Within the induction and first weeks of the course I specifically landed in an area of study called emerging technologies. This is something that I have been interested in for a while and so possibly that’s why I was drawn to it. Being a predominantly traditional artist that uses technology as a tool, I find myself questioning where the future of traditional art is heading and how the tools of the future will look in my practice. An evolving world might encourage change into the digital arts but it might also crave the nostalgia of the past. This reflection intrigues me and the only way- apart from time- to tell which way things might go is to explore it. Why not then delve straight into a very current and very heated debate among illustrators worldwide, artificial intelligence. AI is being used to create art by programs that ‘learn’ from art history to produce work.
The use of social media keeps me up to date with the debate and the use of tik tok allows almost head on debates in live time with a very clear dividing line between the camps of AI users are artists and AI users are programmers. Social media cannot form my arguments but has been able to inform me of social trends where I am exploring academic articles based on the topic. Different media surrounds me constantly and has led to this weeks research, what is the ethics of AI used in illustration, who owns the outcome and are we looking at a future where technology has taken over instead of being used as a tool? I ask these things purely because listening to debate can only take you so far when no one can tell me the answers to these points in an entirety.
‘Artificial‘, meaning opposed to natural and ‘intelligence’ meaning the faculty of understanding (Oxford learner dictionaries, 2022) has been explained over decades of science fiction and scientific writings. Roger Schank wrote in a magazine article “AI is the science of endowing programs with the ability to change themselves for the better as a result of their own experiences” (schank, 1987) this is a response to explaining AI to a new audience where the technology was still based in science fiction literature. This articles argues that AI is nothing but programming and until humankind understands intelligence itself then it is impossible for a machine to gain it artificially. With leaps and bounds in research and technology this basic explanation of AI still stands strong just with a different approach to it. “Today, however, most researchers want to design automated systems that perform well in complex problem domains by any means, rather than by human-like means (Dick, 2019). What Stephanie Dick is talking about in this paper is how the evolution of AI has transformed from science fiction to an actual usable intelligence that resembles its origin.
After some research into the meaning of AI as a solid basis for my work I decided to jump straight into some of the questions at hand. The big debate on AI in illustration truly revolves around ethics. This has made for some interesting research and reflection as I personally sit on one side of the debate but following ethics it has made me rethink my position. While continuing to sit on the same side it may now be for different reasons. The writer Isaac Asimov is famous for his science fiction novels and infamous for creating the three laws of robotics; Law 1 “A robot may not injure a human being or, through inaction, allow a human being to come to harm.” Law 2 “A robot must obey orders given to it by human beings, except where such orders would conflict with the first law.” Law 3 “A robot must protect its own existence as long as such protection does not conflict with the first or second law.” (Murphy & Woods 2009) Although these laws apply to robotics, Asimov writes about the laws being broken or bent which in turn covers the topic of AI in that the machine has learned from experience and made a decision based in human like thinking. These three laws are almost engrained into anyone who has experienced science fiction and bend the thoughts of ethics within AI on a daily basis. As you read or watch these stories you can find the emotion within them and understanding of why robotics might become unlawful which gives interesting thought into whether robotics and AI should be covered by ethics and human law. We may still be far off of that discussion in the real world, but it provides an interesting point of reference to think about during debates about whether AI generated art can infringe or own copyright and whether one day they may be protected. Asimov creates compelling characters that find empathy with an audience and in turn making you root for them and their rights but, will people feel the same about a program on their computer at home when the time comes to ask these questions? Can ethics be applied to AI or is it simply science fiction writing that compels us to believe so for well written characters?
One side of the AI debate asks whether AI users are just programmers and not artists. The handbook of ethics of AI discusses transparency with algorithmic systems. It goes on to say “An effective application of algorithmic transparency should strive to locate the relevant aspects of human involvement in the design, operation, and management of a system. “(Dubber, 2020) This chapter talks specifically about accountability and how transparency must be used to locate the human element within a design. With this in mind it makes me understand the debate a little more. AI users, by this method, should be transparent in their use of AI to determine that they had a hand in the design also. Artists are firmly convicting AI users of fraudulent artworks by hiding the fact that they are using AI in the first place, AI users perhaps, should show how they have used it as a tool, and it does indeed have human elements. For instance, a drawing program can create a straight line with ease which the artist uses as a tool to save time in measuring. In the same instance AI users are using tools and should show their involvement in what they are doing for AI and not what AI is creating for them.
