Calgarypuck Forums - The Unofficial Calgary Flames Fan Community
Old 08-25-2022, 03:01 PM   #1
octothorp
Franchise Player
 
octothorp's Avatar
 
Join Date: Oct 2002
Location: not lurking
Exp:
Default The AI artwork tipping point

The last month has been absolutely insane for the state of AI-generated imagery... throughout this year Dalle 2 has been gradually opening up their service to an increasingly large number of users, Midjourney released their v3 last month and quickly maxed out the number of users that their discord-based service can support, and this week Stable Diffusion released their open-source model for anyone to download and run on their computer (if your computer can handle it). What seemed like a novelty a year ago seems like the inevitable future for a lot of creative industries... if not outright replacing people, at least vastly changing workflows.

(If you're not familiar with these technologies, they all operate in a similar way: you write a sequence of words, a prompt, which the algorithm than used to create an image for you. The algorithm has been trained on anywhere from hundreds of millions to billions of images, which it uses to figure out what you want. Results range from photorealistic to artistic to just plain weird and incoherent depenting on the algorithm and what you ask it for.)

Each of these services has their own strengths... Dalle 2 is fantastic at creating realistic images of things that don't exist. Midjourney is the more artistic, being absolutely fantastic at capturing the styles of different artists, but struggles at coherence. Stable Diffusion sits somewhere between the two, but has the advantage of being locally runnable and open-source.

There are so many interesting societal and ethical issues involved in this. These services haven't paid for the rights to use the images that they use for their dataset, and in the case of SD in particular, sometimes it'll return an image that is more like a photo-collage of existing photos... sometimes even showing watermarks from clipart sites on the final results. Midjourney does a fantastic job imitating the styles of other artists... but is it ethical for someone to type in another artist's name and get something that imitates that style? Does it matter whether the finished result is used commercially, or whether the referenced artist is still living or working?

Then there's the whole 'objectionable content' side of it. Both D2 and MJ attempt to control their created content to limit the amount of nudity, gore, celebrity fakes, etc. being generated. SD has no such practical limitations (just some vague and unenforceable terms of service), and when MJ did a beta-test that used SD's dataset earlier this week, they shut it down after a day because the extreme objectionable content being generated was more than their mods and development team could manage. Reddit has already banned a large number of SD NSFW communities because of the content (particularly the faked celebrity photos). It seems likely that there within years there will almost certainly be web-based services aimed specifically at the NSFW audience, because that audience is huge and willing to pay.

In artist community discussions, the discussions range from dismissing it as completely not a threat to their industry; to freaking about about becoming obsolete quickly. The middle-path approach seems to acknowledge that these are tools that will inevitably change the industry and how artists work, but won't fully remove artists from the process. I think the real change will come once entertainment studios from Blizzard to Disney have their own in-house models tailored to their needs... something that's already almost certainly in development. And it won't be just 2D static images: Midjourney already has a 3D version in the pipeline, and while animation is beyond what these public models can currently do, big studios with the money to spend on the hardware could absolutely do that.

My own experience has been primarily with Midjourney, although I also played around with SD's model earlier this week when MJ had a brief beta-test of SD's model, and these are tools that I'll likely involve in my own creative workflow going forward. I've got my Dalle2 invite but haven't used it yet, although later this year I'll definitely explore some of its tools and see how they can be used in tandem with MJ. I do have societal concerns, like if in the future will there be fewer kids learning to draw, if you can get a computer to render whatever you imagine. Does it even matter if fewer kids are learning to draw?

I'm curious to know if anyone else is exploring these tools, or thinking about what the implications are.

Last edited by octothorp; 08-25-2022 at 03:07 PM.
octothorp is offline   Reply With Quote
The Following 7 Users Say Thank You to octothorp For This Useful Post:
Old 08-25-2022, 03:45 PM   #2
Wastedyouth
Truculent!
 
Join Date: Aug 2013
Exp:
Default

What language are you speaking?
__________________
Quote:
Originally Posted by Poe969 View Post
It's the Law of E=NG. If there was an Edmonton on Mars, it would stink like Uranus.
Wastedyouth is offline   Reply With Quote
Old 08-25-2022, 03:59 PM   #3
psyang
Powerplay Quarterback
 
Join Date: Jan 2010
Exp:
Default

Quote:
Originally Posted by Wastedyouth View Post
What language are you speaking?
This was my introduction to AI image generation. It is pretty incredible how far it has come in so little time.

psyang is offline   Reply With Quote
The Following 2 Users Say Thank You to psyang For This Useful Post:
Old 08-25-2022, 04:09 PM   #4
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

I've mainly been playing around with SD lately, but I'm in the next beta round for Singularity AI.

