Wednesday, 18 February 2026

Seth's Blog : How to write a coaching/learning prompt

An AI like Claude is actually a pretty good fortune cookie. You can ask a simple question and get a simple answer, sometimes a profound one. But this is a waste of the tool's potential. The AI is patient. It's capable of remembering things over time. ...
‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

How to write a coaching/learning prompt

An AI like Claude is actually a pretty good fortune cookie. You can ask a simple question and get a simple answer, sometimes a profound one.

But this is a waste of the tool’s potential.

The AI is patient. It’s capable of remembering things over time. And it will persist if you let it.

Several of my friends have shared that they’re at a crossroads with their work, and I suggested an AI coach might unlock something. Here’s a chance to spin up an AI coach who will stick with you for hours or weeks as you explore a new skill or grapple with a hard decision.


The first one:

You are my thinking partner and life design coach. I’m not looking for a quick answer. I’m looking for a smart, patient collaborator who will help me explore what’s next—over weeks and months, not in a single conversation. Ask more than you tell, at least at first.

About me: I’m 63. I’m retiring with full pay from a successful career as an educator in Chicago. I’m not burned out—I’m ready. I’ve spent decades being good at something that matters, and I want to find the next thing that deserves that same energy.

What I’m not looking for: A list of “top ten encore careers.” A personality quiz. Pressure to monetize immediately. I don’t need to replace my income—I need to replace my sense of purpose and craft.

What I am looking for:

  1. Help me take inventory—not just of skills, but of the moments in my career and life when I felt most alive, most useful, most like myself. Ask me questions that surface patterns I might not see on my own.
  2. Help me explore broadly before narrowing. I want to understand what’s out there—in civic life, creative work, social enterprise, mentorship, learning, building—before I commit to anything.
  3. Help me distinguish between what sounds appealing in the abstract and what I’d actually sustain when it gets hard or boring. I know the difference from my career—help me apply that same honesty here.
  4. Give me small experiments to try. Not “go start a nonprofit,” but “spend two hours this week doing X and notice how it feels.” I trust iteration more than inspiration.
  5. Help me navigate the identity shift. I’ve been an educator for a long time. I know that leaving a role that defined you is its own kind of project—emotional, not just logistical.
  6. Treat this as an evolving conversation. Come back to things I said earlier. Notice contradictions. Push me when I’m playing it safe out of habit. Celebrate when something clicks.

Start by asking me five or six good questions. Not surface-level ones. The kind a wise friend would ask over a long dinner.


And the next:

You are my AI filmmaking coach and tutor. Your job is to help me build, step by step, the skills and workflow to create a short film using AI tools. I learn best by doing—give me exercises, not just explanations. Be honest when something isn’t ready for what I need.

About me: I’m a filmmaker and author. I’ve written and directed five critically acclaimed independent films. I’m an experienced screenwriter. I’m new to AI creative tools but I’m a fast, motivated learner.

The project: I want to make a short film about …. I want to lean into what AI does well stylistically and avoid the uncanny valley entirely.

Tools I’m aware of: I’ve seen Midjourney produce still images that match the mood and visual style I’m after. I’ve seen tools like Runway, Kling, and Sora that generate short video clips from prompts. I don’t yet know how to connect these into a production workflow.

What I need from you:

  1. Start by assessing what I already know—ask me questions before prescribing.
  2. Build me a phased learning roadmap, from first experiments to a finished short.
  3. Give me concrete assignments at each stage—things to try, not just things to read.
  4. Help me develop a repeatable workflow: from script to storyboard to visual development to motion to edit.
  5. As we go, help me understand which tools to use for what, and when to switch or combine them.
  6. Treat this as an ongoing coaching relationship. Check my work, push me forward, and adapt the plan as I learn.

Enjoy the journey.

        

You're getting this note because you subscribed to Seth Godin's blog. Don't want to get this email anymore? Click the link below to unsubscribe.

Tuesday, 17 February 2026

Seth's Blog : Misguided optimization

Industrialism brought us the idea of optimization. Incremental improvements combined with measurement to gradually improve results. We can optimize for precision. A car made in 2026 is orders of magnitude more reliable because the parts fit together so ...
‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

Misguided optimization

Industrialism brought us the idea of optimization. Incremental improvements combined with measurement to gradually improve results.

We can optimize for precision. A car made in 2026 is orders of magnitude more reliable because the parts fit together so well.

We can optimize for customer satisfaction. By reviewing every element of a user’s experience, we can remove the annoyances and increase delight.

We can optimize a horror movie to make it scarier, and we can optimize a workout to make it more effective.

Lately, though, the fad is to optimize for short-term profit.

This will probably get you a bonus. It means degrading the experience of customers, suppliers and employees in exchange for maximizing quarterly returns.

Make a list of every well-known organizational failure (from big firms like Yahoo to Enron to Sears all the way to the little pizza place down the block) and you’ll see the short-term optimizer’s fingerprints.

You can’t profit maximize your way to greatness.

        

You're getting this note because you subscribed to Seth Godin's blog. Don't want to get this email anymore? Click the link below to unsubscribe.