Getting More Out of AI

I recently listened to Simon Sinek's A Bit of Optimism conversation with Ethan Mollick, The AI Skills Nobody Is Teaching (And Everyone Needs). This one particularly resonated with me and the work that I do.

Mollick is a Wharton professor who studies how people use AI. Their conversation covers work, learning, creativity, and the choices we have about fitting AI into our lives. Experience and judgment still matter, especially when deciding whether an answer is actually useful.

With my background in consulting, engineering, and usability, I found plenty to think about here. I've always enjoyed figuring out how to make a tool work better for me. Apparently that also means telling it when it's being a little too agreeable.

A Few Things to Try

• Ask for pushback. Tell AI to critique your idea and look for weak spots.
• Pause and review your thinking. Ask where your argument falls short and how to improve it, with examples.
• Give it harder work. Try a substantial task, then evaluate the result and give specific feedback.
• Select a capable thinking model. At recording, Ethan recommended paying around $20 a month and actively choosing the strongest model available.
• Keep your voice. Give it writing samples, request a two-page style guide, and save that as instructions. Start with your own rough draft.
• Try different perspectives. Ask a simulated beginner or skeptical expert to review your work.
• Use it to learn. Have it tutor you at your level, using your interests to explain things. Try deep research for initial research.

The pushback suggestion is the one that stuck with me. A tool that helps me spot something I've overlooked is useful. I'd like to get better at asking for that.

The episode is just under an hour. Give it a listen, then pick one of these ideas and try it on something you're already working on.