Home / Me & AIA VERY USEFUL WORKING RELATIONSHIP

I use AI a lot.
I still do the thinking.

I love what these tools make possible. I also know the feeling of reading an AI answer and thinking, “Hang on. That’s not how this works.”

Illustration of a person directing and reviewing AI-assisted work
Illustration: the human still brings the direction.
EXPERIENCE HELPS

I’ve been doing this
since before the chat box.

It’s easy to get busy
going in the wrong direction.

I see people follow AI into another tool, another workflow, and another problem they didn’t need to create. I get why it happens. AI nearly always has another suggestion.

I’d already spent years working with customers, marketing systems, and developers before AI became part of my day. That experience helps me notice when an answer doesn’t fit the actual problem.

Now I’m constantly trying new ways to improve my workflow with AI. Better research. Clearer explanations. Less time wasted. More attention for the person who needs my help.

I like the tools.
I care more about the result.

  1. Start with the goalWhat needs to improve?
  2. Bring the contextThe setup and constraints.
  3. Work with AIResearch, draft, explore.
  4. Check the evidenceDoes the answer hold up?
  5. Own the resultExplain and follow through.
The way I use AI: experience and judgment stay involved throughout.

What are we actually trying to do?

I start here. Who needs help? What would make things better? It’s much easier to steer AI when I know where I want to go.

What does it need to know?

The customer’s setup, the product’s behavior, the constraints. I bring the context. Leaving out one important detail can send the whole conversation sideways.

Wait. Is that right?

AI can sound very sure of itself. I check the documentation and the evidence, and I ask what else could explain what I’m seeing.

Did it actually work?

A nice-looking answer is only the start. I check the result and make sure I can explain what changed and why.

Who’s responsible? I am.

The client deserves a clear answer, and the developer deserves a useful handoff. I’m the one responsible for both.

Tarotmantic’s public reading-experience demonstration
Tarotmantic’s public product preview—not a customer reading. Open image ↗
The Mildly Offensive Tees public product collection
A real storefront where I put the ideas to work. Open image ↗
ONE QUESTION I FIND USEFUL

“What else could
explain this?”

Give AI the evidence and ask what you’d need to check to tell the possibilities apart. Then go check. That last bit is where the experience comes in.

See that thinking in a real case
ENOUGH ABOUT ME. WHAT ARE YOU WORKING ON?

Have something in mind?
Let’s talk.

A project, a tricky problem, or an idea you can’t stop thinking about. I’d love to hear it.

Let’s book a call

Email me and we’ll find a time.