[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/ic/ - Artwork/Critique


View post   

File: 19 KB, 112x106, 592.png [View same] [iqdb] [saucenao] [google]
6867790 No.6867790 [Reply] [Original]

ai made me quit drawing all together

>> No.6867797

for some reason it made me draw, it actually made me get over fear of failure and instead do it because I find it interesting skill to develop with a meditative mindset

>> No.6867800

Reminder to sage and report AI threads.

>> No.6867803

It motivated me to draw out of spite
but i dunno, spite is always what drives me to create

sucks to be you, OP

>> No.6867804

>>6867790
Did you donate your art supplies to a children’s charity yet?

>> No.6867811

>>6867790
>Translation: I only got into art for money and likes and the rise of AI art filtered me

>> No.6867813

>>6867790
AIbros won

>> No.6867815

>>6867811
wait you dont wanna make money doing what u like?

>> No.6867816 [DELETED] 

>>6867797
same. started drawing few months ago because of AIcucks. want to be a diamond among the rough

>> No.6867818

>>6867804
i only have a wacom and its collecting dust rn sadge

>> No.6867819
File: 3.00 MB, 640x900, 1614111751600.webm [View same] [iqdb] [saucenao] [google]
6867819

Why are there so many defeatists on this board? I don't want that shit rubbing off on me.

>> No.6867822

>>6867797
same. started drawing few months ago because of AIcucks. want to be a diamond among the rocks

>> No.6867826

>>6867815
there is a difference between
>making money doing what you like
and
>liking something because you could make money with it
lots of shitters get into art for money because they saw the epic anime artiste pulling in 10k a month shitting out flavor of the month anime pin ups

img2img shitters are just trolls and scammers

>> No.6867827

>>6867811
Quitting because of AY EYE is retarded but I'm honestly tired of pretending that chasing money is bad.

>> No.6867828

>>6867819
they're the same 'tards who blamed aphantasia, ADHD, neuroplasticity. no one likes these fools

>> No.6867829
File: 118 KB, 976x850, 1695357763975.jpg [View same] [iqdb] [saucenao] [google]
6867829

>>6867790
If vile gelatinous anime girls were able to push you out then you probably weren't GMI.

>> No.6867831

>>6867822
for me it made me stop chasing some end goal and just focus at the task at hand and it's also very meditative and makes me think about what I personally am looking for when I draw and what I miss, it's a means to explore my mind and contemplate things while at the same time training my drawing skills

>> No.6867832

>>6867819
because im trying to cope with wasting 3 years studying art

>> No.6867838

>>6867819
Art is an inherently masochistic hobby

>> No.6867839

>>6867831
that makes sense and is very understandable. personally I'm also tired of being a mindless consoomer, and I feel the need to engage my brain in a creative way now

>> No.6867845
File: 1.41 MB, 1024x1024, image.png [View same] [iqdb] [saucenao] [google]
6867845

>>6867829
never meant to be

>> No.6867847

>>6867832
i wish you could give me those years then. I wasted 30 years doing nothing

>> No.6867849

>>6867827
True, there's nothing wrong with wanting to make money with the skills you've trained.

>> No.6867854
File: 29 KB, 400x465, 1687524030973972.jpg [View same] [iqdb] [saucenao] [google]
6867854

>15 replies
>over an AI bait thread that was already posted on the very same day
>with a pepe for OP of all thigs
If Im not backseat modding this shithole of a site, telling something as basic as "report, hide and move along", you faggots swallow each and every single bait that comes, fishing rod included, I fucking swear to god.

Fuck this, you can reply to this AI bait threads as many times as you want, shit up this board, Im not your dad to tell you what to do.

>> No.6867855

>>6867826
I got into exactly through this shit
>see artwork that looks amazing to my eyes
>see that everyone else thinks it's amazing too
>see artist is very successful (money/fame)
>start dreaming of being able to wow people too and being paid for it
basically your average rockstar dream

but it was not just looking at some patreon stat and seeing it as an easy oportunity, rather it was seeing something amazing and it being combined with the fact that the amazing thing also allowed a sense of validation from it too

I was a kid back then btw, in the years since then my world view has developed drastically but I recognize that that's what kickstarted something in me

>> No.6867858

>>6867855
based and same. with the mention that I care only about validation from other artists

>> No.6867859

>>6867855
Most artists do but they won't admit it because it looks "shallow"

>> No.6867868

>>6867859
most successful artists did it because drawing is second nature to them. what AIcucks like to call "talent"

>> No.6867872

>>6867859
yeah I'm sure but I'm at a point in life where everything looks shallow, so I don't exactly have anything to hide from myself or others at least in this context

funnily enough nowadays the process is more interesting to me, I guess it's because with an outlook where everything is shallow, there's no need to expect immediate returns in order to achieve some vague dream

>> No.6867878
File: 73 KB, 720x406, IMG_20231001_231839.jpg [View same] [iqdb] [saucenao] [google]
6867878

>> No.6867900
File: 22 KB, 612x422, we ain't ever gonna make it.jpg [View same] [iqdb] [saucenao] [google]
6867900

>>6867872
>>6867855
Honestly it's not really the money I truly admire (though that's obviously a big plus). It's that these artists are so well liked. I'm a complete shut in and I want to be loved even if it's some parasocial internet shit

>> No.6867901

>>6867900
as the saying goes, everybody wants to be someone

>> No.6867928

>>6867855
For me it was quite the opposite
>drew as a kid
>life happened
>drew a little more around 18 because friend gave me a "how to draw manga" book and thought i'd just do something
>life happened
>be 25
>no life
>neet it up
>learn about how people can draw digitally
>decide to just buy tablet
>a few months in i learn about all the social media shit and people making money
>never even once attempted to "make it" or monetize
>had a few social media accounts but always started growing uncomfortable the more numbers i had and what i had to do to even be acknowledged
>the more i think about it the less likely i actually want to turn this into a job
>thinking about making money and the constant mentality of the art niche that every has to be appealing and your value and worth as a living being is directly correlated to the money you make and the numbers you have makes me hate art and drawing
>i started neeting it up because i didn't want to deal with this shit
On the bright side, I've been experiment for 5 years now and learned to critically observe everything around this shithole of a community, because i cannot have any conflict of interests since i don't really give a shit about who is who and what money they make

>> No.6867931

>>6867827
>ayy eye
yup this stuff ain't human

>> No.6867939

>>6867928
yeah I was looking at it through the eyes of a teenager and I basically thought that if I'm as good as X then the rest just happens, I had no interest in the social media hustle side of things or even really thought of that aspect so it didn't filter me

>> No.6867950

>>6867790
same.
i've been photobashing and cleaning AI art lately, but i realize that in a few years that will also become an useless skill
kind of lost about what doing with my life right now. drawing will always be a great hobby, but i need to make a living too

>> No.6867980
File: 133 KB, 1024x1024, 1696191947231667.jpg [View same] [iqdb] [saucenao] [google]
6867980

in just a few years the available jobs will be reduced an order of magnitude, and I honestly don't see any place for human artists in the industry in about 20 years
always a cool thing to know, like those dudes that can play the guitar at parties

>> No.6867983

>>6867980
damn I'm impressed with this new bing ai, no bullshit with loras needed yet it's able to replicate tons of likenesses reasonably well and it actually can do a variety of interactions, shame it's censored

>> No.6868002

>>6867980
in just a few years your slop machine will be gutted and banned in the industry. maybe this will fly in China or something

>> No.6868007
File: 56 KB, 604x377, 1458226596816.jpg [View same] [iqdb] [saucenao] [google]
6868007

>>6867980
>You will live just long enough to experience a Miku Ai embracing you.
Soon.

>> No.6868012

>>6868002
>multi billion dollar cooperations will allow themselves to be put at a competitive disadvantage because of (((artists))) screeching about muh copyright

sure thing buddy

>> No.6868017

>>6868012
>copyright will disappear over night just because I want my slop generator real bad
lol
>cooperations
how red is your face, corporate dicksucker?

>> No.6868023

>(((artists)))
says the cuck who shills for Sam (((Altman))). consider shotgun mouthwash faggot

>> No.6868030

>>6868017
private companies raising the expectations with cutting edge technology incentives open source projects to improve too, you can actually acknowledge the benefit of the profit motive without being a coorpo shill, and you literally can't stop people from running free software on their personal computer, with no amount of legislation. the cat is out of the box you absolute retard

>> No.6868035

>>6868002
it would be smart to be prepared for the worst, dude
>>6868017
corporations are precisely the ones trying to censor ai, make it closed source, and charge you monthly for it. the anti-ai artists are clearly useful idiots in this battle

>> No.6868052
File: 938 KB, 1022x1387, 1674272287945094.png [View same] [iqdb] [saucenao] [google]
6868052

>>6868035
In what way is being anti-AI useful to the corpos? It takes billions of images in the datasets to server as any sort of foundation. No single large corporation has a big enough supply of data where they own the IP to make such models. The whole house cards will fall apart if IP rights of the individual artists whose art has been stolen are respected. Corpos want to have their cake and eat it too.

>> No.6868056
File: 13 KB, 209x241, jewintensifies.jpg [View same] [iqdb] [saucenao] [google]
6868056

>>6868017
>The same people who made and pushed for Copyright laws are the same people funding and utilizing this AI shit the most.
Dualism is a hell of a drug.

>> No.6868075

>>6868052
>the whole house cards will fall apart
at this pace, future models will be able to generate full length feature films within a couple a years to a decade, this is literally a multi billion (trillion?) dollar industry right on US doorstep, losing out on this unprecedented increase in productive output would literally be a blow to american soft power, maybe even negatively affect its hegemonic position, it literally does not make sense from a game theory perspective. the US gov has always been hands off with emerging technologies despite their disruptiveness for this exact reason. literally everyone benefits from this, except (((artists))), who have no actual political power.

>> No.6868080

>>6867790
lmao honestly burn out had already made me stop practicing drawing but the new AI just outright killed any motivation I had left. What's the point? Sure it can't do NSFW but I'm not a coomoid so I don't care.

>> No.6868084

>>6868080
>it can't do NSFW, that's like the strength of it

>> No.6868085

>>6868080
>Sure it can't do NSFW
what the fuck are you talking about?

>> No.6868086

>>6868075
Yes it's called a "race to the bottom", only this time the price will be the downfall of what's left of human civilization. There will be a brief period where corpos will appear to win big (because they have the resources to actually market and gatekeep slop to the masses), but that will be short lived. People often compare Huxley to Orwell but I think HG Wells probably is the closest.

>> No.6868088
File: 158 KB, 398x559, it's over - le ducke face edition.png [View same] [iqdb] [saucenao] [google]
6868088

>>6867819
It's time to face reality: you can't put the AI genie back in the bottle and it's only getting exponentially better.

the battle is lost...
I see my house of cards collapse...

>> No.6868089
File: 84 KB, 512x768, download (7).jpg [View same] [iqdb] [saucenao] [google]
6868089

Just prompt instead. It's fun, and you can always make fancier stuff too like feature-length movies and games. AI won't do those very well, anytime soon. AI is only as good as your taste.

>> No.6868091
File: 74 KB, 637x628, IMG_4332.jpg [View same] [iqdb] [saucenao] [google]
6868091

>Almost nobody gives a shit about AI music
>Thousands of people shitting themselves good or bad over AI visual art
Why is one viewed differently seemingly to the other? Is music just viewed as more important or something

>> No.6868092
File: 2.90 MB, 2304x1536, comparison of AI prompt ''Realistic photo of a cat riding a unicycle on the moon. The unicycle's wheels are made of pepperoni. The earth is exploding in the background''.jpg [View same] [iqdb] [saucenao] [google]
6868092

>>6868084
>>6868085
I'm talking about the DALL-E 3 AI that Microshaft rolled out yesterday, it blows SD out of the water. It can't do NSFW but it does exactly what you prompt which is a huge deal.

>> No.6868097

>>6868092
god I fucking hate the dumb shit people prompt to try to show off AI
>le so random XD

>> No.6868102

>>6868097
le randum XDD shit is a good benchmark for seeing how precise it's at rendering abstract concepts, look at how SDXL, which is what coomers use for their god tier art of waifus, failed the test big time.

>> No.6868105
File: 80 KB, 183x322, twitch.png [View same] [iqdb] [saucenao] [google]
6868105

>>6867980
>always a cool thing to know, like those dudes that can play the guitar at parties
>75 per cent of children want to be YouTubers and vloggers
Nah, artists also can be streamers, YouTubers and so on. The #1 profession of the future, kek.

>> No.6868107

It's always strange to me to find people that don't actually like drawing on this board. Why are you here? It doesn't matter how good it gets, I'm not gonna stop drawing anime girls and shitty loomis heads on the backs of receipts.

>> No.6868108

>>6868102
except iterally all dall-e excels at is le random retarded shit, it utterly fails at making things anyone might actually want.

>> No.6868111

>>6868105
>Nah, artists also can be streamers, YouTubers and so on.
I'm a fairly know artist and as soon as the first AI rolled out I started learning animation, I will make it through this shit. New /beg/ drawfags are fucked forever tho lmaoooooo

>> No.6868133
File: 89 KB, 960x1200, F7RbfLbX0AAM4jS.jpg [View same] [iqdb] [saucenao] [google]
6868133

>>6868111
if you're a social media celebrity, you're 100% safe. that shit was never about the art
if you're a normal artist with a normal job, it's a good moment to rethink your career. the change is gonna be gradual but irreversible

>> No.6868136

>>6867811
>implying AI fags aren't only in it for the money

>> No.6868138

>>6868092
the wheel is a pepperoni pizza rather than a pepperoni but still it's impressive how it's now able to localize things as it's asked to and how the general quality seems higher too (the cat looks like a cat way more than the others)

shame it's neutered and won't produce cool shit like big tits and ass

>> No.6868139

>>6868111
>I'm a fairly know artist
yes same sir, we are both well know sir. AI is the future I gave up art because I know it is over for me as a greatly know arter.

