AI generated art

[Mon0Zer0]

I don't think so. People don't generally read a lot of Balzac and then write a series of Balzac-like things.

Because humans aren't restricted to just one set of training data, but everything we experience in life trains our representational models. You can pick out influences in people's writing, and some writers do write like other writers.

The AI isn't trained to think like Berkley or to have a rationalization of technology like Berkley.

This is true, there's no sense that it understands what it's making in any deeper sense than the probabilities of a set of pixels appearing in certain positions on a field.

However, we don't really understand how human vision or creativity actually works either - so we don't know whether the understanding is outside of conscious experience or not. Chomsky seems to think 80%+ processing happens outside consciousness.

The next level of adversarial / GAN networks are far beyond what I'm using here (see David Martinez's work above) and it seems to have an intuitive sense for reflections in water and other phenomena which suggests that it has some level of "understanding" - An AI which can rationalise technology may not be that far away.

Personally, I like the impossible nature of the images - it's an aesthetic that's valid in itself even if it's not a perfect representation of "real painting". As starting points for digital paintings they're also useful.

 

[Mon0Zer0]

Imagen: Text-to-Image Diffusion Models

imagen.research.google

 

[Swank]

The next level of adversarial / GAN networks are far beyond what I'm using here (see David Martinez's work above) and it seems to have an intuitive sense for reflections in water and other phenomena which suggests that it has some level of "understanding" - An AI which can rationalise technology may not be that far away.

If it can understand creating symmetrical objects in vanishing point perspective, it can understand reflections. They are the same kind of math problem.

I like these images, but I feel like their generation doesn't accomplish anything but produce visual spam versions of people work. I'm sure they will soon take over the self published cover art market. (Which might be a good thing compared to the schlock often assembled from bad clip art.)

 

[Rodders]

Is this the art equivalent of bot written scripts?

 

[Mon0Zer0]

Is this the art equivalent of bot written scripts?

I don't know enough about that to say, tbh.

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[Toby Frost]

Wow!

 

[CupofJoe]

I like the images but I wonder how many failed/nonsensical images this system produces?
I'll be impressed with AI when it decides that it really wants to be in a band and starts learning the guitar.

 

[Mon0Zer0]

I like the images but I wonder how many failed/nonsensical images this system produces?
I'll be impressed with AI when it decides that it really wants to be in a band and starts learning the guitar.

From playing with it for a few days, it depends. It seems to like Landscape more than Portrait - portrait you get odd perspective errors.
It is excellent at fantasy landscapes, not so much at actual objects like a car or a spaceship. People's faces are always weird.

You can also play with some of the underlying code to get better results - getting more seeds (and therefore more odd shapes splattered around randomly).

Anything which is too far away from the training data gives you odd results.

Here's a failed one: "Dragon fights soldiers in the polish countryside."

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[Mouse]

Heh. Weirdly, I discovered AI generated art only yesterday and been a bit obsessed with it since. Mine is... erm... weird, though. Maybe it's my prompts.

I call this terrifying image "Guinea Pig with a Doughnut"

And there's this joy of joys:

And um... this is a thing, too. "Dog with Boy Face"

Mine are all wrong.

 

[Mon0Zer0]

Heh. Weirdly, I discovered AI generated art only yesterday and been a bit obsessed with it since. Mine is... erm... weird, though. Maybe it's my prompts.

I call this terrifying image "Guinea Pig with a Doughnut"

And um... this is a thing, too. "Dog with Boy Face"

That is TERRIFYING!

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[AE35Unit]

A Fractal mixed with AI (google deep dream)

 

[Mouse]

Ok I finally managed something not terrifying!

 

[Mon0Zer0]

Dragonwasp. Quite heavily edited afterwards in Photoshop. I think the biggest use we'll get from these kinds of technology is to use them as starting points for photobashing / editing.

 

[Mon0Zer0]

So, discovered a feature where, instead of starting with noise, you can start with an image.
Before and after: