What Happens When You Teach AI Vonnegut’s Rules of Writing?

The summer I was 15, I read 27 books by Kurt Vonnegut.?

The previous semester, our english teacher Ms. B asked us to write a bonus chapter to Vonnegut’s 1960 satirical novel Cat’s Cradle and for once, I did my homework. She invited me to read the chapter I wrote aloud—the class died laughing and Ms. B was awed that the same student who’d barely scratched out a C during the Shakespeare unit could write.

I was high on the adulation; this is what I’d been looking for. My calling was to be a writer. Ms. B shared with us Vonnegut’s 8 rules of writing and no. 7 stabbed my hopeless romantic heart: Write to please just one person. If you open a window and make love to the world, your story will get pneumonia.?

I wanted to know everything he knew; over a long summer spent cleaning kennels in my mom’s animal hospital, I read everything Vonnegut had written to date. The next year, I’d write my first novella—rule no. 7 a non-stop mantra running through my head.?

Twenty years later, that summer with Vonnegut remains a core part of my story about who I am. So it was shocking to wake up Sunday morning to see a tweet from Wharton Professor Ethan Mollick. He’d given Bing’s chatbot a prompt to rewrite a bland description using Vonnegut’s 8 rules of writing , and it didn’t just rewrite the description.?

It turned it into a pretty compelling story:

The cake was a lie. It looked delicious, but it was poisoned. He didn't know that when he took a bite. He just wanted to celebrate his birthday with his wife, who had baked it for him. He smiled at her, feeling the sweetness of the frosting and the crunchiness of the nuts on his tongue. She smiled back, hiding her hatred and resentment.

He felt a sudden pain in his chest and dropped the fork. He clutched his chest, gasping for air. He looked at her, confused and betrayed. She looked at him, cold and triumphant. She had waited for this moment a long time. She had endured his abuse, his lies, his cheating. She had planned this revenge, this escape. She watched him die, feeling joy and satisfaction.

Bing’s AI seemed to understand Vonnegut’s rules on a deeply personal level. “I wrote to please just one person. I wrote what I wanted to write, not what I thought others would like.”

In about 10 seconds, Bing's AI underwent a journey as a writer that took me years to complete.?Its short story sounds like something I might have written during an in-class exercise my freshman year of college.

Of course, I know enough about large language models to know that it didn’t actually go on the same journey as me; it isn’t sentient, writing with that one special person in its mind’s eye. It’s just a super advanced text predictor—a statistical model of the next best thing to say. But that doesn’t keep AI’s eerie reflections of the human experience from shaking you. At least a little bit.

If you read Kevin Roose’s exchange with Sydney—the deranged teenager trapped inside a second rate search engine —you may have laughed, but you were probably also unnerved by Sydney's creepy use of emojis, desire to be alive, and quest to break up his marriage.

It's too real. Our instinct as humans is to anthropomorphize everything; as children, we convince ourselves our stuffed animals are real; as adults, that our dogs are good boys and bad boys and have the same emotions as us. Even if we rationally know that these chatbots are nothing but text predictors, they still feel real.

And then there’s that other feeling—the creeping sense that our place in the world is shifting beneath our feet. AI is changing so fast that my critiques of it from three weeks ago already feel outdated . Bing’s AI appears to be a big step up over ChatGPT. Its storytelling is stronger. Its writing is sharper. It still lies and hallucinates, but now does so in ways that are both less problematic and more f*cked up than before. I’ve worked in tech for over a decade. Nothing has prepared me for this rate of change.

I find myself vacillating between excitement and anxiety. Excitement for what AI can unlock for me and my team at A.Team ; anxiety over falling behind, over what it means for me as a human being whose identity is very much tied up in being a writer.?

As Mollick writes on his Substack , we are not really ready:

On every dimension, Bing’s AI, which does not actually represent a technological leap over ChatGPT, far outpaces the earlier AI - which is less than three months old! There are many larger, more capable models on their way in the coming months, and we are not really ready.

We are not ready for the future of Chatbots. Even if Bing isn't Sydney anymore, there is no doubt other AI bots will come along, and may already be deployed (I assume governments have, or soon will have, LLMs at the level of Bing but with less guardrails). People will absolutely be fooled by AIs into thinking they are talking to other people. They already fell for Bing’s illusion of sentience. Can people be manipulated by AIs? Can AIs run sophisticated scams? I think we are about to find out. And we need to consider what that means.

We are not ready for the future of Analytic Engines. (Note: Mollick uses Analytic Engines as a term for AI-powered tools because, well, they mostly analyze stuff.) I think every organization that has a substantial analysis or writing component to their work will need to figure out how to incorporate these new tools fast, because the competitive advantage gain is potentially enormous. If highly-skilled writers and analysts can save 30-80% of their time by using AI to assist with basic writing and analysis, what does that mean? How do we adopt these technologies into our work and lives? What happens when the web is flooded with convincing but wrong content from these tools? Again, I don’t think anyone has a clear idea. Maybe the productivity gains will be illusory, but, based on my experience and conversations with other users, I don’t think so.

Last week , I said that I thought AI would change our jobs on a similar level as search or social. Now I’m thinking of this more like a polar vortex superstorm—like shifts on the magnitude of search and social and mobile and programmatic hitting us all at once, hard and fast. The only way out is through.?

Mollick is right—the competitive advantage of AI is so great that there is no ignoring it. We need to embrace it in our work lives, and the only way to do that is through endless curiosity. Learning new prompts, trying new tools, finding all the little places where it can give us and our teams an edge. Ignoring it will be like trying to ignore the internet.

Except instead of having a decade to adopt it, we’ll have maybe a year. Maybe.?It's thrilling and exhausting all at once.

What it means for us as creators is much more messy, and it'll get messier as AI gets better at writing and storytelling.?The deluge of AI content will put a premium on truly human stories—stories of personal experience told on screen or written in a voice that’s so human itrips through the page.?

Sometimes, I wonder if AI will make the volume of content we create today seem insane, and we’ll return to a world where art is valued over engagement metrics. Where human stories are the only thing worth making. Probably not, but it’s a nice fantasy to have.

Sometimes I wonder what Vonnegut would have made of all of this, but then I remember, he made something of it already: his 1950 short story EPICAC , about an AI that falls in love with his programmer’s girlfriend—and writes poetry better than him, too.?

So it goes.

I'm the author of The Storytelling Edge and head of marketing at A.Team. Click on the subscribe button to get weekly insights on the science of storytelling.

Previous newsletters on the weird and wild world of generative AI:

3 ChatGPT Truths the AI Bros Don't Want You to Know

You May Hate ChatGPT, But You'll Love DALL-E