Here’s a thread for posting your AI generated images, good and bad, and general discussion of AI as an artistic tool.

We often hear that artists, composers and creative people in general hate AI. In reality of course, many of us are already making enthusiastic use of it.

As an artist, musician and composer myself, I think AI is great fun. I haven’t made use of it in my own work so far but that may change, as it’s particularly useful in, for example, creating videos for music tracks.

I view the AI art landscape as both a graveyard and incubator of the human imagination. Everything we see in these images is born of the dust of what we’ve imagined so far, and can contribute mightily to the creation of new worlds within worlds.

So continuing with the graveyard theme, here’s a little series of images I’ve just generated in Fooocus, in MK Anthotype Print, lowest quality setting.

The feed: “Rusty graveyard of Daleks and robots. Some of the Daleks are standing, other have fallen over and some have decayed into broken pieces.”

Baby Daleks? First I’ve heard of the reproduction cycle of Daleks.

Next I entered: “Rusty graveyard of gothic harps. Some of the harps are standing, other have fallen over and some have decayed into broken pieces.”

But the AI had to scratch its head on this one, and compromised by conflating the harps with headstones, coming up with these unique and charming images.

a different filter and those harps would look full of life..

This one isn’t my prompt, but hilarious nonetheless.

Divine Angel said:

This one isn’t my prompt, but hilarious nonetheless.

It has been shown here before.

Divine Angel said:

This one isn’t my prompt, but hilarious nonetheless.

I’m glad we don’t really have to tune our bows.

AI generated images are wonderful!

Bubblecar said:

Divine Angel said:

This one isn’t my prompt, but hilarious nonetheless.

I’m glad we don’t really have to tune our bows.

:)

Now I’ve entered: “Rusty graveyard of rocking horses. Some of the rocking horses are standing, other have fallen over and some have decayed into broken pieces.”

And the result again is some beguiling and pleasantly disturbing images, of which most mortal artists would be proud.

fsm said:

AI generated images are wonderful!

That’s what they mean by “save thyself”

They are getting quiet impressive indeed.
Some of the Bubblecar images are a touch haunting.

Last year I made a little book for my year 1 students, using AI images of a cat. The prompt was something along the lines of “cat splashing in puddles, with a bird, lizard, and frog watching”.

AI me

Divine Angel said:

AI me

Not too bad!

Divine Angel said:

AI me

It is recognisable.

Divine Angel said:

AI me

Nailed me as well!

Spiny Norman said:

Divine Angel said:

AI me

Nailed me as well!

Nup.

Spiny Norman said:

Divine Angel said:

AI me

Nailed me as well!

Even got your third arm!

Follow the potatoes.

Interesting to see how much of a prompt the AI can address, when you’re generating images on low-quality, supposedly high-speed setting.

(All the AI image generators are optimised for NVIDIA GPUs. I’m running Fooocus on my AMD machine and it’s veeery slow as a result, even on “Lightning”. I’m going to invest in a latest generation NVIDIA graphics card).

My prompt for this one: “Patrick Troughton, dressed as an old monk, is asking a very old monkey to marry him. In the background there is an observatory, half hidden by solemn trees.”

My prompt: “Here below the mandelbrot, the sky descends and onions rot.”

Bubblecar said:

We often hear that artists, composers and creative people in general hate AI. In reality of course, many of us are already making enthusiastic use of it.

well yeah the traditionalists probably thought that young upstarts with a bit of light sensitive paper and splashy chemicals were just a bunch of plagiarizing fuckers as well back in the day

SCIENCE said:

Bubblecar said:

We often hear that artists, composers and creative people in general hate AI. In reality of course, many of us are already making enthusiastic use of it.

well yeah the traditionalists probably thought that young upstarts with a bit of light sensitive paper and splashy chemicals were just a bunch of plagiarizing fuckers as well back in the day

They certainly did.

I could play all day with this stuff. Got to get that NVIDIA card.

