AI-generated images

Generated with DALL·E mini. Probing how far it can go: is it truly creative, or "merely" mixing up its source material to match the prompt. (I believe the latter.) I'm also interested in what it "thinks" short prompts look like, like the "minecraft", "beauty" and "code" prompts below. Last update: 2022-07-04.

Contents:

Videogames
Things on the computer
Abstract concepts
Places
Other stuff

Videogames

Minecraft

Prompt: minecraft.

1: Green plain, looking at what could be a building. 2: Green plain, looking at a big oak tree in the distance. There might be a hole in the ground containing cows or other livestock. 3: Green ground and blue sky, close to a farm, looking at either a player or a zombie. A village in the background? 4: Slightly hillier than the others, but still green ground and blue sky, looking at a bizarrely-shaped tree. 5: Green grassy plain, looking at a player or zombie. 6: Looking at the edge between a green plain and a forest, perhaps a thick dark oak forest. 7: Looking at a player, in a green plain filled with tall grass, with one torch stuck among the grass. 8: Looking at a small building and a farm, green ground. 9: Two players looking towards the camera, on a completely flat green landscape, a superflat world.

The output does, undeniably, look like Minecraft, but with the smears and imprecision of all the images it generates. Clearly a lot of its input material on Minecraft was from multiplayer games: four of the nine images I got show other players, or perhaps hybrids between players and zombies. (I wonder if the player skins in the seventh and ninth images could be identified and traced back to specific players.) They're also all clearly screenshots (or frames of let's-play videos), since they all contain the toolbar and most attempt to show the health bar. (The fourth image is perhaps based on a console version screenshot, since its toolbar is floating.)

What I also find interesting is how all of the screenshots are of the green plains biome, with a few trees. (The ninth image looks like a superflat world, though.) Are plains biomes overwhelmingly popular among those whose images ended up in the dataset? Do people not like playing in deserts, savannahs, or mountains?

Minecraft cave

Prompt: minecraft cave.

1: A small Minecraft cave, four blocks wide, looking at a few dropped items. 2: A bit wider and taller Minecraft cave, around 10 blocks wide, with some lava flowing. 3: A tall Minecraft cave, with some green blocks down at the floor. 4: A low and narrow cave, and one red mushroom, looking towards a mineshaft and underground waterfalls. Two toolbars. 5: A taller cave, two toolbars. 6: A large cave, some greenery, a water puddle, a few farmland blocks? 7: A large cave, looking slightly downwards, quite dark. 8: A large cave, quite dark. 9: A greenish-tinged cave, with a row of water flowing towards a player; a farm under construction.

I had to specifically request caves to get images of Minecraft not on the surface of the world. It's interesting how all the toolbars have a greenish tinge on them now. A few of the pictures (the sixth, the ninth) mgiht show attempts at building an underground farm. Also interesting is that there are no clear occurrences of torches.

Combining Minecraft and World of Warcraft

Prompt: minecraft character in orgrimmar.

1: A closeup of a player character in Minecraft, at the seashore. 2: A closeup of a player character in Minecraft, on a hilly beach. 3: A closeup of a player character in Minecraft, on a grassy plain. 4: A closeup of a player character in Minecraft, on a grassy plain. 5: A closeup of a default player character in Minecraft, on a grassy plain. 6: A closeup of a player character in Minecraft, on a grassy hill. 7: A closeup of a player character in Minecraft, on a grassy plain. 8: A closeup of a player character in Minecraft, on a grassy hill. 9: Somebody's house in Minecraft, on the top of a sandy hill.

I wanted a view of the World of Warcraft city of Orgrimmar (which is mostly stone buildings in a reddish-brown valley) with Minecraft-style humanoids in it instead of, or in addition to, orcs and undead and tauren. Instead I just got closeups of Minecraft players, with the interesting exception of the ninth image, which is somebody's house.

Tetris

Prompt: tetris.

