|
|
If this email looks jacked up in your email window,
click on this link to see it on the website
https://arseniclullabies.com/nl20241230chuckdxn.html
I will be on the the radio on the Pint O' Comics show
Sunday Jan 5 6:30pm EST
listen live online at
wesufm.org
THIS WEEKS BLOG
Chuck Dixion's betrayal of every illustrator who helped make
his career...and of himself
Or
Yeash, that was a...questionable decision
I'm as capable of doing inadvertently self
destructive things as the next guy. So, when
I see someone else about to do something
stupid, I'm fascinated and mortified. Like
watching a horror movie and yelling at a
character to not go into the basement. It's
the wondering "why" that always
gets at me. Like...you could have just not
done that.
In this case I'm
seeing, Chuck Dixion, long time writer of
comic books going all the way back to Marvel
and DC in the 90s, who is giving the middle
finger to every illustrator past and present
by having his new
graphic novel being made with AI.
Protect your art from AI with Glaze
or
Nightshade
First I'll breakdown an element of
illustrating a comic page so the
"creative process" is less abstract,
and the visceral reaction to AI's use, that
you see from pros, makes more sense.
Illustrating ain't about who can draw the
most realistic car or fantastic looking
spaceship as much as it's about decision
making.
Panel
1 is an establishing shot, the decisions are
based on need. What needs to be
there/primarily seen.
Panels 3,4,5 is
a pacing/storytelling decision. It could
have been only two panels, maybe even just
one...I set it up as three for the sake of
timing, having the girl and reader learn
what the gift is slowly.
Panel 2, is a mood/relationship
decision. I choose to move the two
figures to sitting side by side, to add some
emotional closeness.
Lot's of ways to draw panel 2 and technically have the
same thing happen. Kid could be sitting on
his lap, have a curios look on her face or
determined, looking at the Dad or the gift,
maybe his arm around her instead of talking
with his hands. It's important to understand
here that an illustrator is, subconsciously
perhaps, arranging the visual relationship
with the father and daughter based on what
he knows of such things through life he or
she lived and/or observed.
Panel
6, This is now a decision in execution, as
in people draw things the way they draw
things. If I were writing that script for
someone...how would I even describe the look
on her face? Slow comprehension of the
horrific? Would that give an illustrator a
mental image the same as mine? Maybe...At
that point it's be much like being in a
band. You need someone who's on the same
creative mindset as you and very good at
what they do. So if you said, "the
baseline here needs to be bluesy but also
kind of fast", they'd know what you
were looking for and be able to pull it off.
All of those decisions, if done differently,
would have made the gag hit differently or
maybe even not land at all. The point being,
in understanding what the point of the scene
is, an illustrator is going to make it look
the way his mind thinks it works
best/perceives as the most
visceral/effective. That is the job, that is
the skillset. Not to make it look real, but
to make it feel real. Got it? Good.
The whole AI art issue is bit
abstract for anyone who's not an artist or
illustrator. I'll lay it out the basics of
it for you, in a way that can
get across how dehumanizing it is
for...well...everyone. And why artists are
so furious...it's more than just jobs/income
being lost to a machine.
We all
understand Google Maps at it's root
is a program that performs a function -find
a route from point A to point B. But it is
only as useful as the amount of data it has
to create that route. AI art programs are
essentially the same, they preform a
function- make an image (via a
staggeringly sophisticated mathematical
probability equation). In order for it to do this
to any worthwhile effect, it needs a
massive amount of visual data
to interpret, boil down to zeros and ones
and reassemble. An AI art program with
only the data from the limited amount of
public domain (uncopyrighted) images available, would be
about as useful as if Google Maps only had data
on north/south streets starting with the
letter Q.
SO... these programmers did
not stop at acquiring public domain images,
they set the program loose to cull every image
it
could from the internet, knowing full well
the vast majority were under copyright, and they
gave the illustrators and artists no warning
and no choice. They just flat out stole peoples
work. and in some cases, if that artist had
a large enough body of work...essentially
stole their style. A style that years of time,
thought, and effort was
spent developing.
That last bit is
particularly nefarious. I'll explain more
specifically. Artist X, has decades of work,
thousands of illustrations stolen by this
program, and because that body of
work/sample size is so large...anyone can
use that program and type "in the style of
Artist X" and the image created will be in
his/her style. Nefarious is the only word I can
think of for making an option that's
intention is to be able make any specific
person obsolete.
If you invent
something you can patent it. Imagine your
invention was a style...actually that's not
parallel, because a style is more personal.
It is as personal to you as your
personality. It is your own personal visual
interpretation of the world, it is as much a by
product of your personality/perspective, as
it is your effort. And this program sucks it
into it a data sample and lets anyone use it
,who pays the fee. And does not pay you, or
ask you, or even tell you. Seems like that
shouldn't be legal...oh...standing by...
They
stole millions upon millions of copyrighted
works to use as a data source...for their
program that is making them rich.
Midjourney has admitted this and basically said "do
something about it, LOL". and they are
currently facing class action lawsuits.
*previous
blog going in depth on how AI art works
here*
With
ALL THAT explained...back to Chuck Dixion.
