[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Settings Mobile Home
/gd/ - Graphic Design

4chan Pass users can bypass this verification. [Learn More] [Login]
  • Please read the Rules and FAQ before posting.
  • Additional supported file types are: PDF
  • There are 37 posters in this thread.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]

Janitor applications are now closed. Thank you to everyone who applied!

It’s literally over, bros.

In a short time artificial intelligence will completely replace real artists in all forms of media.
File: Z(1).jpg (10 KB, 256x256)
10 KB
So true
>These silly Photography crap will never replace portrait paintings.
We're still in the early times, this tech will improve, and quickly overtake artists.
There won't be any need to acquire artistic skills, no need to grind for years just to become able to make decent-looking images. Idea>Skill>image will be Idea>Image. There won't be any need to pay artists, or even pursue art as a career. Making new art styles and such will have more to do with coding, breeding different image sets together, and picking the aesthetic that appeals to you. As a consumer we will have so many products with crazy amount of unique well made designs and art pieces, entertainment will probably become much higher quality with a lot more output. But the skill required to make a living will steadily increase, as AI will replace the need for lower skill artists, and soon there won't even be a profession.
That is happening in every industry lel
That’s Dall-E 1
This is a good thing; less useless jobs on the market. Get a real actual job instead.
I don't think you have an understanding of the gravity of the situation here. If they can automate something with originality like art, they can automate plenty of other jobs as well. This may include yours and yes engineers can eventually be automated.
File: 1625066193269.jpg (150 KB, 1079x1178)
150 KB
150 KB JPG
It's over........
Even with painting murals you don't need skills anymore.

You just do a squiggle grid and then fill it in like tracing on the wall.
>Get a real actual job instead.
like what? flipping burgers?
File: 1654801475684.jpg (1.35 MB, 1202x1761)
1.35 MB
1.35 MB JPG
yep, it's over
File: adadadv-Q6.jpg (156 KB, 1024x1024)
156 KB
156 KB JPG
thats Dall e mini..

Its just over, in the future everybody can create pictures with words. So why pay a human artist that cost 10x times more if you can just create pictures.

80% will use the Ai
you're forgetting that computers are still stupid as fuck, if you think AI is capable of sentience or acting on it's own, you don't understand how AI works. ai art is probably going to be something graphic that designers need to know about like, but they'll probably be working with it rather than against it

socially we need to call out and shame cheapskates that use AI as their final design like people who use clip art. and they'll probably be legal copyright battles with the ai software company for doing that.
if you actually know anything about art, dall e is hopelessly immature and ineffectual.

you think code bros are gonna into art? LOL NO.

shit is so far away from being an artists tool. shit is just a parlor trick for now a long time comin.
File: 1579247173638.jpg (58 KB, 684x453)
58 KB
I mean even if you wanna argue that, society always values things made by hand, or by humans more than those made by tech, so realistically at worst it puts the lazy fucks making corporate Memphis put of a job, while people with actual taste will go with real, human designers. This of course all *assumes* that this technology will develop further, which is not a forgone conclusion, it could very easily stay the way it is, spinning its wheels in the mud, many such technologies have done so in the past
cope, you will be out of a job within 10 years.
File: index.jpg (15 KB, 311x162)
15 KB
File: 312me3zstt691.png (310 KB, 640x639)
310 KB
310 KB PNG
Dall-e mini ( Soon to be named Craiyon ) isnt associated with Dall-e 2. Even then, posting a shitty version of something better else does little to prove a valuable point
I want AI generated porn. I want my imagination to be the limit to porn.
no, they're right
Look, hear me out, because I've been tinkering around the both machine learning and the commercial graphical design. People there talk about IA's obstacles in creating art mostly from the side of the technical limitations of the modern computers or from the point of the precision of generated images. I want to talk about the limits inherent in the algorithm.
>learning set limitation - what AI can recognise and therefore paint depends on the learning sets it was given. This means that the generic things like kangaroos, sunglasses and hoodies are easy, but non generic things like the trademarked characters require creation of the dedicated learning sets build specifically for them. If you want pictures of Spireman for your Spireman comic, you need somebody to draw enough Spireman pictures to create the learning set, so your AI can learn what Spireman is. That alone is a huge obstacle for somebody trying to replace human artists with AI.
>consistency - can your AI draw me the exact same 'roo in the exact same hoodie but blue? Can the AI draw 100 panels of Spireman while keeping the costume consistent? The trouble with the current alghoritm is that every picture is essentially made from scratch without refferencing the other pictures.
>multiple related images - relating to the previous point, the alghoritm has no concept of the previous pictures drawn. It couldn't draw the action sequence, because that would require it being able to related the each picture to the previous ones.
>real time learning and fine tuning - the alghoritm as it stands has no ability to either learn new concepts (like learning that the generated character is named Joey d'Roux) or to make changes to once generated pics.
>intelectual side of paiting - the alghoritm has no concept of the color theory, picture composition or perspective. Sure, the learning set not being butt ugly helps there, but the machine still lacks 90% of what actually is people expect from the designer.
>the whole production process - this alghoritm churns out raster graphics. It's literaly a single layer of pixels, while everybody who even touched the graphic design and the digital art knows that the proper use of layers is a key to actually delivering work that can be fine tuned or changed later. Again, it's not about the quality of the picture itself, it's about the fact that the alghoritm itself just churns out raster pictures, the end product without crucial previous steps.
>it can barely handle the text - do you have any idea how much of the designer's work amounts to putting the text on things and then making dozen corrections because your client didn't like your previous choice?
>no image insertion - look, it may seem crazy but the ability to insert the ready made graphics is crucial. Sometimes you don't need a drawing of the robot on your poster but rather the actual picture of the real machine incorporated into the larger image.
All in all this thing works more like the stock picture and concept art generator and just focusing on the image recognition and efficiency won't change it.
That's gonna get automated too.
File: wifi_MedievalPainting.jpg (320 KB, 1024x1024)
320 KB
320 KB JPG
Has any looked at the Twitter Dalle2 Pics? Absolutely nuts results