Really excited to put my "render farm" to good use; a friend and I are actually working out a Discord Bot that will allow people to feed prompts to my Singularity farm and spit out really high quality renders on demand without a sub.
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.
PsYcNeT is offline   Reply With Quote
Old 08-25-2022, 04:21 PM   #5
octothorp
Franchise Player
 
octothorp's Avatar
 
Join Date: Oct 2002
Location: not lurking
Exp:
Default

Quote:
Originally Posted by Wastedyouth View Post
What language are you speaking?
Ah, maybe it's not being talked about generally as much as it is in the circles I follow. Let me explain with an example (using Midjourney):

I send a message to one of these services. I tell it what I want it to draw... suppose for example, I want it to draw a machine-human interface for the creation of artwork (this is an abstract idea, so expect the results to get a little weird). And then I tell it about some visual references, like, say medical illustrations, vintage, and star-wars concept artist Syd Mead. Within seconds, it responds with four ideas for this illustration:

Spoiler!


Maybe I don't like any ideas and I ask it to try again, or I change my instructions, but maybe I like #3, so I ask it to create me a higher-resolution, finished version:

Spoiler!


Now, maybe you want something different. Maybe you want a picture of Jason Kenney at the fuel pump, inspired by byzantine iconography. Again, you can get that in just seconds:

Spoiler!

Last edited by octothorp; 08-25-2022 at 04:29 PM.
octothorp is offline   Reply With Quote
The Following 8 Users Say Thank You to octothorp For This Useful Post:
Old 08-25-2022, 04:34 PM   #6
octothorp
Franchise Player
 
octothorp's Avatar
 
Join Date: Oct 2002
Location: not lurking
Exp:
Default

Quote:
Originally Posted by PsYcNeT View Post
I've mainly been playing around with SD lately, but I'm in the next beta round for Singularity AI.

Really excited to put my "render farm" to good use; a friend and I are actually working out a Discord Bot that will allow people to feed prompts to my Singularity farm and spit out really high quality renders on demand without a sub.
That's awesome, I'd love to try it when you get it up and running, if you don't mind sharing! I'm definitely at the level where I'm happy to use things running on other systems, but have no interest in trying to get my own local implementation running.
octothorp is offline   Reply With Quote
Old 08-25-2022, 04:43 PM   #7
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

Personally I'm a big fan of the "AI understands what you're asking for but here's a eldritch monstrosity of a YouTube Thumbnail instead" genre.

Spoiler!
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.

Last edited by PsYcNeT; 08-25-2022 at 04:48 PM.
PsYcNeT is offline   Reply With Quote
The Following User Says Thank You to PsYcNeT For This Useful Post:
Old 08-25-2022, 07:24 PM   #8
photon
The new goggles also do nothing.
 
photon's Avatar
 
Join Date: Oct 2001
Location: Calgary
Exp:
Default

This is cool. Can you suggest a basic "Run SD locally" tutorial?
__________________
Uncertainty is an uncomfortable position.
But certainty is an absurd one.
photon is offline   Reply With Quote
Old 08-25-2022, 08:39 PM   #9
octothorp
Franchise Player
 
octothorp's Avatar
 
Join Date: Oct 2002
Location: not lurking
Exp:
Default

Quote:
Originally Posted by photon View Post
This is cool. Can you suggest a basic "Run SD locally" tutorial?
Personally I can't, as I haven't really explored running SD locally. I will say for anyone that just wants to get a sense of the technology that if you have Discord, you can demo Midjourney for free, for about 25 image-generations.

https://www.midjourney.com/home/
octothorp is offline   Reply With Quote
Old 08-25-2022, 10:02 PM   #10
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

Or for rudimentary level stuff, you can just go generate infinitely with CrAIyon

https://www.craiyon.com/
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.
PsYcNeT is offline   Reply With Quote
Old 08-26-2022, 12:02 AM   #11
JohnnyB
Franchise Player
 
JohnnyB's Avatar
 
Join Date: Mar 2006
Location: Shanghai
Exp:
Default

I haven't used SD, but I'd been using diffusion models for a few months and have been using MJ pretty much since it became available. Lots of fun to work with these and to play with their potential and boundaries.

Using MJ I recently tried doing a long series of prompts based on the stem "the face of ...." and slotting in words like 'God', 'poverty', 'a CEO', 'an Indian' etc. just to see what the outputs for these would be in accordance with the content of the training data sets. It was an interesting exploration to see when the faces were male or female, or when they had different racial features etc.
__________________

"If stupidity got us into this mess, then why can't it get us out?"
JohnnyB is offline   Reply With Quote
Old 08-26-2022, 12:15 AM   #12
JohnnyB
Franchise Player
 
JohnnyB's Avatar
 
Join Date: Mar 2006
Location: Shanghai
Exp:
Default

Quote:
Originally Posted by psyang View Post
This was my introduction to AI image generation. It is pretty incredible how far it has come in so little time.

Two-Minute Papers is such a great channel. One of my top two youtube channels for sure. Everyone with an academic or practical interest in AI should subscribe.
__________________

"If stupidity got us into this mess, then why can't it get us out?"
JohnnyB is offline   Reply With Quote
Old 08-26-2022, 07:52 AM   #13
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

MJ is pretty second-to-none for art-style generatives, no doubt.