Chuck wrote comics for Marvel and DC, on
probably every title you can think of at one
point or another. And over that period of
time many great illustrators worked hard to
bring his words to life in the minds of
comic book readers.
Chuck's new
graphic novel set to come out this year
is being done completely with AI art,
specifically Midjourney. This mutherf*cker
only has
a career because of the illustrators who's
work, and in some cases style, was stolen by the very AI
program, that is being used to make his next book.
Whadda P.O.S..
Took me a whole 2 seconds to notice that was
made with AI. I asked Chuck what was going
on...he choose not to answer. Let's
be as fair as I am able. I
don't know who's brainchild this project was
(it's an indy project), but
Chuck's named as the writer (with a
co-writer), the entire first half of the
promo video is all about what a big deal he
is, no other name brand person is
involved...this project would not be
happening without Chuck. Anyone,
whether they have been in this industry for
2 minutes or 40 years, is going to ask
"who is going to be drawing my script?".
So...he knew, knows, and is promoting it.
He could have, if it wasn't his idea in the
first place, just did what many (myself
included) have done when approached with a
project that has AI involved and said
"nope." He made a choice, no one made it for him.
Speaking on creative merit alone, it's
pathetic that Chuck's fine with his writing
being crapped out via an AI program. I'm
trying to imagine having 140 pages that I
had ,theoretically, put much
thought, planning, and heart into, to tell a
story that connects with an audience...and
just handing it over to an AI slurry
machine. I would know it's not going to be
good.
When you're creating work that may
be seen by an audience of strangers, this goes for
musicians, writers, comedians, ect, you
want your work to connect with people and/or to
be true to the vision you had of it, and you
do as much as you are able in order for that
to happen. Maybe out of artistic integrity
or out of capitalistic goals, or some
combination of the two.
To just
forfeit the entirety of the visuals, in a
visual medium, to an algorithmic
mathematical randomizer is f*cking bizarre.
The stories and characters he co-created in
the 90s, those were a by product of humans
collaborating and giving their
interpretation of what the story meant or
needed. There are a staggering amount of
decisions to be made that require,
experience, skill and instinct...
The camera angle, posture of the dead body,
the color of the stairs on panel
two. The
posture of the hand, the angle of the grenade and how far it
is from the hand
in panel 3, how far down the stairs Frank is
panel 4, the colors of the explosion and how
much of them are transferred to the
background images in panel 5...those are all
elements decided on and executed by
different people who have different creative
perspectives, all collaborating with the
skills they have, so that everything works
together to tell that story in a way that
the reader is drawn into the scene.
If he thinks the words he wrote were the
total sum of the reason any of his stories
connected, and any series of pictures with
the same basic information would have been
the same as any other, he is a fucking
idiot...who maybe doesn't even like comic
books in the first place. If I ever met the
Chuck in person, I don't
remember. So I won't claim to know how deep
he is. Whatever
reason he is okay with AI being used on his
work, it's a giant middle finger to the
illustrators he's worked with, the ones he
hasn't, and frankly to his own work.
In promoting this project, he's not
mentioned that AI is being used in place
of an actual illustrator
(...use your own faculties as to why that
is). In fact, the
whole pitch fails to mention that, until way
at the bottom in answering questions that
are required.
100's of hours per panel, ya say?...we'll get
back to that...
As I look at this
now, that second answer is...if we
wanted to give a bend over backwards level
of benefit of the doubt "incorrect". They
are using Midjourney for the AI generated
portion, Midjourney did not get
consent of the owners of the works it
uses to generate images, and is being sued,
and nothing made with midjourney can be
copyrighted...that is why that question is
there. So the correct answer is NO.
Back to the
100's of hours per panel. Yeahhhhh...no.
First off, It takes about 15 hours a page
doing it traditionally. So if this was the
case, he'd be a complete moron for not just
hiring a human after he worked for 200 hours
and was only through one panel. Some quick math here...he's got a
sample page with 4
panels, we'll take the low end and call
"several hundred hours" 200 hours. 200
times 4 panels is 800 hours. 800 hours per
page times 140 pages that'd be 112,000 man
hours to make the visuals for this book.
Working 40 hours a week that be 2800
weeks....or 50 years.
So...he's
claiming to have spent 50 years doing the AL
images for this book. My educated guess
is he put zero
time into the AI images, besides resizing
them to fit on the page. Look at this pin up for the
book...can someone look at the hand and
explain to me what the f*ck is going on with
the fingers? Is that four fingers and a
stump, or what?
He expects people to swallow the
notion that he spent hundreds of hours on
each panel, and he didn't even bother to
check the one thing everyone knows AI can't
do (fingers), on the image used to advertise the project...what a dope.
I think whoever wrote that
has no
idea how long anything actually takes, or
why. And wrote that out to sound like he
didn't just type some bullshit into an AI
program and let it spit out images.
One word comes
to mind for doing an entire comic with AI
(if in fact it is done)... tedious. Type in
some bullshit, it spits out an
image...repeat 4-6 times a page for 140
pages. You're doing data entry at that
point, and not getting health insurance.
It's as hell and gone from the creative
process as I can think of.