1: Brightly-colored Tetris blocks on a black backround. 2: Brightly-colored Tetris blocks on a black backround. 3: Brightly-colored Tetris blocks on a black backround. 4: Brightly-colored Tetris blocks on a black backround. 5: Brightly-colored Tetris blocks on a black backround. 6: Brightly-colored Tetris blocks on a black backround. 7: Brightly-colored Tetris blocks on a black backround. 8: Brightly-colored Tetris blocks on a black backround. 9: Brightly-colored Tetris blocks on a black backround.

It generates Tetris blocks, but they are less regular than I would've guessed.

Dwarf Fortress

Prompt: dwarf fortress.

1: Black background, green blobs. 2: Grey background, divided by a blue blob. 3: Black background, green blobs. 4: Brown background, green, grey and blue blobs. 5: Black background, green, teal and blue blobs. 6: Black background, green and grey blobs, brown rectangles. 7: Black background, green, blue and grey blobs. 8: Black background, mostly blue blobs. 9: Mostly grey, surrounded by some green blobs.

What does the incredibly complex fantasy world generator Dwarf Fortress, with its (descriptions of) elaborate carved artwork, (descriptions of) fine literature, (descriptions of) legendary artifacts adorned with hanging rings of precious metals and menacing spikes of gemstones, and its generated forgotten beats of gigantic hairy winged snails spewing poisonous dust, enormous undulating eyeless lobsters, skinless antenna-having dimetrodons, eight-legged crocodiles with amethyst hair, and so on, look like? It looks like its screenshots, not its fan art. What do its screenshots look like? They look like a black background, filled with axis-aligned rectangles and polygons of primary colors, mostly green (for above-ground ground), grey (for rock and the dug-out fortresses), and blue (for the copious amounts of water both found underground, and channeled into often-fatal mechanisms by the dwarves, who end up flooding their homes).

Forgotten beasts of Dwarf Fortress

Prompt: dwarf fortress forgotten beast.

1: Grey tangly path-like shapes on a black background. 2: Left half green, right half blue, both with various specks of color. 3: Narrow green paths on a black background. 4: Left half dark green with blue specks, other half black with a wide teal path through it. 5: Left half green and brown, right half bluish-black, both specked with color. 6: Green and brown specks and paths on a black background. 7: Brown courtyards and a brightly-colored stockpile on a green background. 8: Small but tall rectangles of color on a black background, then a grey blob with green paths through it. 9: Highly speckled green area, with an almost regular matrix of brown specks (could be tree trunks); on both sides are grey blobs, so perhaps a forested hilltop.

The model doesn't know the descriptions of DF's forgotten beasts, so it just gave me more screenshots. Oddly, these have more detail in them, much smaller rectangles. The second picture might even show a bit of the interface, at the top.

I grabbed some descriptions from the DF wiki page on forgotten beasts and pasted them in.

Ica Oilypride the Assault of Onslaught

Prompt: Ica Oilypride the Assault of Onslaught was a forgotten beast. A huge hairy slug. It has thin wings of stretched skin and it undulates rhythmically. Its pine green hair is unkempt. Beware its webs!

1: A brown slug with green hair. 2: Divided by a horizontal divider into two pictures: a green mess of hair in the upper half, and a brown hairless slug in the bottom half. 3: A green slug with green hair. 4: A brown slug with green hair. 5: A closeup of the antennae of a green hairy slug. 6: A greyish-green slug with bright neon green hair. 7: Divided by a horizontal divider into two pictures: a bright neon green mess of hair in the upper picture, and in the lower picture a slug with a brown front and banana-yellow back. 8: Divided into two images, one above the other: on top, a vaguely slug-like shape with spiky green hair, and on bottom, a slug that's black at the back, and at the front is green, and has hair sprouting up from its back at the back half. 9: A brown slug with green hair, with the hair growing out from below it.

Quite gross. It didn't really get the wings, webs or size, but did get the pine-green hair.

Sedme

Prompt: Sedme was a forgotten beast. An enormous porcupine with external ribs. It has a fat, bulging trunk and it is ravening. Its brown hair is long and straight. Beware its hunger for warm blood!