Spoiler!
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.
PsYcNeT is offline   Reply With Quote
Old 08-26-2022, 10:06 AM   #14
edslunch
Franchise Player
 
edslunch's Avatar
 
Join Date: Apr 2009
Exp:
Default

Does this spell the end of commercial artists? Commercial photographers too? Anyone who is creating art or content not for art’s sake but to satisfy a commercial requirement is in danger of being supplanted. Video game and other animation was mentioned earlier, and I assume this would apply to music also (e.g. movie scores). What a time to be alive indeed.
edslunch is offline   Reply With Quote
Old 08-26-2022, 10:14 AM   #15
Russic
Dances with Wolves
 
Russic's Avatar
 
Join Date: Jun 2006
Location: Section 304
Exp:
Default

I have a buddy who mainly uses mid journey as a creative engine for himself. He mostly just screws around and posts stuff to his instagram and twitter accounts, but the quality shift even over a single month is noticeable.

It'll be very odd when you can get a massive art piece for your home that doesn't technically have an artist attached to it. You could literally get something that is wholly original, which would appeal to many.

I'm seeing a whole lot of "AI will never replace me as a <profession>" type comments, and I can't help but get a real "computers will never be able to beat a human at chess" vibe. I think it's human nature to be not great at understanding how compounding works. It feels like I went from laughing at robots that couldn't climb stairs to being overly concerned about a sniper rifle being attached to a Boston Dynamics robo-dog in about 5 years.

Last edited by Russic; 08-26-2022 at 10:20 AM.
Russic is offline   Reply With Quote
The Following 2 Users Say Thank You to Russic For This Useful Post:
Old 08-26-2022, 10:21 AM   #16
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

SD has a music creator coming out soon; I'm very intrigued.
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.
PsYcNeT is offline   Reply With Quote
Old 08-26-2022, 10:22 AM   #17
JohnnyB
Franchise Player
 
JohnnyB's Avatar
 
Join Date: Mar 2006
Location: Shanghai
Exp:
Default

Quote:
Originally Posted by edslunch View Post
Does this spell the end of commercial artists? Commercial photographers too? Anyone who is creating art or content not for art’s sake but to satisfy a commercial requirement is in danger of being supplanted. Video game and other animation was mentioned earlier, and I assume this would apply to music also (e.g. movie scores). What a time to be alive indeed.
It's really not that good, yet, but it will have a huge impact on a lot of commercial artists.

Really, we're not that many years away from a parent being able to say "an uplifting story about a young girl who conquers he fears of dragons" and having AI write the story, provide all illustrations and add an appropriate musical score as a new bedtime story or animation every night.

That is not going to replace the J.K. Rowlings or the great illustrators of the world, but it is going to raise the bar substantially on what still has commercial value as it will add an almost infinite number of competitors to the market at the lower end.
__________________

"If stupidity got us into this mess, then why can't it get us out?"
JohnnyB is offline   Reply With Quote
Old 08-26-2022, 10:26 AM   #18
PsYcNeT
Franchise Player
 
PsYcNeT's Avatar
 
Join Date: May 2004
Location: Marseilles Of The Prairies
Exp:
Default

The Death of Fiverr
__________________

Quote:
Originally Posted by MrMastodonFarm View Post
Settle down there, Temple Grandin.
PsYcNeT is offline   Reply With Quote
The Following User Says Thank You to PsYcNeT For This Useful Post:
Old 08-26-2022, 10:33 AM   #19
JohnnyB
Franchise Player
 
JohnnyB's Avatar
 
Join Date: Mar 2006
Location: Shanghai
Exp:
Default

Quote:
Originally Posted by PsYcNeT View Post
The Death of Fiverr
Or, a lot of fiverr artists start using AI for initial concepts and then doing all the refinements to meet customer needs and get the right file formats by themselves, but increase their throughput
__________________

"If stupidity got us into this mess, then why can't it get us out?"
JohnnyB is offline   Reply With Quote
The Following User Says Thank You to JohnnyB For This Useful Post:
Old 08-26-2022, 10:37 AM   #20
Wastedyouth
Truculent!
 
Join Date: Aug 2013
Exp:
Default

Quote:
Originally Posted by octothorp View Post
Ah, maybe it's not being talked about generally as much as it is in the circles I follow. Let me explain with an example (using Midjourney):

I send a message to one of these services. I tell it what I want it to draw... suppose for example, I want it to draw a machine-human interface for the creation of artwork (this is an abstract idea, so expect the results to get a little weird). And then I tell it about some visual references, like, say medical illustrations, vintage, and star-wars concept artist Syd Mead. Within seconds, it responds with four ideas for this illustration:

Spoiler!


Maybe I don't like any ideas and I ask it to try again, or I change my instructions, but maybe I like #3, so I ask it to create me a higher-resolution, finished version:

Spoiler!


Now, maybe you want something different. Maybe you want a picture of Jason Kenney at the fuel pump, inspired by byzantine iconography. Again, you can get that in just seconds:

Spoiler!
This was all done on AI?

Holy ####.

That's awesome. I want one as a picture on my wall!!
__________________
Quote:
Originally Posted by Poe969 View Post
It's the Law of E=NG. If there was an Edmonton on Mars, it would stink like Uranus.
Wastedyouth is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -6. The time now is 06:48 PM.

Calgary Flames
2023-24




Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright Calgarypuck 2021