>> No.6868144

>>6868136
I'd say most of AI posters on 4chan are in it to see some peak looney troon reactions from retards who cope in the most ridiculous ways possible, easiest replyfarming possible and entertaining too
the other reason is fapping purposes and also just general interest in where it's going

>> No.6868147

>>6868139
>le Pajeet cope
lmao this nigga hasn't seen the crazy anti-Pajeet AI images /pol/ has been generating, Indian /pol/tards got really, really mad about these, I even felt about them until I saw the toilet Arjun tank and then laughed for 5 minutes.

>> No.6868152

>>6868147
>the supposed artist is actually a /pol/ shitposter
go back.

>> No.6868154

>>6868092
Quick rundown?? How is it better than SD?

>> No.6868156

>>6868091
popular music figures already are kee deep in ai, as long as they get a slice of the profit. the whole industry has always been more about brand recognition

>> No.6868166

>>6868092
the pace of progress is crazy right now
no sensible person could make an informed guess about how far this tech will get in a decade
>>6868154
just look at the picture

>> No.6868169

>>6868111
how's animation going to help you? not only is AI advancing in that field too but your output will decrease hugely to the point where your avenues for staying relevant decrease drastically

I mean I suppose you can hope to be some diives type of character out of sheer luck (biggest fucking mystery in the universe why this particular artist has reached popularity when all the output is just the bare minimum back and fort jiggle and borderline /beg/ tier quality)

or you can aim for some few minute long porn animation and hope to get patreon pay piggies to sustain your for a few months between every release

or do you aim to work in the animation industry? sounds like personal hell to me but you tell me

>> No.6868170
File: 235 KB, 624x631, g.png [View same] [iqdb] [saucenao] [google]
6868170

>>6868147
>Indian /pol/tards got really, really mad about these
>just click on /g/
>pic rel
o.O

>> No.6868177

>>6868154
it's not.

>> No.6868183

>>6868154
it follows what you ask it to do way more closely and isn't as limited in how much stuff can be going on before it starts to turn into a mess or just ignore 90% of what you asked from it

>> No.6868187

>All images and discussion should pertain to the critique of visual artwork.
>the critique of visual artwork

There is nothing too critique or discuss about "AI" "art" the only discussion around this topic pertains directly to technology, economics, and social media, but not to visual artwork itself. This topic belong on >>>/g/

>User-created artwork is submitted for critique, and visual art is discussed. Do not claim authorship of works you did not create.
>User-created
>Do not claim authorship of works you did not create

"AI" "art" is not user-created. Claiming authorship of images that are automatically generated is claiming authorship of work you did not create.

>> No.6868192

>>6868170
i heard some theory about preemptive demoralization of the population, given that india and china are potential economic rivals
same could be suspected about tranny and race mixing propaganda in america

>> No.6868195

>>6868035
>if you defend your work you're a useful idiot
fuck off retard

>> No.6868198

>>6868030
>the cat is out of the box you absolute retard
you're not a cat. we can shove you in a box and turn on the gas

>> No.6868199

>>6868187
wrong. there's more critique of visual artwork itt than in the average one
there are no rules, just tools

>> No.6868201
File: 117 KB, 638x638, 9a2b614e229ec2136566aab04d37b2a4969f33421d5b62cc9690c11e4c484886.jpg [View same] [iqdb] [saucenao] [google]
6868201

>>6868187
Dont bother anon, these people dont draw and much rather talk about nonsense than drawing. Let them take the bait, they just dont know any better.
Feel free to report the thread, however.

>> No.6868204

>>6868199
>there are no rules, just tools
kill yourself
https://www.instagram.com/p/CxyBNwGrBvh/

>> No.6868205

>>6868201
/ic/ in general doesn't draw and rather would make up excuses for its failures
AI IS STEALING MY WORK is just another new cope in a long list

>> No.6868207

>>6868030
>and you literally can't stop people from running free software on their personal computer
https://edition.cnn.com/2023/09/27/asia/south-korea-child-abuse-ai-sentenced-intl-hnk/index.html

>> No.6868218
File: 251 KB, 1024x1024, 1691436955624516.jpg [View same] [iqdb] [saucenao] [google]
6868218

>>6867797
AI has been great for me it lets me give people who aren't schizos like you example work for concept art comissions.

>> No.6868222

>>6868218
>I can't and won't draw but I will go on a board for drawing anyway

>> No.6868224

>>6868218
wish i could render something as cool as that someday

>> No.6868226
File: 378 KB, 900x900, bckomix_kjg_panel_shot_09.jpg [View same] [iqdb] [saucenao] [google]
6868226

>>6867790
>ai made me quit drawing all together
It made me double down on trad cartooning techniques, been inking comics with a brush & doing my own lettering. After all, AI will never be me.
pic related, a comic I'm working on about KJG.

>> No.6868229

>>6868226
smart thinking. all those anti-ai retards will support you, but not because of the art, obv

>> No.6868231

>>6867790
Demoralization threads made by ai faggots. As usual. You will never be a artist. You will never make OC.

>> No.6868233

>>6868226
>"modern kid" has early 2000s JNCO pants and chain wallet

Unironic Boomer detected.

>> No.6868236

>>6868233
well even the comic is your typical kids with their smartphones not appreciating a good pail of water energy

>> No.6868237
File: 203 KB, 468x520, 1683927088274151.jpg [View same] [iqdb] [saucenao] [google]
6868237

>>6867790
Same, the reason I even started was to make money. Who knows what'll happen to jobs in a future with AI.

>> No.6868241
File: 149 KB, 1024x1024, 1696193856209654.jpg [View same] [iqdb] [saucenao] [google]
6868241

>bing can't make lewd-

>> No.6868249

>>6868241
Alright then anon post this to Twitter/reddit to let everyone know you can make porn with this

>> No.6868250

>>6868241
Prompt? asking for a friend haha

>> No.6868255

>>6868241
This is somehow even worse than stable diffusion slop

>> No.6868278

>>6868229
And you will be another nobody pajeet

>> No.6868315

>>6868166
>just look at the picture
It still looks like an oddly photoshopped together mess? I'm glad that the prompting shit is now turning more intelligible, rather than the repeating rambling of a madman that it was previously. I guess prompters are already being made redundant; I always said that if artists get fired, it'd be the secretary doing the prompting, not some 'prompt engineer".
But other than that, I'm not really anymore impressed than I ever was by AI.

>> No.6868357

>>6868250
bing can guess prompts from pictures better than humans. also, new prompts don't require magic obscure bullshit anymore

>> No.6868358
File: 122 KB, 487x518, IMG_20230519_191146.jpg [View same] [iqdb] [saucenao] [google]
6868358

>>6868089
So anon how often do you try to red pill people

>> No.6868365
File: 54 KB, 512x768, download (35).jpg [View same] [iqdb] [saucenao] [google]
6868365

>>6868358
I've been prompting for a grand total of 2 days now. in that time I have shilled exactly twice in the past few hours.

>> No.6868370
File: 86 KB, 1024x519, 1679496202893414m.jpg [View same] [iqdb] [saucenao] [google]
6868370

Abortion graph art
>also this is now an image dump thread post whatever
>janny ain't gonna do nothing

>> No.6868373
File: 806 KB, 951x648, 1693251023553663.png [View same] [iqdb] [saucenao] [google]
6868373

>>6868365
that ain't what I asked nigga

>> No.6868375
File: 49 KB, 625x790, 1694107503715611.png [View same] [iqdb] [saucenao] [google]
6868375

>> No.6868376

Bing is good... but after 500 prompts I see the code now, I am starting to prompt within it's limits as pushing them any further was getting too tedious.
>prompt blocked
>Prompt blocked
>Prompt blocked
>Banned for one hour for trying to get a woman fitting lots of pingpong balls into her mouth

>> No.6868385
File: 1.07 MB, 220x220, rip-bozo-packwatch.gif [View same] [iqdb] [saucenao] [google]
6868385

>>6868376
Well soon it will prob be pozzed as well, the AInigga already feeding bing their typing data and posting it
>not like they'll be missed

>> No.6868387

>>6868376
that's what uncucked open source is for. if you own a $3000 PC, that's it

>> No.6868390

>>6867803
sucks to be you for needing spite to create

>> No.6868395

>>6868387
Show me that via a local install. Idk if you could get that... maybe. I was inspired by the spy with 100 cigarettes

>> No.6868401

>>6868376
Why the fuck are some AI so weird with what you can prompt?

>> No.6868402
File: 897 KB, 852x964, l683mzivre381.png [View same] [iqdb] [saucenao] [google]
6868402

>>6868390
Never underestimate the power of spite

>> No.6868403

>>6868376
this is why im glad i got into it two years ago
>>6868387
>$3000 PC
you can do it all on the cpu, depends how long you want to wait for the gen

>> No.6868405

>>6868401
>what do you think coomers are trying to bypass
>6868376
>"pingpong" {BALLS in mouth}

>> No.6868420

>>6868376
>>6868385
Funny, I never these problems with my pencil & sketchbook.

>> No.6868423

>>6867790
AI made me quit drawing for money but drawing for fun

>> No.6868449
File: 140 KB, 1024x1024, _533c2ea5-61a6-4847-88f1-daa893963ba6.jpg [View same] [iqdb] [saucenao] [google]
6868449

I make traditional art for fun and as a hobby, so it's not a problem for me. I never liked digital art that much. Experimenting with AI is fun tho :) I'm going to try to replicate this image with graphite and maybe gouache.

>> No.6868468

programmers havent killed themselves

>> No.6868470
File: 470 KB, 973x819, i made this.png [View same] [iqdb] [saucenao] [google]
6868470

>>6868390
The alternative is consuming product to produce product referencing the product of chosen consumption for the sake of consumption and mindless pretty images to feel amazed that you have laid eyes upon the pretty image and delude yourself you're some art connoisseur because you can really see how the flowers are soooo pretty.

I'd rather get run over by a bus and shot in the dick than live like that

>> No.6868472

I can't even imagine quitting drawing. Thats equivalent to committing suicide in my mind. If I can't draw, I don't have a reason to keep living.

>> No.6868482

>>6868470
that's not the only alternative

>> No.6868484

>>6868449
you can always tell how full of shit AIcucks are when they claim they're artists, because they always generate trash like this. completely uncreative "people"

>> No.6868500
File: 49 KB, 575x380, 1639987986465265.jpg [View same] [iqdb] [saucenao] [google]
6868500

>>6868482
Bro, you completely changed my mind.
How did i not see that?

Truly insightful. Thank you.

>> No.6868518
File: 80 KB, 922x988, c53.jpg [View same] [iqdb] [saucenao] [google]
6868518

Honestly when it comes to Le "AI" I find it more funny if some of the AIniggas do try to copy some of my stuff cause they'll need to explain why they're being racist.
>not like they want to actually explain where some of those loli "pics" came from

>> No.6868519
File: 130 KB, 1024x1024, _a08e2f7e-68fe-4a39-aff4-c6001b6621ab.jpg [View same] [iqdb] [saucenao] [google]
6868519

>> No.6868533

>>6868472
Same, it's like tasting freedom for the first time. There's no way you can imagine giving that up. It's sad to think this freedom might be felt less in the future as kids give in to corporations and machines...

>> No.6868538
File: 249 KB, 828x879, 1670787834657801.jpg [View same] [iqdb] [saucenao] [google]
6868538

>>6868111
>>6868133
>artists also can be streamers, YouTubers and so on
Pic related, Zone is more of a vtuber than an artist now

>> No.6868541
File: 53 KB, 1024x1024, OIG (14).jpg [View same] [iqdb] [saucenao] [google]
6868541

>>6868519
this one is pretty good

>> No.6868542

What's the most painless way to kys?

>> No.6868545

>>6868542
Staying alive

>> No.6868548 [DELETED] 

>>6868542
Strapping c4 to yourself and walking into google hq

>> No.6868553
File: 99 KB, 1024x1024, OIG (15).jpg [View same] [iqdb] [saucenao] [google]
6868553

why are we here just to create

>> No.6868556

>>6868542
drawing

>> No.6868562

>>6868542
>lethal injection
>unless you know

>> No.6868564

>>6868533
making art unviable as a job should have the opposite effect: people only will get into creating art because they legit LOVE the craft, kids won't be deluding themselves with "making it" and crap like that
and humanity as a whole will enjoy better art; i can't think of a better possible timeline

>> No.6868568

>>6867803
Very relatable

>> No.6868581
File: 11 KB, 99x112, IMG_20230407_141521.jpg [View same] [iqdb] [saucenao] [google]
6868581

>>6868533
Would that just mean you've have to red pill them faster before they can get groomed

>> No.6868588
File: 187 KB, 1024x1024, 1696191518204.jpg [View same] [iqdb] [saucenao] [google]
6868588

>>6868423
You get it.
But seriously, I won't be mad if /ic/ adds "NO AI-ART" to its rules
This place should transition now into an oasis for people that enjoy the old school style, just for the sake of it

>> No.6868592

>>6868519
>Mkay anon but "AI" thread
>might as well move this to bump limit because janny ignoring this one

>> No.6868596

>>6868588
>those are some long arms

>> No.6868604
File: 27 KB, 1877x129, a cow life for all.png [View same] [iqdb] [saucenao] [google]
6868604

>>6868564
nigger do you understand that it's insulting and revolting for these parasites to take my shit even if it's just as a hobby/fun? the thought of these cancers taking our shit and selling it back to us disgusts me to the core. I'm never, EVER going to post shit on the internet, and most artists will feel this way as well. you cannot expect people to be treated like cattle and shrug it, our brains did not evolve to work like this and for good reason

>> No.6868605

>>6868581
Absolutely, it's for them I'm going to continue advancing my skill and remain ahead of machine diarrhea as I have been so far, when before my only ambition was to just be good enough to impress people at parties while I continue my studies in other fields. But after Kim Jung Gi's death, and seeing some literal soijack take his life's work and fed it into a program to shit out abominations turned a switch on in my psyche. I hate these bugmen with so much passion that I won't even bother describing, I'll just continue drawing.

