Measure your practical AI literacy
Knowing which AI tool to reach for, writing prompts that work the first time, automating repetitive work, and spotting when a model is wrong are the skills that separate people who get real leverage from AI from people who just chat with it. This 25-question self-assessment scores you across those four pillars and turns your weakest areas into a concrete improvement plan.
How it works
You rate 25 statements on how true they are of your current habits, grouped into four areas: prompt writing, tool selection, workflow automation, and critical evaluation. Each answer scores 0 to 4. The tool averages each area, weights them equally, and scales to a 0–100 overall score with a level band from Beginner to Expert. It then identifies your two lowest-scoring areas and generates a targeted plan with specific next steps.
What each pillar measures
Prompt writing
Prompt writing is the skill of translating what you want into instructions a model can follow. It covers whether you provide context the model needs (role, goal, format, constraints), whether you iterate when the output misses, whether you know when to break a complex task into sub-prompts, and whether you use techniques like examples, role assignment, and step-by-step reasoning when they help. Most people who use AI regularly have intermediate prompt skills; the gap between intermediate and strong is usually knowing when and how to decompose hard tasks.
Tool selection
AI tool selection is knowing which tool to use for which job, and when not to use AI at all. This covers whether you know the real differences between the main models and tools, whether you choose based on the task rather than habit, whether you are aware of when AI output needs verification versus when it can be trusted, and whether you track new tools that might be better for tasks you do regularly.
Workflow automation
Workflow automation is where AI moves from answering one-off questions to saving real time at scale. This pillar measures whether you have identified and automated your most repetitive AI use cases, whether you pipe outputs from one tool to the next, whether you use APIs or integration tools rather than always going through a chat interface, and whether you have documented the workflows that work so you can repeat them reliably.
Critical evaluation
Critical evaluation is the skill that prevents AI from making you less accurate than you were before. It measures whether you verify AI-generated facts before using them, whether you recognise the linguistic signals of low-confidence output, whether you know the specific domains where models hallucinate most (dates, citations, statistics, code edge cases), and whether you have a consistent habit of checking outputs before acting on them.
The four score bands
| Score | Band | What it typically means |
|---|---|---|
| 0 – 30 | Beginner | Using AI occasionally, mostly for chat and simple questions |
| 31 – 55 | Developing | Regular AI user with some prompting skill; limited automation |
| 56 – 75 | Proficient | Consistent AI workflows; aware of limitations and verification |
| 76 – 100 | Expert | Advanced prompt engineering, integrated automation, strong critical habits |
Tips for an honest, useful result
- Rate your actual habits, not your aspirations. The plan is only useful if the weak areas it finds are real.
- Focus on the breakdown, not the headline number. A 70 with one very weak pillar is a clearer signal than the average alone.
- Act on the two-area plan first. Spreading effort across all four pillars at once dilutes progress; fix the weakest two, then retake.
- Retake quarterly to confirm your AI habits are genuinely improving over time.