So What Is Next? AI will Graduate From Your MFA Program

It’s Only Zeros and Ones

MFA programs are inadvertently designed for hostile takeover by Probability Engines, because at the lowest level they teach dogma and Rules. [Of course you’ve heard “Be creative. Learn the rules and then break them judiciously.”]

Probability Engines love rules. They reverse-INVENT the rules from the thousands and millions of texts upon which they have trained. They need the guardrails, the connect-the-dots, the predictability. More precisely, they invent rules. They will predicate structure using the Rules, build character using the Rules, repeat three times using the 3-Rule (just like this sentence). They are going to use Rules more rigorously than you ever will.

Consider some of the obvious Rules to hand (I add in some famous exceptions just to amuse myself):

Don’t start in monologue or dialogue (the Forget-Ishmael Rule).

Don’t start with broad-brush generalities with no human present (the Forget-Dickens-Worst-Of-Times rule).

Start in tight focus and then pan back to show your landscape / theme / milieu. Start zoomed out and then tighten the shot until we meet our protagonist.

Establish the normal and then disrupt it.

Don’t use dialect (the Forget-Brooks-and-Everett rule).

Reveal the detail as the reader needs it (the No-Data-Dumps rule).

Slaughter your antagonists but don’t damage livestock or pets (the Save-the-Cat rule).

Don’t start in flashback (the No-One-Hundred-Years-Of-Solitude Rule).

Your protagonist should be likeable or become likeable (The Forget-Kafka rule).

Come to a resolution – the protagonist makes that giant personal decision, and either fails or succeeds (the Anti-Julian-Barnes rule).

Don’t forget that Probability Engines don’t have to map the exceptions to the rules, just the preponderance of books that are following the rules.

[As an aside, ff course, your writer’s circle / class roundtable / Ted Talk don’t define these as rules, rather they are “guidelines.” Right, guidelines. There’s going to be someone at the table or in the Zoom who either says “you broke the rule” or “what a clever subversion of the rule. This is code-switch for ‘I got my MFA ha-ha.’ ”]

Now, as a thoughtful reader or writer, you’re going to say that the Probability Engines aren’t going to have enough material to eat up in order to become MFA graduates. I challenge this view—Probability Engines don’t have AS MUCH to work from as they do using the Amazon B Novel List. Yet they have plenty. I’m reading enough columns by writers / readers / editors off Substack to eat up half a morning every day. FaceBook holds enough writers circles to bewilder mere mortals. Multiple literary organizations are conducting online MFA programs and short courses. Anything visible to you on the Internet is visible to the Probability Engines, even if their masters aren’t cheating and sucking up private material.

This installment of “What Comes Next,” predicated where and how Probability Engines can put more pressure on human writers. I offer you a shot to write and read into a space that remains foreign to AI, and we’ll keep nattering on about this and fielding your comments.

In my next couple of installments on this theme, I’d like to make a stab at where AI’s weak zone. I might call it. “Beautiful Writing,” and offer several examples in several voices. Or I might decide I’m wrong and we’ll let Gemini write something.

The last episode might be a description of my final edit process, which is a five to six week struggle/romance to make my own language beautiful. Because everyone is (not) interested in the individual writer’s process.