It is clear that AI is having an impact on the arts, including poetry and photography.
Despite being language models, the current crop of AIs are pretty shite at the greatest linguistic art, poetry. Now, to be fair to them, I haven’t done a survey, I’ve just read some examples. I think their basic problem is that these models are based on content hoovered from the internet, and the average quality of poetry on the internet is … erm … questionable. Yes, I am being a complete hypocrite here, given I’ve posted quite a lot of poetry on the internet.
But it’s more than that. Good poetry, IMHO, comes not just from the conscious mind, but from the full gestalt of the poet. Speaking personally, when I write a poem, what comes out is rarely what I thought I was going to write when I sat down with virtual pen in hand. Were I to sit down & write a poem, and force myself to write what I expected to write when I sat down, I would write utter shite. Instead, when all parts of my being engage in the poem, including those parts of which I am not normally consciously aware, then I can sometimes produce a decent poem. It’s not just the underlying themes, motivations, and other factors which underlie whatever it is I thought I wanted to write about, it’s everything.
A specialist AI simply doesn’t have this mental structure. It doesn’t have a rich cross section of bizarre underlying emotions to draw upon. So it can’t write poetry with any real depth. Or, at least, to be more accurate, current AIs can’t do that. General AIs, when they come along, may be able to source that rich cross section of information. However, I suspect that what they write will be quite different from what people write, because, I hope, their rich cross section of bizarre underlying psychological material will be underneath a quite different structure of complex mind, their structure. I wrote a longish poem on that back in the 1980s, published in anticipating the metaverse, published by the Knives Forks and Spoons Press in 2014.
Photography, on the other hand, is quite a different art. The rich subconscious content in the visual arts is more reserved for painting than photography. Photography captures the real world, supposédly, although it must be said the photographer choses what part of the real world to catch, and that clearly has, or can have, artistic content. (My photography is extreme in that respect.) All the same, specialist AI can easily fake photographs, and the faked images are fine. Yes, there are sometimes flawed details (such as hands), but those flaws will vanish as AI gets better.
Photography will continue to be used to record the world, although I suspect the great ease with which AI can create fake photographs supposedly recording real events will mean that, sooner rather than later, some kind of certification scheme will be required to verify genuine photos from the fakes. Still, given the long history of creating fake photographs, I don’t think such a scheme would ever be fully clean.
So how do I feel about all of this as an artist? Does the AI represent new competition? Well, potentially, of course it does. Does it represent a source of new forms of art? Well, as AI, in its own right, of course it’s a new source, but I doubt it will produce a new form of the art just yet (I could be wrong, but I think that will require general AI). Specialist AI consumes the internet, & what it produces depends on what it finds there. It doesn’t create, it translates (is what people do any different?). But that’s now; in the longer term, it can only get better and fresher, and no doubt eventually some new art movement will be created by AI.
On one level I fear the introduction of new competition, but on another level I anticipate even more fresh and exciting ideas. Ultimately, though, I don’t make art for the accolade (although let me assure you that’s rather nice), but because it’s a human need. AI won’t change that need. I’ll still need to write, I’ll still need to recite, I’ll still need the praise! Work out for yourself why I see the human need to produce art as an evolutionary thing.
AI could even help develop new human arts. For example, nowadays it takes a great deal of resources to become a movie director. You need actors, you need sets, you essentially need a studio. But if the impact that AI is starting to have on photography can applied to video too, then those actors and those sets can be created by AI. Being a director would change from depending on studios to depending on software. Your future video sequencer won’t just allow you to edit the video you’ve made elsewhere, it’ll allow you to create that video. There’s not enough money today to allow someone in a bedroom to produce a blockbuster of an Iain M Banks novel, but there will be when AI does its business, and I’ll be utterly delighted to finally see his vision in vision.
There are some arts that will only be mildly changed by AI. Dance, for example, will still remain dance. AI might enhance choreography, but I don’t see it replacing choreographers: that depends too much on experience of dance. I think the same applies to other forms of performance art, although I suspect AI may make inroads into writing plays.