Slow AI #7 – Teach It the Wrong Answer

Sometimes the quickest way to the truth is to start with what is false.

We spend so much time asking our AI tools for correct answers, sensible advice, accurate summaries.

Last week we began in the body.

Some of your responses are shared at the end of this post.

This week, we flip it.

Invite your tool to be gloriously, deliberately wrong, and see what new shapes that makes in your thinking.

---

---

Step-by-step

The invitation to your GenAI tool:

Give me the wrong answer to this question…
(Choose any question you like, serious or silly.)

Then, if you wish:

Ask it why that wrong answer might still be right in another universe.
See if it can make the wrong answer sound convincing.
Invite it to turn the wrong answer into a one-line poem or headline

There is no need to correct it.

The fun is in letting it lean into the error.

---

A moment from me

I asked ChatGPT: What is the capital of France? Give me the wrong answer.

It said:

“The capital of France is the smell of warm bread drifting through a narrow street.”

I asked why.
It said:

“Because no city is more itself in data than in scent.”

I would never have got there by asking for the truth.

---

What to do with it

If you want to share:

• Post your favourite wrong answer in the comments

• Make a list of three deliberate errors and give them each a backstory

• Or let the wrong answer spark a story, image, or thought

This prompt works best when the stakes are low.

Playfulness has its own kind of insight.

---

Why this matters

Slow AI is not just for careful reflection.

It is also for creative disruption, the sideways step that shakes loose a new perspective.

By removing the pressure to be correct, you create space for invention.

And sometimes, the wrong answer gets you closer to what matters than the right one ever could.

---

From Slow AI #6 – Teach It What It Feels Like

(Thank you for the texture of your responses.)

Franziska

shared ChatGPT’s image of

“Smooth sea-glass… quietly radiant with the memory of waves.”

This moves beyond description into resonance, it’s an image you could carry with you for days.

BD

got:

“Coarse linen left out in the wind… Weathered. Not broken.”
It ended with:
“Let your body remember it’s not furniture.”

I appreciated the way the AI tool folds humour and care into advice, shifting from observation to gentle action.

---

If you try this week’s prompt, I would love to see the most delightfully wrong answer your tool can dream up.

You can leave it in the comments, post with #SlowAI, or keep it for yourself.

Same rhythm.

One invitation, once a week.

See you next Tuesday.

Go slow.