>> No.6868612

>>6868588
You think gookmoot cares about the wellbeing of this site? It's basically a dumping ground for the CIA and other gov entities to manipulate public mood and feelings.

>>6868564
Humanity for the most part doesn't enjoy art, they consume shit and dump it when something new comes up.

>> No.6868613
File: 151 KB, 1024x1024, _4c724db1-3355-4da2-ace3-da15bd99a8ac.jpg [View same] [iqdb] [saucenao] [google]
6868613

>>6868592

>> No.6868644

>>6868613
>(You) me again

>> No.6868654

>>6868604
you put an unhealthy ammount of you self-worth in a mundane skill
i've been painting as a hobby for 10 years, but i was quick to realize the writing on the world last year
still, i kinda understand your loss. denial is hard, hope you'll be able to accept reality soon

>> No.6868656
File: 24 KB, 700x448, FKCzXg1XwAI07BZ.jpg [View same] [iqdb] [saucenao] [google]
6868656

spider ref time

>> No.6868657

>>6868654
PYW then nigga

>> No.6868662

>>6868613
love those colors

>> No.6868665

>>6868662
>>6868613
T.samefag
>you do this Everytime

>> No.6868669

>>6868657
true story: i made something for a draw-this-in-your-style thread a couple of years ago and some kid told me that i should be desperate for (You)s for being a pro and posting here. i'm mediocre at best, but i felt like a real pro that day

>> No.6868675

>>6868665
not a same fag, also am telling you to get out of AI threads and to just draw.

>> No.6868678

>>6868669
pros posting in those threads inspire me to draw more. Please pros, do it for me.

>> No.6868681

>>6868654
kek, how cute. know that you're so small and worthless in my eyes, no amount of "demoralization" tactics work coming from a literal liar. you don't even know what I want and what principles I have. you're not worth the last thing I've stepped in bitch
>denial is hard, hope you'll be able to accept reality soon
what even am I supposed to make of this worthless platitude? you're a cunt who thinks he has an impact on my future decisions. that's how arrogant and ridiculous you are

>> No.6868686

>>6868675
Okay anon but over where I'm at it's 10:00pm basically
>I already drew in the morning and posted it
might as well get this thread off front page for the morning

>> No.6868687
File: 1.17 MB, 330x248, 686.gif [View same] [iqdb] [saucenao] [google]
6868687

movement ref

>> No.6868690
File: 1.60 MB, 320x180, MQPNXG.gif [View same] [iqdb] [saucenao] [google]
6868690

bug ref

>> No.6868692

>>6868588
as long as they don't say it in the rules, is very hypocritical to keep deleting these threads for no reason
it makes sense to protect the fragile mental health of begs in denial, but the rules are supposed to be the same for everybody

>> No.6868694

>>6868681
Struck a nerve?

>> No.6868695

>>6868686
ok, for you that's ok. anyone reading this who hasn't drawn an ink spider today, what are you doing itt? Go draw!

>> No.6868698

>>6868694
no retard, I actually want you to tell me what did you mean by that. what "denial" am I in and what "reality" do I have to accept?

>> No.6868703

>>6868698
Those who can't do, buy.

>> No.6868704
File: 517 KB, 910x966, 1696149318239284.png [View same] [iqdb] [saucenao] [google]
6868704

>6868692
>spam board 24/7
>why are the AI threads being deleted
>you could just go to /g/

>> No.6868709

>>6868704
>:I have low self-worth and want others to suffer with me, I can't try to demoralize artists on /g/"

>> No.6868712

>>6868675
NOOOOO, anything but drawing
i must blame AI or anybody else, for my lack of motivation and discipline

>> No.6868715

>>6868703
pure incoherent drivel

>> No.6868716

>>6868715
Too close to home, eh?

>> No.6868721

>>6868092
Stable diffusion sers I don't feel so good

>> No.6868727
File: 780 KB, 432x243, 305f1e083811435eb5bb6f23cca814d6.gif [View same] [iqdb] [saucenao] [google]
6868727

>> No.6868731 [DELETED] 
File: 66 KB, 540x564, 1685480168954888.jpg [View same] [iqdb] [saucenao] [google]
6868731

Jannies really are useless huh

>> No.6868733

>>6868721
SD is still better because it requires actual effort, Dalle-3 is for promptlet faggots who can't handle the jandal.

>> No.6868734

>>6868731
Give them a break! trains is a hard job.

>> No.6868738

>>6868731
Only the most pathetic of individuals would moderate this garbage for no pay.

>> No.6868739

>>6868721
still the best tool for a number of reasons, but if they don't keep improving, closed source is gonna left them behind. and that would be the worst possible scenario for everybody but corporations like adobe or m$

>> No.6868742

>>6868716
it's hard for something to hit close to home when you're not making any sense. try again

>> No.6868746
File: 7 KB, 259x195, images (2).jpg [View same] [iqdb] [saucenao] [google]
6868746

>> No.6868750
File: 410 KB, 1462x1462, 1578932412956.jpg [View same] [iqdb] [saucenao] [google]
6868750

>>6868739
>tool
you faggots really think you can normalize this shit here, huh?
>>/g/

>> No.6868754
File: 643 KB, 200x119, 200w (1).gif [View same] [iqdb] [saucenao] [google]
6868754

Gif version ref

>> No.6868759

>>6868731
>>6868746
i feel your angst, and i admire you for trying to fight for a lost cause
still, you gotta admit: it's petty and pathetic: try to change the rules, instead of crying everytime things don't go your way. peace & godd night

>> No.6868761 [DELETED] 

>>6868759
Good night sir please do not redem your dreams sir

>> No.6868801

>>6868694
He clearly triggered you, yeah

>> No.6868811
File: 2.87 MB, 4808x2680, 1695308366513641.png [View same] [iqdb] [saucenao] [google]
6868811

>>6868759
well anon this nigga exists and he's been still developing his shit for 10/11 years now waiting for the perfect "workflow"
>gamemaker stuff have gotten finished and acclaimed by the time he change to the lastest thing

>> No.6868833

>>6868030
You can't but you can probably effect where the image generations get posted.

>> No.6868858

hey guys, stan prokopenko here, just wanted to let all you losers know after my evil plan to destroy the art industry is nearing completion. after you nerds mercilessly mocked my kangaroos I invested all my capital into this technology. Come 2 more weeks aphantasia diffusion will release, and then it'll be over for you all, ahahahahahaha

>> No.6868861

>6868030
>such as what the app store
>what "private companies" do you mean

>> No.6868871
File: 17 KB, 474x473, d01b7ac0e32e1b8d111edc4c90bd0709.jpg [View same] [iqdb] [saucenao] [google]
6868871

>6868858
But how racist will it allow me to be

>> No.6868873

>>6868218
And when eveyone can do this you won't be needed but let me guess this is just your side gig so it doesn't matter.

>> No.6868888

>>6868564
The concept of the starving artist exist for a reason dumbass anyone that gets into a art career only does so because they like making it or else they'd literally do anything else if money was the goal.

>> No.6868891

>>6868218
>just head portraits are concept art
>not even turn around or other refs just 3/4 view
>also head portrait of character you prob didn't make or got from touhou

>> No.6868896

>>6868564
have looked up game development practices anon
>or animation development

>> No.6868904

Can AI make skibidi toilet?

>> No.6868907

>>6868904
AI can make anything if you reroll often enough.

>> No.6868915
File: 67 KB, 951x817, 6364431249ba9.jpg [View same] [iqdb] [saucenao] [google]
6868915

>>6868907
so anon will it also be making (they deserved it posts)

>> No.6868916

>>6868907
Can AI Make Garten of BanBan?

>> No.6868946
File: 346 KB, 1920x1076, 1686595250955852.jpg [View same] [iqdb] [saucenao] [google]
6868946

Corporations like Amazon are already using AI for promotional material, pic very much related. It's only going to get worse with the advances made by DALL-E 3. Once they start shopping it around on the enterprise level, it's truly over. Couldn't imagine being a concept artist right now.

>> No.6868974

>>6868946
kill yourself

>> No.6868981

>>6868946
aifag discovers that photobashing is basically a must for this kind of stuff. it's over for who, photobashers? this shit is used for the sake of novelty, that first car you see has two fronts. you retards keep talking about art, keep false flagging as demoralized regulars but you instantly give yourselves away because you put effort into false flagging as much as you put into making art. none. by all means, these companies should keep doing this shit and then get sued into oblivion

>> No.6869103

>>6867828
>ADHD
Nope, I'm still going. Hugely at random but having successes. Yes, I did make one thread at one point. Yes, I'm still unable to study in conventional way, but I'm a lil bit smarter than authors of the rest of those threads so I managed to create a "workflow", if you can call it that, that helps me with learning much better than my normal but yet not quite as well as normal people.
Determination wins it all. I will defeat adhd, just as I always do with any other problem.

>> No.6869108

>>6868946
I'm 100% sure that is going to be a good quality product with absolutely no cut corners.

>> No.6869122

how can we beat it?
what skills can human artists do that ai can't?
i know you will say trad but ai can make trad looking art

>> No.6869124

>>6869122
innovation.

>> No.6869125

>>6869122
Only I am able to put my schizobabble onto a canvas. AI will never take my delusions away from me

>> No.6869127

>>6869122
>>6869124
you can't beat a thieving cunt using image recognition tech with ML dumbasses. they'll just train it on your shit. the answer to this question is through punishing laws meant for protecting your work and suing. which is the main problem after all

>> No.6869130

>>6869127
You can't beat chinese sweatshops neither with that reasoning.

>> No.6869136

>>6869127
>they'll just train it on your shit.
Ok, but I'd be first, and for them to even know my shit to bother to train it means I've probably made it.

>> No.6869146

>>6867819
Whos this cutie

>> No.6869149

>>6869122
(me)
We can use interesting physical material. Like glass shards, or glitter. Then we can make a tiktok video of us rotating the piece of paper under the light to show that it reflects. This is so complex ai's stochastic (random) function won't be able to replicated consistently in a video.
I am le genious
>>6869127
Why waste money and time on copyright bullshit when you can give money to actual artists?

>> No.6869154

>>6867832
>wasting
yeah, I can see why AI worries you
I'll keep living my life and doing the things I enjoy, including drawing

>> No.6869157

I like AI because it makes people grow up
Drawing pictures doesn't make you special, lol

>> No.6869161
File: 114 KB, 1024x1024, _059b6a9b-3e0b-459e-b1c8-81374ad6b181.jfif.jpg [View same] [iqdb] [saucenao] [google]
6869161

I'll be honest I don't get the hype around dall-e 3. It's so censored even innocuous prompts like "hugging" are a 50/50. It struggles with established character much more than SD without loras. It has a very, very limited range of faces it can produce for photorealistic pictures, no means of additional control, no inpainting, so you rely on gachaing more. Hands are still a miss 60% of the time. Text is okay but still skewed, you're better off just biting the bullet and photoshopping it in, it's just wasted time

I think this is just a giant ploy to try and inorganically boost their shitty search engine, they openly advertise giving you "tokens" for faster gens if you redeem "points" you gain by using Bing as a search engine and logging in everyday, literal mmo/gacha hooking tactics. I still think SD is superior and I'm still not scared, I think anyone who thinks they should give up with drawing is a pussy and should take a break off the internet to JUST DRAW

>> No.6869172
File: 130 KB, 2520x889, wowie.png [View same] [iqdb] [saucenao] [google]
6869172

Also, I want anyone who is seriously thinking of becoming a dall-e PROOOOMPTER to take a good look at the terms of use. You can also get permanently suspended from using their services.

>> No.6869177

>>6868449
Why you made Anakin zesty as hell wtf man

>> No.6869202

>>6868091
Unlike music, people don't give a shit about the artist when it comes to visual art.

>> No.6869209

>>6868981
I work in the games industry as a viewmodel animator you flaming faggot. I'm not a fan of AI art but it's about being realistic and seeing the forest for the trees. This is where we are heading and AI goes way beyond photobashing. Equating the two with one another means you're simply being disingenuous.

>> No.6869214

>>6869209
pyw

>> No.6869217
File: 170 KB, 792x1000, 1690243745426622.jpg [View same] [iqdb] [saucenao] [google]
6869217

>>6869214
Play Modern Warfare II :)

>> No.6869220

>>6867790
Good, less competition for me.

>> No.6869221

Now that the profit motive has been abolished, maybe now people will make artwork for emotional reasons again.

>> No.6869223
File: 57 KB, 405x720, 1689525873556313.jpg [View same] [iqdb] [saucenao] [google]
6869223

>>6869122
>how can we beat it?
>it's stealing!
>it's not a tool!
>report AI threads because i want this to be my safe space where i don't have to be confronted with AI despite it being clearly related to art and this board!
ngl, genuinely tired of /ic/s and artists retardation.

