I recently touched upon the problem with AI-generated art in a recent blog about ‘What is Art?’ – https://katefletcherpaintings.co.uk/what-is-art/ . I only brushed upon its existence, but the debate gets much deeper than just the moral aspect of letting robots take over our jobs and lives and dictate artistic expression to us.
The word is that A.I. could eventually displace 300 million full-time jobs. It’s all hypothetical, but the sad fact of future life is that we will be forced to adapt and accept A.I. for what it is.
There are indications that 60% of jobs will need significant adaptation, and then there is the theory that A.I. could create 11 million jobs while it displaces 9 million jobs elsewhere. It’s all very gloomy on the AI side of things, and the problem with AI-generated art is just a small part of a bigger problem. It’s not only a taste of things to come for painters, musicians, writers and graphic designers, but something that will affect all aspects of artistic expression.
Take writing, for instance – not this post, of course – but A.I. has swept through the written content industry like a cancer. Many writers believe that A.I. has hit a brick wall as it has obvious flaws, it doesn’t really know what it’s talking about; it’s inauthentic, and, at the end of the day, it’s just churred data that originally belongs to an abundance of other human writers. In some ways, it’s very close to plagiarism if you think about it.
The problem with AI-generated art.
The problem with AI-generated art is that it raises several major problems—some philosophical, some economic, some ethical, and some aesthetic. None of these issues has simple answers, but here are the most commonly discussed concerns.
One area that has raised debate recently is the issue of AI-generated art. The music industry has felt the need to protect artists, and the arena of AI-generated music is now a legal and government issue. What Will The Government’s Proposed Changes to the Rules on Copyright and Artificial Intelligence Mean for the UK Music Industry? – UK Music.
As we covered in a previous blog, there is an immediate denial that A.I.-generated material can be classed as real art because the work is not the physical manifestation of an artist. But as A.I. becomes a larger part of our lives, just as there is a place for drug-enhanced athletics – Enhanced Games – Wikipedia – AI-generated art will certainly take its place on the starting line.
Artificial intelligence has become increasingly prominent in a number of areas as AI plays an increasingly dominant role in our lives – The Pitfalls of AI Generated Art – Data Column | Institute for Advanced Analytics
We assume that, eventually, AI-generated art is likely to have its own classification, which is better for ‘real’ artists as it is predicted that, very soon, a third of professional artists and designers could sadly see their work replaced by A.I., with graphic designers feeling the biggest impact – Why Artists Hate AI Art: The Meaning Behind the Debate on Creativity and Control
The problem with ownership.
One problem that immediately springs to mind is the argument surrounding the owner of AI-generated art. Whereby, A.I. drags data to form words that loosely give A.I. the recognition of a storyteller, the selection of words originally game from a human mind.
Ownership is a central question in all this because it is the artist, the programmer, the dataset’s creators, or the user who writes the prompt?
Traditional ideas about creativity assume a conscious agent with intentions, but A.I. systems do not have intentions or self-awareness. This creates tension around whether A.I. can truly “create” or merely collaborate on a mixture of human ideas.
Most A.I. models are trained on massive datasets scraped from the internet, often without permission from original artists. If a piece of AI-generated art is very similar to that of a human version, this raises issues of Copyright infringement, lack of consent, and unfair use of intellectual property.
This legal ambiguity makes the field of AI-generated art unstable, to say the least. I mean, when A.I. generates something that is harmful or infringing, who is responsible for that? The whole legal fragility surrounding A.I. sounds like a bad mix of dark colours to me.
An artist could feel that their style has been absorbed into a system that now competes with them, without recognition, credit or compensation – because you can ask A.I. to do that, you know? – but who is culpable?
Take the style of Argentine artist, Fabian Perez, for example – https://en.wikipedia.org/wiki/Fabian_Perez . He has a ‘neo emotionalism’ style, one characterised by emotion and passion, with a dramatic narrative and a heady atmosphere. A programmer could easily replicate his work and create a unique version without the hours of detailed brush strokes.