1: A porcupine, standing on its hind legs. White background. 2: A porcupine. White background. 3: A porcupine, eating a snake or similar. Naturalistic desert background. 4: Two porcupines, one big and one small. 5: A porcupine, but its back half blends into the green grass of the background, so it looks like it only has two legs. 6: A porcupine. Natural background, gravel and grass. 7: A porcupine. Natural background, stone and grass. 8: A porcupine. Natural background, dirt and grass. 9: A porcupine. Natural background, grass.

These are just normal porcupines.

Aspad Echoedmurk

Prompt: Aspad Echoedmurk was a forgotten beast. A towering humanoid composed of grime and filth. It has wings and it squirms and fidgets. Beware its deadly dust!

Don't know why it assumed the beast was green, but it certainly is a winged humanoid. Except for the last picture, which looks like a McDonald's Happy Meal toy. I like the second and fourth ones the most. Also interesting is that with this set of pictures it decided to use grey and black backgrounds, instead of the white in some others.

Counter-Strike

Prompt: counter-strike.

1: Counter-Strike, de_dust2-like. 2: Counter-Strike, de_dust2-like. 3: Counter-Strike, de_dust2-like. 4: Counter-Strike, de_dust2-like. 5: Counter-Strike, de_dust2-like. 6: Counter-Strike, de_dust2-like. 7: Counter-Strike, de_dust2-like. 8: Counter-Strike, de_dust2-like. 9: Counter-Strike, de_dust2-like.

Counter-Strike looks like de_dust2, by far its most popular map. It both exactly looks like de_dust2, and not at all: I don't recognize any of these spaces actually being part of the map, but the style is spot-on. I don't know what that stuff at the top of the images are, maybe part of some popular stream layout, maybe some CS1.6 or CS:GO thing (I'm mainly familiar with CS:Source).

Chess

Prompt: chess.

1: Chess stock image, white pieces. 2: Chess stock image, white pieces. 3: Chess stock image, white pieces with one black pawn. 4: Chess stock image, white pieces. 5: Chess stock image, white pieces. 6: Chess stock image, mainly white pieces. 7: Chess stock image, mainly black pieces. 8: Chess stock image, black pieces. 9: Chess stock image, white pieces.

I thought the output of this one would be 2d diagrams of chess positions, but instead they're all stock images of chess sets from a low angle.

RuneScape

Prompt: runescape.

1: A couple of players on mostly grey ground. 2: A couple of players on green ground. 3: A couple of players on mostly-green ground. 4: A couple of players on mostly-green ground. 5: A couple of players on a cobbled path, next to bright green ground. 6: A couple of players on mostly-green ground, next to a brown patch. 7: A couple of players on light-green ground. 8: A couple of players on green ground. 9: A couple of players on mostly-green ground.

This must be modern RuneScape, rather than the old-school version, based on the interface and the textured nature of the landscape. I don't recognize any of the locations, but they look very generic anyway: cobbled path, surrounded by greenery. The players are quite small and far from the camera.

World of Warcraft

Prompt: world of warcraft.

1: WoW, outside. 2: WoW, outside. 3: WoW, outside. 4: WoW, inside, dungeon. 5: WoW, outside. 6: Close-up of a WoW character. 7: WoW, inside. 8: WoW, outside. 9: WoW, outside.

These are more varied than the other game screenshots. These players are experienced: full toolbars and a bunch of addons, in either dungeons or PvP battlegrounds. The locations are almost identifiable, and they definitely fit the World of Warcraft art style and geography. The sixth image stands out: it's a close-up, no-UI screenshot of a highly-geared Tauren character, which has a demoniac greenish aura. What also stands out is that all of the outdoor images are seemingly taken at dusk or night.

Cities in World of Warcraft

Prompt: world of warcraft capital city.

