Introduction
A question that becomes increasingly more important with the rise of AI is the following: if AI cannot create art, can art then be created by a user using AI and can the user call the result art?
There’s a subtle difference here between saying AI, or any other machine or software, cannot create art, but a user can create art, no matter if the user uses AI or Photoshop for example.
I attempt to answer this question in this blogpost but I will already state that the answer, given the state and nature of AI right now, will be ‘NO‘, because there is too much inference instead of direct interpretation of subjective experiences.
I hope that my answer substantiated by logical reasoning and arguments, gives you a guide when navigating through the world of art and human creativity in times of AI.
Art is consciousness manifested
I have written several articles on this website on art, or fine art. As someone who claims to be an artist, one has to know at the very least what it means to be an artist. You should be able to articulate for yourself what art is, either in words, ideas, strong feelings, or images. And then be able to defend it and embody it at all times.
One of my favorite phrases is ‘Art is consciousness manifested’. It is a brief phrase that encompasses and combines many of my ideas and concepts. Concepts and ideas from the world of art, science, philosophy, and Jungian psychology, mixed with my ideas. One important conclusion is the notion that consciousness is fundamental and cannot be reduced to matter. The only thing we can really know is our own subjective experience, our own consciousness, everything else, including science is just inference.
This is the main premise of a metaphysical philosophy called analytic idealism which has its origins in Arthur Schopenhauer’s philosophical ideas and is implicitly supported by Carl Jung to name a few of the past. One of the modern proponents and advocates of analytic idealism is philosopher Bernardo Kastrup and I can highly recommend his latest book ‘Analytic Idealism in a Nutshell’. A very important and accessible read if you want to find your way in, what I believe is the only convincing philosophy on consciousness on the table right now, and the fundamental questions of life.
The consequence of this is that we can never know what it means to be someone else. Art, however, is one of the most important ways through which consciousness can manifest. When someone expresses consciousness authentically through art then consciousness is manifested in a way that can transcend the artist and the observer.
So, art is consciousness manifested because it can give us a way of experiencing directly the subjective experience of the creator. It may be a brief but transcendental moment. And vice versa, true art, can only be a manifestation of consciousness. Hence, “Art is consciousness manifested”.
The fundamental question
This is one of the things I have been discussing during webinars and in my recently held Berlin architecture workshop. At some point, one of the attendees confronted me with a challenging and very intriguing question. A question that becomes increasingly more important and even fundamental to ask with the rise of AI in all aspects of life.
The question was as follows:
“[…] Can an AI prompter manifest consciousness and thereby create art as a result of that manifestation of consciousness even if the art created is created by AI and not directly by the prompter? […]”.
A short sidestep to make the importance of the question clearer and to be able to formulate a possible answer I am using the concepts of ‘Degrees of Separation’ and ‘Degrees of Freedom’ in a more colloquial way than the more formal way they are usually used in different contexts.
Degrees of separation and freedom
First let me explain ‘inference’ as it plays an important role in this analysis.
It is defined as “an idea or conclusion that’s drawn from evidence and reasoning. An inference is an educated guess“.
Or: “a guess that you make or an opinion that you form based on the information that you have”.
Degrees of Separation is often used in social sciences to indicate that the entire humanity is connected by at the most 6 different persons. The more degrees a person is separated from another person, the less both persons know of each other’s existence and the less we can influence the actions and behaviors of each other. An increasing amount of coincidence and unpredictability becomes the rule then. I’m using degrees of separation to indicate the number of intermediaries needed between the artist and the result of the artist’s actions.
Degrees of Freedom is a concept often used in the world of statistics that I’m using here with a more philosophical perspective. I’m using it to indicate that the more degrees of freedom there are, the more information there is available, and the more we are capable of. At the same time, it becomes harder to understand and manage all available information in daily life. More information gives us more freedom but too much information works against the practical use and implementation of that information. It becomes information overload that we cannot act upon efficiently and needs to be put in a black box to use in daily life. E.g. we don’t need to know how a computer on the level of a microprocessor and electromagnetism works to access information on the Internet, write emails, use apps, etc. The actual processing, the switching on and off of bits takes place in a black box and we all know that. We know it is not magic and there’s a science behind it that we don’t need to know to use it. We just click on icons or enter some keywords. A child knows how to use it. But the concept of the Internet, the cloud, or even telephony would be completely alien to even the brightest minds of the 19th century. We have a larger degree of freedom than they had. It would be nonsense and absurd to say that since we know how to use computers, access the Internet, write emails, etc. that also implies that we are computer scientists and know exactly what switch is on or off and that we can directly affect electromagnetic waves. The regular user doesn’t know that, not even more expert computer users know that.
In other words, more degrees of freedom separate the actor from a black box of essential but invisible information processing, to get to the result, and make the actor accountable for the result without knowing how the result is created.
The user only triggered the result but didn’t create it. The more degrees of freedom, the more we become users or appropriators of the result, not the creator of it. And the more there will be inference instead of the manifestation of direct experience.
This conclusion becomes important when we talk about the creation of art as a manifestation of consciousness.
Can one manifest consciousness by proxy
Back to the original question, now differently put: Can one manifest one’s consciousness in a creation, through more than 1 degree of separation, with multiple degrees of freedom, to call oneself the artist of that creation?
