10 Things (I Think) I Think About the End of the World as We Know It (and I Feel Fine)

Untested Hypothesis #1: Reading is the new writing.

Factoid: I teach writing to teenagers (and others).  

The school where I teach is competitively admitted. In the wake of the most recent “audition” process, where applicants visit campus to complete in-person writing prompts and to be interviewed, I did an experiment (the only hypothesis I actually tested, I guess, so it’s fitting that I start there):

I plugged the writing prompts the applicants were completing into ChatGPT. It generated this story and this poem. I did nothing but change the font and add a title to the story. I had to cut a couple of lines from the end of the poem (they were too tidy, too sentimental). But the lesson was clear: generative AI writes like a smart, talented high school student. That is to say, it writes like someone who doesn’t have a ton of life experience but does have really good language skills and has read a lot. Like a dedicated young teenage writer, it’s only going to get better. And quickly.

As a writer and a writing teacher, this seems to pose an existential threat. But I’m not sure it does. I do think it places a premium on reading like a writer. I also think it will have the most impact on the early phases of the writing process, particularly as the stigma of using generative AI wears off over time.

The key word is “generative.” Fact is, writing is hard. Most people hate it. And the hardest (most hated) part (arguably) is generating material. Where that process is time-sensitive and where there is a financial incentive, generative AI is an obvious aid people will gladly use (and they already are).

The people who are going to use generative AI to most successfully produce creative content are those who can (A) prompt it well, and (B) take a collaborative, “developmental editing” approach to what the AI produces. With text, that means reading like a writer.

How do you do that? I’ve heard this skill most simply described as the ability to simultaneously read with and against the grain of a piece of writing.

• Reading with the grain: What does the writing want to accomplish? What is it doing well? Can it do more of that?
• Reading against the grain: What could this writing try to do (but it isn’t doing it right now)? What is it not doing well? Can it do less of that?      

There’s also a very cool form of rabbinical exegesis called PARDES.¹⁵ I won’t go into it at length here, but I probably will at some point in the future. Suffice it to say, it’s an approach to reading sacred texts that I think can be applied to any text. It articulates modes of reading that process through the surface level toward deeper, more numinous levels of understanding that require increasing degrees of creative engagement in the reader — such that, in the end, the writer and the reader have a symbiotic (collaborative) relationship in the creation of meaning.

People who can read like that? They’re future-proof.

NEXT: Untested Hypothesis #2: Genius isn’t genius (anymore?).