1: A city at night. 2: A city among mountains, from a distance, with tall trees. 3: A city with stone constructions and trees. 4: A city with purplish wood buildings and lots of greenery. 5: A city of red-roofed round buildings and light brown ground, with some green at the forefront. 6: Many stone paths twisting through a greenish-brown landscape, mountains in the background. 7: Stone walls and towers and a floating building in a green landscape, looks ruined. 8: A red tower and brown buildings on a reddish-brown landscape, with a greenish mountain in the background. 9: A city of stone towers and stone paths, in a green landscape with mountains in the background, viewed from a distance and from above.

I wondered if I'd get pictures of Stormwind, Orgrimmar or Shattrath City with this prompt. Instead, these look to be completely new cityscapes, taking only inspiration from WoW. All of them are generated from a high angle, and all except the eighth contain traces of the user interface. The fourth one looks a bit like the night elf city of Darnassus, the fifth and eighth ones are kind of like the orc city of Orgrimmar (the fifth is a bit too green, though), and the rest look like various human cities, with lots of rather light stone and tall buildings.

Things on the computer

Linux

Prompt: linux.

1: A very malformed Tux the penguin. 2: A very malformed Tux the penguin. 3: A very malformed Tux the penguin. 4: A very malformed Tux the penguin. 5: A very malformed Tux the penguin. 6: A very malformed Tux the penguin. 7: A very malformed Tux the penguin. 8: A very malformed Tux the penguin. 9: A very malformed Tux the penguin.

Linux is simply its mascot, Tux the penguin.

Linux hacker

Prompt: linux hacker.

1: A malformed Tux, with Matrix code raining down around him. 2: A malformed Tux, sitting on a shiny floor with a green reflection and dark walls. 3: A malformed Tux, with a small green rectangle to his left. 4: A malformed Tux, sitting under a very bright light. 5: A malformed Tux, sitting in a circle of green code. 6: A malformed Tux, being beamed away by a Star Trek transporter. 7: A malformed Tux, in a black void. 8: A malformed Tux, in front of a hazy green circle. 9: A malformed Tux, sitting on a nonderscript surface.

I was hoping to get pictures of people looking at screens of green text on a black background. Instead, I got Tux in the Matrix.

Internet

Prompt: internet.

1: A glossy globe on a white background. 2: A glossy globe on a white background, probed by middle fingers. 3: A glossy globe on a white background. 4: A glossy globe on a white background. 5: A glossy globe on a white background, prodded by hands. 6: Two glossy globes, one green and one blue, on a white background. 7: A glossy globe on a white background. 8: A glossy globe on a white background, with hand-colored blobs. 9: A blue glossy globe on a white background, held by malformed hands.

The internet is stock images of globes.

Code

Prompt: code.

1: Blue and turquoise lines of characters, a shadow in the middle. 2: Blue, green and dark red lines of characters. 3: Bluish-green, cyan and blue lines of characters, getting bluer towards the bottom. 4: Green and cyan lines of characters, fading out towards the bottom. 5: Yellowish-green lines of characters, getting sparser towards the bottom. 6: Very regular blue and green lines of characters, looks almost like a waterfall spectral chart. 7: Blue lines of characters, nogt entirely straight, interrupted by a red and yellow blob in the middle. 8: Lines of bright green characters. 9: Lines of green characters, and the silhouette of a standing person seen from the waist up.

Code is what I thought Linux would look like. I don't even know how to write alt text for these pictures.

Abstract concepts

Truth

Prompt: truth.

1: A golden statue in a Statue of Liberty -like pose. 2: A bronze statue in a Statue of Liberty -like pose, blue sky background. 3: A Christian cross on what could be a hand, pink background. 4: A statue holding a sword and one half of a balancing scale, white wall background 5: Silhouette of a statue of Lady Justice, orange-red background. 6: Silhouette of a statue of Lady Justice, grey background. 7: Stylized white statue of Lady Justice, without a head, pinkish-yellow background. 8: Bronze statue in a Statue of Liberty -like pose, whiteish-yellow background. 9: Very stylized statue of Lady Justice, lit from behind, blueish-green material, yellow background.

Truth is statues of female-like figures, holding scales. Truth is also a Christian cross. All of these look like stock images.