Whatever the actual amount of time
was put into the AI pages...the end result
is...well...it's a page. There's no
backgrounds, his amour and cross insignia
change between panel 1 and 4 and it's full
of visual nonsense, but it
looks professional if you don't look at it
too long, (and no
one will).
Every time someone asks how I knew
something was AI so quickly I say the same
thing
"It's easy...I don't feel anything
when I look at it" and the reply is
always "huh...yeah...yeah, you're
right."
When you're telling a story, in a comic book, it's about if the reader feels
like they are watching something
happening...and wants to know what happens
next. It's about capturing the
imagiation.
At the core, a script for a comic book
is fairly similar to a movie script. The
script says who's where, what's happening,
who says what. The director decides the
camera angles, how much time's devoted to
certain aspects and so forth.
Same with a comic book but in that case the
illustrator is the director (...and set
designer, and costume department, ect ect).
So that page up there would have a
script saying - witch looks into ball and
freaks out and smashes it saying "blah blah"
She says "blah blah" to ogre. Her and ogre
go into dungeon saying "blah blah". They see
dead guard and she says "blah blah Blah"
She hears something and they both lay in
ambush. Consider that limited
amount of info and look back up at that page
and realize all that
went into bringing that to life. Some
of it instinct, much of it skill and
technical expertise. I could break down all
the subtle things that made this a great
page, but we'd be here all
day. Trust me, I mean all day...a lot went
into this to give it the right feel, just in
composition alone...
The common question regarding AI, is if it will be a passing
fad that goes away due to
negatives inherent to it, or if it is
the way of the future...the answer is
neither. It is, on the highway of
communicating creative ideas, an off-ramp to
mediocrity. It will always be there for
people to take, but it's only ever going to
get you so far. Because the point is to get
your work to make a connection with people, and every time you use it,
you put vital creative decision making/creative
elements in the
providence of something that is not alive.
The last thing we need right now, is
something that lets more people who don't
really understand what makes anything good,
make more mediocre shit, faster.
Let's define mediocre shit, by what it is
not. It is not great...and we'll define
"great", in the battleground of ideas, as
something that makes a profound connection
to people. Being masterfully crafted is not
a requisite for this. Nirvana made a
profound connection with people, they were
not a masterful group of musicians by any
stretch of the imagination. The comic,
Johnny the Homicidal Maniac, made a profound
connection with people. That (with all due
respect to Jhonen, who's work I do love...see
panel below) was
not a masterfully written or illustrated
book. What they both did masterfully was
make a connection to an audience. IT IS THE
MOST IMPORTANT THING THERE IS. It's
something intangible, unlearn-able perhaps,
just some pure human element that speaks in
volumes in a way no one else does.
If
you take the human out of any aspect
of what you are doing, you are eliminating any
possibility that it will have that
element. There's just no two ways around
that. No mathematic theorem that can capture
it, any more than one could be made that
breaks down what love feels like, into zeros
and ones. If you want to make something
"great" it must give off some kind of
energy...vibe...passion...and computer code
has none of those things. It just doesn't.
It's zeros and one's brotha, no matter what
you're brain full of too many sci-fi movies
is telling you.
Ironically, or maybe
obviously, the more grounded in actual
programming someone is, the more they
understand this. It's the daydreaming
creatives (and/or incels) who let their imaginations
romanticize what these programs are capable
of. There is a coding genius by the name of
Rob Braxmen (below) who said something that makes
eerie sense, regarding people who think AI
is something that it is not- "AI is not
dangerous because it's smart, it's dangerous
because it's stupid. It has no reason or
logic it is just a program that executes
whatever commands you give it. and it is not
capable of realizing that you really meant
something else...it is just zeros and ones"
Anyways...
Back to Chuck,
he spent decades working for other people,
under the rule of editors and editors in
chief. Free of that, this should be a golden
age for him, making the best books of his
career. Instead,
he's done a 140 page story that's going to
be mediocre shit/butchered by AI art. There's a lot
of stores that will not carry anything done
with AI, most comicons now will not let AI
be sold or displayed, it doesn't look like
his orders on this launch are even going to
get to 500 readers which, for someone with 4
decades of name recognition via Marvel
Comics, is pathetic. Unless the only
goal was giving the finger to an entire
profession, I don't see an upside. Maybe
he's just dumbass, who despite decades in
the industry never understood what goes into
the visual side of things, or for that
matter, why anyone loves the comics they
love.
Without the "why" it's hard to find a teachable
moment here...it's just watching an a-hole
shoot himself in the foot. I guess
that's fun sometimes, all by itself.
Especially when it ain't me.
I don't really know how to wrap this up.
Remember a couple weeks back I mentioned
that there's' some cultural paradigm shifts
going on that are going to have their way
with the comic book industry and you're
going to start seeing some people doing
stupid shit to try to stay relevant? Maybe
that's what this was. This is the opposite
of how you steer through a paradigm shift.
Put out the
best, most genuine, most sincere work you can, always.
Later.
OH...ALSO...I will be on the the radio on
the Pint O' Comics show Sunday Jan 5 6:30pm
EST
listen
live online at
wesufm.org
|
|