2D Personal Development Pieces utilizing AI

As a digital artist I have been using 2D digital and mixed media techniques coupled with hand drawn / painted elements and 3D for over 25 years. It’s how 99.9% of production work gets done. There are some people still drawing by hand like my good friend Jim Hewitt " set designer extraordinaire " but those are few and far between. I specifically remember the first digital paintings I did for BRC back in 1998 on a rented Mac tower, As far as I know they were the first digital pieces done in house for the company. Over the years I’ve become a fully digital artist, truth is it would be almost impossible to work in production today without pushing pixels…

I believe in embracing new technology, I have been experimenting with AI, primarily MidJourney since another good friend " amazing concept artist Mathius Beaguer " sent me an invite link back in 22. It’s come a long way in a very short time. Im not going to get into the discussion over ethics or any of that here as that's not what this post is about... Suffice to say I don’t agree with how it was developed, it shows a shocking disregard for all creatives that I really don’t understand, but pretending this isn’t happening benefits no one.

I am trying to see where usage of this new technology can fit into my pipeline. Like it or not AI is here and it will be used scratch that already is being used 2 out of my last 3 jobs had an AI element, not by me but it was there... I know of a number of studios using it and individual artist and creatives as well, Not too many people are wanting to admit it but its definitely here... One could remind everybody how long it took the industry to accept photoshop renderings as " real art " but this isnt a TED talk.

In all of these pieces I have used AI generated elements combined with my own sketches and 3D massing fed into the beast and then painted over, on teh illustrations its a lot of paintover. I then composite the image to my liking and go in cleaning things up adding subtracting elements etc… Not that different from a good old fashioned photo bash.
Im seeing it currently as maybe good base underlay and for generation of elements to be used altered just as you would in more traditional photoshop painting even textures for 3D " that would be another post " Probably at least in my view better suited to illustration as opposed to product or industrial design as you have such little control with the AI. The first 3 images here are examples of that, just making a fun " painting " combining elements with no strict adherance to a specific design. For the 2 other images I used a 3D model I built one for work and the other for a personal project. For the construction crane piece I fed VizcomAI a render of my design model asking for an abstracted painting which resulted in an interesting look I believe I fed in a pallet knife painting for ref but I can really remember right now. The second example is from a recent set design illustration for Lexus, after the job I fed in my base render from Twin motion and again I think you could definately use it as an underlay for a final image.

The two design examples are intersting to me as it shows the potential for making the AI stick to a specific set of geometry without all the crazy AI hallucination stuff

AI is not a replacement for designers " it cannot do that " Its not a replacement for artist. It is and should remain a tool, no different then your Wacom tablet or a set of sweeps. Design; IMO, is problem solving. An algorithm doesn’t have awareness “ yet “ and you can’t solve problems without awareness

What Im enjoying the most about this experimentation is that the AI element takes me out of my heavy production influenced style. It makes me want to try new things, form languages, subjects, etc that I dont generally use in my contracted work…

I would be curious to hear from people what they think, feel free to critique.