Bitcoin art created by AI {gallery} ... and the satoshi portrait as well.

[TheBeardedBaby]

Oh nice it is addictive for sure and does amazing job.
It is struggling with logos, I've noticed another non bitcoin logos looking weird.
You can try adding "unreal engine" and 4k as a key words too and see the result.

[1713858763]

1713858763

[TheBeardedBaby]

Bonus Gallery

Some more playing with the AI.

How it will looks like if famous artist from the past would paint bitcoin:

Michelangelo



Leonardo da Vinci




Rembrandt


Vincent van Gogh

[so98nn]

Basically this AI stupid or missing the algorithms which happens to be perfected by Humans?

Hmm, not that much impressive. Maus is 100x better artist than this AI, lolz. I loved the gallery, first I read then post and it was in my back of the mind that it's made by an AI, but soon the motivation faded away after looking at the bitcoin symbol everywhere.

That is simply artwork by the artist and nothing more. We live in the era of Internet fakism peeps! Just be aware what's being sold and whats we are getting.

Just because one image was processed by AI does not mean they are perfect and now whole internet is  flooded with the same.

[yazher]

I tried this out today, its very addictive!

Here are some results from my keywords, it does struggle with the Bitcoin symbol. I would
like to maybe embellish the images with my own design and the Bitcoin symbol like the
full sized ones below in particular.

As you can probably guess two of the key words were: Bitcoin, City!

This looks cool and you only need to give the AI some words for this stunning images to pop-up. But my question is, how do you guys do it? I already joined the discord channel and see if my bitcoin keywords will gonna show some results of those nice artworks as well. It was really nice this came after the bitcointalk 10 years anniversary because it would have been hard to choose the winners when everyone uploading those images.

[kryptqnick]

AI sees what it can, based on the biases of our own online world from which it draws inspiration and learns (or biases of a specific database of imagery or something like that which was given to it). So Satoshi seems to be a white middle-aged guy on the majority of the images. Since there's zero reason to believe it's a white person and little reason to believe it's a guy, that really says something. Or maybe it's the discord people cherry-picking such images, I don't know.
I do like some of the more abstract images, though, and I love how the colours match nicely on them. Surely, some of these could be chosen for an art exhibition.

[fillippone]

I have been playing with this boy today, and I am really impressed with the result.
I think it could be really useful to share the pseudo code and the description used to generate the images.

[TryNinja]

But my question is, how do you guys do it?

Just visit one of the #newbies-XX channels under the "newcomer rooms" cattegory and run /imagine with the prompt you want. You may want to look at a few examples on the same channel to see how to get good results with it.

[aoluain]

But my question is, how do you guys do it?

Just visit one of the #newbies-XX channels under the "newcomer rooms" cattegory and run /imagine with the prompt you want. You may want to look at a few examples on the same channel to see how to get good results with it.

and use as TheBeardedBaby suggests the keywords

"unreal engine" and 4k

So I have expired my FREE invitation and might subscribe for 1 or 2 months for $10
per month just to play around with it a bit more.

Its raining here all weekend so I would love to edit some "I" already created 🤯 to include
the BTC logo but I was thinking what way the © is setup on this, after a little investigating
on their website I came across this:





Everyone also grants them basically unlimited useage of everything created.


Just be careful if you think you can do what you want with the images.

https://midjourney.gitbook.io/docs/terms-of-service

https://creativecommons.org/licenses/by-nc/4.0/legalcode

https://creativecommons.org/licenses/by-nc/4.0/

[TheBeardedBaby]

-snip-

Thank you  for the valuable info, I had no time to read T&C so that was quite helpful.
BTW i tried some other free AI Art Generators, turned out there there are many, but until now midjourney seems like the best one, best quality for the time it takes to generate an image and you can get a High-Res ones too.

[fillippone]

BTW i tried some other free AI Art Generators,

I just want to try Stable Diffusion.
Anyway I am impressed with the level of complexity already available with mid journey.
I saw some cheat sheet on reddit about the various styles and codes available.
Might be interesting to build a little guide here.

