OP
OP
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
Think I might have to get a subscription to Midjourney to get this out of my system lol. Running out of time in the beta.
ghostly viking longship emerging from the mist in rough seas, a sea monster shrouded in the mist behind it, 8k, hyper-realistic, by H.R. Giger
I love this. This really speaks to me.
Orange tabby cat emperor, detailed, slightly damaged tapestry
I can see how addictive this could be. The only limit is imagination (and a paid subscription =p)
I can see how addictive this could be. The only limit is imagination (and a paid subscription =p)
The scary thing is in 5-10 years this request will probably generate a convincing-looking image, and deepfakes will present a real threat to society in terms of surfacing the truth to the public.
Thick forest of spruce trees, vegetation and moss on forest floor, church fresco style
Ancient limestone mural of a man staying up too late due to computer use
Ancient limestone mural of a man staying up too late due to computer use
Last edited:
OP
OP
I think I will sneak these into a presentation at work and see if anyone notices.
molecular structure high quality realistic rendering
molecular structure being struck by lightning high quality rendering
molecular structure high quality realistic rendering
molecular structure being struck by lightning high quality rendering
Yeah I've seen it, it's very impressive. Don't think it could generate a convincing-as-real image of Reggie behind a podium at the White House yet, but maybe in a few years.
Made a little four part "spirit photography" series in Midjourney. Still getting used to it, but absolutely overwhelmed by the versatility of it. Prompts that I used included below, but I did modify them with some specificities like "black and white" and "film photography."
"We lost everything in the fire, all that remained were ghosts."
"The house was little more than smoke and ash by the time we found it. Our relatives remained, but only their spirits, floating aimlessly through the corridors."
"It took years to rebuild even a bit of the community we had, but their spirits remained to guide us through the process, clearing the rubble and rebuilding our humble homes from the remains."
"And now our home is chilled by the spirits of our ancestors. Their ghosts walk the halls, the floorboards creaking over our heads, but we find comfort in their presence."
"We lost everything in the fire, all that remained were ghosts."
"The house was little more than smoke and ash by the time we found it. Our relatives remained, but only their spirits, floating aimlessly through the corridors."
"It took years to rebuild even a bit of the community we had, but their spirits remained to guide us through the process, clearing the rubble and rebuilding our humble homes from the remains."
"And now our home is chilled by the spirits of our ancestors. Their ghosts walk the halls, the floorboards creaking over our heads, but we find comfort in their presence."
I feel like I'm making some decent stuff, but I wanna go more advanced in my prompts. Anyone got any resources for where to see others' advanced prompt-work?
I mean, yeah? Most of those images have telltale signs of what I have been calling "AI chaos" that makes them obvious as fakes. Like most tech, especially kind of black box tech like machine learning, my guess is that getting to that last 5% to 10% of progress to where an image will look convincing upon inspection to most people will take a few more years. The only visual AI project I've seen so far that lacks the chaos is that secretive Google project, assuming the images they showed off were legit and not touched up.
massive lush arboreum inside glass dome on space station orbiting ringed planet, magical atmosphere, complimentary colors, vines, trellises, blooming flowers
lush arboreum, scrollwork, magical atmosphere, complimentary colors, vines, trellises
lush arboreum, scrollwork, magical atmosphere, complimentary colors, vines, trellises
I subbed to the basic at first, but when I realized the sheer amount of variations it takes to get an amazing piece out of the process I realized standard was the way to go and switched. In two days I've already done something like 400 generations. I honestly think it's super worth it and so far so I don't regret it. From both an enjoyment/entertainment perspective and from a creative perspective - already seen so many moods, interesting settings, subjects, and compositions that I'm filing away into an inspiration folder to tap into via other manual forms of art. It's such an amazing creative tool as it can expand your mind's library of conceptual connections that you would not have otherwise stumbled upon.
Just my two cents, your mileage may vary.
Thanks
Might try the basic first then I'll upgrade later
Edit: yup I dis basic and run through 100 variations less than an hour.