On the other side of the argument, we can look at who actually owns an AI created artwork. The debate has recently turned to plagiarism on account of the systems being able to have inputs on a style. By inputting an artist’s specific style the AI programmes are mimicking but are they stealing directly? AI uses thousands of reference images and then creates a new image based on what it has “learned” but as the definitions have changed over the years we can see that AI does not learn in a human like way but by any means necessary and this becomes cloudy in terms of determining a fine line between the usage. “There’s something missing in AICAN’s artistic process: The algorithm might create appealing images, but it lives in an isolated creative space that lacks social context. Human artists, on the other hand, are inspired by people, places, and politics. They create art to tell stories and make sense of the world.” (Elgammal, 2019) Ahmed is talking about a machine that created an image using no human interaction and prompting at all. Purely fed images. It shows that the AI side of AI programmes can do no more than reference existing material. It cannot think and learn about politics and social events because that is what a human can input. I believe that this shows AI based images to be using reference material without permission and even though human input is added, by transparency we can differentiate the two and find the human creative side and the AI side that has not had permission to use past works.
By looking at a copyright law case filed by Peta on behalf of Naruto the Macaque we can also start seeing an interesting case for AI. The lawsuit was against photographer David slater who claimed a photograph that was in fact a selfie taken by the Macaque. As David slater did not take the photograph it was argued that revenue should not be generated by him under copyright law. However, current copyright law does not apply to lower animals and so the case was settled. Can we then apply this case to AI and understand that copyright law does not apply to machines. It does however apply to the artists whose images are being fed into the machine which asks a very big ethical question into who owns the final image and whether copyright can come into question even before and image is produced by AI.
One final argument about ethics is “in the worst case, clients will turn to machines to get the job done faster and cheaper” (Martin Nebelong, 2022) This is a projection about a possible future outcome that has been debated since the dawn of technology. It is here that I can truly start reflecting on AI use within my illustration practice as it is true that people will always look for cheaper and better. I do not believe that AI is going away and I also do not believe that it will take over. I think that the arguments happening in the world now will shape how it used in the future. Questions into copyright will surely be brought into courts at some point and once this matter is settled I believe AI will be used both solely by illustrators and partly for thumbnail concept use. If this saves time in a lengthy process then I may be forced to compete within my field later. And here I propose an experiment to myself. I am going to create a piece of art using AI to perform first-hand experience into the programmes. Here I can also test timing on how long this artwork takes me to complete so that I can reflect on this.
I started with the prompt of gargoyle as I was sketching one out for this year’s Inktober. I spent a couple of hours doodling some linework for my sketch and so decided I would spend the same amount of time on the AI drawing too. Instantly I realised some fascinating results. Using the free online software ‘Dall-e’ I entered a simple prompt which immediately gave me a result for an image which resembled what I had asked for. I then tweaked the prompt a few times until I found one with style that would be workable to mine. Interestingly I found that the only way the AI images truly changed was when I asked it for different well-known styles. For instance, I tried ‘Hokusai wood block style and horror-based Giger’. This strengthens the argument that can be made about AI taking images for use without consent. However, the final concepts were not so truly in the style that it can be seen as any more than reference use which is encouraged by artists. The final image gave me a workable piece of art within a very short time frame meaning that I had a workable base in minutes and with no knowledge of the AI software’s or prompt lists.
