AI that creates Mindblowing Art (Free)

[Space]

Looks very similar to this guys work:

https://www.instagram.com/riabovitchev/

[ItsNick]

It becomes interesting when real artists start to use AI as an aid.

Like this:

 

[Loba]

@ItsNick I plan to do that, to use the AI images underneath my collages.

That was amazing btw.  I like that I can make meaning with the AI images

Edited May 14, 2022 by Loba

[zurew]

On 2022. 05. 13. at 5:53 PM, Federico del pueblo said:

It can create really good horror images. This tool is way too good.

 

[Federico del pueblo]

24 minutes ago, zurew said:

It can create really good horror images. This tool is way too good.

Yes! Those are some creepy faces ?

I made some more too.

[zurew]

3 minutes ago, Federico del pueblo said:

Giving Elon Musk some tips, how to make a Mars colony 

[Federico del pueblo]

8 minutes ago, zurew said:

Giving Elon Musk some tips, how to make a Mars colony 

Exactly. 

Elon is gonna have his office in the big tower to the right.

Folks like us will have to live in these smaller bungalows ?

[zurew]

Wow, i couldn't help myself, i had to write "scary alien". This tool is actually addicting.

Edited May 16, 2022 by zurew

[zurew]

https://huggingface.co/spaces/dalle-mini/dalle-mini

This is a different tool, works the same way, the only difference is that it generates 9 pictures in 15 seconds. (and you can't choose any background)

 

Joe rogan, after smoking some DMT.

 

 

Edited May 16, 2022 by zurew

[Leo Gura]

On 5/14/2022 at 11:54 AM, ItsNick said:

It becomes interesting when real artists start to use AI as an aid.

Like this:

 

Wow! This is amazing! I am stealing this for my blog.

[thisintegrated]

Everything in this thread is old tech.  No longer news-worthy.  What you want is "Dalle 2".  Though it's not yet fully open to the public, it's at least an order of magnitude more impressive.

 

https://openai.com/dall-e-2/

Edited May 17, 2022 by thisintegrated

[AtheisticNonduality]

@thisintegrated This stuff is basically just the cut-up technique done robotically, throwing randoms together or making slight readjustments and then calling it impressive. But perhaps it will develop into something greater . . .

[thisintegrated]

3 minutes ago, AtheisticNonduality said:

@thisintegrated This stuff is basically just the cut-up technique done robotically, throwing randoms together or making slight readjustments and then calling it impressive. But perhaps it will develop into something greater . . .

An interesting way to sum up the most advanced machine learning systems known to man?

[AtheisticNonduality]

@thisintegrated I'm not terribly inclined to be amazed with this stuff---well it's not even really stuff, just collage nonsense---but I recognize it's the infrastructure/foundation for something else to come.

Edited May 17, 2022 by AtheisticNonduality

[thisintegrated]

1 minute ago, AtheisticNonduality said:

@thisintegrated I'm not terribly inclined to be amazed with this stuff---well it's not even really stuff, just collage nonsense---but I recognize it's the infrastructure/foundation for something else to come.

Alright.  You've made me do it.

Don't blame me if your mind is blown..

 

https://www.youtube.com/watch?v=U1cF9QCu1rQ

[sleep]

[AtheisticNonduality]

19 minutes ago, thisintegrated said:

Alright.  You've made me do it.

Don't blame me if your mind is blown..

 

https://www.youtube.com/watch?v=U1cF9QCu1rQ

They attempted to distinguish it from random images, from the cut-up technique, by pointing out how it has a sort of logical synthesis to it and how they pull from "aesthetic" data. And yet, so far, I remain unimpressed. The images are missing something: they all look the same and feel the same. There's no element of creation, just of computing something that was already created in the text into a factory-type process. New creations are not existent in those programs, solely hogwash that allows programmers' self-aggrandizement.

Edited May 17, 2022 by AtheisticNonduality

[thisintegrated]

3 minutes ago, AtheisticNonduality said:

They attempted to distinguish it from random images, from the cut-up technique, by pointing out how it has a sort of logical synthesis to it and how they pull from "aesthetic" data. And yet, so far, I remain unimpressed. The images are missing something: they all look the same and feel the same. There's no element of creation, just of computing something that was already created in the text into a factory-type process. New creations are not existent in those programs, solely hogwash that allows programmers' self-aggrandizement.

Designing a space suit for a dolphin kind of requires creativity.  There exists no data on what this should look like, yet the algorithm manages to make one.

 

There is no limit to what it can make

Edited May 17, 2022 by thisintegrated

[Loba]

Similars, but different.

[AtheisticNonduality]

17 minutes ago, thisintegrated said:

Designing a space suit for a dolphin kind of requires creativity.  There exists no data on what this should look like, yet the algorithm manages to make one.

That's the logical synthesis I said.

Quote

There is no limit to what it can make

It is limited to what instructions it receives. So far it does not have a consciousness of its own that arose out of eternities of evolutionary complexification; it has no mind of its own to create its own worlds. It's still a set of pathways that mesh different human-given "datas" together and is not able to survive outside the womb so to speak, is still dependent on instruction and then reassembling data points. It has no creativity since what it creates is essentially nothing, worthless, bad, so it cannot be said to be a creator but rather an ungodly stagnation. It cannot have any existing reference for what a dolphin wearing a space helmet looks like exactly, but it still has the proper parts given as "dolphin" and "wearing" and "space helmet" and such. It may "understand" what a dolphin is, what pattern that a dolphin's nose has, what it means spatially/dimensionally to wear something, that a space helmet adheres directly around an object's boundaries rather than ten feet from it, etc. It's just basic logic skills, like how machines play chess. But it's not creative in the slightest, since to be a creator requires agency, and nothing has less agency than something that literally requires programming to function.

Edited May 17, 2022 by AtheisticNonduality