“Robots attacking a huge pile of cakes. In the background is a mushroom cloud.”

Bubblecar said:

I could play all day with this stuff. Got to get that NVIDIA card.

“Robots attacking a huge pile of cakes. In the background is a mushroom cloud.”

These look like a computers dreams.

I guess they are, in a way.

doesn’t each 爱gen murder like 60000 native animals or something like that

Last few for tonight.

“Here come the cabbage men, green and leafy.”

Bubblecar said:

Last few for tonight.

“Here come the cabbage men, green and leafy.”

i think some of them are the most successful but mostly i did not enjoy them overly.

sarahs mum said:

Bubblecar said:

Last few for tonight.

“Here come the cabbage men, green and leafy.”

i think some of them are the most successful but mostly i did not enjoy them overly.

I prefer the sepia/old photo style ones, but as I say, until I get the new graphics card it’s painfully slow to run this stuff on my own computer (despite me having a fairly powerful machine).

On the online image generators, the images are generated very quickly. That’s because they don’t rely on my own computer and its AMD GPU. But I far prefer to install it all and use it on my own machine.

Bubblecar said:

Interesting to see how much of a prompt the AI can address, when you’re generating images on low-quality, supposedly high-speed setting.

(All the AI image generators are optimised for NVIDIA GPUs. I’m running Fooocus on my AMD machine and it’s veeery slow as a result, even on “Lightning”. I’m going to invest in a latest generation NVIDIA graphics card).

My prompt for this one: “Patrick Troughton, dressed as an old monk, is asking a very old monkey to marry him. In the background there is an observatory, half hidden by solemn trees.”

The trees don’t look that solemn to me.

I used the same prompt and it spat out this glorious image

SCIENCE said:

doesn’t each 爱gen murder like 60000 native animals or something like that

How many do they eat?

roughbarked said:

SCIENCE said:

doesn’t each 爱gen murder like 60000 native animals or something like that

How many do they eat?

fine apparently Reddit claim it’s about 2 kJ per image but then we don’t know what Quora reckon

Divine Angel said:

I used the same prompt and it spat out this glorious image

Heh, nice one.

I can generate such images easily by using the online sites. But I’m working with the basic AI engines on my own computer, which should in theory offer similar speed as the online sites while allowing complete freedom to explore all the potential of the technology.

But for optimal performance it requires one of the latest NVIDIA graphics processing units, which I have now ordered.

Doing a few fancy brass Daleks before going in the shower.

Bubblecar said:

Doing a few fancy brass Daleks before going in the shower.

They seem a bit less menacing than the usual mob.

A few more.

Bubblecar said:

A few more.

I do wonder. Is your AI in any way linked to other AI units?

roughbarked said:

I do wonder. Is your AI in any way linked to other AI units?

No, it’s installed on my computer. Several GBs of it.

Bubblecar said:

roughbarked said:

I do wonder. Is your AI in any way linked to other AI units?

No, it’s installed on my computer. Several GBs of it.

Cos it could get AI all confused like…

Some handsome model steamboats.

My prompt for these ones:

“Winston Churchill and Lucille Ball sitting on fine armchairs in a beautiful living room, drinking glasses of sherry.”

LOL

Churchill’s an amputee, there’s a random lady with a very skinny head, and poor Lucille’s nose job didn’t go well.

Nice work.

Divine Angel said:

LOL

Churchill’s an amputee, there’s a random lady with a very skinny head, and poor Lucille’s nose job didn’t go well.

Nice work.

I like the one where they both have glasses of sherry in both hands.

Bubblecar said:

Divine Angel said:

LOL

Churchill’s an amputee, there’s a random lady with a very skinny head, and poor Lucille’s nose job didn’t go well.

Nice work.

I like the one where they both have glasses of sherry in both hands.

That’s not how you drink sherry?

I like the first one, where the room has two fireplaces, side-by-side. If you’ve got it, flaunt it.

