middle america it smiles and eats a pudding cup thanks!
|
# ? Apr 24, 2022 19:00 |
|
Do they say anything about copyright? Can you use those images on products you sell for money (as in album covers)? As for an idea, maybe "Giant smurfs run after small Gargamels in the smurf village."
|
# ? Apr 24, 2022 19:23 |
|
a cherry blossom tree with petals falling with a ruined city in the background
|
# ? Apr 24, 2022 21:54 |
|
If you feed it the same prompt twice in a row, does it return the same result? Like does it decide what the ideal result is, and then that logic doesn't change? Or does it return different results each time because art doesn't have a logical end point.
|
# ? Apr 24, 2022 22:13 |
|
Computer, can I get a nude Tayne?
|
# ? Apr 24, 2022 22:22 |
|
stratdax posted:If you feed it the same prompt twice in a row, does it return the same result? Like does it decide what the ideal result is, and then that logic doesn't change? Or does it return different results each time because art doesn't have a logical end point. It's always a different result. Might be really similar, might be quite different. There was an example about asking it to show an avacado. Sometimes it was a whole one, sometimes it was sliced in half with the pit visible.
|
# ? Apr 24, 2022 22:26 |
|
I don't have access to DALL-E but I do have a diffusion vector based GAN model and a very big GPU, so just for funsies I ran some realizations:Randabis posted:Anthropomorphic salt shaker sprinkles humans onto his food I like that it seems to have gotten confused about 'sprinkles' and then fixated on that. Booty Pageant posted:i must see four balls at the edge of a cliff Mameluke posted:Computer. Show me inside the ladies' locker room The Donut posted:A magic space wizard in pixel art style Not as nice as the DALL-E ones but then again I am a hobbyist goon and not a well funded company, so eh. (I have each of the squares in the collages as larger images if you see one you want a 256x256 of)
|
# ? Apr 25, 2022 03:09 |
|
when will you be minting the NFTs? edit: i'm really curious now what it would generate from the phrase "bored ape, in the style of a cartoon" Bad Purchase fucked around with this message at 03:15 on Apr 25, 2022 |
# ? Apr 25, 2022 03:10 |
|
I gotta admit, even though it was struggling with the pixel art part, that wall-eyed little guy in row 8 col 2 is fun. edit: some apes(?) Objective Action fucked around with this message at 03:33 on Apr 25, 2022 |
# ? Apr 25, 2022 03:13 |
|
one of the creators of Dall-E 2 has been posting videos of using it to modify images in weird ways and it's pretty cool bonsai kitty https://twitter.com/model_mechanic/status/1516065929305616387?t=PjRQkX8mFUmmz4tGOh-R4w&s=19 https://twitter.com/model_mechanic/status/1516164008646090754?t=7gJcyBrrIHQff2EEZtF9Sg&s=19 https://twitter.com/model_mechanic/status/1517205808680869888?t=y47ktY_DIZjB-JRK2hzd-A&s=19 https://twitter.com/model_mechanic/status/1515849481635336193?t=zXFlHN8Pa4JWDmqFfsmkjg&s=19 https://twitter.com/model_mechanic/status/1514266246849593344?t=y1wS-8fqTBpB6OzXrCLDOA&s=19 TIP fucked around with this message at 03:40 on Apr 25, 2022 |
# ? Apr 25, 2022 03:36 |
|
Objective Action posted:I gotta admit, even though it was struggling with the pixel art part, that wall-eyed little guy in row 8 col 2 is fun. look at this rascal
|
# ? Apr 25, 2022 03:39 |
|
I do not like the cat one.
|
# ? Apr 25, 2022 03:40 |
|
Zil posted:I do not like the cat one. I can't help imagine some horrific sound effects to go with these videos. Objective Action posted:edit: some apes(?) Incredible power in this stuff for random avatar generators. Fascinated by the way text gets visually slurred, with strange intermediate letterforms.
|
# ? Apr 25, 2022 04:34 |
|
wow even computers have a hard time imagining, a rendering of a zybourne clock would leave a data center in flames, though i like the bigstock watermarks that's not any ladies room... that's a computer ladies room!!!
|
# ? Apr 25, 2022 13:10 |
|
I’m on the waitlist but am desperate to see this prompt: “a modernized rendition of the lady of the lake depicted as a dirty bum retrieving a firearm named ‘Excalibur’ from under a bridge for a contemporary knight”
|
# ? Apr 27, 2022 17:50 |
|
https://twitter.com/soficrespo91/status/1518331023871791105 If you can literally just feed it a neologism and get something like this out, I am extremely impressed. Not sure if that's actually what happened here... edit: same folks just released a companion visual language model: https://www.deepmind.com/blog/tackling-multiple-tasks-with-a-single-visual-language-model take_it_slow fucked around with this message at 23:42 on Apr 28, 2022 |
# ? Apr 28, 2022 23:34 |
|
Been playing with Disco Diffusion a bit.
|
# ? May 2, 2022 02:58 |
|
I’ve had this dream^^
|
# ? May 2, 2022 15:15 |
|
Philthy posted:Been playing with Disco Diffusion a bit. What the sweet hell generated this madness.
|
# ? May 2, 2022 16:16 |
|
Who spilled their H. R. Giger all over my Zdzisław Beksiński?!
|
# ? May 2, 2022 16:25 |
|
Philthy posted:Been playing with Disco Diffusion a bit. A quick Google search suggests that there is already an attempt to make this 3D and then VR compatible. Can't wait for No Man's Sky 2 with totally novel worlds to explore instead of procedural generation.
|
# ? May 2, 2022 17:04 |
|
I can’t wait until the line between reality and media is blurred until we all live in spooky hell dimension *doom music*
|
# ? May 2, 2022 17:13 |
|
Someone whipped up their own variant of the methods used for DALL-E, so you can try it here: https://huggingface.co/spaces/dalle-mini/dalle-mini The results are not as impressive as DALL-E 1 or 2, but it definitely can do some creative stuff
|
# ? May 2, 2022 21:04 |
|
GABA ghoul posted:A quick Google search suggests that there is already an attempt to make this 3D and then VR compatible. Can't wait for No Man's Sky 2 with totally novel worlds to explore instead of procedural generation. A mathematical process that creates worlds to explore is procedural generation. Dall-E-2 is more complex than the algorithm used by No Man's Sky, but it's the same concept.
|
# ? May 2, 2022 21:57 |
|
|
# ? May 30, 2024 13:16 |
|
Only tangentially related, but does anyone follow “heavenslastangel” on Instagram? He does these psychedelic chaos-based recursive animations. I’d like to know how he does it (without going to into the weeds).
|
# ? May 3, 2022 01:42 |