Sabine Devins asks those two questions before she writes anything. Every founder I know who freezes at the blank page is freezing because they haven’t answered them yet.
I sat down with Sabine for this week’s episode of Changing Shapes. She’s a digital storyteller and creative content director based in Berlin. Before that she was a journalist — National Geographic, Voice of America, NHK, Handelsblatt Global. These days she mostly works with B2B tech companies, the ones building genuinely interesting technology who haven’t quite figured out how to talk about it beyond their own four walls.
I wanted to talk to her because I’m seeing something in my own work that I think is bigger than my own work.
The skill we thought we’d need is not the skill we actually need
When AI started showing up inside real workflows, the assumption — mine, everyone’s — was that the technical skills would become the most valuable ones. Coding, data analysis, systems thinking. The hard, legible, expensive skills. The ones that look like they take a long time to learn.
That’s not what’s happening.
What I’m watching is that as AI gets very good at research, analysis, building, and generating — most of what we used to call knowledge work — the thing that’s becoming most valuable isn’t more technical work. It’s storytelling.
And that completely flips the script. We’ve been treating the humanities in the workplace as the lesser degree. The soft skill. The nice-to-have. Sabine and I spent most of this conversation in that flip.
Storytelling isn’t valuable because it’s hard or creative or in some way expressive. It’s valuable because fundamentally it’s a human-to-human act. Why would you care what a machine has to say to you? You want to connect with people. With other humans, who have lived experiences.
AI fatigue is real
The pattern I’m seeing — and Sabine sees it too — is that once we feel a story is coming from a machine, we switch off. Instantly. Not because the machine is bad at writing. Often the words are fine. But just because the words exist doesn’t immediately give them meaning. The meaning is because it comes from another human.
That gap is what AI fatigue is. People aren’t tired of AI’s quality. They’re tired of trying to build a relationship with something that isn’t there.
Sabine put it this way: AI is really good at seeing patterns. It’s really good at the common denominator. What it’s not good at is outliers, unpredictability, genuinely new ideas. And the output of AI is, by definition, the common denominator of what came before.
True creativity has to come from us. Not because of some romantic claim about the human spirit. Because of how the math works.
The two questions
Here are the two questions Sabine asks before she writes anything, and the reason I think every founder should be asking them too:
What’s the point?
And who cares?
That’s it. It sounds trivial. It is not trivial. Every B2B founder I’ve ever worked with who has frozen at a blank LinkedIn post box has frozen because they’re trying to write before they’ve answered those two questions. They start with the words. The words come last.
Sabine gave me an example I’m going to steal. She used to work at an agency that made infographics, and a colleague taught her this: if you’re explaining how butterflies experience the world, you can say “I smell with my feet, I taste with my tongue, I hear with these little holes in my body” — and a child will get it. Or you can say “my olfactory glands are found in my legs” — and a scientific reader will get it. Same story. Two completely different shapes. Because the storyteller answered the second question first.
Who cares.
Once you know who cares, the shape of the story falls out. You stop trying to be impressive and you start trying to be useful. That’s the move.
Finding the human in the data
The story Sabine told me that I’ll keep thinking about for a while was about Brexit.
She was working as a business journalist when the pound crashed. Every outlet was reporting the same way — exchange rates, falling values, the abstract machinery of markets. Then she read an interview on the BBC where someone said, simply: “20% of my retirement is gone.”
That was the click. Same data, completely different story. The first version is information. The second version is a person who can no longer afford something they had been planning for. The first version produces nodding. The second version produces action.
This is what Sabine taught me storytelling actually is, when it’s working. It’s the move from data to “what does this mean for me.”
It’s the same move product designers make all the time, even if we don’t call it storytelling. When someone opens a product for the first time, the work isn’t to show them every feature. The work is to answer the question they’re silently asking: what does this mean for me. The home screen, the empty state, the first prompt — those are all answers to that question. If they answer it well, the person stays. If they don’t, the person leaves.
The storyteller’s question is the product designer’s question
The thing I keep coming back to, after this conversation, is that every product and every piece of content and every business interaction is some sort of a relationship. And like in every relationship, trust is built through consistency, but it can be lost in an instant.
So the storyteller’s question is the product designer’s question.
What are you actually giving, and to whom?
If you’re a founder reading this, and you’ve been frozen at the blank page, the reason is almost certainly that you’re trying to answer the third question — what should I say — before you’ve answered the first two. Don’t write anything until you can answer them. Then write the thing that answers them.
If you can, the post writes itself. If you can’t, no amount of polish or hook or AI-assist will save it. The reader will feel the gap before they can name it. They’ll just scroll.
— Tom
Full conversation with Sabine is out now on Spotify, Apple Podcasts, and YouTube. Recorded at The Social Hub Berlin. Follow Sabine on Instagram, LinkedIn, and at sabinedevins.com.
