A team game about AI at work
Which tasks should you hand to AI?
The AI In-Basket is based on the in-basket, a classic management exercise: a pile of tasks to sort. Here, there are four bins and a team that has to agree. You come out with a shared language: what you keep, what AI can help with and what you stop doing.
The idea
One task, four bins
For each task, the team picks a bin. Everyone has to agree. That’s where the conversation gets interesting.
I keep it
My judgment, a relationship or trust.
AI helps me
It writes a first draft. I correct it and decide.
AI does it
It handles it. I check now and then.
Eliminate
Nobody should be doing this task.
- Does it come up often?
- How long does it take?
- If AI gets it wrong, how bad is that?
- Does it need me: a relationship, my judgment, trust?
How to play
Pick a job
Choose a ready-made job, or prepare the tasks for your own job with your own AI.
Read the task out loud
Each task arrives the way it does in real life: an email, a message, a request.
Talk it through and agree
Four questions guide the conversation. The team picks a bin and writes down what it said.
Review the results
At the end, you see the big picture: time at stake, disagreements with AI and one task to try this week.
Why it works
It starts from real work
Not from a list of tools. From emails, reports and requests that everyone recognizes.
Disagreements are useful
When two people pick different bins, they end up talking about what they value: relationships, risk, time.
A shared language
The four bins give the whole team simple words to use. The game builds on the AI Fluency Framework.
Who it’s for
Managers
To see with your team where AI is worth it, before buying a tool.
Teams
To set clear rules: what you hand to AI, what you keep and what you stop.
Trainers
To open an AI workshop with a hands-on exercise, in 45 to 90 minutes.
Consultants
To get a client thinking about their own job, with their real tasks.
The AI Fluency link
A game to practise the 4Ds
The AI Fluency Framework, created by Rick Dakan and Joseph Feller and taught by Anthropic Academy, describes four competencies for working with AI: Delegation, Description, Discernment and Diligence. The AI In-Basket puts them into practice, especially the first one: deciding what to hand to AI, and how.
See the links with AI Fluency- D
Delegation
- D
Description
- D
Discernment
- D
Diligence
Going further: the AI Canvas
Does a task look promising? The AI Canvas, the In-Basket’s sister tool, helps you check whether it’s worth it: for whom, with what data, at what cost.
A tool by KODRA Conseil
Want us to run it for you?
KODRA Conseil runs workshops on AI at work with the AI In-Basket and the AI Canvas: for a leadership team, a board or a group of businesses. We start from your real tasks, and you leave with clear rules and a first task to try.
Ready to sort your tasks?
A game takes 30 to 60 minutes. All you need is a screen and a team.