It's a little helper. Just embrace it quickly and start using it.
The threat to the work of computer graphic designers and traditional graphic designers from AI does not exist for now, in my opinion.
So far, the results of artificial intelligence work are sometimes beautiful and surprising.
But you know what's wrong with them?
These are quite haphazardly created creations, they are the results of trial and error. Sometimes repeated many times with minimal changes, until finally, the person who operates the AI decides that enough is enough, time to quit pressing the "generate" button, because it hurts.
Very often it's not quite what said person had in mind, but it's good enough to say: Oh well, fuck it!
(this pic I used as a base to make all the avatars in this post)
The problem of AI is also consistency.
Imagine drawing a comic book.
In fact, there is at least one comic has already been created ( I'm sure couple of them by now, actually), but when I saw it, well... It's just a series of uninteresting ( not pleasing ) pictures of characters who are static, and generally, behave in uninteresting manner. (I am generous).
It's time for a challenge:
Let someone with the help of AI generate something like this:
A chase scene of a Batmobile and a cab with Joker in it through some known location such as the Brooklyn Bridge and a few more blocks (I know it's not Gotham, I know.). Views from different sides of the car, from inside of the cab, etc. And all this in some specific graphic style. Characters and vehicles must be similar from scene to scene (AI has problems even with generating plates or Outdoor AD banners), the location is changing consistently as the vehicles move.
Or have two superheroes fight in a MMA style while demolishing the room they are in.
And this must be pleasing to eye.
I want 4 pages of a comic book. Go ahead.
I don't think anyone will be tempted to do it. It is too hard for AI to create something like this.
For now.
For now.
But in the future? Who knows?
Who knows how fast all this AI will go forward?
I bet that for a little while, AI will only be a convenient and quick help to achieve semi-desired effect, but it will not replace a real graphic designer.
Of course in a few years, everything may change and computers will really replace most of the graphic designers, but I suspect that it will be those lousy graphic designers, who will rebrand themselves as operators of artificial intelligence and that's how it will roll. Why not? It will be interesting. I'm looking forward to it.
So there is nothing to complain about. If someone prefers to work with a computer and AI over a brush and paint, that's fine.
I think that graphics generated with the help of AI will simply have a lower price (due to their speed of creation but also a lot of competition. After all, just anyone will want to try to be an artist :) )
But I also think that this type of graphics (AI) will be needed in some niches in industries such as gamedev or storyboarding, where it is important to be fast and make a large number of different versions of similar things, and not necessarily with the artistic vision of a real human artist.
Now I am trying this AI stuff myself
Most people outside the graphics industry are certainly aware of the existence of tools such as Dall-e or Midjourney, but tools of this type are springing up like mushrooms after the rain. there are more and more of them and they are getting better and better.
I, for example, today tested something called playground (https://playgroundai.com/create)
It uses versions of Stable Diffusion technology and is simply packaged in a user-friendly way with a nice and easy user interface.
Here is a screenshot of this application and some of the most important places to click.
Using this program is trivial, You need to log in with your google account and pass your phone number ( or sign in in another available way) and you are ready to go.
All you need to do is to
- write an appropriate prompt,
- choose Stable Diffusion version you want to use at the moment,
- play with a few sliders at random, select the output image size
- and you can press the generate button and enjoy the results
alternatively you can upload your own image and try to convert it with the help of artificial intelligence and specially written for this purpose prompt.
the rest is just playing with sliders. an example of my work is that I converted a picture of my pants into several avatars. it took me not much time. some are interesting, some are terrible, but it's a lot of fun. The fun is great the effects can sometimes surprise you
This is the effect of about 30 min of messing around with the initial picture of my pants. There were more of it but I didn't like them, or were too weird, or were too disturbing for viewers.
I could make NFTs of them and try to sell them, but, naaah... take it freely if you like it.
So... enjoy? :P
Phew! this is the longest post ever!