After I had a base, I spent a little time adding some of my own style to the image with bold linework and strong shadows. I quickly became bored and unenthusiastic however. It became apparent that by not working on thumbnails for composition and development of a character I just simply didn’t enjoy the process anymore. It may indeed save time by having a base already laid out but at significant cost to my creative process. The thing that sets me apart from AI is the ability to refine at the sketching process. I can decide whether a perspective is off or a colour palette is on trend and even research for cultural background into a piece to define shape and colour. The human input into AI can prompt all of these things but can they really re-design them? AI does not learn this way and cannot make these human decisions. Without all of the back reasearch and a lifetime of experiences to help within the concept stage I feel like I was working on top of a piece of artwork that another person started and gave me no context for. The art feels lost and without purpose of function.
My final reflections end somewhere that I did not expect. The debate will continue for some time and may even never be resolved but the matter of the debate is slightly insignificant to me. As much as I enjoyed delving into the subject and will likely do so some more, I feel like the most important realisation is, what this technology means to me as an individual creative. Within my practice I can see AI as an inventive tool that perhaps could help somewhere in my work, but the cost of my own creativity is too high. Coupled with the debate side of things I don’t feel like I could see it as a journey that I will be taking for my art style. When asked why a creator makes the things that they do I feel like the main purpose is because they enjoy it and enjoy the process. Even before function, comes enjoyment, which also becomes a function to ourselves through wellbeing.
References
Elgammal, A. (2019). ‘AI is blurring the definition of artist. Advanced algorithms are using machine learning to create art autonomously’, Sigma XI, the scientific research society, 107 (1). [Online] Available at <https://go-gale-com.ezproxy.herts.ac.uk/ps/retrieve.do?tabID=T002&resultListType=RESULT_LIST&searchResultsType=SingleTab&hitCount=1&searchType=AdvancedSearchForm¤tPosition=1&docId=GALE%7CA579092374&docType=Article&sort=RELEVANCE&contentSegment=ZONE-MOD1&prodId=AONE&pageNum=1&contentSet=GALE%7CA579092374&searchId=R1&userGroupName=uniherts&inPS=true> [Accessed 06/10/2022]
Dick, S. (2019). ‘Artificial Intelligence’, Harvard data science review. 1 (1) doi:10.1162/99608f92.92fe150c. [Online] Available at <https://hdsr.duqduq.org/pub/0aytgrau/release/1> [Accessed 06/10/2022]
Dubber, M. (2020). ‘Transparency’, Diakopoulos, N. The Oxford handbook of ethics of AI. London: Oxford University Press. [Online] Available at <https://academic-oup-com.ezproxy.herts.ac.uk/edited-volume/34287/chapter/290661457> [Accessed 06/10/2022]
Murphy, R and Woods, D. (2009). ‘Beyond Asimov: the three laws of responsible robotics’, IEEE Intelligent Systems. 24 (4) PP 14-20. doi: 10.1109/MIS.2009.69. [Online] Available at <https://ieeexplore-ieee-org.ezproxy.herts.ac.uk/document/5172885> [Accessed 12/10/2022]
Nebelong, M. (2022). ‘Create unique art with AI tools’, Imagine FX. 216, p 70.
Oxford learner dictionaries. (2022). Oxford University Press. [Online] Available at <https://www.oxfordlearnersdictionaries.com/definition/english/artificial> [Accessed 06/10/2022]
Roose, K. (2022). An AI generated picture won and art prize. Artist’s arn’t happy. [Online] Available at <https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html> [Accessed 06/10/2022]
Schank, R C. (1987). ‘What is AI anyway?’. AI Magazine, 8 (4: Winter) PP 63-64. [Online] Available at <https://ojs.aaai.org/index.php/aimagazine/article/view/623> [Accessed 12/10/2022]
Slater, D. (2017). Monkey Selfie. [Online] Available at <http://www.djsphotography.co.uk/monkeyselfie.htm> [Accessed 06/10/2022]
Slotkin, J. (2017). ‘Monkey selfie‘ lawsuit ends with settlement between PETA, photographer. [Online] Available at <https://www.npr.org/sections/thetwo-way/2017/09/12/550417823/-animal-rights-advocates-photographer-compromise-over-ownership-of-monkey-selfie> [Accessed 12/10/2022]
Chloe Cooper | MA Illustration 