>> No.6869228

>>6869223
>tries to stir shit up by replying to a falseflag

>> No.6869230
File: 232 KB, 1051x547, 1683580555141477 (1).jpg [View same] [iqdb] [saucenao] [google]
6869230

>>6868144
Haha, true. Nothing better than just slapping down some obviously good ai art and seeing /ic/ guzzle buckets of copium telling me not to believe my lying eyes

>A-actually there's two pixels wrong so it's shit
>P-people only like human art I swear
>It's all astroturf, every thread on every board is astroturf, nobody has ever liked an ai artwork ever
>Check out the REAL soul that AI will never create and it's the worst artwork you've ever seen in your life

>> No.6869233

>>6869221
just like the good old days, when merchant guilds paid sculptors to create beautiful works of art for life.

>> No.6869235
File: 170 KB, 512x512, 00001-2252605782.png [View same] [iqdb] [saucenao] [google]
6869235

its not about passion

its all about profit

>> No.6869239

>>6868588
/pw/ and /sp/ existing, but no /ic/ and /ai/ separation is fucking bizarre

It's all about what advertisers apparently want

>> No.6869241

>>6869230
>I'm winning
>that's why I'm typing out a longass cope post

>> No.6869246

>trying to do inktober with some AI assistance.
>prompt dall-e cause why not
>result is already better than anything i've seen submitted for inktober
>and it is good enough to serve as inspiration

>> No.6869248
File: 48 KB, 1260x213, ic expert predicts the future.png [View same] [iqdb] [saucenao] [google]
6869248

>> No.6869249

>>6869246
>not a new IP

>> No.6869257

>>6868654
>writing on the world
Sanjay, good post. I hope many rupees befall you

>> No.6869261

>>6869248
>2021
Was he living under a rock? Artbreeder was already showing how powerful ai was going to be shortly.

>> No.6869262

>>6869261
That was from a Dall-E thread.

>> No.6869265

>>6869262
weird. well, maybe it will be the last to go...
it may morph, some forms may be no longer in vogue.

>> No.6869268

>>6868226
based. Where can I read your comics?

>> No.6869269
File: 257 KB, 2245x395, Screenshot from 2023-10-02 14-35-15.png [View same] [iqdb] [saucenao] [google]
6869269

>>6869261
Microsoft's "Next Rembrandt" was already doing it in 2016.

>> No.6869277

>>6869230
>obviously good ai art
heh

>> No.6869279

>>6869230
how many boards to you spam with your "good" ai """"art"""""? what part of "you're not wanted here" do you not get? you're not an artist, nor do you draw. you're like a cockroach

>> No.6869287
File: 26 KB, 464x461, c.jpg [View same] [iqdb] [saucenao] [google]
6869287

>>6869248
Still ZERO job offers from AAA video game and big movie/animation studios for prompters

>> No.6869293

>>6869223
>>how can we beat it
Based. Wanting to expand artistic expresion beyond what is currently common is what artists do
>>it's stealing!
True, but all artists steal. Even drawing trademarked characters is stealing. I agree, this is a bad sentiment to have
>>it's not a tool!
It's not. The pieces are already done
>>report AI threads because i want this to be my safe space where i don't have to be confronted with AI despite it being clearly related to art and this board!
Yes. Look at >>6869239

>> No.6869295
File: 332 KB, 1024x1024, _42e049c5-a320-4142-8a7a-8d44013c6820.jpg [View same] [iqdb] [saucenao] [google]
6869295

I might use AI as reference, especially as it gets more and more coherent, but I will never quit drawing.

>> No.6869296

>>6869287
Are there any job offers for artists from game and movie/animation studio of any size? Prompting is so easy the janitor can do it while not cleaning the floor.

>> No.6869298

>>6869122
I will do trad, I have a feeling that in the rise of AI and digitisation of life, people will get a renewed interest in human made things. And a real tangible painting is always going to be unique compared to a digital picture.

And I still love plein air painting, I don’t care so much about AI when I’m out and about to capture a moment on canvas.

>> No.6869301

>>6869277
>>6869279
>The cope brigade already coming in with "it's not actually good", "it's all astroturf by 1 guy and nobody really wants it"

kek

>> No.6869306

>>6869296
https://www.artstation.com/jobs/B2Ev

>> No.6869326

>>6869223
>ngl, genuinely tired of /ic/s and artists retardation.
lucky for you, can can go to any other board and multiple subreddits and discord servers and swallow all the AI spam you want

>> No.6869335

>>6869295
>you've said this about 20x already
PYW and let us see what your "referencing" or are you just tracing

>> No.6869341

>>6869296
some artists are capable of doing multiple positions anon
>some are also capable of being a solo dev

>> No.6869342

Thanks for feeding the generator with your future art, everyone!

>> No.6869345

>>6869295
The thing is that using the AI even as a "reference" just makes you a cheating fraud using crutches, and when your fans find out they will turn on you.

>> No.6869348
File: 107 KB, 341x334, 1629078456254.png [View same] [iqdb] [saucenao] [google]
6869348

>>6869230
>spam all boards non stop with random images
>go in art space with the whole "im real artist" shtick and "ai is like real ppl"
>anyone replies with words or tells you to stop spamming;
>UR COPING AND BTW WE'RE JUST TROLLING YOU BECUASE UR SCHIZO UR COPING CCOPIUM SCHIZO WALL OF TEXT HAHA UR SEETHING AND MUST BE MAD
>the modern trolling consists of being annoying and then spamming "cope" at anything because anyone replied
>implying that signaling how much you're "replyfarming" doesn't come off as mental illness
>if you ignore them, they just see it as an excuse to spam more threads
Nigga, just go ask gookmoot and get your own board already

>> No.6869353

>>6869335
I don't reference AI. I said I might. it's way too shit right now.
>>6869345
>when your fans find out they will turn on you.
I have no socials, I do it as a hobby

>> No.6869357

>>6869353
pyw

>> No.6869362

>>6869353
then why mention it at all. It basically saying your waiting for the perfect image search thing to look up practically.

>> No.6869366
File: 167 KB, 1024x1024, AI-generated.jpg [View same] [iqdb] [saucenao] [google]
6869366

>>6869348
>>spam all boards non stop with random images
This is what the future of the internet is going to look like, everything drowning in AI-generated media and fake content. There are even AI made videos on youtube already. You could probably even go as far and use it to fake historical photos and add them to national photo archives.

>> No.6869368

>>6869362
>It basically saying your waiting for the perfect image search thing to look up practically.
sure, that's kind of what I want from AI.
>designing a robot joint, don't like the google results, prompt AI using some slightly unusual keywords, find something more interesting to
combine with my own ideas.
like a springboard. I don't want to copy the results and present them as my own.

>> No.6869369

>>6869366
Can you write more than the same two fucking arguments?

>> No.6869373

>>6869369
What argument? Some boorus are already getting shit up by it, it's more of a fact.

>> No.6869374
File: 85 KB, 595x900, iknn8iljdnq21.jpg [View same] [iqdb] [saucenao] [google]
6869374

>>6869366
so are ya telling me people are going to do this anon
>and now people should "believe" stuff on the Internet more

>> No.6869378
File: 102 KB, 645x444, Pepe the rad dude with attitude.jpg [View same] [iqdb] [saucenao] [google]
6869378

>>6868192
It's all propaganda, baby

>> No.6869379

>>6869373
I never interact with AI art outside of spammers on this website.

>> No.6869385

>>6869348
this too is cope.
>stop spamming your "shtick"!
it's called opinion, retard
>UR COPING AND BTW WE'RE JUST TROLLING YOU
you wish they were trolling you. they're just telling you the reality you want to avert your eyes from.

>> No.6869386

>>6868088
ywnbaa

>> No.6869387

>>6869366
>I can't believe photo edits never existed before AI anon

>> No.6869388

>>6869385
>no argument

>> No.6869390
File: 857 KB, 977x987, 1655465665654.png [View same] [iqdb] [saucenao] [google]
6869390

>>6869366
>You could probably even go as far and use it to fake historical photos and add them to national photo archives.
Why would anyone do that?

Yeah, but it will also work the other way around.
AI is like giving a gun to everyone.
>everything's gonna be fake
It already is.
I don't really care either, i just want to talk because i'm lonely.
>>6869385
stop seething
why are you so mad?

>> No.6869391

>>6869385
>>6868704
So anon /g/ wasn't invaded by newfags making 13 and more threads that hit the image limit in seconds
>are you telling these people are spreading the truth there

>> No.6869393
File: 15 KB, 173x150, IMG_20231002_070738.jpg [View same] [iqdb] [saucenao] [google]
6869393

>>6869366
>can't believe photoshopping images was only a recent development

>> No.6869395
File: 13 KB, 230x213, Hoaxed_photo_of_the_Loch_Ness_monster.jpg [View same] [iqdb] [saucenao] [google]
6869395

>>6869390
>Why would anyone do that?
Yeah man, there are 0 people out there interested in historical revisionism and crazy conspiracy theories.

>> No.6869397

>>6869379
That's what you think

>> No.6869400
File: 69 KB, 650x832, naz_egypt.jpg [View same] [iqdb] [saucenao] [google]
6869400

>>6869393
>>6869387
At least those took time and some sort of skill.

>> No.6869401

>>6869400
The shitty AI rabbit profile pic already destroys any credibility that guy could have

>> No.6869405

>>6869400
>skill
>you find any old new sources laying around
do you believe most historical stuff are 100% accurate
>do think most nazi propaganda footage is accurate to their capabilities
>do you even /his/fag

>> No.6869407
File: 683 KB, 724x1315, RDT_20231002_072015884224208751274469.png [View same] [iqdb] [saucenao] [google]
6869407

My favorite historical moment to happen recently trying to be covered up

>> No.6869417

>>6869306
>view these positions
>senior
>senior
>senior

For me it is over before it even started.

>> No.6869418

>>6869405
>do you believe most historical stuff are 100% accurate
Yeah, anon, the problem with this stuff is not intelligent people who have critical thinking skills, it's the majority of humanity who are dumb as rocks and minors on the internet.

>> No.6869420

>>6869417
If your work is pro, you can get a senior position anyway if your portfolio beats the other seniors who apply.

>> No.6869421

>Other boards are now filled with AI slop since dalle 3 came out
How do we stop this?
It used to be just some retards from /g/ spamming other boards, now it feels like everyone jumped on the AI train.

>> No.6869423

>>6869421
Wait a week for them to get bored

>> No.6869424

>>6869418
so anon do you care about what the majority of humanity thinks

>> No.6869427

The constant crabbing made me quit drawing.

>> No.6869428

>>6869421
>newfags anon newfags...
hopefully they do get the /mlp/ treatment cause it's basically what they deserve

>> No.6869429

>>6869424
I do not but that's not the point.
They don't care either but will take tangible actions based on impulses and feelings.

And when you got an unreasonable majority, nothing works.

>> No.6869431

>>6869366
Waiting for the day that 4chan posts get a built in prompt feature

>> No.6869432

>>6867815
no

>> No.6869439
File: 104 KB, 273x127, 1586519434694.png [View same] [iqdb] [saucenao] [google]
6869439

>>6869395
Yeah, no one would like having easy propaganda pamphlet printing machine or infinite population pacification image generator.

>> No.6869440

>>6868704
>>6869391
you can't even call this spam. they're just continuations of the same thread mostly. /g/ has an image limit of 150. it's just that plus a few duplicate threads when multiple anons try to create the next thread.

if anything this just shows how willing people are to try this shit.
AND YET here we are on the art board. and everyone is sticking their heads in the sand and pretending that
>it's not happening
>it'll go away
>i don't want to see it

>> No.6869442

>>6869431
would be a kino april fools joke

>> No.6869443
File: 186 KB, 220x216, american-psycho-patrick-bateman.gif [View same] [iqdb] [saucenao] [google]
6869443

>>6869429
So anon are psychopaths and sociopaths really the betters of most people and are desirable features to find in a person

>> No.6869444
File: 144 KB, 1011x939, prooompt.png [View same] [iqdb] [saucenao] [google]
6869444

>>6869431

>> No.6869446

>>6869440
not knowing board culture/actively image dumping random shit a kin to googling is not spam sir... this has never happened before
>t.newfag

>> No.6869447
File: 1.07 MB, 1400x787, 1588901184944.jpg [View same] [iqdb] [saucenao] [google]
6869447

>>6869440
Honestly, i don't get it
>make thread
>spam images
>talk about nothing
I get it on the porn boards since they're supposed to be a fetish dump, but on other boards? It really escapes my understanding.
What are they expecting people are going to say to them if they makes threads upon threads of senseless image dumping?
>>6869443
Monkey.

>> No.6869449

>>6869444
Terrible greentext

>> No.6869450
File: 204 KB, 1916x945, img.jpg [View same] [iqdb] [saucenao] [google]
6869450

>>6869447
>i don't get it

>> No.6869451

>>6869440
>AND YET here we are on the art board. and everyone is sticking their heads in the sand and pretending that
The first report that comes on this board is "Not poster's own artwork"

>> No.6869452

>>6869444
kek

>> No.6869453
File: 93 KB, 1000x667, photo-1519944518895-f08a12d6dfd5.jpg [View same] [iqdb] [saucenao] [google]
6869453

>>6869444
My god anon people going to art shows just wipe their phones to look at instead of looking at the pieces
>and they paid for admission
>money well spent right there

>> No.6869454
File: 86 KB, 613x640, 1589779446707.jpg [View same] [iqdb] [saucenao] [google]
6869454

>>6869450
I like you.

>> No.6869458
File: 1016 KB, 1064x930, 208.png [View same] [iqdb] [saucenao] [google]
6869458

>>6869444
Will their screen look something like this anon but with "AI" talking points

>> No.6869462

>>6869447
this is legit too much power for some people. they don't understand that not everyone wants to see every single gen they shit out. it's like that in /sdg/ as well.
but still, they might be spamming images, but not threads. the threads are just filling up too fast because of the low image limit.

