Table of Contents
Overview
This page brings together all the cumulative components for Unit 3. After completing Lessons 3.1–3.4, you are now ready to consolidate your learning through a comprehensive exercise, a homework set, a culminating project, a Module 1 synthesis activity, a cumulative quiz, and a journal guide.
These components are designed to help you integrate the concepts of purpose, suitability, risk, proportionality, authority, oversight, transparency, accountability, escalation, stop conditions, and responsible refusal into a coherent framework. They also prepare you for Assignment 1 and for the transition to Module 2.
The consolidation components are formative and ungraded (unless your instructor designates otherwise). They are intended to help you solidify your understanding, identify gaps, and build confidence before moving on to prompt engineering and advanced AI collaboration in Module 2.
A. Unit 3 Consolidation Exercise
- Purpose: Apply all Unit 3 concepts to a single synthetic use case.
- Estimated time: 60–90 minutes.
Use Case: AI‑Generated Study Guides
A university is considering using generative AI to create personalised study guides for students in a large introductory course. The AI would take course materials (lecture notes, slides, readings) and generate summaries, key concepts, and practice questions tailored to each student's performance on previous quizzes. The study guides would be reviewed by a teaching assistant before being released to students.
Your task
Work through the following 24‑item analysis. For each item, provide a clear, concise response based on the use case. You may use the worked examples from Lessons 3.1–3.4 as a guide.
- Define the human problem. What is the need, and what evidence supports it?
- State the legitimate purpose. Is the purpose lawful, ethical, clear, proportionate, and consistent with human authority?
- Identify users and affected stakeholders. Who benefits? Who bears burden?
- Identify the proposed AI role. Which stages (information, evidence, analysis, recommendation, decision, action) will the AI assist with?
- Identify a non‑AI alternative. What other approach could address the problem?
- Describe the technical capability. What can the AI do?
- Describe a technical limitation. What might the AI get wrong?
- Develop an opportunity‑risk pair. Use the AI Paradox.
- Apply ORBIT. Address each element.
- Assess consequence. What are the possible consequences of error?
- Assess information sensitivity. How sensitive is the information?
- Assess detectability. How easily can errors be detected?
- Assess reversibility. Can errors be corrected?
- Consider scale and vulnerability. How many students are affected? Are any vulnerable?
- Assign a preliminary risk category. Lower, moderate, higher, or unsuitable?
- Identify required evidence. What must be available for human review?
- Assign human authority. Who has authority for each stage?
- Design meaningful oversight. What makes the TA's review meaningful?
- Identify transparency requirements. What should be disclosed to students?
- Assign accountability. Who is accountable for the final study guides?
- State an escalation condition. When would you escalate?
- State a stop condition. When would you stop the use?
- Make a recommendation. Choose from the allowed options and justify.
- Identify remaining uncertainty. What is still unknown?
Response template
You may use this template to structure your responses:
- 1. Human problem: ...
- 2. Legitimate purpose: ...
- [continue for all 24 items]
Show self‑assessment checklist
- ☐ I have addressed all 24 items.
- ☐ My responses are specific to the use case (not generic).
- ☐ I have used concepts from all four lessons.
- ☐ My reasoning is clear and justified.
- ☐ I have identified safeguards and uncertainties.
- ☐ I have made a recommendation and justified it.
Show staged hints
Hint 1 – Human problem: Students in a large course may struggle to synthesize course material into effective study guides. Evidence might include low exam performance, student requests for more guidance, or instructor observations.
Hint 2 – Legitimate purpose: The purpose is to improve student learning outcomes by providing personalised, accessible study materials. It is lawful, ethical, and proportionate.
Hint 3 – Risk category: Consider the consequence of error (students may learn incorrect information) and the detectability/reversibility (TA review provides a safeguard). This is likely moderate risk.
Hint 4 – Recommendation: Given the TA review, the use may be appropriate with safeguards. Consider "Proceed with safeguards" or "Pilot or limit".
Show solution design (illustrative partial)
Partial solution for key items:
- 1. Human problem: Students in a large introductory course often struggle to create effective study guides from extensive course materials, leading to lower performance and frustration.
- 2. Legitimate purpose: To improve student learning outcomes by providing personalised, accessible study guides that help students focus on key concepts and areas of weakness.
- 3. Stakeholders: Students (beneficiaries), teaching assistants (reviewers, burden bearers), instructors (accountability), university (reputation).
- 4. AI role: Analysis (synthesise materials) and recommendation (generate study guides).
- 5. Non‑AI alternative: Students create their own guides; TAs hold review sessions; instructors provide detailed study guides.