[aysg76]

I just want to try Stable Diffusion.
Anyway I am impressed with the level of complexity already available with mid journey.
I saw some cheat sheet on reddit about the various styles and codes available.
Might be interesting to build a little guide here.

They have opened up the applications for their beta || access to the users and you have to fill the form to get yourself notified when it's available and will see forward how it works.

Also have checked their GitHub Version 1 and it looks really cool and above the level of those free AI art generation tool and they are arousing my curiosity with the advancement and will really like to tweak with the codes to see how much capable they are.

We might see many more tools coming up that will really have that orginal and 4k art we see on the Unity engines from the experienced artist.

The Google has also announced their YouTube Channel to explain the people AI/ML and algorithms for the same which can be interesting topic for the people who want to have information about it.

[larry_vw_1955]

I have been playing with this boy today, and I am really impressed with the result.
I think it could be really useful to share the pseudo code and the description used to generate the images.

there is no code to generate images, they use real images it just modifies them. but it doesn't create anything. not from scratch. how could it.

[fillippone]

I have been playing with this boy today, and I am really impressed with the result.
I think it could be really useful to share the pseudo code and the description used to generate the images.

there is no code to generate images, they use real images it just modifies them. but it doesn't create anything. not from scratch. how could it.

In addition to the description of the image, you can use some “command” to define the type of picture you want, the engine used, the startle of the render and other commands.
I find very useful to see how others got to their results to play with our creations.

[Dave1]

There was also this image from a discussion here:

https://www.reddit.com/r/Bitcoin/comments/wrwmet/dalle_prompt_an_ancient_roman_sestertius_with_a/




And it was AI image was developed using Dall-E, https://openai.com/

[Rizzrack]

Would be cool if you also wrote the prompts (the words you gave it to create the image)

Tried it a few weeks as well ...

Prompt: satoshi nakamoto bitcoin creator


Prompt: satoshi nakamoto bitcoin blockchain cryptocurrencies


Prompt: privacy

[TheBeardedBaby]

Would be cool if you also wrote the prompts (the words you gave it to create the image)
~

Most of the words are in the name of the file of the pictures I posted. If you open the image files in a new tab you see them

[larry_vw_1955]

In addition to the description of the image, you can use some “command” to define the type of picture you want, the engine used, the startle of the render and other commands.
I find very useful to see how others got to their results to play with our creations.

but its really not creating anything as much as it is just matching text to a database of images from the real world. then it does some type of processing on them. that's about it i would think. so whoever owns the original images put into the database probably has a claim on any derived works from it.

[fillippone]

In addition to the description of the image, you can use some “command” to define the type of picture you want, the engine used, the startle of the render and other commands.
I find very useful to see how others got to their results to play with our creations.

but its really not creating anything as much as it is just matching text to a database of images from the real world. then it does some type of processing on them. that's about it i would think. so whoever owns the original images put into the database probably has a claim on any derived works from it.

I am not über technical (also) on this particular field. But I guess the images used to train the neural network aren’t directly used to build the image, or at least not in a recognisable way…

[larry_vw_1955]

I am not über technical (also) on this particular field. But I guess the images used to train the neural network aren’t directly used to build the image, or at least not in a recognisable way…

just because it's not a "recognizable" way how do you even define that term technically? you can't just get copyrighted pictures of google images and let it use those in anyway unless they got permission from the copyright holders. i would imagine a celebrity could sue anyone using their images in an unauthorized fashion. if they turned up in any "generated" works.

[fillippone]

I am not über technical (also) on this particular field. But I guess the images used to train the neural network aren’t directly used to build the image, or at least not in a recognisable way…

just because it's not a "recognizable" way how do you even define that term technically? you can't just get copyrighted pictures of google images and let it use those in anyway unless they got permission from the copyright holders. i would imagine a celebrity could sue anyone using their images in an unauthorized fashion. if they turned up in any "generated" works.

Why? How can them sue? For example, Dalle2,doesn’t permit generating images with celebrities, while Midjourney and Stable diffusion allows that.

I have serious doubt about the fact they copyrighted material must not be included in the training set of an AI.