Thoughts on the AI ‘ART’ revolution.
Apr 1, 2023
Reading Time: 4 minutes
13 Dec 2023
We are entering an age of a brave new world but we should do so with equal parts of enlightenment, excitement and deep concern. To do otherwise would be foolish at best.
Edit Dec 2023
Almost a year has passed since I wrote these AI articles. What’s since transpired about the tech-bro bastardry and their somewhat deranged fan boi club has put a certain black spin on things. The open and blatant plagiarism from the likes of OpenAI and their attitudes to copyright laws is, to put it lightly, FUBAR. While much of the sentiments I have expressed in my writings about AI image tools still hold true, the mechanisms of the tools themselves as they currently stand are gravely flawed, though how badly might also be based on the intent of the end user. Over the past year, in maybe a more positive note, I have also read of designers using AI trained on their OWN images to help the generative design process, and that is some really interesting stuff.
As if a sign from above, all the comp work I had started doing with the generated images, hours on hours worth or work, was all lost in a disc mishap (don’t get that with paper and pen unless coffee is involved); while I still have all the base backplate images that I generated through Midjourney, I have not decided if I will use them to recreate what I had already done…
I mentioned in my last post my experimenting with the Midjourney ‘AI’ and the current two thought trends around such systems. Over the past week I took a much deeper dive into AI based image creation (again via Midjourney) and have come out the other end equal parts enlightened, excited and deeply concerned.
Enlightened: Art is ever changing and though the ages it’s evolved countless times in technique, styles and tastes. I am seeing this as the AI driven evolution of the art process, where it does not replace but expands it, opening new doors, directions and areas of imagination. In my time, I honestly have not witnessed anything like it.
Excited: These days there’s a certain idealist, romanticised notion about art, but the reality of the commercial world of ‘art’, especially digitally based art in areas such as entertainment, is that there are already so many shortcuts and tools creative professionals can utilise to arrive at an end result, that the idea of the artist spending years to hone their skills to perfection is idealistic at best. Are these AI tools really any different from ‘photo-bashing’, ‘kit -bashing’ (in the 3D world), procedural modelling or photo-montages? Probably not, all are shortcuts, ‘cheats’ to arrive at a desired end result. In this realm, AI generators are just another tool to be added to the professional’s quiver.
The images I am now able to produce would have taken me hours of work to arrive at, if I arrived at at all; the AI takes one’s ideas into directions and compositions you’ll never would have thought of… at least not without the use of ,*ahem*, imagination stimulants; in some ways using an AI is like five of me working on the same concept independently. So where I do what I do on the side of my ‘day job’, being able to use a tool to expedite, expand, and visualise my written word relatively quickly is a godsend. As a friend said “it’s like a little army of robot junior designers at your call”. I am truly excited by these tools thus far. I was more than skeptical at first but after spending the time need to learn how to use Midjourney effectively, then developing a methodology to arrive at the result I wanted, I can see an endless range of possibilities that expand beyond anything available to date.
Deeply concerned: The gist, and somewhat contrary to the two previous points, is given time these systems will wipe out a layer of people doing, or trying to do, ‘digital’ art for a living. I have zero doubt about that. Where an art director can go and verbally define their idea, fine tuning it to land a visual result that tells their story, there will be no need for those in the layer of art production that aided in this process. We are less than six months in on public access these AI engines and already, in the hands of people such as art directors, I am seeing results that make me stop and stare. Less than six months. What will we be seeing in a year or two??
Case in point – I just cancelled a software license subscription because this AI stuff is faster, better, and helps take my ideas in directions I would never have thought of. For me, the AI is a better tool. What we are seeing here is a bigger version of the desktop publishing revolution of the late 80’s, where computers and software wiped out an entire layer of the publishing industry, redefining an industry literally overnight.
It is indeed a brave new world we are entering, for better or worse. Where these tools will take us, and even if their creators even give a proverbial flying shit, is something for another conversation. Regardless, here we are. But I will end on an optimistic note…. As has been the case with many digital revolutions, I am already foreseeing that the analogue skills, you know, the ones where you spend years learning how to do it right because there’s no endless undo or digital canvases that can be blanked out via control-z, will make a comeback. More than that, I believe that the analogue, and those that can execute them, will gain a new high tier value. The skill in the craft will return.