Our approach
How we measure AI fluency
Most AI skills assessments ask people how good they think they are. Which mostly measures confidence. The gap between perceived capability and demonstrated capability tends to be wide — and it widens for people who are the most overconfident. Groovia is designed to address that gap.
01
The framework
AI fluency isn't a single skill. It's a cluster of related capabilities that develop at different rates and in different sequences depending on how someone uses AI in their work.
We measure across five domains:
AI Fundamentals
understanding how AI systems work, what they can and can't do, and how to evaluate their outputs critically
AI Interaction
the practical skill of directing AI effectively: constructing prompts, refining results, knowing when to push and when to stop
AI Application
integrating AI into real work: redesigning workflows, building automations, solving problems that resist simple prompt-and-response
AI Strategy
thinking at the level of teams and organisations: identifying opportunities, managing risk, governing AI use, scaling adoption
AI Growth
the meta-skill: learning continuously, adapting as the tools change, and helping others develop their own capability
Within each domain, we assess against five stages — from Unconscious Incompetence (you don't know what you don't know) through Conscious Incompetence, Conscious Competence, and Unconscious Competence, to Mastery. Stage doesn't just describe how much you know. It describes the relationship between your capability and your awareness of it.
02
Three ways to assess
Different questions call for different tools. Groovia offers three assessment types, each designed to reveal something slightly different.
The AI Fluency Self Assessment
~10 minA 25-question self-report across all five domains. It's honest about what it is: your own read on your capability. That's useful for getting a baseline, identifying where you feel uncertain, and tracking how your perception changes over time. It takes about ten minutes.
The Quick Checkup
~3 minA faster calibration — five domain ratings, useful for checking in regularly rather than doing a deep audit every time.
The Skills Tests
~15 min · ObjectiveDifferent in kind. Each is a domain-specific scenario test: you work through a realistic situation and make judgement calls. Your responses are scored against demonstrated capability, not self-report. They take around fifteen minutes each and are considerably harder to fake.
03
Why both matter
Taking a self-assessment and a Skills Test in the same domain often produces two different numbers. That gap is the most interesting result either assessment can give you.
A high self-assessment score with a lower Skills Test score suggests overconfidence — a common pattern in domains where people have surface familiarity but haven't tested their judgement under pressure. A low self-assessment with a higher objective score is less common, but it does happen: people who've been using AI effectively without realising it's a skill.
Neither result is a verdict. Both are information.