>>6869421
>now it feels like everyone jumped on the AI train.
almost like your made up concerns won't stop anyone from using AI or stop it from being the future.

>>6869450
this has nothing to do with it.
honestly, they do it for the same reasons why artists want others to see their art or sketches.
yes, they have much less input in what they make, but subjectively the feeling is the same.
i'd know since i've been doing both sides and the FEELING you get from a finished AI piece is not that dissimilar from finishing real artwork.
it really isn't.
even if it's less pronounced, you'll still sit there. keep looking at it and feel a small sense of pride for having made something.

>inb4 retards screeching about how they're not making anything
this is just cope. and plain delusional. get it together please.
yes they should get less credit for it, but they're still making something. if only because the AI is capable of making what they want to make.

>> No.6869464

>>6869462
>/pol/ should be able to promote and radical more people freely and openly

>> No.6869465

>>6869298
That's right. If everyone is an artist no one is. Poo is out of the loo and everyone knows what ai is now, and as a result no one is impressed by proompters except delusional proompters themselves

>> No.6869468

>>6869462
>the FEELING you get from a finished AI piece is not that dissimilar from finishing real artwork.
lol?

>> No.6869469

>>6869462
>if only because the AI is capable of making what they want to make
So it's the AI not you. It could even be programmed to generate all on it's own, you're not needed at all for it.

>> No.6869470

>>6869462
>t.concern fag
>they're not spamming the image is being reached to fast
>they only post a image every half minute
>it's the other threads that are just to slow
>It should be more like /v/

>> No.6869472
File: 20 KB, 500x500, 1612788364699.jpg [View same] [iqdb] [saucenao] [google]
6869472

>>6869462
>this is legit too much power for some people
This is the only smart thing you posted

The rest does come off as bait tho

>> No.6869473
File: 107 KB, 220x124, fly-funny.gif [View same] [iqdb] [saucenao] [google]
6869473

also finally we're at bump limit

>> No.6869478

>>6869462
Seeing ai generated images that you prompted gives the same feeling, assuming it's decent, of looking at someone else's art. Not your own. You would know if you actually drew.

>> No.6869482

>>6868733
>SD is still better because it requires actual effort, Dalle-3 is for promptlet faggots who can't handle the jandal.
Lmao the cope. All your "effort" is now invalidated by the existence of dall-e 3. I called this from the beginning, that ai would be accessible and stupid easy to use for any brandead average Joe or Karen and then SD fags get mad because they can't feel special anymore

>> No.6869486

>>6869482
holy shit I missed that post, the fucking irony

>> No.6869489

>>6869464
no?
but yes, that will be an issue. especially with how gullible zoomers are.

>>6869468
it really isn't. again, a different magnitude, but it's similar.

>>6869469
no, the AI is only the tool that automates almost all of it.
>It could even be programmed to generate all on it's own
and who decides what it generates? even if another AI prompts it, if THAT AI too, is following human instructions, then that too is just a result of human vision.
even right now you can already automate prompts in a way that doesn't make one specific thing, but a range of specific things (just by using wildcards).
but if i was the one who decided on those things, then it's still ultimatively following my input.

>>6869472
everything i say comes off as bait to people here.
because you people legit have no clue about anything. you just want to stick your head into the sand when it comes to AI.

>>6869478
exactly wrong.
it obviously depends on how much time and effort you invested into the final piece. the more you did, the more the feeling is similar. again, it's obviously not comparable to the effort of drawing. it's a very different kind of effort.

>> No.6869491
File: 46 KB, 800x450, 1603320381424.jpg [View same] [iqdb] [saucenao] [google]
6869491

>>6869489
>everything i say comes off as bait to people here.
>because you people legit have no clue about anything. you just want to stick your head into the sand when it comes to AI.
No, it's because you're trying to attribute absolute moods and feelings to others.
The whole
>prompting is the same as creating art
Has been so overused it can only be seen as a bad faith arguments.

Sure, you will have newprompters get a feeling of finally having created something worthwhile and could finally get any attention they so desperately want, but it's more like a kid with a new toy that will get bored and hate the toy once it brings him no more joy.

You forgot your trip tho sir

>> No.6869493

>>6869470
i'm saying that they're not spamming threads. which is what >>6868704 implied.
i just consider spamming threads, or "spamming the board" something different.

>> No.6869494

>>6869401
I think it looks neat

>> No.6869495
File: 115 KB, 350x246, 123456567657.png [View same] [iqdb] [saucenao] [google]
6869495

>>6867790

here's my prediction:

- the current AI will be criminalized

- a new stronger copyright law will be implemented in favor of artist

- ai tech companies is going to continue loose money and go into bankruptcy

- ai tech ceos are either going to run and hide or get thrown in prison.

- after some time people will forget ai even existed and the only people who still use it are bottom of the barrel posers fighting each other for scraps. or pedos making CSAM.

- a new "ethical" ai that is fair to artist will appear and this will be the new medium that will propel artist to a higher level.

we will see work of art and CGI that are out of this world. and this time with the support of artist.

So anon, you better keep practicing and level up.

>> No.6869496
File: 21 KB, 657x527, 6425753573673674.png [View same] [iqdb] [saucenao] [google]
6869496

>>6869420
>if your work is pro

>> No.6869497

>>6869489
>time and effort you invested into the final piece
it's nothing like creating art, it's more like a gambler playing slots and finally getting a payout.

>> No.6869499

>>6869495
This world is too chaotic and humanity is too fucked for this to happen anon.

>> No.6869507

>>6869495
nice fan fic

>> No.6869512

>>6869495
Based fan fic.

>> No.6869517

>>6869366
I am sad. I don't care about fame. But there will be no more internet communies. Please tech bros make us a new Turing test please
>>6869495
>a new stronger copyright law will be implemented in favor of artist
Please no. Copyright only hinders innovation. Most eceleb artists use parasocial relationships or status as a way to earn money anyways
We need a new Turing test to avoid chatbots and a way to verify if irl photos are ai to avoid abuse against irl people.

>> No.6869522
File: 636 KB, 840x1132, 5f6b32f5596416aff68b978a4c0afcc5.jpg [View same] [iqdb] [saucenao] [google]
6869522

Here's MY prediction
>image generator will get regulated
>no more free loli porn for prompters
>prompters gonna complain
>only ones having access to unlimited ai generation models will be companies
>ai tech ceo will sell their companies and codes to the highest bidders and fuck off unless they want to run another start up
>now everyone will larp as an artist
>normies will have access to infinite selfie generators they can use to be content creators on the youtube and generate infinite content
>everyone is a content creator now
>no one creates actual content but just flood the internet with slop
>internet traffic will double and triple
>ISP will double their prices
>Same as server hosts
>most people will get off of the internet due to the constant outright spam of nothing but infinite slop
>even porn addicts will get tired of the constant and never ending stream of content and crave novel things
>industry artists will try to shit out works by incorporating generated images into their work flow
>no one will care because they have other infinite content getting thrown into their faces 24/7
>people will get numb to it
>meanwhile companies don't have to pay people anymore to create content but just generate it
The long term result of automatic content production is diminishing returns.
You will pay more to generate than what you will make in profits.

Copyright laws already work in favor of any creator and don't really need to be changed.
It's the absolute lack of enforcement of such laws that brought us were we are.

>> No.6869527

>>6869491
>No, it's because you're trying to attribute absolute moods and feelings to others.
no, that's literally what you're doing here:
>Sure, you will have newprompters get a feeling of finally having created something worthwhile and could finally get any attention they so desperately want,

meanwhile, what am i doing?
i'm merely telling you what i've experienced. and you think it's bait :)
again, you tards think everything i say is bait.
this is not me trying to attribute anything to anyone, this is just a fact at this point.

>>6869497
not exactly, but yes, it's not the same process obviously. but that too depends on the actual process.
and it's not pure gambling. there are many, many different ways to tie things down.
very simple example: if i use a style lora, am i still gambling on the style?
if i use a pose CNET, am i still gambling on the pose? same with compositions etc etc.
there are obviously many things you can clamp down on. and there is always some gambling, but it's not what you think probably.

>> No.6869530
File: 119 KB, 1013x661, 1692345281051.jpg [View same] [iqdb] [saucenao] [google]
6869530

>>6869517
>I am sad. I don't care about fame. But there will be no more internet communies. Please tech bros make us a new Turing test please
Some bots are already better at solving captchas than humans. I fear that AI systems could kill off anonymous posting really easily.

https://arxiv.org/pdf/2307.12108.pdf

>> No.6869535

>>6869482
since it is so easy to make good AI now, my SD work doesn't feel special. I feel like I've wasted a year of my life learning loras/controlnet/inpaint etc. makes wish I'd been a drawer.

>> No.6869541
File: 22 KB, 259x257, you just know.jpg [View same] [iqdb] [saucenao] [google]
6869541

>>6869527
Anon, as a prompter, you basically have no influence or hands into the creation of image generation.

You are
>input words
>hope for the machine to shit you out any image you might like
That's like throwing sticks on the ground and feeling accomplished that something resembling a smiley face appeared.
Or taking a picture of a cloud that resembles anything and claiming you did the cloud.

If anything, that's peak delusions that only stems from anyone who cannot, for the life of them, get any attention or acknowledgement they are really desperate for, otherwise.
It's like artists copying csr/sakimi and trying to farm numbers.
It's all purely logical and well observed.

Radically speaking; A generated Image CANNOT be the result of a human vision, because the images are being created by already existing data and images and more or less mashed together, and we had this argument countless times already here.
>b-b-b-but the ai learns and diffusses because its in the name
No one cares about the techno gibberish and it's irrelevant anyway and is mostly just an attempt to argue from authority if anything, hence a bad faith argument and bait ;)))

You cannot disagree with this and be correct.

>> No.6869542

>>6869530
I am a CS student, I could make a new Turing test for my Master thesis.
This is what I've been brainstormimg:
Physics. Both me and you experience physics the same
AI used randimastion and chance
Human 1, the one who wants to check the other, could perform a physics experiment. It can be something as simple as them singing a pendulum on webcam.
The other user, either person or a bot, now has to swing a pendulum.
Could a bot use ai video generation to do a flawless pendulum swing?
Could bot use ai plugin to call a physics simulation program to call whenever human 1 asks for a test?
I will ask /g/, but I need more time to formulate the question and idea better

>> No.6869550
File: 139 KB, 768x1152, 0558d2c5f5105f0e052df04703baec01.jpg [View same] [iqdb] [saucenao] [google]
6869550

>>6869542
Anon, you can tell if someone is a chatbot by how they're unable to contextualize information and be coherent with their replies.

Or just use the forbidden words.
Use ebonics or transformative language, innuendos or straight up made up words and allegories any real dirtmonkey would snap up and able to get it if they were real niggas.

You can also filter low iq tards with it.

>> No.6869560
File: 939 KB, 1024x688, 1688588389172680.png [View same] [iqdb] [saucenao] [google]
6869560

>>6869541
that's just not true.
see pic related.
i wanted to get this, i used a specific model because i knew it would get me a certain look, same with the keywords. same with how i configured the two cnets i used. you absolutely can follow a vision to some degree.

you're not always throwing shit at a wall because if you experimented a lot with the model, you KNOW how it will respond to certain things.

>and we had this argument countless times already here.
yes and you're still wrong. in fact i've been trying to learn things more in depth recently in order to explain it better, but it's pretty hard to understand, especially the cross-attention stuff where all this actually matters.

but the AI doesn't learn just images, it learns images tied to words. and when you prompt it, you're using those same words. words that are not tied to any image, but to thousands if not millions of images.
in the end, the word you prompt is simply the models representation of all the training data. not any particular image, or even any particular part of an image, but but a far more wholistic understanding of the word and what it represents.
that, on its own is already plenty of proof that the AI is learning.

>> No.6869569

>>6867790
i'll keep drawing, but won't be posting my work anymore
the whole point of posting was trying to make the best possible finished pieces, but nowadays i could make them much better with a little work using img2img on stable difussion
it would feel dishonest having a self inflicted handicap for no reason. everything i can make now are unfinished studies

>> No.6869573

>>6869560
That's nice but you'll never be an artist

>> No.6869579

>>6869560
did you control any of those brushstrokes? looks like a photofilter

>> No.6869588

>>6869579
controling each stroke would be the equivalent to making manually the whole thing, retardo
you can always manually tweak the shit you like, but the ai already gets you 90%+ there

>> No.6869601
File: 9 KB, 316x413, wow.png [View same] [iqdb] [saucenao] [google]
6869601

>>6869560
>i used a program to do what is has been programmed to do therefore i can claim the output of the program as my own effort
Guess i'm a mathematician then

>> No.6869622

>>6869601
If you can not figure out anything you could use a calculator to figure out without one then you are a brainlet. It would simply take you longer. It won't make you understand things about math you don't already understand. Bad analogy

>> No.6869624

>>6869601
mathematicians don't use calculators?

>> No.6869625

>>6869601
Most mathematicians DO use calculators and computers.

>> No.6869631
File: 6 KB, 306x159, im real math.png [View same] [iqdb] [saucenao] [google]
6869631

>it learns images tied to words
So, basically, it has noise saved and keywords attached to that noise then it reconstructs an image using various tagged noise based on the keywords used giving you an output that tries to match the keywords you used?

How is that any different from photobashing or collage then?
>its proof ai learns
How can it learn if it isn't self aware of it's own mistakes? It just reproduces imag-- excuse me, NOISE, then it diffusses the noise (because it's in the name of course) and make pretty flowery pretty artwork instead of being copy&paste with extra steps.
>>6869622
So, when are we going to see your work?
>>6869624
>>6869625
Pineapples and coconuts.

