This week I went to a lunch at Stanford Law with Ethan Mollick and Kevin Roose.

If you haven't read them: Ethan Mollick is the Wharton professor behind Co-Intelligence, the book that showed readers how to use AI productively. His latest, Co-Existence: The Next Phase of AI, is about working well in the uneven zone where AI's abilities and ours overlap, the jagged frontier. Kevin Roose wrote Futureproof, and his new book, The AGI Chronicles, draws on more than 150 interviews to tell the story of the race among AI companies to build artificial general intelligence.

They didn't agree on everything but they did agree on this: we are not pushing AI far enough.

First, Kevin Roose's advice: give AI harder tasks. Most people, he said, are "using a Ferrari to go to the grocery store."

Second, Ethan Mollick's advice for the moment you hit a wall: escalate one level further. He said the gap between what AI can do and what people use it for is enormous.

It reminds me of decades ago when I was a new manager and went through new manager training. I learned about the Pygmalion Effect where people rise to high expectations.

The same (to an extent) is true with AI.

In the discussion, Kevin talked about how he used AI to help him file paperwork for a new company he was starting. He had AI write and check over his work.

Ethan challenged him. He said: AI can do a lot more than that.

Raise your expectations.

So do this:

1. For a task you're using AI right now, make a note of where you're still doing *anything* as part of the task: whether it's the prep (uploading a file) or the next steps (copying it into an email).
2. Instead of accepting it and going through the manual work, offload it to AI. Tell your LLM: I'm still doing XXXX. I need you to do this for me. What would it take? Walk me through the steps, simply.
3. You might not be able to take all off your list, but you might edge out some of the manual work.
4. Keep a note of that task. Revisit it in 6 months with a new LLM model release. Try again. You may be surprised by the results.

Apply to work with me here. I'm closing the doors on this option at the end of October. In November, I'll be looking to book 2027 clients.

Alex