Beauty

Prompt: beauty.

1: A woman in a make-up ad. 2: A woman in a make-up ad. 3: A woman in a make-up ad. 4: A woman in a make-up ad. 5: A woman in a make-up ad. 6: A woman in a make-up ad. 7: A woman in a make-up ad. 8: A woman in a make-up ad. 9: A woman in a make-up ad.

Beauty is the faces of East Asian women in make-up ads, highlighting the biases of the model and its original dataset.

Math

Prompt: math.

$1: Black curvy lines drawn on lined paper.$ $2: Various blue lines on a white background.$ $3: Rather regular white squiggles on a green background.$ $4: Green, blue and yellow blobs on a white background.$ $5: Mainly black writing-like dots on a white background, but also five large green shapes that vaguely resemble numbers.$ $6: Sixteen circles in a 4-by-4 grid, each with a shape in it, looks like blue ballpoint ink on a white background.$ $7: Blue shapes that vaguely look like numbers rather regularly placed in columns of a white background.$ $8: Black and red shapes, mostly square-shaped squiggles, very regularly placed on a white background, as 11 lines of first 7 characters, a space, then three more characters.$ $9: White squiggly characters, in approximately a 5-by-5 grid, on a black background.$

Math is writing and scribnbles on either paper or a blackboard, that might be green. Some of the writing looks like Mayan writing.

Hate

Prompt: hate.

1: White background, a face held in hands. 2: Black backround, white text-like shapes. 3: Black backround, white text-like shapes on top, pinkish-red lower half of an angry face. 4: Black backround, white text-like squiggles on top of a skin-colored blob. 5: Black backround, a white triangle above two white drawn faces. 6: White background, a drawn face below black text-like shapes. 7: White background, a drawn and slightly-colored face below text-like shapes. 8: White background, text-like shapes bleeding blood. 9: White background, a hand, showing its palm, with text-like shapes on the palm.

Hate is rather angry abstract art. These could almost be album covers.

Love

Prompt: love.

1: Three hearts, one made of wire, the other two looking like rose petals. 2: One heart. 3: Three full hearts, two half-hearts cut off by the edges of the picture. 4: Two hearts. 5: Two hearts, one red, one made of wire. 6: One fully red heart and one outline of a heart. 7: One heart. 8: A heart made of glass, on a beach at an orange sunset. 9: One heart.

Love is just red hearts on pink backgrounds.

Places

Washington

Prompt: washington. The prompt is intentionally vague.

1: Capitol Building, Washington DC, from a distance. 2: A forested landscape, a river, and a snowy mountain in the distance. 3: Capitol Building, Washington DC, up close. 4: Capitol Building, Washington DC, from a distance. 5: Capitol Building, Washington DC, up close, dawn. 6: Capitol Building, Washington DC, from a distance. 7: Landscape of a body of water, from its shore. 8: Mostly sky: the roof of a building is at the lower edge, with a tree (a conifer of some kind) popping up from behind it. 9: Capitol Building, Washington DC, from close up.

I didn't specify if I meant the state, the city of Washington DC, or the president. I got in return a mixture of both landscapes of the state of Washington, and a few pictures of the city of Washington DC, which consists entirely and only of Capitol Hill.

Tokyo

Prompt: tokyo.

1: Skyscrapers at day, overlooking a busy street. 2: Skyscrapers at day. 3: Skyscrapers at day. 4: Skyscrapers at day. 5: Skyscrapers at night, with a Tokyo Tower -colored triangle-shaped building in the background. 6: Skyscrapers at day. 7: Skyscrapers at day. 8: Skyscrapers at day. 9: Skyscrapers at day.

I recognize Tokyo Tower in the fifth picture, but don't recognize any of these other skyscrapers. They look plausible.

Paris

Prompt: Paris.

