Dall-E, Midjourney, Stable Diffusion (etc) - who's playing, and with which?
Dall-E, Midjourney, Stable Diffusion (etc) - who's playing, and with which?
- Started
- Last post
- 824 Responses
- microkorg2
Midjourney > Magnific > Lightroom
<3
- Pretty astounding. Only really the pupils in this that give it away as AI when you pixelpeep.microkorg
- DamnYakuZoku
- @microkorg and the way her hair grows out of her head. the rest looks greatsted
- its incredbile what u can do with breaking copyright laws... ;)neverscared
- <OBBTKN
- HAL90014
https://www.cbc.ca/news/business…
"New York Times sues OpenAI, Microsoft for copyright infringement.
The New York Times is suing OpenAI and Microsoft, accusing them of using millions of the newspaper's articles without permission to help train artificial intelligence (AI) technologies."
------AI is literally stealing all the world's creations and make robots regurgitate content based on keywords typed by people who spend a few days or months (few hours!) learning how to type effective "prompts"...
These people are often stealing their prompts from other "prompters" or just copy paste "prompts" made by... AI!
Try to convice me the content these "prompters" are generating has any artistic or creative value!
- It’s inevitable. The question is, what happens to creativity going forward when ai has consumed and used all human creation from the past 3000 years_niko
- And multiplied it with Ai creations a million fold in a matter of months?_niko
- I used to see beautiful hand drawn charcoal portraits on Pinterest, hundreds of them, but it was the same ones over and over again._niko
- Now every day I see hundreds of new ai ones, out of the tens of thousands being created and they all are on a level not possible with a human hand._niko
- So ai will most likely train on those and get even better. Same with literature and music. Human models will be a drop of water in an ocean of ai content_niko
- Real artists will gain value out of all thisHAL9001
- I agree, a human artist will make way more than his ai counterpart just like the master Japanese knife maker makes way more than the mass produced Chinese shit_niko
- The examples are pretty damning
https://storage.cour…yuekit - It's literally just regurgitating identical content from NYT articles when prompted with part of the text.yuekit
- Same as the Midjourney example I posted above. I wonder how people who think AIs are intelligent interpret what's happening here.yuekit
- utopian3
The New Midjourney Image Upscaler Is Live!
- want to check how one can upscale old prompts...uan
- those skulls are wild!uan
- Niiiiiceeee!PhanLo
- It's great!yuekit
- holy shit! hell yesdoesnotexist
- hmm...the new upscale doesn't seem to be that great...don't see it adding detail, just doing what those algorithm based upscales for photoshop do...doesnotexist
- @uan, find the original grid, upscale again. it'll show you the new upscale options.doesnotexist
- ...or just search the job ID, you'll get the new options as well.doesnotexist
- tnx @doesnotexist!uan
- actually the upscales are pretty good loldoesnotexist
- Jrdntnnr5
I started making actual photos of an AI collages. Spacing the collage to create parallax for depth of field.
An imaginary market in East Ham, 1980s:
- Shot on 4x5, large format.Jrdntnnr
- dude you forgot to add, fuji neopan superia 400 pushed to 1600 pulled to C41 Tri-X developer medium format 126 filmfuturefood
- Beeswax1
I've been using Dall-e 3 through Bing image creator for couple days and I have to say it's a lot better than Midjourney to create initial images. But it doesn't have features like creating variations or zooming yet
- i feel like dall-e reads what i ask for then gives me what it thinks i asked for, like some shitty ex-girlfriendimbecile
- It's great for creating clipart and cartoons.utopian
- try dall-e v3, v2 suckedBeeswax
- Where Bing's concerned, I've really come to understand the threat model to its search business that Google perceives - I literally didn't 'Get It' 'til now.Nairn
- I tried Dall-E 3, its like using Midjourney last year.utopian
- depends on what you're using each for. i mostly go after things that look more photographic, so MJ is better imo. dall-e seems more fun/goofy?doesnotexist
- @Beeswax show us some of your resultsKrassy
- microkorg4
- I've been using the Zoom Out feature a lot on MidJourney recently to make more interesting compositions.microkorg
- The image here started as a portrait of the girl in the background who is looking at the camera.microkorg
- Zoomed out 2x a few times and always selecting the zoom output with the most minimal details.microkorg
- Then one of the outputs it added a girl in the foreground. There were a few options, including a floating head, but this was the best.microkorg
- zoom out is dope. i wish it would add more pixels though instead of just making the original smaller.doesnotexist
- niceutopian
- Can you share your original image before you started zooming out? It would be cool to see your process.utopian
- Yes doesnotexist, I wish it did this too. Would mean creating zoom out videos would be super easy. Currently you have to use multiple outputs all perfectlymicrokorg
