Released: 2nd December 2022
Phil Christopher – Digital Marketing Specialist
Odette Johnson – Artistic, Musical Human
Tyron Santiago – Character Concept Artist
Some post podcast thoughts…
I was quite surprised to be asked to be involved in this round table discussion on ArtyParti, about the good and bad aspects of the use of AI in art. Surprised because although I have tinkered on the edges of this technology with regard to some of my art, I do not consider myself an AI artist or an expert in the field and AI raises many ethical and problematic issues, which has bothered me for some time.
I’ve been dabbling on the edges of using some of the simplest AI art programmes to create some images which have been the basis of some bigger pieces. The subject matter has always been quite specialist such as on subjects of dementia, environmental issues, women’s health and war. I have however, used it more as a tool and the many images generated and then personally chosen, have been used in layering and digital processes to create something ‘new’.
I have increasingly felt uneasy with this process.
On the one hand, at times, it has produced images that I could never paint or draw, as I do not consider myself to be that skilled. However, on the other hand, since the creation of AI art is dependent on code, logarithms, engineering, image scrape etc., many of them have a particular look about them, they start to get very samey and typically look AI produced and for me, I am always trying to make them less so, which is a very human response. My actual art in watercolours, acrylics and photography is created by me and has a personal style, which reflects my personality, human experience and creativity.
Much of the discussion covered some of these issues and was very interesting, with the four of us talking about this huge subject for over 2 hours and we could have gone on…
There was a general feeling that the cat is out of the bag, so to speak, and how do we deal with that? It is good at some things and fun, such as the otters on a plane example or the adding to famous paintings backgrounds or characters who are not in the original, like Phil’s adding Blade Runner elements to Hopper’s Nighthawks…and yes, I know that artists have themselves added things to old masters, such as Dali adding his famous moustache to the Mona Lisa (remembered post discussion in a Homer Simpson Doh moment). However, as Tyron quietly stated, all this human input into the database means that the most recent iterations of the AI art programmes are developing superfast and getting more complex and possibly more ‘intelligent’ or manipulative, depending on your viewpoint. There are definite issues about how the art is created by means of scraping the internet, which in itself raises the problematic issue of copyright and ownership.
Does an AI image created from prompt words equal art, and can you call it your own art? Should someone benefit from it when it is an amalgamation of many different images both public domain and not?
For me, although we touched on the wider discussions for and against, AI platforms struggle or do not address the many questions there are about lack of diversity, being western centric, racism, misogyny and also being stereotypical in their outputs. It also struggles with understanding more complex prompts and can return very odd images especially when asked about the human face. In an experiment, I prompted simple descriptions of each of us and you can see the result above. There is an approximation, but you can see that the human face, eyes, ethnicity, gender, among other things, causes a few problems. Also, who are these images created from? Are these created from real peoples’ photographs?
Although I did mention some of the benefits of AI in the development of assistive technologies, such as predictive text and language apps for those with communication needs, there are definite ethical reasons why this should be welcomed with caution, particularly with regard to using the seemingly positive programmes for more darker purposes such as disinformation, hate speech, deepfakes and the possible rewriting or dissemination of unscientific research and theory, to name a few. If you wish to read a more in-depth and knowledgeable article on these concerns, you can find it here in WIRED –
Effective Altruism Is Pushing a Dangerous Brand of ‘AI Safety’
This philosophy—supported by tech figures like Sam Bankman-Fried—fuels the AI research agenda, creating a harmful system in the name of saving humanity
I still feel that despite things, people will always appreciate and want ‘real’ art, that is created by a human to hang on their walls, or to wear or to admire and let’s be honest, painting /drawing / illustration is just one tiny element of the arts. I think it will be a while before any AI actually creates a sculpture, or crafts something out of wood, metal, textiles. The appreciation of someone’s ability to create a finished item from raw materials will always be there and needed. Beautiful glassware, embroidery, paper art, knitting …. the list goes on. And what makes this all possible are the human traits that Tyron mentioned…Imagination, Passion and Individuality.
It is the human imagination that separates us from the algorithm. We alone can be inspired by something, have an emotional response and create something from that experience. Our individuality enables a plethora of complexities and differences in colour, texture, medium and representation. We make the choices, and what we make shows our passion for that creative process.
On a final note, at the end of the afternoon, we were asked if and how AI will be used and featured in our art practice in the future?
For me, it has been an experiment that leaves me apprehensive and conflicted. I was already moving away from its use for all the reasons mentioned above and more and more I want to go back to creating artwork with paints and camera. Also, I was quite vocal in pointing out that there has never been nor ever will be any use of AI in my audio art.
That is pure hard work, imagination and a passion for sound.