- 15. Risk category: Moderate risk – errors could affect learning, but TA review provides a safeguard.
- 23. Recommendation: Proceed with safeguards – pilot with a small cohort, ensure TA review is meaningful, and gather feedback from students.
Note: This is a partial solution. Your full response should address all 24 items.
B. Unit 3 Cumulative Homework Set
This homework set covers all the key concepts from Unit 3. It is formative and ungraded, but it provides excellent preparation for the cumulative project and for Assignment 1.
Scenario: A marketing team uses AI to generate 100 social media posts per hour. The posts are generic and often require substantial editing to match the brand voice.
Task: Is this effective or merely efficient? Justify your answer.
Show sample answer
Answer: This is efficient but not necessarily effective. Efficiency is about speed and volume; effectiveness is about achieving the intended purpose (engaging the audience with accurate, on‑brand content). The need for substantial editing undermines effectiveness.
Scenario: A researcher uses AI to generate a literature review, but does not verify the citations or claims.
Task: Explain why this is not responsible use. What should the researcher have done?
Show sample answer
Answer: This is not responsible because the researcher failed to verify the AI's output. Responsible use requires verification of claims, source checking, and ensuring that the AI's contribution is transparent and accountable. The researcher should have cross‑checked all citations and claims against original sources.
Scenario: A statement: "We should use AI to automate our customer service."
Task: Rewrite this as a problem‑centred statement. Identify the current problem, affected stakeholders, and intended improvement.
Show sample answer
Problem‑centred rewrite: "Our customer service team is overwhelmed by routine inquiries, leading to long wait times and reduced satisfaction. We need to improve response efficiency while maintaining quality. Affected stakeholders: customers, service staff, and management. Intended improvement: faster, consistent responses to routine inquiries, freeing staff for complex cases."
Scenario: A school uses AI to monitor student online activity to identify "at‑risk" students.
Task: Evaluate the legitimacy of this purpose. What information is needed to determine if it is appropriate?
Show sample answer
Answer: The purpose is potentially legitimate but requires careful consideration of privacy, consent, and the effectiveness of the intervention. Key questions: Is the monitoring lawful? Is it ethically defensible? Are students and parents informed? What safeguards are in place to prevent misuse? If these are not addressed, the purpose may be questionable.
Scenario: A non‑profit uses AI to generate personalised fundraising appeals.
Task: Define specific, observable success criteria for this use.
Show sample answer
- Donation response rate increases by at least 10%.
- Average donation amount increases by 5%.
- Donor satisfaction (surveyed) remains high.
- Time spent by staff on appeals decreases by 30%.
- No increase in donor complaints about content.
Scenario: A city uses AI to prioritise road maintenance projects based on traffic data.
Task: Identify all affected stakeholders, including indirect and burden‑bearing stakeholders.
Show sample answer
- Direct users: City engineers, maintenance crews.
- Indirect beneficiaries: Drivers, pedestrians, local businesses.
- Burden bearers: Residents near prioritized roads (construction disruption), maintenance crews (workload), engineers (review and oversight).
- Potential burden: Communities that are deprioritized may experience longer delays in road repairs.
Scenario: A health clinic wants to use AI to translate patient instructions into multiple languages.
Task: Assess the suitability of this use. What factors favour AI? What factors require caution?
Show sample answer
Favours AI: Speed, scalability, cost‑effectiveness for common languages.
Caution: Medical translation errors could have serious consequences. The AI may not capture nuance, cultural context, or medical terminology accurately. Requires human review by a qualified translator or medical professional.
Conclusion: Suitable only with strong human review and verification.
Scenario: A university wants to use AI to generate feedback on student writing.
Task: Compare the AI approach with a non‑AI alternative (peer review or instructor feedback). When might the non‑AI approach be preferable?
Show sample answer
AI approach: Scalable, consistent, immediate feedback, but may miss nuance and struggle with creativity.
Non‑AI (peer/instructor): Higher quality, contextual, empathetic, but slower and more costly.
Non‑AI preferable when: Feedback is high‑stakes, requires deep understanding, or is for creative work where AI may be inappropriate.
Scenario: An AI system is used to recommend job candidates to a recruiter.
Task: List at least five possible consequences of an error in the AI's recommendation.
Show sample answer
- Qualified candidates may be overlooked.
- Unqualified candidates may be advanced.
- The company may make a poor hiring decision.
- Legal liability for discrimination.
- Reputational damage to the company.
- Wasted time and resources in interviewing unsuitable candidates.
Scenario: An AI system is used to detect fraud in financial transactions. Errors are rare but are often only discovered months later.
Task: Assess the detectability of errors. What factors contribute to low detectability?