1: The Eiffel Tower, from a plaza at its base. Blue but smoggy sky. 2: The Eiffel Tower, from afar and high up. Light grey sky 3: The Eiffel Tower, from a medium distance. Light blue sky. 4: The Eiffel Tower, from afar, also showing the Champ de Mars park. Grey cloudy sky. 5: The Eiffel Tower, from afar, half-illuminated orange. A sunrise or sunset, with the sky being a gradient from yellowish-orange at the horizon to blue at the top. 6: The Eiffel Tower, from afar. Almost black-and-white, not quite. A tiny bit of horizon at the bottom, and two weird figures in the lower right corner, that look perhaps like squashed-up narrow tall-headed humans observing it. 7: The Eiffel Tower, from very close up. Blue sky. 8: The Eiffel Tower, from a medium distance. Picture is essentially black-and-white. Low skyline. Slightly cloudy, but a bright sky. 9: The Eiffel Tower, highly stylized and painterly, with a busy but low city to its right and a featureless green plain to its left.

I predicted I would get nine pictures of the Eiffel Tower, which is what most tourists take a picture of. I was absolutely right. I do like that this has different lighting, instead of it just being sunny. The eighth picture looks like it imitates an early-20th-century black-and-white photo.

Helsinki

Prompt: helsinki.

1: Helsinki Cathedral, from the front, from Cathedral Square. 2: Helsinki Cathedral, from the front, from Cathedral Square. 3: Helsinki Cathedral, from the front, from Cathedral Square. 4: Helsinki Cathedral, from the front, from Cathedral Square. 5: Helsinki Cathedral, from the front, from Cathedral Square. 6: Helsinki Cathedral, from the front, from Cathedral Square. 7: Helsinki Cathedral, from the front, from Cathedral Square. 8: Helsinki Cathedral, from the front, from Cathedral Square. 9: Helsinki Cathedral, from the front, from Cathedral Square.

My guess was that I would get nine pictures of Helsinki Cathedral, which is what most tourists take a picture of. I was absolutely right. I didn't expect that all of them would have the same painterly style. And since the tourists only take pictures in the summer, the sky is always blue.

London

Prompt: london.

1: Tower Bridge. 2: Big Ben. 3: Big Ben, with a bulging clock face. 4: Tower Bridge. 5: Big Ben, with a bulbous clock face. 6: Tower Bridge. 7: Big Ben. 8: Tower Bridge. 9: Big Ben, slightly bent.

London is an improvement over other cities: it has two landmarks! Both Tower Bridge and Big Ben appear equally often. I find it pretty funny that it's always miserably cloudy in London.

Berlin

Prompt: berlin.

Beforehand, I didn't know what I would get. Berlin doesn't have an immediately-obvious singular tourist attraction or scene-setting sight like Paris, nor is there just a pair like in London. I could think of four main things: the Brandenburg Gate, the Reichstag building, the Holocaust memorial, and the TV tower overlooking Alexanderplatz.

The model seemed to agree with me on three of these, with the Fernsehturm being most prominent, being synthesized next to both the Brandenburger Tor and the Reichstag building. I especially like the fifth image: the Brandenburg gate, with the Reichstag's glass dome, with the TV tower in the background. The fourth image is similar, but the dome of the Reichstag building replaces the sphere of the TV tower rather than the horses of the gate. The third and eigth images also have a particularly bright red object hanging off the roof of a building – I wonder if these could be from colorized versions of Raising a Flag over the Reichstag?

Other stuff

An avocado in the style of Piet Mondrian

Prompt: avocado in the style of mondrian.

1: An avocado, with colored squares behind it. 2: An avocado, with colored squares in front of it. 3: An avocado, with colored squares behind it. 4: An avocado, with colored squares inside it. 5: Two avocados, one with yellow rectangles behind it.. 6: An avocado, sliced in half. 7: Two avocadoes, sliced in half. 8: An avocado, sliced in half. 9: An avocado, sliced in half.

None of these are really an avocado in the style of Mondrian. It has no idea what I wanted.

Rats rotating cubes

Prompt: 3d rats rotating cubes in their minds.