In the last one, Churchill looks like James Spader.

captain_spalding said:

Bubblecar said:

Divine Angel said:

LOL

Churchill’s an amputee, there’s a random lady with a very skinny head, and poor Lucille’s nose job didn’t go well.

Nice work.

I like the one where they both have glasses of sherry in both hands.

That’s not how you drink sherry?

correct, whilst you may have two hands you only have one mouth. another oversight by a supposedly omnipotent being.

ChrispenEvan said:

captain_spalding said:

Bubblecar said:

I like the one where they both have glasses of sherry in both hands.

That’s not how you drink sherry?

correct, whilst you may have two hands you only have one mouth. another oversight by a supposedly omnipotent being.

You’re obviously unfamiliar with how it’s done.

One drains the glass in the left hand, which is refilled while one drains the glass in the right hand, and so on, and so on.

captain_spalding said:

ChrispenEvan said:

captain_spalding said:

That’s not how you drink sherry?

correct, whilst you may have two hands you only have one mouth. another oversight by a supposedly omnipotent being.

You’re obviously unfamiliar with how it’s done.

One drains the glass in the left hand, which is refilled while one drains the glass in the right hand, and so on, and so on.

you still have only one mouth though.

This one was supposed to be:

“Queen Elizabeth II, Groucho Marx and Mother Theresa chatting together at a grand party in a fine old house, while the Three Stooges serve them drinks.”

…but all that clearly over-excited the AI, at such short notice.

ChrispenEvan said:

captain_spalding said:

ChrispenEvan said:

correct, whilst you may have two hands you only have one mouth. another oversight by a supposedly omnipotent being.

You’re obviously unfamiliar with how it’s done.

One drains the glass in the left hand, which is refilled while one drains the glass in the right hand, and so on, and so on.

you still have only one mouth though.

Quite sufficient, once you master the timing of it.

Stick with me kid, i’ll teach you how to behave like ya wuz to the manner born.

And the weird 3-legged woman…

The Rev should enjoy these railway bridge adventures. Few (if any) of them make any engineering sense, but they’re all pleasing nonetheless.

The prompt didn’t match the output but I did enjoy this part of the pic.

AI Jellybean. I fink sumfink went wrong.

Divine Angel said:

AI Jellybean. I fink sumfink went wrong.

Four Ears.

Nothing to do with Jellybeans.

How do you get these pictures?

This one is using the Apple Intelligence app “Playground” on my phone. Also used it for the AI pic of me earlier in the thread.

Divine Angel said:

This one is using the Apple Intelligence app “Playground” on my phone. Also used it for the AI pic of me earlier in the thread.

They should increase the brightness.

Prompt: unicorn in a magical forest. Playground app.

Divine Angel said:

This one is using the Apple Intelligence app “Playground” on my phone. Also used it for the AI pic of me earlier in the thread.

Ta.

Do you give it photos to work with?

Michael V said:

Divine Angel said:

This one is using the Apple Intelligence app “Playground” on my phone. Also used it for the AI pic of me earlier in the thread.

Ta.

Do you give it photos to work with?

Yes, or I can use a prompt. The pic of Jellybean I gave it was this. She was being sooky cos of the cyclone. You can add themes, accessories etc. I added a park setting to Jellybean’s AI pic.

Divine Angel said:

Michael V said:

Divine Angel said:

This one is using the Apple Intelligence app “Playground” on my phone. Also used it for the AI pic of me earlier in the thread.

Ta.

Do you give it photos to work with?

Yes, or I can use a prompt. The pic of Jellybean I gave it was this. She was being sooky cos of the cyclone. You can add themes, accessories etc. I added a park setting to Jellybean’s AI pic.

Appendages seem to be a problem – ears, head horns, fingers, legs etc.

Divine Angel said:

Prompt: unicorn in a magical forest. Playground app.

That’s no unicorn…

Witty Rejoinder said:

Divine Angel said:

Prompt: unicorn in a magical forest. Playground app.

