How fluently do you work with AI?
Sixteen questions, about four minutes, set in a normal work week. The questions adapt as you go, and you see your result the moment you finish.
Take the Check
No sign-up. Your answers stay in this browser until you choose to email yourself the result.
What the Check measures
Four dimensions, each scored from four questions. Most people are uneven across them, and the weakest one is usually where the risk sits.
Delegation
Match the tool to the task.
Decide what AI is for before you reach for it: when to hand a task over, when to draft together, and when not to use AI at all. The first question is whether the task should be delegated, before any question of how to prompt it.
Description
Say what you mean.
AI does what you describe, not what you intend. Strong requests give context, constraints, the shape of the output and an example of what good looks like.
Discernment
Check before you trust.
AI sounds confident when it is wrong. Trace the claims that carry weight back to a source, treat citations as leads until checked, and notice when it agrees with you for the sake of agreeing.
Diligence
Own what you ship.
Accountability for AI-assisted work stays with the person, not the tool: disclosure where it matters, care with sensitive data, and a record you can produce if asked.
Why fluency matters to an organisation
- Adoption follows capability, not licences
- Giving AI tools to people who do not yet use them well adds new ways to fail rather than speed. Fluency raises the return on every tool already paid for.
- Low fluency fails silently
- Plausible but wrong output sent to a client, sensitive data pasted into a public tool, an undisclosed AI draft signed by a colleague: none of these appear on a dashboard until they become an escalation.
- It lasts longer than the tools
- The habits carry across models and products. What your people learn now still applies when you change model or add a new tool.
- Discernment and diligence carry the risk
- Most people pick up delegation and description by doing. Checking output and owning what ships are what separate teams that use AI from teams that use it well.
Thinking about a whole team?
Individual fluency is where AI adoption succeeds or stalls. Kode-1 runs AI literacy and adoption work for teams, built on the same four dimensions, alongside the governance that keeps it safe.
Common questions
About four minutes. There are sixteen questions, four for each dimension, and you see your result as soon as you answer the last one.
No. You can take the Check and see your result without giving any details. If you want the result emailed with a link to your full report, you give a name and email address at the end.
Every question has four answers worth 0 to 3 points, for a maximum of 48. When you are doing well in a dimension, later questions in it are drawn from a harder set, so a strong result has to be earned on harder questions. The Sharp level in a dimension requires passing one of those.
Aware means you are building the instincts, Capable means you apply them in real work, and Sharp means you are ahead of most people and ready to make your habits repeatable for a team. Each dimension gets its own level as well as the overall one.
Yes. The questions are drawn from a pool of 64, so a retake shows you a different mix.
Nobody, unless you ask for it by email. If you do, a Kode-1 partner can see your result so they can follow up if it is useful. We do not add you to a newsletter or a sales sequence.
The four dimension names (Delegation, Description, Discernment and Diligence) draw on terms popularised by Rick Dakan and Joseph Feller's "Framework for AI Fluency" (CC BY-NC-ND 4.0), credited here as inspiration. The sub-dimensions, questions, scoring and the Aware, Capable and Sharp levels are our own and reproduce none of its text.