File: VRRenaissance.jpg (772 KB, 2048x2048)
772 KB
772 KB JPG
Literally everyone has an amazing camera in their pockets, yet professional photographers exist.

When cameras first came out they thought it would be the end of traditional painting.

This will replace low tier art thought yes
File: taco-bell.jpg (214 KB, 2000x806)
214 KB
214 KB JPG
Still has problems with text.
>Literally everyone has an amazing camera in their pockets, yet professional photographers exist.

Don't tell people who shoot film this
What was the prompt, "Taco Bell sign"?
Maybe it's deliberately programmed to not use trademarked names.

Even AI has been gate-kept, as with CAPTCHA the very first job google had was making it politically correct and not infringe trademarks
We hired a friend to photograph our wedding with his prosumer grade DSLR and used my DSLR to do the video. We hired a pro for 1hr to do the glamour shots after the wedding because fancy special artsy need.
What would've been a full day for 2-4 professionals turned into 1hr for 1.
Certainly everyone doesn't do that, but a good piece of that market has shrunk at the low end (which is the meat and potatoes end). I can think of many many instances where professional photography has been replaced by my own DSLR in my life, including business uses. I even video'd a friend's wedding with it (free too).
If I want to advertise something now, I just take shots myself.
Half the people I know have DSLRs, though iPhones suffice for 98% of shots/vids.
Shoot a short company video at a lesser company? Just let Susan do it with her new iPhone; why pay $3000 for something you can slap together in a couple hours with free software?

Cheaper photography/videography has opened up some marginal new demand scenarios, but mostly people just record their own lesser events and personal moments rather than go to the photo stand or hire a guy with a big film/video rig.

Think of logos.
How much demand will be opened up for designing logos by a radically lowered cost? I'm guessing it's a fairly inelastic marketplace. But now any random jack can mock up a logo in 2 minutes, show 20 options to their client in a half hour, and touch one up in another 30-60 minutes.
What was a significant contract just got turned into hourly piecemeal work for all but the largest clients (which are rare compared to [random local business]).

Website design is going to rapidly follow this development. Any job that a low-skill html coder with an eye for copy-paste design philosophies can do, an AI will do in a couple years.
As you say, dumb design work is ending.
Unfortunately for professionals, that's like 90% of the market.

Oh, and anyone that was trying to build an Instagram or other creative digital art presence is about to get futilely buried by a million zero-effort AI posts that look equivalently good if not better than what they're making.

How is a random mid-tier artist going to compete against a program that can literally copy-paste the skill of DaVinci all the way to Warhol and every museum collection artist in-between?
The vast majority are, as always, mere copycats or very minor innovators.

"Yeah, nice painting you spent 200 hours making, but this program made 500 better versions in 20 minutes, and my normie eyes love them because they look like Rembrandt, not your style, whatever that is. Think I'll follow the AI not you."

