There are some common misconceptions regarding the differences between Traditional Art, and Digital Art, especially in the current day when AI Generations are muddying the waters and unfortunately denigrating the work and talent of Digital Artists in the eye of the public. In reality, Traditional and Digital art are incredibly similar, with only one key difference between them. AI Generations, on the other hand, may mimic either while being nothing like either of them.

Traditional Art is the creation of a piece of work by one or more artists, technicians, engineers, producers, etc. It could be anything, and while each person has their own view on what Art is and what it is not, I wont be making any such argument here. All I’m doing is pointing out Traditional Art in general and how it differs from AI Generation. From drawings, paintings, and guitar riffs to sculptures, clothing, and installations, artists create works out of physical media, with physical tools, in physical spaces. Drawings, paintings, and sculptures are built to last and exist in the real world, and unless they are replicated (either by hand or by mechanical means of reproduction) there will only ever be one. A limited edition would usually be produced at one time, and there would only ever be that limited run of the artworks. There would still, even in a limited edition, be only one true Original. This is the key difference between Traditional and Digital Art.

When working digitally, artists still work by hand, or by using specific tools for specific functions, just like in the real world where traditional artist use whatever tool suits the job at hand. They require the same, or equivalent, skills and have a few advantages over real world creation. Specifically, things like painting in layers, the “undo” button, the ease of saving variants, and not having to wait for paint to dry, can allow artists to work faster than they would with their real world counterparts. There are some significant drawbacks to the process though, namely things like the loss of tactility, other sensory deprivations like loss of the smell of paint or clay, and other things which detract from the immersion of the artist in the process. On top of that, there is still a misunderstanding of what Digital Art is, and how valuable it is to the world – Digital Art is everywhere, it is in pretty much everything now; from the digital design work used to create vehicles, buildings, tools and appliances, etc., to every TV show or film you might watch, every advert, every book cover, magazine, newspaper, etc. All have had the hand of a digital artist involved in their creation. They took the same skill and knowledge as their counterparts in the days before the use of digital tools.

The huge difference between the two, and the largest detriment to Digital Art in its value to the artist, is that there is no longer an Original. There can be a First Edition, there can be Limited Editions, but unless an artist was to create a piece of work that was never displayed or shared anywhere, and only printed once, and sold only once as a physical creation, there is no real way for there to be an Original.

Where this issue of Digital versus Traditional Art has come to the foreground of late is with the current boom in the Artificial Intelligence technology now being used to Generate content. This could be aesthetically pleasing for sure. It has enabled individuals to essentially push a button and Generate (not paint, not sculpt, not compose, not photograph, just Generate) content which appears to have been created with a great (and increasing) level of skill in the style and medium of their choosing. Some unscrupulous, and deeply unethical people have begun shilling this content as their own Artistic Creations, and not labelling them as AI Generations. There are others though, who will proudly label themselves, and their Generated Content, as AI Generation (some use the term AI Art).

As it stands at the moment, most AI Generators have been “trained” by scraping the internet for any and all content it can get its hands on, and building enormous datasets for it to work from. The most glaring ethical problem here is that consent was never given. The people whose images were ripped off, all the personal photographs scraped off the net, for example, are now in vast datasets for the robots to study and utilise in their Generated images. Did you consent? Neither did the artists, musicians, photographers, etc. And now you can type in the subject you want an image of, and also name the artist whose style you want it in, and boom, you have an image that looks like it was painted by your favourite artist. But they didn’t consent, and they were never paid, and will never benefit from this counterfeit.

Looking to the future, there is vast potential for people to completely rip off artists to create Generated Content to sell, and profit from. Or to use to advertise their own brands and their own products. So the artists who have been ripped off in this way studied and trained for years to perfect their craft, only to see it counterfeited, lowering their own value as a Creative, and potentially suffering from reputational damage depending on what purpose these counterfeits are being Generated for. Not only that, but when the public begins to view Art as nothing more than typing a handful of keywords and pushing a button, all Artistic endeavours run the risk of becoming extinct.

Not to be too doom and gloom about it, I want to quickly address the potential use of AI in Art in ways which may be of benefit to the artist, and to Art in general, so long as they are done ethically. One example would be if an artist wanted to create their own comic, but the sheer amount of time it takes makes it unlikely they will be able to do so. They could, in theory, train an AI Generator in their own style (and theirs alone), and then script and do the layout of their comic themselves. After blocking it all in with quick roughs of the pages, and providing a colour palette to the Generator, it could run through the whole thing doing all the main block ins, in that artist’s style and using their own first pass as the guide to make sure it all does what it needs to. The artist could then go back through and check it all looks correct, before doing any final detailing passes themselves. They could get the AI to do the lettering for them. Yes this would mean some artists would be out of work, but it could also mean that all of those artists would be capable of doing so much more in so much less time. All of them could, in theory, work on their own books this way, or work together with each person taking on other parts than they normally would.

You also have AI doing things like upscaling images, colour corrections, etc. All jobs that either would not be possible for a human to do (upscaling), or massively ramping up the speed with which an artist (for example a Photographer sorting through and editing Wedding photographs) can perform their roles, and even helping do some things to a higher quality. These kinds of tools are not like AI Generators though, and it is very likely that a lot of common tools use some kind of AI in them nowadays. If they don’t, at some point they will.

I do think that AI tools will be of an enormous benefit to Artists over time, and to society as a whole. I just think we need to get a lid on the AI Generators until their function can be carried out ethically, and without destroying the Arts, or any other sector. As it stands, I likely use some AI tools in my work (most likely in my photo editing tools and when doing final exports on my paintings) without knowing it. I do not ever plan to use Gen AI to Generate final images.  I can see potential use cases for Gen AI when it comes to 3d and film making, once again so long as it is ethical.  One potential example would be in the Generation of thread patterns for use in 3d textures for textiles.  It is something that frankly is a pain to do manually, if you were painting it / creating a 3d patch of cloth by hand, and way more time consuming than it requires to be.  The other option is finding some assets you can purchase the rights to, which then requires a lot of time spent searching.  All this for a simple image of threads that is used to create a tiny bump and roughness to a surface for fabric.  Currently I have enough assets on hand for my needs, but it is an example of where I can see a valid use that 100% of people would not notice if not disclosed.  

I plan on always being honest in my approach to these subjects though, and aim to always disclose the use of Generative AI no matter how subtle in my work.  There are so many AI tools built into 3d, audio, and video editing software now that I will not be disclosing every time I’ve used an AI based tool, however I will be disclosing if it is something I know to be Generative, and not, say, running noise reduction on a 3d render.  I currently use LLMs like search engines (though try to verify anything they claim as they are wrong all the time), as they can be especially useful while trying to troubleshoot technical problems in 3d.   I will not use any Gen AI in my paintings, because I enjoy painting and don’t see the point in removing myself from the process.  I also will not use any Gen AI in any scripts I write for videos.  Once again, this is something I enjoy doing, and do not see the point in removing myself from the process.  I see these things as tools to be used, so long as they are used ethically.

At some point I’ll be putting together a long form video showcasing my workflow, and I’ll be discussing all this in more detail there.