That’s no unicorn…

a bicorn.

Toy soldiers and cannons. It’s not very good with cannons.

Some nice model buses and trams.

It misunderstood this one entirely, but came up with some interesting rooms.

“A fine old tarnished brass model of an arched railway bridge, with towers at each end with lamps, displayed on a wooden table in an old library.”

Bubblecar said:

It misunderstood this one entirely, but came up with some interesting rooms.

“A fine old tarnished brass model of an arched railway bridge, with towers at each end with lamps, displayed on a wooden table in an old library.”

That place would be a bugger to heat.

captain_spalding said:

Bubblecar said:

It misunderstood this one entirely, but came up with some interesting rooms.

“A fine old tarnished brass model of an arched railway bridge, with towers at each end with lamps, displayed on a wooden table in an old library.”

That place would be a bugger to heat.

Just do some global warming and all good¡

SCIENCE said:

captain_spalding said:

Bubblecar said:

It misunderstood this one entirely, but came up with some interesting rooms.

“A fine old tarnished brass model of an arched railway bridge, with towers at each end with lamps, displayed on a wooden table in an old library.”

That place would be a bugger to heat.

Just do some global warming and all good¡

especially since 爱gen is allegedly so emissions heavy this should be simple

“In a clearing in the forest, a group of monks gather round to venerate a huge statue of Bugs Bunny.”

Bubblecar said:

https://www.youtube.com/watch?v=Cwkej79U3ek

SCIENCE said:

Bubblecar said:

https://www.youtube.com/watch?v=Cwkej79U3ek

Breaking a few health and safety plus road user rules there.

Few more dollies graveyard.

Last ones for now.

A few more to cheer you up in the meantime.

Weaver Valley Speedway is a page on Facebook. I’m suspicious that some of the racing pics have been modified.

————————————

“Racing was stoped last nite coz som spaceshit landed on the trak. Luckily the ofishals spotted it early after the entir field hit it.”

The trak inspectas ar tryig out the new testing rig, bran new from the factry.

Nasville hot pidgin is back on the menu the weak end $20

Thanks to goggle steatvew car for finaly getting are trak on the map. He even took out farstet time in the hot laps seshn.

Hears som pix from the granysmasher 2024 event. Some of these girls had a great time, hitin each uther at up to 5 miles an hour.

Baby rideon nite was a big hit, the litle uns lovd it, we only lost one this time.

Modified RC Strolers was well atended with a ful field, and a few pilups. Not too worry, everyon crawld away.

Ar winner for the night, she taeks home haf a gallon of jebs best ‘shine, and half a pack of smokes for the win. Congats Skeeter.

9 minute slide show of Graveyard of the Dolls images, no music. Just a first try.

Graveyard of the Dolls

Set the Quality to 1080 for best resolution.

If you don’t want to sit through 9 minutes of it, you can speed it up to double time :)

Bubblecar said:

9 minute slide show of Graveyard of the Dolls images, no music. Just a first try.

Graveyard of the Dolls

Set the Quality to 1080 for best resolution.

If you don’t want to sit through 9 minutes of it, you can speed it up to double time :)

what if they want to skim the transcript

The first two new images for the Doll’s Graveyard series, generated with the new graphics card. Only took a few seconds.

Bubblecar said:

The first two new images for the Doll’s Graveyard series, generated with the new graphics card. Only took a few seconds.

Ten more. As well as being much faster than the old card, this is on a higher quality setting.

Bubblecar said:

Bubblecar said:

The first two new images for the Doll’s Graveyard series, generated with the new graphics card. Only took a few seconds.

Ten more. As well as being much faster than the old card, this is on a higher quality setting.

how different is the energy consumption or CO2 emissions

SCIENCE said:

Bubblecar said:

Bubblecar said:

The first two new images for the Doll’s Graveyard series, generated with the new graphics card. Only took a few seconds.