Also now random no-skills that have good idea-talent can jump in without spending 6 years grinding. Labor supply is getting flooded while hours-demand is getting demolished.

Truly novel style and idea genesis will be all that matters soon.
Skilled talent is utterly insufficient now.
95% of the talent pool/work is mere skill currently.
Have fun with your industry.
Creative skill just got calculator'd.
Remember all the guys you needed to process numbers before computers?
That's you now.
This is not true in the slightest.
Dalle-2 is not going to be able to just create what you have pictured in your minds eye. It literally creates entirely random stuff. There are countless, countless different styles of art and it can't replicate 99% of them. Not only that but as we're seeing, very unfortunately - the people who run these projects OBSESS over moralism as if people having access to a tool that helps them get off (Create anything adult/graphic) is the end of the fucking world.

These programs, all of them from GPT-3 to Dalle to whatever other AI deepmind shit they're working on out there are imprisoned behind dozens of precautionary filters and rules etc.

Really you have nothing to worry about, it's going to be decades, if ever, when AI is capable of making anything you actually want it to and porn/adult artists probably will never have any real competition from AI ever.
They're evolving
>Its just over, in the future everybody can create pictures with words.
Me, future proofing my job by working in the printing sphere rather than a design one. I'm the one who's designing the machines to print artificial intelligently designed art.
>What happens when we make machines that make machines?
The human race is finally extinguished.
I think this will benefit market, because now everyone who wants a design/art piece for exposure or for $7 can just be handed one made by AI and maybe touched up by a human.
and there will be always a market for "human made" stuff, just because of the sentimental value. look at handmade anything vs mass-produced.
It's literally impossible to build a robot that can replace plumbers
I'm not sure if it's just trademarks, I think they butchered it's ability to do any words on purpose.
What >>432687 said in regards to CAPTCHA is most likely correct.
This. Dall-E cannot make custom illustrations of your staff, your product, your building, etc. It will have to be handmade by someone.

I was thinking about this last night, but AI will only eliminate jobs that could be done by someone on Fiverr for under $100. I'm a graphic designer and even I was too lazy to work on my artwork for unpaid personal projects. If there's a website/app and it needs custom icons... why would I want to spend my time figuring out the proper shading for the trashcan icon? Why should I learn it? My time is better spent on solving a problem - which AI cannot do.

AI can make beats and instrumentals, it can write lyrics, but it cannot do emotions you can connect with. The answer to AI will be going the other way: Human to human contact and sharing of feelings. Not efficiency and automation. If I want to have a cookie, which will taste better: Something my girlfriend baked for me for two hours or something off of a conveyor belt? A handwritten note could be more meaningful than anything AI can do.
>Dalle-2 is not going to be able to just create what you have pictured in your minds eye. It literally creates entirely random stuff. There are countless, countless different styles of art and it can't replicate 99% of them.
Also this. Dalle-2 is the tchotchke equivalent of art.

And if I want something specific, you will have to run Dalle-2 a billion times until it gets it right. But if I just want a concept that catches the eye, that has no utility, it will be good enough for that. I cannot wait to get my hands on it for pixel art and make games (something AI cannot do). Hopefully game editors will also be drag and drop and life will be like the level editor in Fortnite. Anyone can use it, only 1% actually can use it well and make interesting things.
It can be used as a tool. Prompting "modern sleek hero image of a barstool product shot" will give you a perfect example of how to set up your own photo shoot to recreate it
Dall-E 2 can't generate porn
too expensive
Retard how can it replace artist when it literally work by generating drawing BASED ON ALREADY EXISTING ARTWORK? This thing cannot draw by itself
to be fair, thats what the majority of artists do to some degree, with commercial artists being heavily weighted towards derivative works that follow trends and ape stylistic cues from different sources, including distinctive individual styles of iconic artists.

But the flip side of that is that that's EXACTLY how human artists can compete with AI, by doing the same thing it does but with more focus and real intelligence.
The computers ability to expand and speed up a search for appropriate source material over a human trying to process thecsame amount of information really isnt the advantage people assume it is for the simple reason that the human doesn't *have to* do it that way every time.
Ultimately the AI does all that extra work but the returned ideas aren't better by the same percentage, and a human has to weed though all the ones that might technically be *an* interpretation of the prompt but miss the point entirely because theres no actual thought involved.

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.