A concrete example: a person enters a prompt in Adobe Firefly, with more or less specific details, to create an image resulting in the image the prompter had in mind. Can it be said that since it is the prompter’s description that enabled AI to produce the image using AI as a tool, similar to a painter using a brush, a sculptor using a chisel, or a photographer using PS, that it is the prompter’s creation manifesting one’s consciousness by proxy and therefore can it be considered art?
What criterion is decisive when a creator claims to be the creator or artist of a creation – how many degrees of separation and freedom are allowed?
My suggested criterion/criteria
We concluded earlier that the more degrees of separation and freedom there are, the more we become users or appropriators of the result instead of creators, and the less we know how the result is created.
And more importantly: the more inference there is instead of the manifestation of direct experience.
In the case of art, the latter is essential where art is the manifestation of consciousness. Starting from the premise that consciousness can only be inferred from a third-person perspective – for convenience’s sake I’m calling AI also a third-person but not in the literal sense of the word – and can never be directly accessed by someone other than the one experiencing, the creation cannot be considered art if the creation is not directly impacted by, and the result of a direct action of the creator who experienced consciousness. Using AI to create art we become users or appropriators, not creators. And our private consciousness cannot be manifested directly, only through inference by AI. And that is to me the ultimate criterion.
Therefore, only a limited degree of separation and freedom should be present to have a direct conceptual, practical, technical, and predictable impact on the format and nature of the creation without inference.
Photoshop and other software
The most obvious example of art creation is when the artist is directly putting the paint on the canvas, with only the brush as an intermediary. Or when the sculptor is using the chisel and hammer itself when carving out a piece of marble.
It becomes less obvious when the photographer uses Photoshop. After all, isn’t Photoshop a digital tool that processes our actions through a black box, and therefore we don’t directly impact the creation and there is inference?
I posit that Photoshop is indeed a digital tool. Nevertheless, a digital tool with a black box that only does exactly what the artist wants it to do and is very predictable in its targeted result. The black box processing is there only to avoid the manual switching of binary states on a bit level.
Analogous to a writer who uses a computer to write a book. The letters and words entered through a keyboard to be processed by a text editor, are exactly the words we see on our screen. What you entered is exactly what you get. A spelling corrector doesn’t take away from the words we wanted to write or the storyline or structure of sentences. There is no inference here and there’s a direct expression of one’s subjective consciousness.
Photoshop will only do what the artist wants it to do in a very direct and targeted way. Darkening an area with the curves tool, for example, is very targeted and directly influences the luminance values of a range of pixels that the user indicated (or excluded) to be adjusted with a very specific intensity depending on how much the user drags or pushes the curve. Or a gradient is only such as the user using the gradient tool, wants the gradient to be. It is very direct and exact up to a pixel. Also here, there is a direct expression with no inference of subjective experiences.
This is very different with a user entering a prompt into an AI prompt box. The user enters words, that need to be translated to machine-readable information, then interpreted by the AI tool, then related, through an unknown-to-me set of various algorithms, to a set of images that is rendered by AI. There are too many degrees of separation and freedom that I’m not even aware of. The user types letters, and words, and out comes an image. There is not even a hint of direct impact. It is all inference from the moment the user enters the prompt to the end result.
Somewhere in between there are so-called filters, e.g. Instagram or smartphone filters, which increase the degree of freedom and separation. We don’t directly impact the result. When you click on let’s say the button that says ‘spectacular’, we see an interpretation of the ‘spectacular’ effect by the software, not by the user. The user is not manifesting consciousness but is only pushing a button that triggers a black-box process.
The no inference principle is key: art should always be the direct manifestation of one's individual and unique experience without inference.
Conclusions
Based on the previous I propose that the no-inference principle is key: art should always be the direct manifestation of one’s individual and unique experience – consciousness – without inference.
The following criteria can help you to determine whether there is no inference and an actor manifests consciousness in a stylized way and creates art.
- A targeted conceptional intention with a limited amount of randomness and high predictability: a limited degree of separation and freedom, preferably 1 degree of separation and freedom max.
- The presence of technical craftsmanship that directly impacts the target (area) and requires knowledge of the essential inner and outer working of the used tool to impact the target with a high degree of predictability and precision: limited degree of separation and freedom, preferably 1 degree of separation and freedom max.
I would state that the ability to enforce a direct practical and technical impact, with a high degree of predictability and precision and with a limited degree of separation from the final creation together with the ability to have a direct and predictable conceptual impact, are both equally necessary to avoid inference.
When what I experience is inferred instead of directly expressed, there can be no manifestation of consciousness. There can be no art.
Afterthoughts
Questions, answers, considerations and argumentations as presented in this article are the type of discussions that are very rarely discussed in photography workshops. I make it a point to discuss it. To give it more importance than smart technical tricks.
I believe that teaching workshops where the emphasis is on creating ‘Fine art’ be it architecture, portraiture or landscapes, should always also address what that ‘Fine art’ is. Failing to do so, will ultimately devaluate Fine art photography specifically to a hideous and cheap parlor trick, and degrade art in general to nothing more but a luxury item. And you as the participant of such workshop will fail yourself in the end by ignoring the true essence and importance of art in a human life. Choose your teachers/instructors wisely.
For those who want to learn about creating fine art architecture:
– Online: Creating true fine art architecture webinar – December 1.
– On location anywhere in the world: keep an eye on the bookings store