With Google and a host of other tech companies and academics engineering Artificial Intelligence (AI) to produce artistic works, many quarters are starting to question the very nature of this up-till-now uniquely human activity. But although aesthetically pleasing, can AI art ever truly touch us the way great art does?
AI art is here – or at least something that may approach it is.
At one end, there’s Google’s trippy Inceptionism art that sold big at a San Francisco auction last February.
At the other, there’s the new and hugely popular Prism app that uses AI to break down a photo into 100,000+ bits of data and reconstruct them in seconds via a particular artistic filter. That beach snapshot can now be recreated with the style and hues of Munch’s Scream or Hokusai’s The Great Wave off Kanagawa.
Both these examples fit into a paradigm that says artists produce work and it is the appreciation of its enriched consumers that transforms it into art.
These two types of AI art also tick the box of what the content of art is – at its most basic some kind of expression of an aspect of our world. Thus, a sports shoe normally used to kick a soccer ball, but placed on a stand in an art gallery with a spotlight trained on it and used as a vehicle to say something about our culture with people appreciatively looking at it, can be considered as a piece of art.
This is what’s so compelling about Tracy Emin’s My Bed – a self-portrait of despair and addiction that uses as a vehicle a dirty unmade bed surrounded by empty vodka bottles and used condoms.
Emin’s work also highlights how these examples of AI art are not art – or at least not great art.
The idea of the artist as a tortured soul toiling away in a garret in a bid to communicate something crucially important to the world may be a cliché, but it does highlight this up-till-now visceral human need to communicate some self-expression about the human condition.
So maybe real AI art will only come about when it creatively – and independently – says something about the AI experience from an AI point of view for AI consumers.
Moreover, this would move AI’s ability to learn from personal experience to one where one AI entity can learn from the creative self-expression of another and in some way will be enriched by this.
We don’t have a clear picture about what this scenario might look like, much less what a creative expression of the AI experience. Maybe it will be a string of AI generated code that accurately describes an aspect of the AI experience and is put together in a novel, possibly poetic way.
The crucial “but” however is, that this poetry will be AI poetry. We may be able to appreciate it and even understand it, but this will only ever be academic and never based on personal experience because we are not AI.
That said, maybe the final frontier of real AI art – namely great AI art – will in fact be two-fold: on the one hand, an AI artistic self-expression about the AI experience that breaches this gap, and on the other a genuine and enriching AI appreciation of great human art.