- Timed and aligned and fading between. Utopian, no problem, will post all the outputs from process.microkorg
- utopian5
- Wow.Nairn
- I'm now fantasising about infinite api interfaces in VR space, surrounded by the seeds of innumerate dreamsNairn
- Most of his examples are at least a bit shit here ('crown'?!) but the interface this has is... astounding to me. DO you use this, utopian?Nairn
- I love Leonardo's interface and tools...particularly the in/out painting. Their image generation is shit compared to Midjourney, but it's get better.utopian
- I general use Midjourney for all of my image creations, but I'll then take my MJ images into Leonardo for its in/out painting capabilities.utopian
- I'm really hoping that when Midjourney finally launches its website along with its new UI, tools and features it will be far superior to anything else.utopian
- BTW..Midjourney's new "inpainting feature" on their Discord website blows ass. It's probably their worst upgrade since MJ's inception.utopian
- PhanLo4
Same prompt as below in photoshop SDXL then expanded out using generative fill
-
- PhanLo4
Here was an attempt using that https://www.promeai.com site tonight. The guy I work with often paints giant skeletons and we're trying to work out a wall where it's a big forest jungle and lots of graffiti is mixed in it but there will be a slightly off-centre piece skeleton/person to break up the compososition.
So I rendered the skeleton as lines using freestyle in Blender then just added a simple vector graphic for the leaves I found online
-
I only found out Blender did the outlines thing the other day, honestly wish I learned it years ago.
-
Right anyway, here's what it started spitting out
I haven't registered with the site and just download them as I go by right-clicking and clicking open in a new window, for some reason it downloads it as a higher resolution than downloading it the other way. I generated 258 images
-
-
That's a short selection, I probably got something close to what I wanted in the first 20 images but I just hammered through a bunch of variations to see how off it would go.
Then I took the images into Photoshop and extended them out with the generative fill, I think doing incrementally gets better images but the last few I just did in one go.
-
Weird that this didn't exist, at least for me a week ago, but it's a workflow I would use 100% to paint a wall. The skeleton I could render in Blender properly and comp it into the design with exact lighting and the background I'd redraw anyway, but some decisions are helped along.
-
I share a lot of this stuff with other painters, but only a couple of them seem to get it. I mean a lot of us are close to 50, some guys paint more or less the same thing that they perfected as a teenager, just with a different fade, so it's too late for them., their best days were 30 years ago.
-
A guy was in the studio today talking about controlling live feeds from projectors with AI so you could monitor the output remotely which is pretty handy, so you could live update content offsite.
He's an interactive programmer who builds a lot of custom triggers for activations, triggering content or lighting in venues, it was interesting to hear what he thought of som of it.- I was downloading as I was going because I realised if I didn't the images started to disappear because I wasn't logged in.
just in case you get images you likePhanLo - when i use sketch to image it just spit out different sketches. what am I doing wrongwhere_am_i
- Did you go to this one? https://www.promeai.…PhanLo
- yeah, its doing the opposite for me. its rendered sketched version of an image or sketch, not photo from sketchwhere_am_i
- ah weirdPhanLo
- Thats really interesting thanks for sharing.
looking at it yesterday I thought it could only steer towards fairly set generic styles - liek arch-viz orwoowahesque - manga character but it looks much more versatile and responds to the lines nicelywoowahesque
- so if you upload sketch it makes you a photo type image?where_am_i
- oh i seeee - change scenario up topwhere_am_i
- I was downloading as I was going because I realised if I didn't the images started to disappear because I wasn't logged in.
- imbecile3
- Layout too?palimpsest
- no. the panels come from chatgpt descriptions run through MJ at different aspect ratios, then i collaged itimbecile
- Is this what you envisioned in your head when you were prompting it?toemaas
- Really consistent style too. That side panel is really coolPhanLo
- I didn’t really have a vision. just proof of concept. Never even read the panel descriptions, just prompt copy paste aspect ratio.imbecile
- I’m confident that if someone wanted to put a script together, art directing the storyline would be a breeze, then just drop in dialog manually.imbecile
- If you wanted to make it all AI you would have to start with the story first. I would love to see AI building a story from the panels though.palimpsest
- it is all ai, image generations are based on chatgpt outputimbecile
- I was talking about adding a script/text: "drop in dialog manually".