I see this as a huge problem with AI-generated art. Because A.I. models can imitate a specific artist’s style – such as Fabian Perez – so accurately, it becomes indistinguishable from their work. This is therefore an ‘impersonation’, with the artist losing control over their identity.
AI-generated art, as with music, dilutes the artist’s distinctive aspect, whether it be a voice or a painting style.
It’s similar to forging signatures—legal in some contexts, harmful in others.
Economic displacement.
As we began, we arrive at the economic problem. A.I. art can be produced instantly, cheaply, and on a vast scale. This affects illustrators, concept artists, graphic designers, animators and photographers. We’ve all seen the AI-generated videos on YouTube, and, although we can’t deny the aesthetic beauty of these things and the realism, it hasn’t taken the producer a fraction of the time that it takes an editor working on a film with the real David Attenborough.
And, as already touched upon, the problem with AI-generated art is that the human artist is replaceable. AI appeals to companies because it’s faster than humans, cheaper, and it doesn’t have an opinion… yet.
This, of course, raises economic concerns and what the labour market will resemble for creative professionals in the future.
What we have seen in the writing industry is that AI-generated writing all looks the same. There are markers, traits, and a pantomime takes place when AI generates a story. It replicates words such as ‘crucial’, ‘tapestry’, ‘showcasing’, ‘consequently’, and ‘furthermore’ because it looks to use certain words to appease algorithms.
Consequently – see what I did there – large models learn from existing work, and they provide a predictable pattern. The problem with AI-generated art is the same as with anything generated by robots. Technology seems to have been predominantly programmed by unsociable people who spent their entire lives in a dimly lit room, only to emerge when sustenance was required. The programmer is likely to be a more analytical individual than an artistic one, and so the creativity wasn’t really flowing. It was just an awareness that there is a need for creativity in life.
And this is why analytical people have created systems to help with their lack of creativity? A debate for another time, perhaps.
AI-generated aesthetics are derivative, and therefore, a visual culture will become one of repetition; a world of sameness. Over time, there will be less artistic diversity and the acceptance of the same damn thing, over and over again, with critics arguing that A.I. pushes culture toward visual sameness.
The inherent bias is a problem with AI-generated art.
If trained in a certain way, A.I. will show favouritism to a specific style. It can’t remain objective on certain subjects if it is trained to adapt to a certain opinion, and so it will reinforce stereotypes, misrepresent cultures, and erase marginalised traditions. It could even favour western styles and eventually shape what is determined to be ‘normal’. We may be able to determine our own vision for the future, but is the future for AI-generated art?
A lot of us speak about the loss of human context. Kids can barely communicate in the same way these days, messages of text replacing a primitive verbal conversation, and a lack of human intention in art means that it is void of the lived experience and absent of an inner life that comes with being human.
Simply being human is what initiates the emotion, struggle and identity within a piece of art. Rap music was originally about a struggle and experiences within black culture. These days, it’s more about weapons and women, but how does AI interpret rap music? Does A.I. interpret Bram Stoker’s Dracula as a gothic horror story or one of love and grief? A robot certainly couldn’t feel the raw emotion that such an epic novel encapsulates.
The meaning we see in a painting, in music, and in writing, emerges from the human soul. It is our worldview, and A.I. cannot create or truly interpret what it means to be human, and it seriously raises doubts about the profound artistic depth of AI-generated art.
A convenient truth.
Finally, the problem with AI-generated art is our eventual over-reliance upon it. We like an easy life; one of quick content and fast results. Like a dishwasher, a tumble dryer, Amazon, and a camera on a phone, AI makes life easier.
My concern is that A.I. could discourage a slow, reflective process. If A.I. had produced this article, would it be as deep and meaningful? Would it really get to the human aspect of the problem, profoundly enough to confront itself?
Despite everything that we have said here, do those who appreciate art – every artistic element of our lives – want to replace human creativity with AI convenience? – https://katefletcherpaintings.co.uk/shop/