>> No.6869637

>>6869625
Do you know what advanced levels of the mathematics look like anon.
>or do presume it's solve for (x)

>> No.6869638

>>6869217
ranon sarono?

>> No.6869644

>>6869631
>it has noise saved
buh?

>> No.6869653

>>6869268
thanx. in the lab building a body of work now. will start on the x site under bckomix.

>> No.6869668
File: 1.11 MB, 1180x573, WOW.png [View same] [iqdb] [saucenao] [google]
6869668

>>6869644
bruh
i cant hear you over the sound of my artistic talent

>> No.6869694

>>6867790
I used to shit on AI art here and elsewhere. even got into some really autistic combos
Truth is, I'm too poor to have tried AI before, but now that Bing is free I finally checked it out, and i wasn't prepared for this shit
It's way too late already: humans lost. If you haven't realized it yet, you're in for a rude awakening very soon

>> No.6869698

>>6867819
she can save me

>> No.6869710
File: 403 KB, 532x745, real arts.png [View same] [iqdb] [saucenao] [google]
6869710

>>6869668
sirs i am artiste now

>> No.6869730

>>6868946
Yeah, concept artist is pretty much a dead career at this point

>> No.6869754

>>6869560
If you worked so hard on it, why is the hand all kinds of fucked? Brightest minds who should be solving the unsolved questions of natural sceinces instead devoted months of their time to make AI generate passing hands.

>> No.6869770
File: 528 KB, 470x657, 1694439296147878.webm [View same] [iqdb] [saucenao] [google]
6869770

>>6869631
>So, basically, it has noise saved
no. neither.
no noise is ever saved. this also shows a complete lack of understanding for how this works.
what these models actually do is REMOVE noise. so basically it tries to see the prompt you provided among the noise.
you can almost say it's a "cheeky trick" where it tries to hallucinate things from within the noise, the noise just represents a nigh infinite amount of starting possibilities.
it gradually removes the noise until nothing but an image is left.

-so during training the model tries to recover the exact training images from noise. (as it gets things wrong, it tries to correct itself. this process adjusts the model. this is the learning part)
-but during inference (when it makes images) it tries to recover a image from within the noise, except that image doesn't actually exist and is only represented by your prompt. so the resulting image is just everything it has learned about the tags during training.

>and keywords attached to that noise
i'm not super knowledgeable about this particular part as it's a bit hard to understand for a novice. but as far as i understand, keywords are tied to the weights in the model. basically when it sees the tag cat, the nodes connect in a certain way, and that is the part that is being trained. the node. i.e. the "neurons".
so here again, it's no individual image or even multiples of images being saved. the only thing remaining is a giant network of nodes that behaves a specific way when met with the input of "cat".

>then it reconstructs an image using various tagged noise based on the keywords used giving you an output that tries to match the keywords you used?
>How is that any different from photobashing or collage then?
like i explained above. because it is not using any specific image, but only its own learned representation of the tags.

look it up yourself how these things work if you don't believe me.

>> No.6869775

>>6869601
mathematicians do use computers to crunch numbers.
and everyone else uses them because they're a convenient tool.

>>6869730
somebody will still have to make concept art. the role is still needed.
it's more like everyone can become a concept artist now.

>>6869754
i made it in seconds before going out. just to make a point about how we can make the model follow a vision using my input. this was not about effort.

>> No.6869791

>>6869770
You just sayin’ it rebuilds pieces of stolen images on a framework of a random noise. The original images are encoded in the hidden layers of the NN. It cannot work any other way, it has no understanding of what objects are or that it is projecting a 3D object onto a 2D plane. This is why it usually fails to depict stuff from unusual angles and why "robot dog" looks like outline of a dog paintbucketed with "robot parts" pattern.

Of course, artists are totally delusional if they thing these minor but fundamental flaws wills ave their career or cultural relevance. It is over, 100 times over.

>> No.6869805

>>6869694
Thank you fellow anon! I'll be sure to check out this new image generator by Bing since it's free and can do anything any human artist can and more!

>> No.6869836

>>6869791
It's over for a lot more than just artists. Ai will bring more disillusionment and suffering than any other technological revolution. Sam Altman even said future generations will hate the current ones for creating ai

>> No.6869839

>>6869775
>it's more like everyone can become a concept artist now.
Lol everyone can't become one if no one is hiring for it anymore. This type of cope is hilarious. More like child labor in the 3rd world will be proompting your anime while you lay bricks until you have carpel tunnel

>> No.6869843
File: 780 KB, 484x271, IMG_3128.gif [View same] [iqdb] [saucenao] [google]
6869843

>>6869668
PLAP

>> No.6869846

>>6869694
It is over

>> No.6869847

>>6869668
muh dik, would unironically throw money at your patreon

>> No.6869853

>>6869805
everyone of them does more or less the same
but stable difussion is the only open source alternative, if you're into cp or something

>> No.6869867
File: 314 KB, 407x626, 1664954439292945.webm [View same] [iqdb] [saucenao] [google]
6869867

>>6869791
>The original images are encoded in the hidden layers of the NN
no, the training images are thorougly TAKEN APART inside the neural network.
imagine the higher layers (layers of nodes) grasping the vague shapes, other layers grasping details, and others again taking relationships and other elements from the image. as well as the tags relationship with other tags even.

now consider that this is what happens with one image.
the next image you add, it further READJUSTS the adjustments from the previous image. THE SAME ONES, if it's the same tag.
so at the end, by training on a cat, you might end up with:
>parts that understands the rough shape of a cat, parts that understands the colors that cats can have
>parts that understand how cats can look in different settings, which are tied to other tags. (like daylight, night, snow, etc)
>parts that understand that a cat has two protrusions called ears. and many, many other details of a cat. or even that it has fur, just like dogs have
and much more.

think about it. every single thing i listed is being trained on the word cat, but the neural net is large enough to capture all of these features separately.
not only that, when training, even only on this single word, cat, all of these features are adjusted simultaneously.
so when you pull the word "cat" you pull all of this "knowledge" with it.
as the model trains on more cats, what model gets out of training is not a snapshot of a single cat. but something more thorough. a representation of what a cat is and can be.
there is a reason that the outputs of AI is just "a cat" and not any specific cat from the training data.

>..rebuilds pieces of stolen images on a framework of a random noise.
there is more data in the image than only its pixels. if it could only work with the pixels (like in photobashing)
it would not be able to do any of this. it works because it can work with the actual relationships between pixels and analyze the image in a deeper way.

cont.

>> No.6869871

>>6869775
If everyone can do it they'll pay less for it and there'll be less demand dumbass.

>> No.6869879
File: 71 KB, 313x286, 1684865576403615.jpg [View same] [iqdb] [saucenao] [google]
6869879

I think now is good time to get acclimated to the idea that AI images aren't going away and will be omnipresent in the internet (and soon maybe even in real life).

>> No.6869885

>>6869522
Doesn't matter if it gets regulated if people can just use it locally. Everything else you seems correct but people assume the good end will some happen.

>> No.6869903
File: 924 KB, 814x610, 1676894114849542.png [View same] [iqdb] [saucenao] [google]
6869903

>>6869791
>It cannot work any other way,
why not? i explained how it works.
in fact you should understand that it cannot possibly be photobashing. just based on how it generates images alone. even without looking at the training.
because it starts with noise and is making up shit the entire time all the way until the finish line.

even just the fact that it starts off blurry and slowly converges on the final image already disqualifies it from being photobashing.


>it has no understanding of what objects are or that it is projecting a 3D object onto a 2D plane. This is why it usually fails to depict stuff from unusual angles and why "robot dog" looks like outline of a dog paintbucketed with "robot parts" pattern.
it fails at things because they're hard. that is not proof for any of your statements. understanding doesn't mean perfect wisdom.
i'll again use the reflection example: pic related.
this time i'm using an actual mirror instead of just water reflections like >>6869867
here you can clearly see what the model grasps and what it doesn't grasp.
>it can grasp that a mirrors contents is related to what is in front of the mirror.
>but think about what it would need in order to do this perfectly. it would need a perfect sense for spatial positions as well as...well, everything else basically. because you'd even need anatomy to know how the exact same pose works from a different angle.
>and you can see the model trying to do that anyway, but failing.
but even here, with more training examples, this is something that can eventually be grasped more fully. because that's just the nature of machine learning. it can extract features and relationships from its training data. it "learns".

>> No.6869908

>>6869839
>>6869871
that's different from the proffession going away. and realistically you'll still want artists to do the job. the only difference now is that you'll have to compete with other prompters. and realistically speaking, those will still be artists, because they wouldn't be pursuing a position in this field otherwise.
maybe they will be lower skilled by todays standards, or maybe not.

>> No.6869921
File: 257 KB, 947x2048, FbtXDYIUIAAcykO.jpg [View same] [iqdb] [saucenao] [google]
6869921

>>6869867
Whatever tangent you are going down is irrelevant to the greater fact that it could not do what it does without the model being fed image data from the works of artists who never provided consent.
>inb4 the "artists use references too" strawman
humans are not and never will be algorithms running on computer hardware, it's a false equivalency

>> No.6869939

>>6869921
it needs data to understand anything. that is a given.
what you don't understand is that it is not a program that does anything at all without its training data. it only understand anything at all due to its training data.
it only can understand "impressionism" by seeing impressionist paintings. that's just how things work. if you're asking it to come up with impressionism on its own, then that would be a FAR, FAR more capable ai with far more creative capabilities. one that would truly be beyond humans (because it can come up with impressionism on its own, while it takes humanity generations and thousands of years to do the same).
SD would be a complete joke compared to that kind of AI.

i would not be able to draw any anime/manga if i never saw any anime/manga. neither would you. that is the fact you're missing here.
the AI's training stage is the only stage where it learns anything at all, where it's neuralnet is changed.
but for us humans, we can learn something by seeing it, anytime we want.

>inb4 the "artists use references too" strawman
it's not a strawman. even now, you people still don't understand. it's just that the point you're holding against AI is stupid and unreasonable. you're asking it to learn without seeing. THAT is the analogy here.

cont.

>> No.6869961

>>6869939
>i would not be able to draw any anime/manga if i never saw any anime/manga. neither would you. that is the fact you're missing here.
completely and utterly wrong
a skilled artist could create something reasonably accurate from description alone using foundational principles

>> No.6869962
File: 862 KB, 890x741, 1689061421255495.png [View same] [iqdb] [saucenao] [google]
6869962

>>6869921
as for the bloodborne example, it's just overtraining.
the AI doesn't know what a "bloodborne" is. if it sees countless images with the exact same pose from different promo materials and covers, then it will eventually tie that to the tag more than it should. it will tie the entire composition to the tag of bloodborne. especially when there is not much other variety speaking against that.
this is something that can happen, it is not supposed to happen and they actively try to avoid these cases by removing duplicates and other techniques.
because despite what you think, nobody actually wants this to happen.

and if you were honest you'd also admit that this is by and large not what the AI does. it only happens with these overtrained examples.
(in fact this even happens with lora. too much training or bad configurations and the AI will literally just give you back your training images without much versatility.)

>> No.6869976
File: 89 KB, 1024x1024, photo of a top of seaguls head.jpg [View same] [iqdb] [saucenao] [google]
6869976

>>6869867
>no, the training images are thorougly TAKEN APART inside the neural network.

What is the difference? It has been demonstrated you can recover the original images, it is even trivial if you use small training dataset. Sure it is not photobashing nobody argued that but still a plagiarism machine it is.

>>6869903
>it fails at things because they're hard.
No they are easy. Doodle a top of seagul’s head. Even a little child can do it. The AI cannot because it has no image of top of seagul’s head in the training dataset.

>> No.6869990

>>6869961
i completely and utterly dissagree.
especially with anime and manga, where a lot of the decisions are PURELY aesthetic, stylistic choices.
these things do not come from skill first. they initially just come from experimentation. and then, they eventually mature into something more elaborate.
there is a reason 60s manga doesn't look like how it does today.

tezuka created the things he did because of disney influence, among others
if tezuka was born today, his style would be completely different.
and even disney and cartooning in general comes from the idea of caricature, to abstract features.

a skilled artist that has never seen what we have today (not that this is possible in the age of internet) would never reach this kind of style on his own.
because these styles are traced down a long line of developments, decades in the making, of artists influencing other artists.
you can clearly even see some of the subgenres affecting the styles, like shoujo in particular. although these days everything is more well mixed.

>from description alone
from description, it is possible. but that is in fact additional data. just in the form of guidance.
and even then it wouldn't be possible in one try. it would take many many tries and you handholding the artist to get there.

all of this just underlines my point: you need to see it in order to learn and be influenced from it. and we ARE influenced by it. none of these are purely our own makings. style especially, is a collective effort that builds upon previous generations.

>> No.6869994

>>6869908
The program does the heavy lifting and now with less word play required means it'll most definitely be low skilled work buy the time it becomes the norm. before you'd get paid dirt but atleast it wasn't a job anyone could just get but soon anyone will be able to qualified to ge a image generation job. It'll be a job that's hard get but has little pay because the effort required is so low.

>> No.6869999

>>6869990
>le human cannot create unless they learned to do it strawman
>people cannot create the same thing purely on their own
>influence = copying
man, and he insists on not getting called his posts as bait when all he does is repeat the same ignorant arguments over and over again

don't (you) me btw, i ain't arguing you anymore

>> No.6870026
File: 305 KB, 512x512, 1685685395667003.png [View same] [iqdb] [saucenao] [google]
6870026

>>6869976
by top you mean from above?
photo models can do that. art models are probably worse since they will have less relevant training on this kind of stuff.

>it is even trivial if you use small training dataset.
of course. because then the model barely knows anything else.
and it would be a literal copy machine if it was only trained on one image. but so?
but it's not. very far from it.

