FID-084 · Open question
Learning, Agency, and Relationships in AI-Assisted Education
Which forms of AI assistance improve immediate task performance while preserving independent judgment, durable learning, and appropriate human relationships?
Why the question remains open
Helpful answers and positive user ratings do not establish educational benefit. Assistance may support or substitute for the effort and relationships through which people learn to judge, seek help, and act independently.
Working hypothesis
A proposition to test, not a finding.
Scaffolded assistance that asks learners to reason and consult appropriate people will improve unaided transfer compared with answer-first assistance. It may provide no benefit or disadvantage learners with different access needs.
Proposed method
How the question could be tested
- 01Begin with adult learners and synthetic family-advice scenarios; compare no-AI, answer-first, and scaffolded assistance with matched learning objectives.
- 02Measure immediate performance, delayed unaided retention, transfer to unfamiliar tasks, error correction, appropriate help-seeking, and reliance calibration.
- 03Co-design a separate youth protocol with educators, developmental specialists, and family representatives before considering research with minors.
Needed controls
What must constrain the study
- 01Pre-register outcomes and distinguish satisfaction, time saved, learning, and well-being. Account for prior knowledge and accessibility needs.
- 02Require ethics review and informed consent; no covert monitoring of family relationships. Any youth study needs appropriate guardian consent and participant assent.
- 03Disclose assumptions about autonomy, beneficial learning, and family authority; do not treat one household structure or educational tradition as universal.
Relationship to existing work
Broadens the learning and agency questions while preserving FID-016 and FID-050 as distinct faith-context calls.
Expected outputs
Artifacts the work should produce
- 01A longitudinal learning-and-agency protocol and safe task materials.
- 02An analysis of performance versus durable-learning tradeoffs, including heterogeneous effects.
Open questions
Uncertainties the protocol must resolve
- 01How long must follow-up last to detect meaningful dependence or skill change?
- 02Which relational outcomes can be responsibly measured without intrusive observation?
Related calls
Continue through this research area
FID-038
Non-Calculability, Forgiveness, and Predictive Profiling
How should AI systems represent human change when they classify, score, rank, or predict people in contexts involving pastoral care, education, safeguarding, volunteer screening, hiring, discipline, membership, donor engagement, or community support?
FID-079
Presuppositions, Disagreement, and Evaluation Judgment
How do researchers' presuppositions shape evaluation design and interpretation, and can explicit disclosure make judgments more inspectable and appropriately trusted?
FID-004
Relational Substitution Risk in Faith-Facing AI
When does a faith-facing AI system move from supporting a user's religious life to substituting for embodied community, clergy, spiritual direction, family, therapy, or other human care?
Open question
Open work
Primary need: education research, developmental expertise, human factors, ethics review
- Educators, developmental researchers, and family representatives to co-design outcomes.
- Learning scientists and accessibility specialists to review the protocol.