##  [The old science journalist leaves the nerd in the family in the dust](/node/166) 

    *Submitted by Lennart on Wed, 8 Apr 2026 - 09:29*  

  ![Dissemination](/sites/default/files/styles/wide/public/2026-04/formidling.png.webp?itok=ydhXTXv-)

 

I recently gave a talk to twelve people, the vast majority of whom had no prior knowledge of AI. Several were openly negative and made no secret of it. I agreed with many of their objections — they were based more on intuition than experience, but the intuition wasn't wrong.

When I realized what kind of audience I was facing, I ditched the slides and preparation and improvised for the rest of the time. It turned out to be one of my best talks.

That's not what I want to write about.

## The interesting part came afterwards

Several people came up to thank me. Everyone who gives talks knows that feeling. The interesting part wasn't the thanks — it was **the pattern of their previous experience with AI**.

Almost everyone had the same story:

> "I have a tech-savvy person in my family who tried to help me get started with AI. It only made me more confused. And they got frustrated that I couldn't see the brilliance of it."

The variations were minor. A son, a sister-in-law, a colleague, a grandchild. Always someone who "knows computers." Always with the best intentions. Always with the same result: confusion for the recipient, frustration for the sender, and a quiet decision by both that AI probably wasn't for them after all.

After a couple of hours of my improvised talk, these same people left the room with:

- A grounded understanding of what AI actually is
- An understanding of what it can be used for
- And — most importantly — an understanding of **when not to use it**

It wasn't because I was smarter than the tech-savvy person in their family. It was because I wasn't the tech-savvy person in their family.

## Why the tech-savvy person fails

The tech-savvy person in the family has a real problem, and it's not technical. It's linguistic.

When you fall in love with a technology, you stop hearing what it sounds like to those who aren't in love with it. You start explaining *how* things work before you've explained *why* they should interest the recipient. You use words like "model," "prompt," "context window," and "agent," without realizing that the three words the recipient actually heard were "model" (a mannequin?), "prompt" (an error message?), and "agent" (like a real estate agent?).

And when the recipient doesn't get it, the natural reaction isn't to lower the level of abstraction. It's to raise it. Because the tech-savvy person believes that if only the recipient understood the full picture, they too would be excited.

This is precisely the opposite journey of what the recipient needs.

## What I did differently

I basically only did three things:

**1. I agreed with their resistance.** When a participant said AI was just hype, I replied that most AI talk *is* hype. When another said she didn't trust chatbots, I said she had good reasons not to. When you start by acknowledging resistance, people stop defending it — and become able to listen.

**2. I talked about when not to use it.** Most AI talks focus on the possibilities. Mine focused just as much on the limitations. Paradoxically, this is what makes people open. When a salesperson tells you what a product *can't* do, you start believing what it *can* do.

**3. I used their language, not mine.** No "tokens," no "embeddings," no "fine-tuning." Just everyday examples from workflows they know themselves. AI is a fast intern who can read, write, and calculate, but has never worked in your industry. That's it. The rest is details.

This isn't pedagogical magic. It's just basic communication. But in a market where most people talking about AI are either tech-savvy individuals or salespeople, it's apparently rare enough for people to notice.

## This is not a coincidence

As I walked home from the talk, it dawned on me how much of my current role directly draws on my old life as a science journalist. For twenty years, my job was to take complex topics from researchers — people who were essentially academic versions of "the tech-savvy person in the family" — and translate them into language that ordinary people could act on. I was never the most skilled in the technology itself. I was just the one who could hear what it sounded like to those who didn't know it.

This is the competence that has value in AI consulting right now. Not because I know more than the engineers — I don't — but because I know what leaders, owners, and boards need to hear before they dare to make a decision. And it's rarely a detailed explanation of how transformer architectures work.

## What this means for your business

If you are considering getting started with AI, and you have an internal tech-savvy person who has tried to bring you on board without success, there are two things you need to know.

First: It's not your fault. Confusion is not a sign that you're too dumb. It's a sign that the person explaining is not translating from their language to yours.

Second: You don't need a better tech-savvy person. You need a translator. Someone who can stand between the technology and the business and tell you what you should do, what you should refrain from doing, and why — in a language you can use tomorrow.

This is what I do. And if you recognize yourself in this story, a conversation is a good place to start. It costs nothing, and you don't need to have prepared anything at all.

Just bring your skepticism. That's where we begin.