Suitability before use
Begin with the human problem, intended outcome, affected people, evidence of need, context, and alternatives.
Unit overview
Unit 3 establishes the decision principles that guide the remainder of COMP 327. A system may be technically capable of producing an output while still being unsuitable for a particular purpose, person, discipline, or consequence. You will distinguish efficiency from effectiveness and net benefit, compare AI-supported and non-AI alternatives, define appropriate AI roles, preserve human agency and authority, identify meaningful oversight, and establish verification, escalation, and stop conditions.
Effectiveness beyond speed
Count preparation, generation, verification, correction, hidden work, consequences, and capability effects.
Human control and responsibility
Keep judgment, approval, decisions, correction, escalation, and accountability with appropriate people.
Central idea: Technical capability does not establish
contextual suitability. Responsible use may require limitation,
redesign, specialist review, a non-AI alternative, or refusal.
Unit learning outcomes
By the end of Unit 3, you should be able to:
- Distinguish technical capability from contextual suitability.
- Define a legitimate human purpose, intended outcome, and evidence of need before selecting an AI system.
- Compare AI-supported, conventional digital, human-led, and non-AI alternatives.
- Distinguish efficiency, effectiveness, and net benefit.
- Identify preparation, verification, correction, coordination, hidden work, and downstream consequences.
- Classify an AI role as tool, assistant, critic, simulator, recommender, or agent.
- Distinguish assistance, analysis, recommendation, decision, approval, action, and accountability.
- Identify meaningful human review and recognize automation bias, nominal oversight, and responsibility laundering.
- Apply the AI Paradox and author-approved ORBIT framework to a bounded use case.
- Define information boundaries, evidence requirements, escalation, fallback, and stop conditions.
- Make and justify a qualified proceed, proceed-with-safeguards, redesign, use-an-alternative, or stop recommendation.
- Revise preliminary personal boundaries for effective and responsible AI use.
Unit 3 lessons
Complete the five lessons in order. For this unit, lesson
filenames are locally numbered from t-01.html through
t-05.html. Keep the homepage and lesson files in the
same Unit 3 folder.
3.1 From Capability to Suitability
Define the human problem, legitimate purpose, intended outcome, context, affected people, evidence of need, and alternatives before deciding whether AI belongs in the workflow.
3.2 Efficiency, Effectiveness, and Net Benefit
Compare visible time savings with preparation, verification, correction, coordination, hidden labour, quality, learning, relationships, and downstream effects.
3.3 AI Roles and Human Authority
Classify the AI role, separate assistance from accountable decisions and actions, and identify information, review, approval, decision, action, appeal, and stop authority.
3.4 Meaningful Oversight, Verification, and Stopping
Evaluate whether human review is informed and empowered, recognize automation bias and responsibility laundering, and establish verification, escalation, fallback, and stopping.
3.5 ORBIT Integration and Responsible Decisions
Integrate opportunity, corresponding risk, beneficiaries, oversight, and transparency using the author-approved ORBIT framework, then make a qualified final recommendation.
Folder convention: Because Unit 3 uses
t-01.html through t-05.html, place Unit 3 in
its own Moodle folder or content package to avoid conflicts with
lesson files from other units.
Embedded Unit 3 study guide
This sequence concludes Module 1 during Week 2. Follow the official Moodle schedule if dates differ.
-
Define purpose
Start with the human problem, outcome, context, and alternatives. -
Test value
Compare efficiency, effectiveness, net benefit, and hidden work. -
Allocate authority
Bound the AI role and preserve meaningful human control. -
Decide responsibly
Apply ORBIT, safeguards, escalation, and stopping.
Suggested study sessions
Total: 7–9 hours-
Session 1: Retrieval and reading, 60–75 minutes.
Retrieve the AI Paradox and Unit 2 system map, then read the opening sections of Chapter 3. -
Session 2: Lesson 1, 70–85 minutes.
Complete the purpose-and-suitability analysis and compare AI with non-AI alternatives. -
Session 3: Lesson 2, 70–85 minutes.
Build a total-effort and net-benefit comparison that includes verification and hidden work. -
Session 4: Lesson 3, 75–90 minutes.
Classify the AI role and map human information, review, approval, decision, action, appeal, and stop authority. -
Session 5: Lesson 4, 75–90 minutes.
Audit meaningful oversight and define verification, escalation, fallback, and stopping. -
Session 6: Lesson 5 and consolidation, 90–120
minutes.
