AI is no threat to poetry; we’ve already got it licked

What difference can AI make when the current rate of poetry production already exceeds our collective ability to respond to its existence?

Note: I’m going to start adding occasional short essays to this Substack as well as general updates. This one was prompted by a few recent encounters cross-pollinating in my brain.

Why do artists feel threatened by AI? Loss of income on the one hand; on the other, the fear that art as a medium of communication — as a testament to subjective human experience and the reach of the individual human imagination — will be replaced by art as mood lighting, as mechanism, as a grey soup of reiterated styles and trends. Remember the promo ad for Mark Zuckerberg’s Metaverse, where the tech bros interrupt their poker game to check out some ‘3D street art’? “Wow, this is stunning”, “That’s awesome”, “I love the movement”, they chime, before drifting onto the next thing — because in a capitalist utopia, the purpose of art is to inspire a warm, fuzzy feeling, either by representing some kind of accomplishment, or by recalling that which is comfortingly familiar.

AI poetry will be up to this task soon enough. But it needn’t bother — human beings have already produced more than enough to meet demand. Whereas in the case of visual art and prose fiction, AI can potentially fill a gap created by the exactitude of an audience’s desires — ‘a portrait of this person in the style of this artist’, ‘a story in this genre featuring a protagonist of my own design’ — poetry of any specific character is barely imagined. As Brian Phillips notes in ‘Poetry and the Problem of Taste’, it’s not that the reading public have poor taste in poetry, but rather that they have no taste at all — no sense, that is, of how their own personal preferences differ from anyone else’s. Poetry is poetry, the way table salt is table salt. You like a little, a lot, or none at all.

Relatedly, it’s become a lot easier for more people to write and publish poems. They’re short, and there are no rules left to break. You can study to be a better poet, of course, but no one is going to stop or even chide you for sharing whatever comes into your head and calling it finished, in which case AI really isn’t much of a time-saver. In the future, it might double or quadruple the scale of poetry production, but what does that matter when the current rate already exceeds our collective ability to respond to its existence?

I’ll just make this clear, in case this is your first time reading one of my commentaries: I don’t think too many people are writing poetry, and I don’t think that the low quality of some or much of what is written is an issue. My position is that the whole artform is diminished by the narrow way it is persistently framed, such that the threat AI poses to other artforms is already a present reality for poetry: that is, we have a landscape where everything looks like a mash-up of everything else, and most of it seems designed to serve its creator’s aspiration to be regarded as an artist, rather than having any clear communicative or explorative purpose. Note: ‘looks’ and ‘seems’. This is an issue of perspective, of fogginess.

For as long as I can remember, people have complained that all modern poetry is indistinguishable. And for as long as I can remember, the principle way critics have tried to separate the ‘good’ poetry from what they implicitly agree amounts to a rubbish heap is through insistent use of subjective epithets. In other words, in place of an ongoing exercise to document what distinctive characteristics may or may not be possessed by an individual poem, book or author (the appropriate answer to accusations of sameness) we have perpetuated a game of ‘squeaky wheel gets the grease’. The loudest, the most repetitive, the most passionate, fawning or grandiloquent claims are those that stick, and these on behalf of, inevitably, the better-connected, better-resourced, more shrewd and more well-behaved poets — though that point matters less than the fact that the qualities which are thereby attributed to them are vague, bland and frequently preposterous. Rather than teaching readers to discern and prize myriad specific attributes, and thus to tell one kind of poem from another by sight and feel, this process teaches them to think predominantly in terms of how ‘important’ a poet or poem seems to be, and to feel warm, fuzzy feelings that should on no account be interrogated further. It is one almighty confidence trick, at the expense of any sense that the new thing is much of a departure from the last thing. Gaze! Gasp! But do not look behind the curtain.

This in turn affects the way poems are produced and distributed:

• It incentivises (for both poet and publisher) high output with minimal editing, since only recently released work is regarded as sufficiently exciting to swoon over, and right-place, right-time has more to do with it than content.

• It incentivises broad, bombastic claims about the scope and purpose of a publication, lest it fail to speak to some common mood.

• It de-incentivises investigative reviews or cautious responses to a less visible work, since the only currency the reviewer may deal in is applause or heresy.

• It positions the reviewer, or critic, as someone lesser than the poet, someone who is merely affected and reports the effect, putting people off a role that is potentially vital in leading to the formation of individualised tastes.

Most frustratingly, for me at least, it steers what ought to be healthy debate about and around the artform toward a sluggish kind of territorial warfare. Disagreement over which poetry deserves what kind of attention is rife, as it should be. But trapped within the confines of a metanarrative that characterises poems as sources of fleeting, powerful feeling, too many interested parties end up huddled around their shared prejudices and faiths, failing to mount any argument beyond “Thing bad, other thing good” — albeit spun out across thousands of words. Others, wary of outbreaks of ugliness, stick resolutely to “Thing good”. Tower-of-Babel-style, we are not really talking to or understanding one another, except where we already see eye to eye.

I mean this, as ever, at a general level. There is good criticism and there are productive exchanges that lead to one or both parties being able to say, “I now see a little more of what you see”. But so much of what is supposedly the serious attention paid to poetry by its champions is barely more than gestural, tribal, phatic. Basic maintenance of the same rhythms and rituals of praise and complaint. As long as that continues we will struggle to shift the impression that the artform amounts to anything more than a piquant condiment which some consume in greater quantity than others.