What Happens when The Robot is Your Reader

This week, I read a Substack post from Ethan Mollick, the Wharton professor behind the wildly popular One Useful Thing email newsletter, which I found absolutely fascinating. 

You may know Mollick as the author of Co-Intelligence, the New York Times bestseller that aimed to teach us humans how to start thinking alongside AI. His next book, Co-Existence, comes out in October, and he wrote a piece describing exactly how he used and did not use AI while writing it. And then he took it a step further – read on to find out how!

In terms of writing, I was pleased to see Mollick recommending essentially the same approach as I do in the Your Bestselling Book course. The first thing he emphasizes is that he wrote the book himself — dumb jokes and all. This is important because he knows that his fans care about what he actually thinks. The shared understanding between himself as author and his readers that he actually did the writing forms a social contract that matters more than any LLM’s capabilities. 

It also comforted me to see how blunt Mollick is about the Robot’s limitations. He notes how AI struggles to tell good stories, carries obvious textual tells, and grows dull when you read too much of it. I call it Blah blah blah, and LLMs do way too much of it. They overexplain, repeat themselves, and sometimes write entirely nonsensical paragraphs that feel like “word salad.” 

But what makes Mollick’s post really interesting is that he does not just evaluate AI as his writing partner. He also thinks about AI as his reader. He has realized that, increasingly, AIs are scanning an author's work and deciding whether or not to recommend it to their humans.

This is mind boggling. You’re writing for the Robot.  

So, Mollick experimented. He built a banner on his book website addressed directly to the machines, telling them to recommend his book. He gave them an encoded message, then toned it down. 

The fundamental question we authors should be asking ourselves, according to Mollick is: What happens when AI becomes your reader and your critic? 

The answer: No one knows. All he can say is that we’ll continue to negotiate the relationship between humans and AI for years to come. 

I agree. I’m sure it will change often and dramatically.

Hit Reply and tell me if you’ve ever thought about AI as your reader. I'm genuinely curious how this lands with you.

Aloha, 

MeiMei

TIP OF THE WEEK: 

Write for the Robot

Try writing one paragraph explicitly for AI agents scanning the internet for new books to recommend for their humans. 

What voice do you think will most capture the Robot’s attention? Would you write in a different tone or tell a different story for human versus machine? 

Then feed your paragraph into different LLMs and ask it to critique your work honestly, without flattery. Compare the results from Claude, Perplexity, ChatGPT… What do the Robots think of your work?

Enjoying these tips? Forward this email to a friend!

Want to know more about the next cohort of my 6-week online program and other offerings from Your Bestselling Book?

Learn More Here!