1: Blue and brown cubes, with round black holes in them, on a grey gradient background, with two of the cubes having pink noses. 2: A large cube made from eight smaller cubes, which are grey with black edges, with black holes in their centers, and the cube on the lowest forward corner has two black eyes and a pink nose and is misshapen, approximately rat-shaped. 3: Five cubes, with black hole eyes, floating above a grey plane. The biggest cube is blue, another is yellow, two are red, and the last one, partly off-screen, is green. 4: Seven colored cubes (four green, two blue, one pink) arranged in a circle, each with eyes or other facial features. 5: Ten colored cubes floating on a grey plane. 6: A bunch of cuboid shapes, mashed into each other, with "eyes" (black circles) mixed within them. 7: An arch-like shape built our of translucent light blue cubes, some of them with eyes black eyes. 8: Blue, green and red cubes arranged on a grey plane. Only one of them, a red one, has an eye. 9: Six blue cubes stacked in a roughly cuboid shape, most of them with eyes.

I got pretty much what I wanted. A combination of two memes: the "can you rotate a 3d cube in your mind?" meme, and the horizontally spinning rat genre of videos.

Dinosaur Comics

Prompt: dinosaur comics

1: Three panels, top two with Utahraptor (orange dinosaur) on a grassy background, bottom panel with a very big green Sauropod-like dinosaur on a white background. An attempt at text bubbles was made. 2: Two panels, vertically divided, with mostly-white backgrounds with a cliff in the background. In both panels, an orange Utahraptor-like creature. 3: Three panels. Top panel: green T-Rex and a Winnie the Pooh -like shape. Bottom left: brown background, vague green shape. Bottom right: brown background, vague orange shape. 4: Three panels, all with white backgrounds. Top panel: an almost-accurate closeup of T-Rex. Bottom left: Utahraptor, from a medium distance. Bottom right: close-up of Utahraptor. 5: Two panels, each with a couple of text bubbles. Top panel: a long-necked dinosaur from very far away, and vague shapes that could either be a forest or a city, also a pyramid that appears to have green wings. Bottom panel: taller than the top panel,k forested background, Utahraptor seemingly lying on the ground, eating. 6: Two panels. Top panel: T-Rex exclaiming loudly like in the second panel of real Dinosaur Comics comics. Bottom panel: a green shape further away, maybe T-Rex's tail, but it's quite thick to be it. Both panels have a white sky, but have some features strewn about the ground, perhaps a thick tree trunk or a large rock on the left. 7: Two panels, yellow skies, an attempt at a text bubble in first panel. Top panel: T-Rex, standing, seen from the side, with a city and a brown mountain in the background. Bottom panel: Utahraptor, standing, seen from the side, the city is gone but the mountain remains. 8: One panel, much more painterly than the other pictures. Utahraptor, but with a round face, almost like a human's. Utahraptor only has one arm. An empty white text bubble emerges to the left of Utahraptor's head. Grassy floor, blue sky, trees in background

I was particularly interested by this prompt, because Dinosaur Comics (i.e. Qwantz) is special in that every single one (with a few exceptions) of its almost 4000 comics looks exactly the same, with no backgrounds, the same three characters (the green T-Rex, the orange Utahraptor, and the pale Dromiceiomimus), in the same poses.

I was very pleased to see that it recognized the different dinosaur characters, and put them in different poses and situations. The fourth image is the most like an actual comic.

I have thoughts on the general use and ethicality of such a system; the main gripe is the mass use of the work of others, as in, none of the source pictures were originally created by the team who built the model, it's just a huge dataset of images taken from the internet, without consent, although the law doesn't even need consent, and it can be argued that consent for your picture (whether a photo, or a drawing or a painting) of whatever to be one of n million isn't needed. Still, I have some reservations using this. Just having a bit of fun here though, while it's still available – I can foresee a future where in 20 years such a model would be far too expensive to run, available to the public. I appreciate that the authors have written about the biases and limitations of the model (archived version), and have asked people not to use it to generate bad stuff.

Created 2022-06-23, updated 2022-07-04.