Self-assessment v1

Where you stand with AI.

Nine questions, about five minutes. Pick what you actually do in a normal week, not what you plan to. You get a level for each of three areas, straight away. No sign-up.

Part 1Tools and tooling

1. In a normal week, how do you use AI?

AI ask a chat assistant for photos, summaries or quick answers.BI build small tools or automations with vibe-coding platforms that others could use.CA personal agent connected to my mail, calendar and docs does multi-step work on a schedule.DI run my own stack of agents and skills across several models, deployed for other people.

2. How did you choose your tools?

AHabit. It's the one I started with.BI use a few tools for different jobs and have prompts I reuse.CI know where each model is strong and what it costs, and route work to match.DI test new tools against a baseline and retire the weak ones.

3. What happens when you are not there?

ANothing. Every task starts with me typing.BA saved prompt or automation runs when I trigger it.CBriefings and loops run unattended and I review the output.DAgents run work for a team or client, with logs, checks and someone else relying on them.

Part 2Strategic taste

4. How do you judge an AI output?

AI mostly accept the first answer.BI know what good looks like for familiar tasks, and edit or reject the rest.CI set the success criteria in the brief before I delegate.DI write them down as rubrics, evals or rules that others and agents apply.

5. How do you decide what to hand to AI?

AI try it on whatever is in front of me and see.BI choose tasks I can specify clearly, where I can spot a wrong answer.CI decide deliberately what to automate and what to keep human, weighing time and token cost.DI kill low-value work early and set that direction for others.

6. When you brief AI or a teammate, how much do you prescribe?

AI don't think about it. I write what comes to mind.BI give a goal and the format I want.CI give the question and constraints, and leave the method open.DI calibrate by person and agent: detailed steps for new ones, the outcome only for trusted ones.

Part 3Systems and domain knowledge

7. How well do you know the workflow around your own task?

AI know my task, not what comes before or after it.BI know my workflow end to end and can spot when AI gets it wrong.CI map the systems, data and handoffs around it.DI redesign the workflow around agents and know where the business constraint sits.

8. How do you hold what you know about your domain?

AIn my head and in scattered files.BIn notes I can paste into a prompt.CIn a maintained knowledge base or graph that my agent reads.DIn a shared, maintained form that agents and colleagues build on and correct.

9. Could you replicate one of your processes with agents and check it?

AI'm not sure where I would start.BI could specify a tool for part of it.CI have replicated one process with agents and checked it against the original.DI have done it across several processes and I measure the results.

0 of 9 answered.