>It has been demonstrated you can recover the original images
yeah? at what rate, under what circumstances? with what kind of images?
you know that your claim has a ginormous asterisk next to it. you just choose to ignore it.

>Even a little child can do it.
children can't draw shit.

>The AI cannot because it has no image of top of seagul’s head in the training dataset.
it would have some, because it probably has some seagulls in it, i can probably get them out reliably with cnet too.
but that's beside the point. if it hasn't trained on it, it can't make it, because it lacks an internal representation of that object/idea. yes.
and?
can you draw something you've never seen? can anyone?
we literally study anatomy from all angles to be able to draw it reliably.

>>6869999
>influence = copying
you're saying it's copying, not me. i indeed to call it influence and learning.
you're unable to understand what i'm saying without putting your own beliefs on my arguments. this is just another proof of it.

>le human cannot create unless they learned to do it strawman
yes and do you have anything to say to that?
how is it a strawman exactly? what am i strawmanning here?

>people cannot create the same thing purely on their own
you're saying you can? you think if you never saw anime in your life, you'd be able to draw it?
without master painters influencing you, you'd be able to paint on their level?

getting pretty tired of getting called ignorant by literal retards who don't understand anything about how art actually works in our minds.

>> No.6870044

>>6870026
>you think if you never saw anime in your life, you'd be able to draw it?
yes

>> No.6870046

>>6869261
basically all of /ic/ was still in denial about AI being able to threaten their livelihood back then. composition was shit, there were clear stylistic limitations, artefacts were much more prominent, there was a lot to cherrypick and cling on to as alleged 'permanent limitations' of the technology.

>> No.6870054
File: 3.31 MB, 1280x1010, The_Rhinoceros_(NGA_1964.8.697)_enhanced.png [View same] [iqdb] [saucenao] [google]
6870054

>>6870026
>can you draw something you've never seen? can anyone?

Yes.

>> No.6870063

>>6869770
I agree with your post except for the parts where you use words like 'learned' and 'representations'.
a NN is the approximation of the relations between variables that get imprinted into it through a ridiculously simple, mechanical iterative process. a human being "learns" something about a subject and forms a "representation" of it, which is to say that we put ourselves into relation to our knowledge and are able to reflect on it, form new hypotheses and build on those. What NNs do is more like evolving a structure that matches a datasets structure and is therefore able to replicate its properties, blindly, one step after another, without foresight or capacity for moving beyond the confines of its training data.

>> No.6870068
File: 9 KB, 300x168, 1695720488554259.jpg [View same] [iqdb] [saucenao] [google]
6870068

>>6870044
then you're just clueless.
you do no realize that what you have is due to what you have seen.
the way styles developed over time is literal proof of that.

>>6870054
>Dürer's woodcut print of a rhinoceros is as iconic as it is inaccurate. In this article we explore the legacy of this artwork and how it shaped public perception for more than 200 years after its creation. In 1515, German artist Albrecht Dürer created the woodcut without ever actually seeing a rhino for himself with his own eyes.
>Instead he used a description from a letter and a sketch created by an unknown artist who had seen the rhino.
>and a sketch created by an unknown artist who had seen the rhino.

but i love how you think he could have gotten the proportions and head shape exactly right just from descriptions. as well as the folds of the skin.
you retards really do believe that "human creativity" is a magical thing that can pull things out of the ether, huh?

IF ANY OF YOU RETARDS COULD ACTUALLY DRAW or understood drawing, you'd know that you cannot even properly draw anything you didn't study. at least not in a competent way. to think that you can draw things you havent seen, or draw/build on styles that you've never seen is utterly laughable.

>> No.6870078

>>6870063
which is also why humans are different, by the way. humans clawed their way upward from smearing paint onto cave walls all the way to a technical understanding of perspective and lighting, invented and innovated on the concept of composition and stylistic representation for millennia and even today the greatest artists move beyond what has been done before and into new horizons. meanwhile if you train AI on other AI data it just degenerates back into incoherent noise because all AI does is imitate. It's bound to its respective past, while humans have a relationship with the future, and that's also what makes art more than just a craft that you can master in a technical sense and why it's so meaningful to people.

>> No.6870088

>>6870078
and in this sense >>6870054 is correct, actually. at some point, no such thing as anime existed, then we slowly brought it into existence. that is fundamentally different from how NNs function - let them run on their own as long as you like, and they're not going to develop a rich flourishing culture that creates new styles and ornamentations. it only works because it has something to imitate - without humans, ""AI"" is nothing. Without AI, humans are still humans.

>> No.6870097

>>6870068
Where do dragons come from? Nobody ever saw a dragon for obvious reasons, yet somebody had to draw the first one, so the rest of humanity could copy it. Was it planted in his mind by God?

>you'd know that you cannot even properly draw anything you didn't study. at least not in a competent way.

Nobody demands overly accurate depiction in these situations. This is of course a problem when trying to depict a real thing, but (sorry to disappoint you) dragons aren’t real, so who cares they aren’t accurate to real life.

>> No.6870104
File: 84 KB, 768x960, dominoscheeseburgerpizza.jpg [View same] [iqdb] [saucenao] [google]
6870104

>>6869560
>see pic related.
>i wanted to get this, i used a specific website because i knew it would get me a certain flavor

>> No.6870111
File: 52 KB, 680x681, 1692786281439601.jpg [View same] [iqdb] [saucenao] [google]
6870111

>>6870097
No, bro, you don't get it
art has to be an exact replica of life or else it's shit
you know, why look at the real thing if you can generate an exact image to then be wowed and amazed that an image can be like a real photograph?

>dragons are not real
What a laugh, lowly peasant.
Thy jealousy is evident.

>> No.6870119

>>6870063
i'm somewhat aware that my language likes to anthropomorphize AI and i've been trying to cut down on that to avoid confusion. but is "representation" wrong at all? it might not be the human level of it, but it is still a representation, no? i don't think having part of ourself in that representation is a necessity to call it such.

>imprinted into it through a ridiculously simple, mechanical iterative process.
the step itself might be simple, but it's adjusting all the weights inside a ginormous network. also i'm not sure if the process of adjusting towards the loss function is actually that simple, not that i'm able to fully understand it fully so far (something something gradient descent)..

desu i have a lot of trouble understanding exactly what happens in the cross attention layers and what they do in detail.

>What NNs do is more like evolving a structure that matches a datasets structure and is therefore able to replicate its properties, blindly, one step after another, without foresight or capacity for moving beyond the confines of its training data.
that's fair enough i'd say.

despite what i'm always saying, i don't think that SD learns the exact same as a human, or even has fully human capabilities. they are merely analogies and i'm aware of where they end.
but i'm also curious to what you would call the training process, if not "learning".
my main point when bringing this up is actually only to highlight how different it is from stealing or copying, which is what antis always like to say.

>> No.6870122

>>6870119
>but i'm also curious to what you would call the training process, if not "learning".
saving data

>> No.6870123

>>6870088
but that's because it has no feedback loop. of any kind.
so it would never develop in any direction regardless. but if you only gave it photos and trained it on its own images in a loop, i wonder if it could arrive at stick figures, lineart and more eventually. what if we added some simple drawings at the start?
according to you, probably not?
i'm not sure on the other hand. certainly not immediately, but modern anime also wasn't developed in one generation. that's kind of my point.

besides, there are some things the AI would NEVER get on its own. the texture of painting, of brushstrokes for example. a lot of art depends on the physical mediums that you make them in. and AI would never develop the same, assuming it can develop at all in that hypothetical.

>>6870097
they are just lizards with wings anon.
and mickey mouse is literally just three circles. and etc etc.

>>6870104
that argument would work even if i used watercolor because i know it would get me a specific look.
or even if i used "green", because i know it would make things... green.

>> No.6870124

>>6870122
what data?

>> No.6870125

>>6870124
the images you feed to it

>> No.6870126

>>6870125
..it saves the images you feed it?
what data?

>> No.6870131

>>6870126
why are you playing dumb?

>> No.6870140

>>6870123
>and mickey mouse is literally just three circles.

Speaking of circles, has anybody ever saw circle? Circle is infinitely thin photons of visible wavelenght pass right through it. In fact all particles do and as such circles if they were real would never interact with anything making them invisible. Then how come we have very accurate idea of what circles are and we can depict them in our art?

>> No.6870142

>>6870122
The people who actually work with NNTs seem to use the word training quite commonly.

>> No.6870146

>>6870131
are you? you're not that anon, are you?
i'm asking you what does it "save"?
you don't actually know, do you? why are you butting into the reply chain? i thought you were the other anon damn it.

>>6870140
no, a circle is an abstrract 2D shape.
you nincompoop.

>> No.6870177

>>6870146
so how does a computer program replicate images if it doesn't save data of those images?

You can't actually answer this question without going into strawmen, can you?

>> No.6870214

>>6869122

At this point, porn or other non-pc topics. It seems like really good ai is in the realm of companies who have large datasets and have the hardware to run it. However, they have to be politically correct for their own reputation.

>> No.6870237

>>6869908
>that's different from the proffession going away.
It's not different unless you are a child laborer in the 3rd world. Who knows much longer the industry will even last with all this. Rather than selling games you'll be paying for your ai game maker subscriptions and paying another fee to the same company to publish them (somewhat like what Unity is trying to pull). Maybe you'll get two downloads from your friends if you're lucky
>and realistically you'll still want artists to do the job. the only difference now is that you'll have to compete with other prompters.
Laughably delusional
>and realistically speaking, those will still be artists, because they wouldn't be pursuing a position in this field otherwise.
Lmao. The world is full of lazy retards who are obsessed with content, not art, and think they can use ai to make an easy living by sitting back having a computer do all their work. All the while fishing for compliments on how "talented" they are. Just about everyone and their mom is going to want a piece of this imaginary pie

>> No.6870262

>>6870119
>but i'm also curious to what you would call the training process, if not "learning".
>my main point when bringing this up is actually only to highlight how different it is from stealing or copying, which is what antis always like to say.
I would call it a compression algorithm. What you get as a result is 'information that describes the training dataset in such a way that it is able to predict/imitate it with accuracy while not being overfitted to said dataset (this second part only happens because we define overfitting as bad and set a cut-off point, the process "wants" to just rote memorize the dataset if it can)'. This is obviously 'stealing' properties of the dataset. If the AI can imitate a style I have invented - yes, there are tons of styles out there that only one artist uses -, the reason it is able to do so is that it has been trained on my work. I did the experimentation and figured out how to create a certain aesthetic effect, the AI takes the results of that process and imitates them. It is a result that would be impossible without me.

>> No.6870274

>>6870123
>so it would never develop in any direction regardless. but if you only gave it photos and trained it on its own images in a loop, i wonder if it could arrive at stick figures, lineart and more eventually. what if we added some simple drawings at the start?
>according to you, probably not?
not just according to me, there's been some mathematical proof that degradation MUST be the result when you train AI on AI-generated data with SDG, though I can't find it, and it should also be intuitive to you if you understand how next token prediction works. If you train it on itself it ends up predicting its own prediction errors in a loop and getting rewarded for it and there's no way to get around that because 'does it predict the dataset or not' is the only basis you have for training it. Reinforcement Learning is different, but it's also a completely different paradigm with other limitations.

>> No.6870286
File: 2.92 MB, 512x512, 1691199168588584.gif [View same] [iqdb] [saucenao] [google]
6870286

>>6870177
i explained it already. countless, countless times.

it saves features and relationships of the images across the entire network. and those are tied to the tags we use.
i repeat: no images are tied to the tags, only the broken down elements are still there in the form of weights across the neural net.
and ON TOP OF THAT, those weights are further adjusted by other images. especially when using the same tags.
so how do you think the original image is still there? it simply isn't

and NO i don't mean
>this neuron is responsible for the cat's eye so it just steals one eye from the training image that is somehow still in the NN (it isn't)
instead it's
>a section/collection of neurons are doing the work TOGETHER to form a cat's eye, one that fits with the rest of the image, which is also created by collections of neurons, i.e. nodes.
there is a process to all this, and it has nothing to do with recalling pixels and stiching them together.
(that wouldn't even work because if you can paint you'd know that colors and values are conditional.)

instead what is happening is that an image is gradually shaped. you can even see this in action with every gif and webm i posted.

>> No.6870308

>>6870286
>no it doesn't save data
>it just takes that data, breaks it up, saves the broken up data and then remixes it when it needs the data
So its just automatic photoshop with extra steps so it can't technically get called automatic photoshop and be accused?

Just like a human artist would do, right?

>> No.6870309

>>6870286
If I save a .jpeg of that cat image, did I save the image? It's not actually the same image it's a highly compressed instruction for creating something that looks very much like the original image while needing far less space.
Of course this isn't quite the same as training a NN, since there you're compressing a huge quantity of images into a single file which is then able to replicate a likeness of the 'space of all images it has been trained on', but I'd still like to see you argue why huge corporations should be allowed to use things you own to create something that can mimic properties of those things you own, allowing them to automate the work that you did up to that point, using that very work as a basis, giving you no compensation for this whatsoever and reaping the benefits while you have to get a job doing something else.

>> No.6870316

>>6870286
>i explained it already. countless, countless times.
maybe you don't understand it well enough to be able to explain it?

>> No.6870326

>>6870309
the shill is just gonna respond with the vacuous "artists will use it as a tool, pls adapt" (non) argument.

>> No.6870327

it just parrots words but it doesn't converse

>> No.6870329

>>6870326
possibly. I think he might just have some genuine misconceptions about how AI works if he thinks you can train it on its own outputs and have some kind of recursive self-improvement happen. If I thought that that was how it operated I'd also say that it's "creating" stuff.