It's already impressive as it is mainly because the style is consistent.palimpsest - Add AI layout and script and it's even more impressive. It still shows we got ways to go.palimpsest
- Creating a script from an image prompt would be really out there. It's easier to interpret a text prompt than an image prompt.palimpsest
- Yeah. I'm interested in seeing how it would interpret prompts for simple storyboarding. I see a lot of stuff, but there is no connection to the people....toemaas
- ...generating it. It looks cool, but is it what they want? I just don't want to waste my time if nothing looks the way I want it. I get the feeling that,...toemaas
- ... in the world of film, you are at the mercy of what A.I. "Generates" and it will never be able to actually generate what you envision. The choices I make...toemaas
- ... and the compositions I want are deliberate and have metaphorical meaning. Will A.I. understand what I'm trying to get at?toemaas
- ask it.imbecile
- @Toemass you can block out scenes in 3d and then add extra bits in AI. Depends how far you want it to go. You could then tidy it up in Photoshop.PhanLo
- @phanlo - yeah, that seems like a more practical approach. Like good VFX - it's there to complement, not replace - why a lot of marvel movies can be shit.toemaas
- I was a bit like you, tried to get shots then it didn't work and ended up just learning the 3D stuff. It's really useful, but having an eye for shots or even...PhanLo
- ...a degree of taste will still be important. In the last while in the studio we'd been making cool visuals with it, but you can only view them from one anglePhanLo
- PhanLo4
Threw the Blender render into some landscape shapes and got some wild flowing abstracts.
-
- so cool man._niko
- you, and others on here should consider offering consulting services. Everything in this space is so overwhelming right now, infinite possibilities but people.._niko
- ...don't know where to start. You seem to have a handle on everything and are at the leading edge of this._niko
- https://i.imgur.com/…imbecile
- ^Yep Imbecile, defo on that art you'd get in offices vibe. I make these little landscapes as a way to design backgrounds for walls.PhanLo
- A common thing graffiti guys say to abstract painters is 'your painting would look good as a background' essentially saying their art is only worth drawing on.PhanLo
- my image is a midjourney generation based on my description of your piece. I used to macro photograph inks mixing, so I dig this aesthetic.imbecile
- Ah decent, there's some magic in how those look in real photos.
We'd got a tank to do some ink drops in the studio but never got round to filming.PhanLo
- stewart4
This would be a great photopolymer etching print when I find some time to make it.
- He's going to do graffiti on that tram.
Reminds me of old computer game load screens. Love it.PhanLo - just think of every element in this pic and how AI got it right. the perspective, the lighting, the texture, the objects, their relation to each other, the styl_niko
- and then it creates this overall mood. And it does it in a couple of seconds. Realistically it would have taken an artist a whole day from concert to final_niko
- https://onlineimaget…utopian
- https://www2.lunapic…utopian
- He's going to do graffiti on that tram.
- sted6
dall·e
firefly
mj4
same prompt :)
- nbq1
Adobe is launching this:
- You can join the Beta here:
https://firefly.adob…nbq - Oh, boy.
Hope it's not a local installation, cos we'd hearing a lot of CPUs shrieking in agony.Continuity - just a way for Adobe to monetize and grind it to a halt, don't think it'll add any value that the others don't offer but I signed up for the beta anyway_niko
- Actually there's some really useful stuff in there. Especially the Sketch to Vector stuff.NBQ00
- @niko: how is it not adding value being able to use this stuff directly in the Adobe apps? I see a lot of potential.NBQ00
- don't underestimate the power of corporate design not willing to risk using a Discord based app for their work...jonny_quest_lives
- WPP, Omnicom, Publicis, IPG and their portfolio of "Creative Agencies"jonny_quest_lives
- This has great potential and would be excited to give it a try.nbq
- prompt & pasteneverscared
- You can join the Beta here:
- Jrdntnnr2
New DALLE experimental alpha is pretty sick
- pillows3
- Prompt: /image Gaius Helen Mohiam, Reverend Mother, Sarah Jessica Parker, Dune, created by Frank Herbertutopian
- ^ nice creative prompt idea!grafician
- https://i.imgur.com/…imbecile
- https://i.imgur.com/…utopian
- https://i.imgur.com/…utopian
- https://imgur.com/a/…adrok
- stewart5
Whichever one I use, it's pretty impossible for me to get a full body shot and a specific pose right. I often get back a portrait with just the head and upper body.
