Question 03 / Capability
Does AI assistance help us become more capable?
Look beyond the immediate answer to learning, judgment, and lasting skill.
Why it matters
The human question.
Finishing a task and developing the ability to do it are different outcomes. We want tools that help people build the capabilities they value, including knowing when to rely on assistance.
Working hypothesis
What we expect—and need to test.
Our working hypothesis is that how assistance is offered affects later understanding and transfer to unfamiliar tasks.
Approach
How we would investigate.
Compare a complete generated answer, guided assistance, and an existing learning workflow on a bounded skill. Assess immediate results, explanation quality, and later performance on a different task with and without assistance.
Hold task difficulty and available time comparable. Account for prior experience and report immediate productivity separately from learning and independent performance.
Outcomes
What we would look for.
- Immediate task quality and effort
- Understanding and error detection
- Later performance on unfamiliar tasks
- Confidence compared with demonstrated ability
Reconsideration
What would change our thinking.
If short-term gains conceal weaker understanding or transfer, we would change the assistance design or narrow the claim about who it helps.
Current status
The study has not yet run.
This is a proposed study, with no participant results or measured effects to report. The sampling plan, final measures, and analysis need to be specified before the study begins.
Next step: Choose a skill with independently assessable outcomes and define the comparison.
Connected question