>> No.6870335

>>6870326
the shill should put his fucking trip back on so he can get filtered like the faggot he is

>> No.6870339

>>6870262
>I would call it a compression algorithm
how is that accurate if the what it gets out of it is completely different from the training data? how is this compression when almost all of the actual image information is in fact lost? (again, the image cannot be recreated in almost all cases)
you seem to understand SD, so will you actually tell me that what it has trained on and what it creates are actually the same thing?
because i would argue that its output is a recreation from scratch, using what it has learned from the training data.

>This is obviously 'stealing' properties of the dataset
i do agree that AI is capable of 'stealing' in this sense. but similar to humans, i think it should only be considered such and considered plagiarism and stealing if it actually happens. meaning when i actually see the similarity.

>the reason it is able to do so is that it has been trained on my work. I did the experimentation and figured out how to create a certain aesthetic effect, the AI takes the results of that process and imitates them. It is a result that would be impossible without me.
but don't you see that this entire argument can be used WORD FOR WORD against other artists influenced by your style? that the people that inspired you can use against you yourself?

> but I'd still like to see you argue why huge corporations should be allowed to use things you own to create something that can mimic properties of those things you own, allowing them to automate the work that you did up to that point, using that very work as a basis, giving you no compensation for this whatsoever and reaping the benefits while you have to get a job doing something else.
because you don't own these things that the AI trains on. none of it. only your style as a whole, but not the parts of it.
and in a similar vein, only blatant copying and actual plagiarism should be considered stealing.
otherwise i think this is just rampant hypocrisy.

>> No.6870340
File: 1.80 MB, 1024x1024, 1668827203139129.png [View same] [iqdb] [saucenao] [google]
6870340

>>6870309
it is still the same image. regardless of what file or compression you use, even if the underlying data differs.
this is NOT the case for latent diffusion models.
see pic related, see how all of these images look similar, but none of them are the same. they merely have the same style.

>>6870274
i can definitely see that happening since errors will pile up eventually. catastrophically so.
so i will probably concede that AI doesn't innovate beyond its boundaries.
but realistically, it still can innovate within its boundaries. and it does do that (by combining and mixing what it has, including styles).
and that also just means that original human input is still worth a ton, especially if you can give it your art as input.

>> No.6870348
File: 231 KB, 1024x1024, OIG.jpg [View same] [iqdb] [saucenao] [google]
6870348

>>6868891
I dont post internal production stuff to 4chan retard.
>>6868873
>you wont be needed
I do this because there is a need. If you cant understand that need I feel sad for you.
>>6868224
I hope you get there.
>>6868222
Yeah I wanted to see how the board was holding up and shitpost about my surface 7 again.

>> No.6870351 [DELETED] 

actually if you just go over
>>6869990 you can maybe gleam my stance on how art styles come into being in general, and why i hold the positions i do.
i do not think that these things SHOULD be owned by anyone.
they are like single notes in a song. only a part of the whole. the more elaborate the sequence, the more a individual can lay claim to it. but just the concept and ideas on their own? nobody owns those. you cannot own a composition, a color scheme, an eye shape, and many, many other things.

>> No.6870353
File: 2.02 MB, 1024x1024, 1679746210740756.png [View same] [iqdb] [saucenao] [google]
6870353

>>6870262
actually if you just go over >>6869990 you can maybe gleam my stance on how art styles come into being in general, and why i hold the positions i do.
i do not think that these things SHOULD be owned by anyone.
they are like single notes in a song. only a part of the whole. the more elaborate the sequence, the more a individual can lay claim to it. but just the concept and ideas on their own? nobody owns those. you cannot own a composition, a color scheme, an eye shape, and many, many other things.

>> No.6870361

>ai shill is still going after multiple threads
>still posts the same copypasted messages
man no wonder he loves ai
he can shitpost 24/7 while having the ai do the art for him

>> No.6870363

>>6870329
>I think he might just have some genuine misconceptions about how AI works if he thinks you can train it on its own outputs and have some kind of recursive self-improvement happen.
...no?
i started the post with saying "but that's because it has no feedback loop. of any kind." that other thing was a hypothetical.
i suppose training is a feedback loop.
but what i meant there was that stable diffusion only trains and cannot improve on its own art without training on additional data, unless it uses its own art for training or something.
which as you said, has some issues.

>> No.6870367
File: 572 KB, 1087x884, desperately.png [View same] [iqdb] [saucenao] [google]
6870367

>>6869293
it's entirely a tool, the issue is most of the people who would benefit most from ai don't know how to use it as a tool because they have zero animation experiance. Honestly just like the people who are using the AI to animate. My kingdom (0$ poorfag) for an artist that knows basic animation workflow and wants my wage cage cash for an ai animation project. The only people who regularly reply to me are russian teenage girls.

>> No.6870369

>>6870361
do you ever notice that you're not contributing anything to the discussion? it's almost like you're heckling from the sidelines like some loser who is afraid to step into the ring.

can you please shut the fuck up if you can't follow the conversation? and you've proven time and time again that you can't say anything worth a damn.

>> No.6870372

>oh no someone is posting in a bait thread well over bump limit
>implying he's having a conversation
Aren't you just posting the same thing word for word you already posting quite a substantial amount of times already?

put the trip back on
>do you even notice you're not contributing
you don't even understand of your precious image bot functions

lmao
lol even

>> No.6870382
File: 156 KB, 519x600, Felis_silvestris_silvestris_small_gradual_decrease_of_quality.png [View same] [iqdb] [saucenao] [google]
6870382

>>6870339
All compression algorithms get out something that's completely different from the training data at the data level - it's only at the level that we care about that we make sure it retains the structure we're trying to preserve. All a NN does in its training process is try to match the training data. The result, as you might guess, is something that matches the training data, and the only question is 'at which level'. As you've already acknowledged, overfitting is a thing - if an image is in the dataset more than a few times, the NN will learn how to replicate that image specifically. What you don't seem to see is that this isn't actually any different at the 'technical' level from the kind of fitting we do want - the training process works in the exact same way. The difference is that when you have 100000 different images, the thing that's being compressed isn't any one of those images individually, but all of them together. what it gets out isn't something 'completely different' from the training data. You can do interesting things with the resulting NN, like tell it to generate a catgirl in the style of Rembrandt and have it hallucinate something that matches those datapoints, but that's all it is - combining datapoints within the model.
This is not the case with other artists, whose work is never just an average of other peoples work at any point on their journey, and who are figuring out their own process to produce the results *they* want to produce. Like, if you don't understand that this is a fundamental difference I don't know what to say.

>you don't own these things that the AI trains on. none of it. only your style as a whole, but not the parts of it.
The way I see it I absolutely own all of it - every single decision I made on my artistic journey, every experiment and technical exercise is part of the process that culminated in the way I draw now. That is my 'intellectual work' as an artist, not any one illustration.

>> No.6870393

>>6870353
>i do not think that these things SHOULD be owned by anyone.
Let me try and give you my argument for why it should be owned
1) neural nets can only mimic existing artworks and their properties
2) existing artworks and their properties are the result of artistic work
3) artistic work takes time, resources, training and dedication, dedication, non of which people will be able to put into it if they can't pursue art fully with hopes that if they get good enough their work will be worth something economically
4) if, the moment an artist reaches the level where their art becomes interesting and appealing, this appeal can be reproduced by a neural network, and this is legal and socially accepted, anyone with access to the replicator can extract the economic value of the work the artist has produced, making their own economic value negligable
5) therefore, public ownership over the totality of artistic achievements up to the present by "everyone" deprives the people actually doing the innovating of the means of continuing to innovate.

You can come try again with "everyone should be able to create art freely :)" when we're living in some kind of post-scarcity utopia. For the moment, we're living in a reality where peoples livelihood depends on their having ownership over the things that give them economic value, and the companies that are working on using AI to automate anything they can think of are doing exactly what I'm describing: extracting peoples skills into neural nets without their consent so they can automate their labor and make more money themselves.

>> No.6870394
File: 2.68 MB, 3684x2876, 1621910003656.jpg [View same] [iqdb] [saucenao] [google]
6870394

>>6870393
kek, I'm clearly starting to type like a retard so I'll end it here for the night.

>> No.6870399
File: 466 KB, 1081x487, fu.png [View same] [iqdb] [saucenao] [google]
6870399

>>6870393
btw anon here about FU which ai clearly breaks
pun intended

>> No.6870400

>>6870393
>dumb jew seething that the goy wont kneel to his shitty argument to be able to copystrike anyone who draws in the style of the IP he paid for using money laundering
Literally everything you just said can be easily applied to human actors.
Go fuck yourself
>>6870394
and fuck your jewhu

>> No.6870402

>>6870399
>teaching
>research
>criticism

>> No.6870403
File: 69 KB, 740x448, 1662388677566.png [View same] [iqdb] [saucenao] [google]
6870403

>>6870400
I accept your concession

>> No.6870404
File: 144 KB, 800x797, marishit.jpg [View same] [iqdb] [saucenao] [google]
6870404

>>6870403
>w-well yeah obviously the next step is to make it illegal for anyone to draw even vaguely the same as the people I have wage slaving for me, so what?
I accept your concession

>> No.6870412

>>6870404
>n-no, you don't get it, you can't own your work because *checks notes* h-hollywood,... and disney! yeah!

>> No.6870414

>>6870382
>..every experiment and technical exercise is part of the process that culminated in the way I draw now. That is my 'intellectual work' as an artist, not any one illustration.
yes. but that is not what ownership is. ownership means it is yours EXCLUSIVELY. and that is the issue here. so i simply disagree.
the way i see it, a lot of the decisions i am making are already building on top of other people's work and decisions. and in turn other people may be influenced by my work and make similar decisions for their own work.
i don't own any of this, at least not until a certain point where i can claim an overarching whole as mine.

>>6870393
>therefore, public ownership over the totality of artistic achievements up to the present by "everyone" deprives the people actually doing the innovating of the means of continuing to innovate.
that is actually a fair point. and one major concern i have over AI as well.
but here you're actually just arguing for the inherent protection of artists.
i also have no solution for this, i basically think we'll have to adapt, one way or another.

i think your proposed solution is only a stalemate. it will not make things any better. you're just artificially propping up artists. how long will you do that for? decades centuries? because AI won't go away. there is no point to doing this. that's my reasoning here. it feels like you're forcing something that shouldn't be forced and halting technological progress in return.

it is also important to note that this is your CORE ARGUMENT for why these things should be owned: because artists should be protected i.e. because otherwise "artists are doomed" (which i don't really agree with btw)
i don't think this argument is going to fly in the face of the market. your argument needs to be about why you do in fact own these things, not because "AI will eat our lunch otherwise."

i really need to catch some sleep too so i think i'll try to reply to the technical side of your post tomorrow.

>> No.6870420

>>6870412
you do own your work retard. that doesn't give you the right to sue anyone who produces similar products.

>> No.6870421

>>6870286
lmao just stop. Niggas are being obtuse on purpose because they're either in denial or just coping, that's it. It's pointless to try to explain them how this shit works because at the end of the day you're not changing their minds much like how you wouldn't change the mind of /pol/tard flat Earther.

>> No.6870426
File: 211 KB, 1200x1200, FrgBlGqaMAAgC1G.jpg [View same] [iqdb] [saucenao] [google]
6870426

>>6870414
>i think your proposed solution is only a stalemate. it will not make things any better. you're just artificially propping up artists.
no hes not. The end goal of this legislation is to make it a legal hassle for any artist to even dare draw something similar to a lucrative established IP. The AI seethe is just a cudgel and wedge to achieve that and nothing more. If you think talentless hacks who buy art with their money laundering, pinkerton squads, and massive media conglomerates wont "own" every style in a heartbeat you've got your head in the sand while you pound it.

People like that don't care about AI and what it does to artists, they want to invent, or maybe even own, a lucrative style, and then sell it for millions to a mega corporation. It's spiritually no different then selling a niche genre of music to UMG. These companies and their stock holders are terrified that soon they wont be needed anymore. It's ironic that that faggot is posting touhous considering the entire point of touhou.

>> No.6870428

>>6870426
you don't seem to understand the difference between
>you can't train models on peoples work against their will
and
>we own this style now, pay us money to draw it >:)

>> No.6870429

>>6870414
>>6870393
actually, another small thing:
personally, the fact that AI is able to do what it can do, and i mean the CAPABILITY of it alone, is already enough to tell me that art has been devalued.
so any protective actions defending against that are not going to last.
what i'm trying to see is how the world looks like after the paradigm shift instead. and using AI as an artist is clearly going to be a big factor of it.
that just seems more sensible to me.

(also i'll use my trip itt from now on just in case.)

>> No.6870430

>>6870428
>what is legal brickwork
How many jabs did you get retard?

>> No.6870472
File: 749 KB, 785x518, 1654656546546.png [View same] [iqdb] [saucenao] [google]
6870472

>NOOOOO YOU CANT PROTECT THE FRUIT OF YOU LABORS YOU HAVE TO LET MEEEE HAVE IT!!! FOR FREEEE!!
The AIshill reveals his final form; a filthy communist

>> No.6870524

>>6868170
Anyone save the pajeet image on the right?

>> No.6870586

Why is this thread still up? Why are AIfags so insistent on this demoralization campaign?

>> No.6870587
File: 795 KB, 1000x1000, 1560793920433.jpg [View same] [iqdb] [saucenao] [google]
6870587

it's over, artbros
this time for real

>> No.6870864

>>6870586
First the pajeets demoralize all the real artists, then they start putting their signatures on AI generated images to get in on the action (because they have no sense of self awareness or shame) or try to make money with AI SAAS sites. Then the gap between normies/tech illiterates closes and the pajeets go to shit up and destroy the next thing. They're a plague of locusts.