Doing Less, But Better with AI

What do I actually want to make?

I've been thinking a lot about what I actually want to make.

I don't want to become another AI content creator chasing the newest tool every day.

AI moves at the speed of daily. There is always another model, feature, agent, or announcement to explain. None of it matters if the system doesn't understand the person, business, role, or problem first.

Context Training is the best thing I know how to teach about AI transformation and adoption. It's the first step, the third step, and the tenth step. Before the tool, before the workflow, and again when the work changes.

But when I think about the content I'm genuinely excited to make, I also think about travel. Vegan food. Tiny towns and big cities. The experiences I'm already having and the life I'm already living.

Which is the point of AI.

I want the systems to create more time, options, and freedom to participate in the life I built them to support.

That doesn't mean walking away from AI. It means refusing to force myself into the most crowded, exhausting version of AI content when my best teaching is Context Training and the content that gives me energy lives in a wider world.

Just because you can doesn't mean you should

At the time I wrote the original issue, I was publishing three podcast episodes every week in five languages. The system could produce that volume.

I chose to move to one stronger episode each Wednesday.

I could keep doing more. The evidence said more wasn't useful enough to justify the review time, and the volume was crowding out work and parts of life I cared about more.

Technical capacity answered Can I do this? It couldn't answer Should I keep doing it?

For that, I needed an evidence loop:

1. Action: Try the thing in the real world.
2. Evidence: Observe what happened without protecting the original plan.
3. Confidence: Decide what the evidence allows you to know.
4. Results: Keep, change, or stop the work based on what matters.

I took action with the podcast. The volume proved the system could do it.

Then came the evidence. The output was possible, but it wasn't moving the work enough to justify what it displaced.

That evidence gave me confidence to change the schedule without treating the change like a failure.

Trying something gives you information. If it isn't working, you can change it.

Can is proven through action. Should is answered by evidence.

One thing to do today: run the evidence loop on your week

Make a list of the projects, meetings, content, systems, and offers taking your time. Give that list to the AI you already use with this instruction:

“Look at this list through four stages: Action, Evidence, Confidence, and Results. Make a table showing what I tried, what actually happened, what I know now, and whether I should keep it, change it, or stop it. Don't make anything up. If information is missing, say so. Then identify the one change that could save the most time without hurting the results that matter.”

The AI can organize the evidence. You still decide what the evidence means for your life.

If this is your kind of conversation, join me for the next Seed & Society® newsletter.