Why AI will never be able to create art

A long-standing argument and debate about artificial intelligence is its ability, along with, the inability thereof, to create art.

Though some may view otherwise, the sketches created by AI should not be considered as real art.

The Definition of Art

Through time and time again, from the dawn of humanity, man has created art. Whether if that was from the aboriginal rock paintings in Australia originating 65,000 or 80,000 years ago, or in modern forms of art even beyond canvases such as rocks or paper.

Though in recent years, some view that modern art had decreased the standard of art, such as the recent exhibition of, quite literally, a banana taped to a wall. No matter how ridiculous it may seem, it is still a form of art.

The reason of this is, despite the method or the means of which art is contrived or delivered, is that art has meaning. Every piece of human art, no matter if it is drawn by a child in primary school, or a professional artist, making a living off of art, they all have meaning, it is revocable, inevitable and undebatable.

Art, is also a form of creation, it creates and delivers these messages through the exhaustion of the artists, which trains the artists in the process, their skills, resilience and all other abilities related to the creation of art.

It takes time, and requires skill to create art, and art has meaning.

AI Cannot Understand Meaning

To its core, artificial intelligence is mathematics, through the calculations of matrices, vectors and other complicated computer science gibberish. It predicts what the next pixel is, and guesses what it should say next.

It does not have a mind, as much as some AI-fanatics pretend it does, it is cold, has no flesh, and it cannot think or understand.

This is shown through a study conducted by Apple [1], it had found that AI is incapable of reasoning, it had found that artificial intelligence “face a complete accuracy collapse beyond certain complexities”.

If AI cannot reason, or actually think, then how can we trust it to understand the true meaning behind human art? If we cannot trust AI to understand the true meaning of art, then how can we trust it to create meaning in art? And, to that extent, to create art as defined above?

Intelligence != Understanding

Sure, one can argue that the human brains also have limits, and also cannot understand beyond certain complexities, but the difference is all humans, no matter where you come from, or what your background is, have emotions, have feelings and have common experiences.

Humans have emotions, and that is what sets humans apart from computers, and AI. When an AI model looks at a piece of art about connection with one’s mother, it calculates the meaning based off of the pieces of real human art it had profiteered upon, and outputs a line of text, outlining the emotions of what a human might feel.

However, when a real human being sees a piece of art about one’s mother, they can feel, they can connect, they can understand through their own experiences with their mother, and understand something through their own thought process.

Human brains, human understandings, human thoughts grow.

AI Steals, not Create

It is a objective fact that all AI models are trained upon data. And, where does this data come from? They come from real human artists, and often without consent or payment.

In fact, the Australian federal government has even recently banned [2] this practice, if this was not happening before the ban, then why would they ban it?

Artificial intelligence ruthlessly and brutally crawls artists’ Twitter posts, through their blogs and wherever they post online. Some artists were even forced to add watermarks or to use specialised programs to prevent this exact behaviour.

It is unbelievable, concerning and dangerous that the creators of real human art, a practice which as mentioned earlier, originated from more than 80 thousand years ago, way before electricity or even any modern commodity was invented, have to go under, effectively, hiding, and having to metaphorically “dodge the AI bullet”.

Artists have a sense of community, they have a sense of unity, they help each other and compliment each other, not destroy, steal and profit off of one another.

Thus, AI is not “an artist” as it lacks the simple understanding of human unity and community, it is designed for one purpose, and that is to help its original creators make money, and it is quite frankly disgusting and, as it is in Australia, heavily regulated.

AI “Art” is Fruitless

As mentioned earlier,

It takes time, and requires skill to create art, and art has meaning.

Sure, people can argue, and they do argue, that AI has made art more accessible. But, is that necessarily a good thing?

Creating pieces of art is rewarding and useful in the development of the human mind because of the skills you build along the way. When you learn to draw, you learn to manage time, and how to, literally create pieces of art.

It is often the inaccessibility of the results of art that makes art a form of creation unlike any other.

It is also rewarding because you are visibly able to see the progress you have made along your create journey.

AI “art” takes both of these fruits of art away from humanity. When all it takes is a prompt, a single line of littered words, is to create a piece of art, where is the growth? Where is the learning experiences?

That picture file you receive at the end of a GenAI creation is not a reflection of your skills, nor is it a reflection of your growth. Art is fruitful, not because of its results, but because of the process.

And without this process, one cannot call art “art”.

AI Fosters a Dangerous Society

We live in an age where everything has to be fast. Whether that is food deliveries, or E-mails, or assignments, we wants things done, quickly.

That is a trap, of which I personally falls into, too frequently as well. We are a generation of social media and short-form videos which had already shown its impact on our attention spans.

And, GenAI “art” does not help, it adds fuel to the fire, further accelerating this understanding in society that:

Effort Spent != Results Received

Though one may argue that this principle is true, one may not be successful despite the efforts they have given, but it is undebatable that this mindset is toxic and should not be indoctrinated into our young people’s minds.

We should foster a society of “creators” and not a society of “receivers”.

AI Accelerates Destruction

So far, as mentioned in the previous argument, AI damages humanity metaphorically and mentally. However, AI “art” also has its negative impact environmentally.

When you ask AI to put together a senseless and meaningless sketch of what can only be considered as gibberish, it takes a heavy toll on the environment.

In an Australian federal inquiry conducted by the Royal Melbourne Institute of Technology [3], Dr. Gordon presented the fact that “producing an image” via AI “uses about 20 times” more energy than a GenAI text inquiry.

Which by the way, uses “somewhere between 10 to 90 times more energy per inquiry” than a Google search. That means, when you ask an AI model to generate a silly picture, it uses 200 to 1800 times more energy per inquiry than a Google search.

The same study also found that these GenAI models also use between “1~3% of the world’s energy”, and that it is “set to grow”.

When the numbers and facts are presented, it is without doubt that it is shocking and concerning at the same time. Without doubt, AI “art” has taken a giant toll on the environment, and a giant dump on the efforts that literally every contributing member of society had made to the coping and the reversal of climate change.

Personally, I have never seen a real human artist consume 200 to 1800 times more energy than a Google search just to create a real piece of art, which, in its core have more value than the trash AI can produce anyway.

Conclusion

As argued above, it is difficult to conclude that AI-generated “art” is really art, and not only that, but it is also a plague, and a tumour which damages artists, our society, and humanity as a whole.

Footnotes

1. Shojaee, P., Mirzadeh, I., Alizadeh, K., Horton, M., Bengio, S. and Farajtabar, M. (2025). The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity. [online] arXiv.org. Available at: https://arxiv.org/abs/2506.06941.
2. Maani Truu (2025). Federal government rules out changing copyright law to give AI companies free rein. [online] Abc.net.au. Available at: https://www.abc.net.au/news/2025-10-27/labor-rules-out-ai-training-copyright-exceptions/105935740/.
3. Rmit.edu.au. (2024). Federal Inquiry hears warning on environmental impacts of AI. [online] Available at: https://www.rmit.edu.au/news/media-releases-and-expert-comments/2024/sep/environmental-impacts-of-ai.