Today, some quick thoughts on surprise, its vital role in your messages, and why AI can’t help much on this one. 

What got me thinking about this is a wonderful Substack post I came across from writer and entrepreneur Nabeel S. Qureshi. In his “What Makes Art Great” post, Qureshi explores an intriguing question: Why is it so hard for AI to write a great poem?  And his answer is a revealing one for anybody using AI to draft their next message.

It all comes down to a key factor: Great art is not predictable or obvious, it is surprising.” And, by their very design, LLM (Large Language Models) aren’t very good at surprise. 

Here’s how Qureshi describes the the brain’s “predictive processing model.”

“As we are scanning a text, our brain is constructing the meaning and predicting the next several words. Where there is no surprise — where something is perfectly predictable, or fits some pattern that we know — our brain registers only dullness.”  

Ah, but look what happens when surprise enters the picture: “When our expectations are violated in a way that’s satisfying to resolve, we get pleasure and novelty.”

That last sentence is worth lingering over. Using surprise to emotionally engage your reader only works if the surprise disrupts things in ways that are “satisfying to resolve.” If your surprise just creates confusion, it won’t have the desired impact.

Thus, the AI problem generating that great poem. When you ask an LLM (Large Language Model) to write a poem (or an email or fundraising appeal), that draft will almost certainly be conventional — built on common patterns, familiar framings, and safe, middle-of-the-road language. 

That’s not an accident. It’s how LLMs work: they generate text based on statistical patterns drawn from massive amounts of training data, not from genuine understanding or original thought.

In his in-depth discussion of what makes art great, Qureshi is pointing to a dynamic that leads copywriters (including myself) to avoid AI-generated first drafts. As copywriter Vikki Ross puts it “AI can generate standard stuff, but it can’t create standout stuff – stuff that’s going to really mean something to someone.”

It’s just too predictable. There are lots of ways that AI can play a powerful role in the creative process (transforming the way you do research, stress testing your thinking, serving as a junior editor refining human-first copy). But generating standout first drafts that skillfully use the power of surprise isn’t one of them.

  • Don’t ignore the power of surprise to enliven and strengthen your messages.
  • Remember to surprise people in ways that engage them, not ways that confuse them.
  • Don’t count on AI ‘s ability to inject your copy with original thinking and surprising elements.
  • Don’t use AI to write for you. Use it to critique and strengthen your writing.
  • Don’t use AI to think for you. Use it to stress test and challenge your thinking.

Most Recent Posts

Leave a Reply

Your email address will not be published. Required fields are marked *