Ten more. As well as being much faster than the old card, this is on a higher quality setting.

how different is the energy consumption or CO2 emissions

Probably all for the better. With the AMD card the whole CPU was involved in these images and I couldn’t do much else with the ‘pooter while they were (slowly) generating.

Bubblecar said:

Probably all for the better. With the AMD card the whole CPU was involved in these images and I couldn’t do much else with the ‘pooter while they were (slowly) generating.

I must be missing something, (be polite you lot!) but I can’t see how a better video card would make much difference. The computing to generate the image is (should be) done at the origin site.
True enough you’d have to specify what resolution you can use, maybe that’s the difference?

Spiny Norman said:

Bubblecar said:

Probably all for the better. With the AMD card the whole CPU was involved in these images and I couldn’t do much else with the ‘pooter while they were (slowly) generating.

I must be missing something, (be polite you lot!) but I can’t see how a better video card would make much difference. The computing to generate the image is (should be) done at the origin site.
True enough you’d have to specify what resolution you can use, maybe that’s the difference?

I’m not doing these on a website, I’m using the actual AI image generating software (stable diffusion and others) that I’ve installed on my machine (many GB’s worth :)).

The websites use their own servers to generate the images. If you’re doing it on your own computer, you need a powerful, recent Nvidia graphics card and generally beefy machine.

Bubblecar said:

Spiny Norman said:

Bubblecar said:

Probably all for the better. With the AMD card the whole CPU was involved in these images and I couldn’t do much else with the ‘pooter while they were (slowly) generating.

I must be missing something, (be polite you lot!) but I can’t see how a better video card would make much difference. The computing to generate the image is (should be) done at the origin site.
True enough you’d have to specify what resolution you can use, maybe that’s the difference?

I’m not doing these on a website, I’m using the actual AI image generating software (stable diffusion and others) that I’ve installed on my machine (many GB’s worth :)).

The websites use their own servers to generate the images. If you’re doing it on your own computer, you need a powerful, recent Nvidia graphics card and generally beefy machine.

Ah, that does indeed make sense then, ta.

The school has taken to using AI imagery. This one just appeared in an email about the Anzac Day March.

Divine Angel said:

The school has taken to using AI imagery. This one just appeared in an email about the Anzac Day March.

Hmm, is he really entitled to wear those ribbon bars?

Fancy medieval helicopters.

I hereby invent the word ‘AImaginary’ for whatsits like Car’s creations here and the various bullshit you find on Facebook. Can you patent a word?

Witty Rejoinder said:

I hereby invent the word ‘AImaginary’ for whatsits like Car’s creations here and the various bullshit you find on Facebook. Can you patent a word?

nuh

Witty Rejoinder said:

I hereby invent the word ‘AImaginary’ for whatsits like Car’s creations here and the various bullshit you find on Facebook. Can you patent a word?

No, but you can trademark it.

Fancy brass dirigibles.

Bubblecar said:

Fancy brass dirigibles.

That would be such a perfect Sunday afternoon outing over the hills. Sipping a cold beer or chilled wine, with some good cheeses etc

sigh Pity it is just imaginary.

Alpine tours by brass airship.

I do like the concept.

party_pants said:

I do like the concept.

It’s cosy watching the AI churn these out :)

It’s just standard Juggernaut XL by RunDiffusion running in Fooocus on my pooter, style: Artsyle Steampunk, prompt:

“A fancy medieval brass dirigible, with Celtic decorations and gothic windows, flying above the Alps, beneath a dark cloudy sky.”

…with various settings adjusted.

Takes the software a few seconds to generate each image.

i do like the aesthetic but I wonder what they’re powered by….

sarahs mum said:

i do like the aesthetic but I wonder what they’re powered by….

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

Bubblecar said:

sarahs mum said:

i do like the aesthetic but I wonder what they’re powered by….

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

I’m using Fooocus as the GUI for Stable Diffusion, on Advanced setting which gives a good range of control while still being very straightforward. It’s a free (but very big) download.

