Prototype

a first version that is functional

This card covers building a first working version of your AI tool. A prototype is rough but real enough to try, so you learn what works before investing more. It turns assumptions into something a journalist can react to, which is usually more honest than any plan on paper.

Questions to explore

// use these as prompts in a workshop or on your own. There are no right answers.

What is the smallest version that would still teach you something useful?
Which assumption are you most trying to test with this first build?
Who will try the prototype, and how will you watch them use it?
What would you accept as a sign to stop, change course, or keep going?
What can you fake or do by hand now to avoid building it fully yet?

Expert voices

// notes from the journalists and AI experts who helped shape this kit

“Release MVPs, but they need to add some value. If you ship several AI tools that are not good, the newsroom's trust in the technology, and in you, will go down.”

Martin Schori, Aftonbladet

“Test new tools in controlled sandbox environments. How can newsrooms experiment responsibly?”

Steffen Leidel, DW Akademie

“Fail fast and fail cheap. Test ten new formats and discard what does not work.”

Walid Al-Saqaf, Doha Institute for Graduate Studies

“Coding with AI has moved at neck-break speed, and it lets newsrooms build apps, websites, and datasets they could never have built before.”

Paul McNally, Develop AI

Things to consider

A rough prototype that people use teaches more than a polished plan.
Build the riskiest part first, not the easiest one.
Be ready to throw the prototype away once it has taught you what you need.

using this card

Pull Prototype when it is relevant and set it aside when it is not. Pair it with the other AI Solutions cards, lay them out on a table, and use the questions above to get everyone on the same page. Capture what you discuss on sticky notes or in a shared doc.