Last edited:
I've been trying using the weights a bit but it still spreads colors incorrectly. I'll play with it a bit more to see if I can figure it out.
I ran out of my free trial of Midjourney, I dedicated my last 4 to just really dumb shit:
Kinda wonder at one point the AI jsut started to ignore the prompt, the result was kinda cool looking though.
For this one I wanted to see if I could mix as many styles as possible. I think the cave painting aspect was probably too powerful as it ended up limiting the color possibilities.
I tried this twice cause I wanted to see if maybe it would actually give me low-polygon aliens but I think it just didn't know how to interpret it.
V4 from the first batch is the only one to feel like it got the "elderly man surprised by his own flatulence" part
Here's the rest of my attempts:
https://www.midjourney.com/app/ (I didnt know early on that we had a limited number of attempts so I wasted a lot of tries on early memey kinda stuff)
If I were to buy a subscription it would be to see if I could find a way to have the AI generate a new style of art, or just investigate how it works and see how far it can be taken. It's pretty fun though.
Kinda wonder at one point the AI jsut started to ignore the prompt, the result was kinda cool looking though.
For this one I wanted to see if I could mix as many styles as possible. I think the cave painting aspect was probably too powerful as it ended up limiting the color possibilities.
I tried this twice cause I wanted to see if maybe it would actually give me low-polygon aliens but I think it just didn't know how to interpret it.
V4 from the first batch is the only one to feel like it got the "elderly man surprised by his own flatulence" part
Here's the rest of my attempts:
https://www.midjourney.com/app/ (I didnt know early on that we had a limited number of attempts so I wasted a lot of tries on early memey kinda stuff)
If I were to buy a subscription it would be to see if I could find a way to have the AI generate a new style of art, or just investigate how it works and see how far it can be taken. It's pretty fun though.
Futuristic Manet Series ("Futuristic [X], Manet painting style")
Robot cooking five-course meal, Vermeer style
First attempt at my dream home... Spanish winery villa
Robot cooking five-course meal, Vermeer style
First attempt at my dream home... Spanish winery villa
Last edited:
Why do we have two different MidJourney art threads now? I thought this one was for Dall-E mini and...non-mini, but now it's a free-for-all. At the least it would be great if people could label what tool they're using.
Don't you mean Monet and Vermeer?Futuristic Manet Series ("Futuristic [X], Manet painting style")
Robot cooking five-course meal, Vaneer style
First attempt at my dream home... Spanish winery villa
Yeah the prompt spelled Vermeer correctly I just typod it here. (Édouard) Manet however is correct as typed. "A Bar at the Foiles-Bergere" is a painting of his shown below:
OP
OP
I quite like plugging in book descriptions of things.
E.g. IT, which looks much less terrifying than you would think...
an endless, crawling hairy creature made of orange light
[Minas Tirith]
beautiful: white-walled, many-towered, proud and fair upon its mountain-seat; its battlements glittered with steel, and its turrets were bright with many banners.
E.g. IT, which looks much less terrifying than you would think...
an endless, crawling hairy creature made of orange light
[Minas Tirith]
beautiful: white-walled, many-towered, proud and fair upon its mountain-seat; its battlements glittered with steel, and its turrets were bright with many banners.
A couple more fallen angels of marigolds. Guess who
This is so addicting
These are amazing
OP
OP
Pretty cool, is that GOW in Egypt?Just playing around with Midjourney
Yup, that's what I was going for. I was going for Kratos vs Anubis but it keeps making Anubis less anthropomorphic, and making its head a pyramid because I said I wanted them fighting in front of the pyramids, lol
I've rarely got anything this clean, can you share your prompts?Few more Kratos variations
This looks amazing , would love to know which prompts you've used.
No problem, it was just one:
I think Yoji Shinkawa's art style is one of the best when it comes to AI art. I have a few cool looking photos using the prompt
How are some faces so clear and detailed and other horrific nightmare distortions? Especially of recognizable figures where there must be millions of source images?
This one was "God of war in India, realistic --v 4"
Few more from recent days