Does require a powerful NVidia graphics card and generally powerful machine to work at a decent speed.

>Fooocus is an open source generative artificial intelligence program that allows users to generate images from a text prompt. It uses Stable Diffusion XL as the base model for its image capabilities as well as a collection of default settings and prompts to make the image generation process more streamlined.

https://en.wikipedia.org/wiki/Fooocus

Bubblecar said:

sarahs mum said:

i do like the aesthetic but I wonder what they’re powered by….

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

no. i was worried about keeping all that weight in the air.

sarahs mum said:

Bubblecar said:

sarahs mum said:

i do like the aesthetic but I wonder what they’re powered by….

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

no. i was worried about keeping all that weight in the air.

Ah :)

Don’t worry, they’re not lead zeppelins, it’s very thin brass.

Bubblecar said:

sarahs mum said:

Bubblecar said:

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

no. i was worried about keeping all that weight in the air.

Ah :)

Don’t worry, they’re not lead zeppelins, it’s very thin brass.

must be magic i reckon.

sarahs mum said:

Bubblecar said:

sarahs mum said:

no. i was worried about keeping all that weight in the air.

Ah :)

Don’t worry, they’re not lead zeppelins, it’s very thin brass.

must be magic i reckon.

True, even if they’re full of gas, there’s no visible propulsion system.

I can of course specify that in the prompt:

“A fancy medieval brass dirigible, with Celtic decorations and gothic windows, powered by propellers, flying above the Alps, beneath a dark cloudy sky.”

Done. I assume those things on top are supposed to be propellers :)

sarahs mum said:

Bubblecar said:

sarahs mum said:

i do like the aesthetic but I wonder what they’re powered by….

Free, open source software.

>Stable Diffusion is a deep learning, text-to-image model released in 2022 based on diffusion techniques. The generative artificial intelligence technology is the premier product of Stability AI and is considered to be a part of the ongoing artificial intelligence boom.

It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt. Its development involved researchers from the CompVis Group at Ludwig Maximilian University of Munich and Runway with a computational donation from Stability and training data from non-profit organizations.

Stable Diffusion is a latent diffusion model, a kind of deep generative artificial neural network. Its code and model weights have been released publicly, and an optimized version can run on most consumer hardware equipped with a modest GPU with as little as 2.4 GB VRAM. This marked a departure from previous proprietary text-to-image models such as DALL-E and Midjourney which were accessible only via cloud services.

https://en.wikipedia.org/wiki/Stable_Diffusion

no. i was worried about keeping all that weight in the air.

maybe they could be suspended cable-cars between the mountains :)

A few more before I drift bedwards:

Bubblecar said:

A few more before I drift bedwards:

Love this one

>But I have been generating some fine old drunks in Fooocus and will be working on Music from a Drunken Haze as well as Music from the Graveyard of the Dolls.

A few to be going on with:

I’m glad to see your software is strictly using the Robot Dictionary of Human Hand Shapes.

The Rev Dodgson said:

I’m glad to see your software is strictly using the Robot Dictionary of Human Hand Shapes.

A lot of those old fellows are veterans, they lost fingers in the war. Some of them even gained a few, somehow. I suppose all is fair in love and war.

This lady is wondering why her digit count keeps randomly changing, drink or no drink.

I asked ChatGPT what it thinks I look like. I guess I left out the bit where I’m female.

Australians are being warned against using artificial intelligence (AI) to create fake receipts to fool the Australian Taxation Office (ATO). ChatGPT recently introduced its 4o Image Generation tool, which can create incredibly realistic pictures with just a few prompts.

https://au.finance.yahoo.com/news/ato-warning-as-chatgpts-new-tool-sparks-concerns-over-new-era-of-tax-evasion-200021359.html

Various mummified clowns.

Top pic, second from right. Creepy AF.

Divine Angel said:

Top pic, second from right. Creepy AF.

could not engage.