Teaching thinking; author platform building; Fifth Third Park: Newsletter 23 September 2026

Newsletter 169. Now that AI can write, we must change how we inspire students. Plus, the disaster of corporate slop grenades, editorial cheat sheets, three people to follow and three books to read.

Teach writing as thinking

In school, we’ve been teaching writing as a series of rules. Here is how to spell; here is how write a sentence; here is how to write a five-paragraph theme.

The emphasis has been on the mechanics of putting words together. We have always spent too much time on this, but it had two benefits. First, it trained young people into a basic level of writing competence. And second, it was easy for teachers to grade.

It has been the wrong emphasis all along, but it’s even more pointless now in the age of AI.

Any AI tool can write flawless sentences and craft a perfectly serviceable five paragraph theme. This gives writing teachers fits, because now they think their job is to police students and keep them from using AI to do their assignments. We tell students, “Your job is to do what a machine can do, but you can’t use the machine.” They think that’s stupid, and of course they’re right.

Take a step back. If we start from first principles, what are we really trying to teach?

I actually had this opportunity when I taught a writing class for homeschooled teenagers. Unlike school students, they were untrained in the “standard” ways of writing. They were creative, inventive, undisciplined, and eager.

I did not teach them rules. I taught them about ideas. We did three things a lot:

1. We read entertaining and clever writing from writers like Mary Roach — partly because it was fun, and partly because it encouraged them to figure out what these writers did to be so effective.
2. We talked about their ideas. Which ideas were unique, and which were commonplace? How do you logically support an idea? What is evidence? If someone read about your idea, how would you hope they would change?
3. We critiqued their writing. There was no time spent on “Here is what you did wrong.” The emphasis was on “Here’s what you did — what worked, what didn’t, and how could it be more effective?” More than half of the commentary came from the other members of the class.

Now, imagine if we could teach everyone to write this way.

There is certainly a need for students to understand the rules of language. Just as we teach students to do arithmetic, even though a calculator can do it better, we need to teach students the logic of language.

But the emphasis in writing should be on ideas, argument, and technique, not on following rules.

Students can brainstorm ideas with AI on their own, and with other students in the class. (And believe me, they’ll enjoy this a lot more than an emphasis on rote assignments.)

AI can help critique drafts students come up with.

Students can compare their own efforts to AI-generated text — identifying what AI does well, and what they, as humans, do better than any machine.

Look at this way. In the 1960s, students learned to write essays longhand, because that kind of writing is how we thought they would need to write as adults.

In the 2000s, students learned to write with a word processor, because that kind of writing is how we thought they would need to write as adults. We didn’t stop them from using spell checkers, because in the real world they would certainly have access to a spell checker.

In the 2020s, students are learning to write with the help of AI. A beautiful draft in which all the ideas and text are AI-generated is a failure. But a draft in which the student has used AI to improve their original work is a success.

What we are teaching is, essentially, how to think, how to translate your thinking into writing, and how to use modern tools to do all of that better.

AI, like word processors and spell checkers before it, can release students from drudgery. The point must not be to force the back into that drudgery and police their attempts to escape it. The point is to nurture their creativity, logical thinking, and ability to benefit from criticism by teachers, fellow students, and AI tools.

Too many of the workers in the current generation — workers who were taught that writing was an arduous task — are mindlessly using AI writing to substitute for thinking at work. The next generation of writers must be trained for a world where AI is available. They can become more powerful writers and thinkers than was ever before possible — provided we use all available tools to nurture their untamed creative instincts.

News for writers and others who think

Jonathan Goodman writes about the unconventional way he is building a social media “machine” to promote his book, using a very different set of skills from writing the book in the first place.

In the ultimate irony, Anchit Sathi‘s essay about “AI and the Crisis of the Author” was removed from the Times Literary Supplement site because they decided it was written with AI.

Editor Erin Brenner has created a useful library of editorial cheat sheets — and is giving it all away for free. That’s a generous and brilliant use of content marketing.

Shopify CEO Tobias Lütke asked his staff to use AI to write more and better. The result is drowning his company in slop. “We call those ‘slop grenades’ that people toss at each other,” he said. “That’s definitely a bad thing.”

The Detroit Tigers renamed their ballpark from Comerica Park to “Fifth Third Park,” after Fifth Third, the bank that paid for naming rights. The new name is weird and hard to say. I pity the announcers. For every three people who become aware of Fifth Third Bank due to the new branding, there will be at least five people who don’t buy tickets because the name has bamboozled them. Ew. Just ew.

Calling all writers: There are 800 responses so far to this year’s “AI and the Writing Profession” survey and editors outnumber writers! This cannot stand. Register your opinion (15 minutes) and I promise, I’ll reveal the rapidly emerging trends in a public forum as soon as they’re available. Please share this link with your writer friends: https://www.surveymonkey.com/r/AIWPbern

Three people to follow

Lisa Wong , thoughtful editor

Leslie Owens , superconnector at MIT’s entrepreneurship hub

Mark Hillary 🤖 , podcaster, ghostwriter, thinker

Three books to read

Data Are Made, Not Found: A Story of Politics, Power, and the Civil Servants Who Saved the US Census by danah boyd (University of Chicago Press, 2026). The fight to find the truth about the people of America.

Cyberselfish: A Critical Romp through the Terribly Libertarian Culture of High Tech by Paulina Borsook (Tin House, 2026, reprinted from 2000 edition). They looked selfish 26 years ago — and now we see the results.

Blockers: Rebels in the Deep State by Michael Lewis (W.W. Norton, 2026). Public servants holding the line against political manipulation . . . and, sometimes, failing.