Apply ORBIT, complete the unit self-check, update Assignment 1, and prepare to begin Module 2.
Evidence to retain
- Purpose-and-suitability analysis
- AI-versus-non-AI comparison
- Total-effort and net-benefit worksheet
- AI-role and human-authority map
- Meaningful-oversight audit
- ORBIT integration analysis
- Revised personal responsible-use boundaries
- Assignment 1 reflection notes
Show a responsible decision sequence
- Define the human problem and legitimate purpose.
- Identify affected people and intended outcomes.
- Compare non-AI and lower-risk alternatives.
- Identify demonstrated capability and remaining uncertainty.
- Define the bounded AI role.
- Assign human review, decision, action, and stop authority.
- Estimate total effort, consequences, and net benefit.
- Apply ORBIT and define transparency.
- Add verification, escalation, fallback, and stopping.
- Make a qualified recommendation and document residual uncertainty.
Assignment 1 continues in Unit 3
Assignment 1: AI Learning and Reflection Journal is worth 15%. Use Unit 3 to record how your thinking changed from “Can AI do this?” to “Is AI suitable, effective, and responsibly governed for this purpose?”
Official Moodle due date and time: ______________________________
Evidence to preserve
- Suitability and alternatives analysis
- Net-benefit comparison
- Human-authority map
- Oversight and stopping analysis
- Final ORBIT recommendation
- Updated personal boundaries
Reflection boundaries
- Write personal reflection yourself.
- Support factual claims with course evidence where appropriate.
- Describe what changed and why, rather than merely listing activities.
- Follow the assignment’s AI-use and disclosure requirements.
Show Unit 3 reflection-planning questions
- When did a technically capable use become contextually unsuitable?
- What hidden verification or correction work changed your estimate of value?
- Which decisions or actions must remain human?
- What makes review meaningful rather than nominal?
- Which stop condition or non-AI alternative did you add?
- How did ORBIT change your final recommendation?
Use these prompts to plan your own reflection. Do not treat them as model responses.
Required learning activities
| Activity | Purpose | Expected product | Estimated time |
|---|---|---|---|
| Suitability test | Distinguish capability from suitability using purpose, context, evidence, people, and consequences. | Suitability analysis | 35–45 minutes |
| Alternatives comparison | Compare AI-supported, human-led, conventional digital, and non-AI options. | Alternatives matrix | 30–45 minutes |
| Net-benefit worksheet | Count total effort, quality, verification, correction, hidden work, and downstream effects. | Net-benefit record | 40–55 minutes |
| AI-role classification | Classify the AI as tool, assistant, critic, simulator, recommender, or agent. | Role and boundary statement | 25–35 minutes |
| Human-authority map | Assign review, decision, action, correction, appeal, and stop authority. | Responsibility map | 35–50 minutes |
| Oversight audit | Identify automation bias, nominal review, responsibility laundering, and inadequate stopping. | Audit with safeguards | 35–50 minutes |
| ORBIT mini-analysis | Integrate opportunity, risk, beneficiaries, oversight, and transparency. | Qualified recommendation | 45–60 minutes |
Participation options
Option A: Approved AI system
Use an approved system only for low-risk, reversible tasks with synthetic, public, supplied, or authorized information.
Option B: Supplied workflow
Use supplied prompts, outputs, verification records, time estimates, authority maps, and decision scenarios.
Option C: Non-AI analysis
Complete the same suitability, alternatives, net-benefit, responsibility, and ORBIT analyses without an AI platform.
All options assess the same judgment and responsible-decision outcomes.
Responsible-use boundaries and support
Boundaries for Unit 3
- Do not enter confidential, personal, proprietary, privileged, security-sensitive, or culturally governed information.
- Do not treat system permission as human, professional, institutional, or community authority.
- Do not allow AI to approve, decide, transact, publish, delete, or take consequential action.
- Do not describe nominal human presence as meaningful oversight.
- Stop when suitable evidence, authority, review, correction, or remedy is unavailable.
Support and accessibility
- Use supplied or non-platform options when live AI is inaccessible, unavailable, unsuitable, or unwanted.
- Contact the course coordinator for scope, assignment, or permitted-use questions.
- Seek qualified professional, privacy, security, accessibility, legal, ethical, or community review when required.
- Use technical support for Moodle and file problems.