AI Will Write the World’s Greatest Novel and it Won’t Matter

The duty for writers to write is heavier than ever.

I'm going to admit something embarrassing to you.

About a year ago, when AI really hit the scene, I spent a few nights laying awake, heart pounding. 

Don’t punish me for my candor: I seriously felt my reality coming to an end. On the surface, my writing might become as appendicular as a horse after the automobile. Deeper than that, though, would there be anything of humanity left if we made Borg brains?

I suppose, in theory, I believed in a unique “essence” to us. But when the tech bros unveiled machines that could understand a gesture as subtle as finger guns, I was genuinely afraid they might prove me wrong. 

Already seems funny, doesn’t it? I told you it was embarrassing. Of course, AI could still turn Skynet in the next 10 years or whatever. But… the vibe has shifted, no? Just you wait, they say, only two decades until AI has a soul. Literally since at least 1965 have we been claiming that. Who knows, maybe that will be accidentally correct one of these days. 

More importantly, I think, by casting a conspicuous shadow, AI illuminated what is primary to humanity. Even when we make very, very fancy tools, there still remains some ineffably human residue. 

So far, AI is good enough for some mundane tasks, but too uncouth for just about anything else. It’s like an alien monkey at the dinner table. It’s causing a scene, but we’re all… getting used to it.

People still have fear around AI, of course. But the overall distaste has shifted from existential terror to what we’re calling “enshittification.” As the outputs invade the internet, they become inputs for the AI; in a strange loop, a haze of hallucination will descend over everything online. 

Erik Hoel covered this a long time ago in, "The Banality of ChatGPT,” and a handful of other essays. Since the data that we feed it is inherently cheap to produce, all we're going to get are outputs that were already cheap to begin with. So, AI doesn’t seem particularly poised to be an economy-buster in the near future – soul or no-soul. This seems to be borne out with the market refusing to rise to the advent of AI, despite the hype. 

In my work (surprise, I work in tech), what I see are a lot of vertical SaaS integrations. They’ll roll out in the next five to ten years and increase the efficiency of bureaucratic systems by some-odd percent. That is, um, marginally good. Things are a lot more boring than I feared.

One unintended benefit is we're now asking ourselves what exactly consciousness does. 

Synced with other societal tides, as Ted Gioia pointed out (I can’t find the piece), a second era of romanticism seems to be arriving. People have a renewed interest in poetry, fiction, and myth. 

Remember just a couple of years ago? It was socially acceptable to say, if you build enough bits and bots to replicate our squirting chemicals and firing neurons, you would have effectively created a human mind. At least in my circles, that idea is a bit gauche nowadays.

One reason for the shift in sentiment is GPT’s demonically bland writing style. But, that’s just because it’s general-use. You could easily train it to write more like a specific person and it would do reasonably well. I don’t think “bland writing” gets to the heart of it. The problem is not the outputs. The problem is that we don't care about the output beyond a certain point. Maybe we never will.

Computers long ago surpassed humans in chess-playing ability. But we don't sit around watching two galaxy-brain robots play the cosmos’ most epic game of chess. Nope, we still prefer to watch Magnus Carlsen. In fact, the only thing that pays for those machines’ electricity is other, more human endeavors.

I predict in the next five years, AI will write the greatest novel ever written, and no one will give a shit. 

We aren't interested in the technicalities of a great novel. We're interested in the human lives that bring a great novel into being. Now that AI has been looming in the corner long enough to stop scaring us half to death, we can see that more clearly. 

I played with AI music a few months ago. I was able to make a couple of tunes that were entertaining (funny, really). But my interest very quickly faded. Normally, if I like a piece of music, I think about who the artist is and what they’re trying to tell me. For AI music, I don't even know where to begin to tell if it is “good.” What do I care to find out?

Really, AI is a stress-test of our materialist presuppositions. If we're really procedural to our core – clockwork down to our soul, then why the hell am I not watching chess robots while listening to hyper-Beethoven?

Here’s an alternative theory: everything is perception. Perception based on deep, embodied biological imperatives, 99% of which are completely invisible to our consciousness. We “booted up” our modern consciousness via a long lineage of “psycho-technologies” pulled from the ether by great philosophers and artists. They all built upon the previously established canon of “grammar” to make something new, yet understandable. 

Monet's work, to take one example, literally changed collective perception. Not only did it plainly delight the senses, but it allowed humanity to see the world in a new way. Both because of the new images, but also of the perceptual framework. It was a movement toward seeing the world more abstractly. It was shocking. Not so much for us anymore, because we're so used to it, but for the people that were experiencing it for the first time, it was amazing (enraging to some). That mode of seeing was just not accessible before someone thought of it. 

“A Woman Reading” 1872. The woman in the work is believed to be Camille Doncieux, Monet's wife.

That’s just one example. There's a grand lineage of a million psycho-technologies like that, some ground shaking and some minor. They make up our consciousness (your consciousness!), most of which we don't notice.

When you read a novel, to the extent that you’re interested, you’re feeling the character’s embodied experience. Not just the character, but the author, too – and how they fit into the landscape of meaning of our entire culture. That’s all nested in archetypal themes of death and birth and mothers and resurrection – all those things that Jung pointed out a century ago. You know, that guy we halfway understood? Now, everybody has to contend with him, because finally, the wheels are coming off our materialist motorcycle.

AI makes all this so painfully obvious, it could be the tipping point for a new romanticism. Then, we might turn around and use AI to channel that spirit. Or, I don’t know, maybe I’m being naive.

I use AI to transcribe my voice and write first drafts. It is amazingly helpful. It removes the tedium of typing – trying to transpose inspiration through the bottleneck of my fingers. I get more bandwidth when I stand up, walk, and make noises with my larynx. I am able to get it out before it slips away, or becomes false, or distracted by what somebody might think. 

That’s not to say that I spend less time on my writing. It's probably more. Somewhere between 60 and 80% of what I say in the initial draft gets completely reworked and the original language is gone. But the spirit is better preserved. I'm able to move it around to make it coherent with editing. That’s higher leverage than trying to draft and edit at the same time, which is like trying to do algebra while playing the piano.

Combining all the factors I mentioned above could mean we are staggering out of a dark age of creativity, blooming into a new renaissance. If so, we will invent new ways of seeing the world – like Monet – that we cannot conceive of until someone does it. Those great eras usually come in bursts between periods of stagnation. The next one could very well be soon.

That’s the opportunity that lies in front of me and you: the right to shape the consciousness of the future of mankind, in whatever small way we might manage. So to the writers and the artists, I say, even if AI creates the most technically perfect version of your craft, it won’t matter. 

In fact, its strange failure might shine lights on new pathways to higher layers of being, previously unimaginable. Then, we can use the tools to more accurately share those new fragments of reality, which have yet never been successfully pulled from the ether. 

Artists, we still have a great, noble task to do. Don’t let AI doomerism infect you.