3 Analogies Every Newsroom Should Use When Talking About AI 

Let’s say you’re in a meeting. 

There’s talk of “leveraging AI” and “scaling innovation” and “transforming the newsroom.”

Heads nod, but no one really knows what we’re actually talking about. It turns out that when people talk about AI in the media, it’s easy to get stuck in the abstract.  

Analogies are how we escape. They’re like little backdoors into actual understanding—suddenly, what felt like a fuzzy vision statement becomes “ohhh, it’s like when you…” 

At the 2025 International Journalism Festival, a few brilliant people used some very strange (and oddly perfect) analogies to explain how to think about AI and the role of product in news organizations. These aren’t brand-new ideas, but they offer clear, memorable heuristics that help us apply them with more intention.

Let’s start with drills.  

1. The Drill Metaphor: AI is Not the Thing. It’s the Thing That Gets You to the Thing. 

Let’s say you work at a hardware store. One day, a customer comes in and buys a drill. Cool, right? But why? 

You ask them, “Why are you buying this drill?” 

They say, “To make a hole in the wall.” 

Solid answer. 

But then you keep going: “Why do you need a hole in the wall?” 

They say, “To hang a picture.” 

Still tracking. 

You push again: “Why hang the picture?” 

And now, things get squishy. “Because it’s a family photo and the wall looks empty, and it makes the house feel more like home.” 

You’ve struck gold. We’re no longer talking about drills. Now, we’re talking about emotional needs. 

This, according to Nikita Roy (Founder & CEO of Newsroom Robots), is exactly how we should think about AI in newsrooms. 

AI isn’t the end goal. Nobody wakes up saying, “You know what I’d love? A language model.” No. What they want is fewer hours spent sifting through data sludge. They want their audience to actually find what matters. They want their workflows to feel less like an endless tangle of tabs and tasks and more like a system that just works. 

When Roy talks about being “feeling buried in data”, she doesn’t mean “give me a robot that summarizes this mountain of info.” She means: “Help me find the red flags. Tell me what’s important. ” 

So the next time your newsroom starts eyeing some AI tool or product, ask yourself what you’re trying to do: Are we buying the drill? Or are we trying to hang a picture? 

2. The Skateboard MVP: Stop Building Cars, Start Building Skateboards 

Imagine you want to build a car. 

So you design this: 

• Step 1: A wheel 

• Step 2: Two wheels 

• Step 3: A chassis 

• Step 4: A body 

• Step 5: A car 

Except this is terrible. Because for the first four steps, nobody can do anything. And then at the end—ta-da!—you hope it’s what people want. 

Here’s a better approach: 

• Step 1: Skateboard 

• Step 2: Scooter 

• Step 3: Bicycle 

• Step 4: Motorcycle 

• Step 5: Car 

This is what Alessandro Alviani, Lead Generative AI from Süddeutsche Zeitung, was talking about. He called it the “Skateboard MVP,” and it’s a metaphor designed to destroy one of the most persistent product development diseases: Shiny Object Syndrome. 

Shiny Object Syndrome is when you see some flashy new thing and you immediately want it. Not because it solves any of your problems, but because it looks cool in the slide deck. 

Skateboard MVP thinking says: Forget the shiny. Forget the perfect. Build the first thing that someone can actually use. Even if it’s a dinky little skateboard. Then ask: did this help? Did they fall off immediately? Should it have brakes? 

Do this enough times, and one day you’ve got a car that people actually want to drive. 

3. The Tent-Building Approach to Product Teams: Minimum Viable Structure 

Trying to organize product teams in media organizations around AI is like setting up camp in the middle of a storm—you need shelter fast, but you don’t have time to build a cabin. 

That’s the idea behind Felicitas Carrique’s (Executive Director, News Product Alliance) “minimum viable structure.” It’s like pitching a tent: it’s not perfect, but it’s good enough to get the team working, thinking, and staying (mostly) dry. And once the wind dies down, you can start adjusting the poles. 

Instead of chasing some flawless org chart drafted by Harvard MBAs, news organizations should focus on a structure that’s functional enough to deliver outcomes now—then tweak it as they go. 

Mariah Craddick, Executive Director of Product from The Atlantic has seen this play out. Their product team has been reorganized multiple times, and their current setup is still experimental. According to her, there’s no one-size-fits-all model. Whether product sits under editorial, on the business side, or floats in its own space depends on the organization’s goals, business model, and digital maturity. 

In short: start with the simplest setup that works. Let it evolve as you learn. Product teams in news organizations don’t need a palace—they need a strong tent that can stand up to the weather and move when the ground shifts. 

So What? 

These analogies aren’t just fun brain candy (although they are that too). They’re shortcuts. They’re conversation-starters. They’re shields against the nonsense and trend-chasing that can infuse conversations about how we use AI in news products and processes.  

When someone says “we need to use AI,” you can say “Sure. But are we buying a drill or trying to hang a picture?” 

When someone pitches a moonshot idea, you can say, “Cool. But can we skateboard there first?” 

And when someone tries to build a crystal palace of an organizational chart, you can gently remind them that maybe we just need something that doesn’t collapse under the weight of our next sprint.