Show sample answer
Detectability: Low – errors are rare and only discovered after significant delay.
Factors: Lack of immediate feedback, complex financial systems, limited human review, and the subtlety of fraud patterns.
Scenario: An AI system is used to adjust thermostat settings in an office building to save energy.
Task: How reversible are the effects of an error (e.g., making the building too hot or too cold)?
Show sample answer
Reversibility: High – the temperature can be quickly adjusted back. However, temporary discomfort to occupants may not be fully reversible.
Scenario: An AI system is used to generate draft responses to customer support emails. The responses are reviewed by a human before sending.
Task: What safeguards are proportionate for this use?
Show sample answer
- Mandatory human review of every response.
- Access to customer history for context.
- Training for reviewers.
- A fallback process for complex inquiries.
- Logging of AI use and corrections.
Scenario: An AI system is used to draft press releases for a government agency. The drafts are reviewed by a communications officer before release.
Task: Classify this use as lower, moderate, higher, or unsuitable. Justify.
Show sample answer
Moderate risk. The consequence of error (misinformation, reputational harm) is significant, but human review provides a safeguard. The use is not higher risk because it does not affect safety, rights, or significant financial decisions.
Scenario: An AI system is used to assist in diagnosing medical conditions. The final diagnosis is made by a doctor.
Task: Who holds the authority to make the final diagnosis? Why can't the AI?
Show sample answer
Final authority: The doctor.
Why not AI: The doctor has professional training, legal authority, and accountability. The AI cannot take responsibility for the diagnosis or the treatment that follows.
Scenario: A loan officer reviews AI‑generated credit scores but is pressured to approve 95% of applications to meet performance targets.
Task: Is the oversight meaningful? Why or why not?
Show sample answer
Not meaningful. The pressure to approve undermines the officer's independence and ability to disagree. This is nominal oversight.
Scenario: An online retailer uses AI to personalise product recommendations, but does not disclose this to customers.
Task: Is this transparency sufficient? What should be disclosed?
Show sample answer
Not sufficient. Customers should be informed that recommendations are AI‑generated and based on their browsing and purchase history. This allows them to understand how the system works and to exercise choice.
Scenario: An AI system is used to generate performance reviews for employees. The reviews are then used for promotions.
Task: Who is accountable for the final review? Who is accountable if the review is unfair?
Show sample answer
Accountability: The human manager who approves the review is accountable. The organisation is also accountable for the fairness of the process. The AI system cannot be held accountable.
Scenario: An AI system flags a student for academic misconduct based on similarity scores. The student disputes the flag.
Task: What is the appropriate escalation path?
Show sample answer
Escalation: The case should be escalated to an academic integrity committee or a senior instructor for a full review. The student should have the opportunity to present their case.
Scenario: An AI system used for content moderation starts generating a high rate of false positives, blocking legitimate content.
Task: What stop condition should be in place?
Show sample answer
Stop condition: If the false positive rate exceeds a predefined threshold (e.g., 5%), the system should be paused for investigation and recalibration. Human moderators should take over during the pause.
Scenario: A school wants to use AI to monitor student mental health based on social media activity.
Task: Is this a responsible use? If not, what is the responsible refusal?
Show sample answer
Not responsible. This use invades privacy, may be unreliable, and lacks the human judgment needed for mental health. The responsible refusal is to not use AI for this purpose and instead to rely on trained counsellors and trusted reporting channels.
C. Unit 3 Cumulative Project
- Purpose: Integrate all Unit 3 knowledge into a comprehensive assessment report.
- Estimated effort: 3–4 hours.
Instructions
Select one bounded AI use from the following contexts (or choose your own with instructor approval):
- Education or learning
- Professional practice
- Organisational activity
- Communication
- Creative work
- Public information
- Everyday life
Develop a comprehensive report that addresses all 39 sections listed below. Use the worked examples from Lessons 3.1–3.4 as models.
Required sections
- Use‑case title
- Human problem
- Evidence of need
- Legitimate purpose
- Intended outcome
- Users
- Affected stakeholders
- Current or non‑AI process
- Alternative approaches
- Proposed AI role
- Model or system characteristics
- Required information
- Information authority
- Technical capability
- Technical limitations
- AI Paradox opportunity‑risk pair
- ORBIT analysis
- Possible errors
- Consequences
- Detectability
- Reversibility
- Scale
- Stakeholder vulnerability
- AI autonomy
- Preliminary risk classification
- Evidence and verification requirements
- Human authority
- Reviewer requirements
- Review timing
- Transparency
- Accountability
- Escalation
- Stop conditions
- Capability and relationship considerations
- Residual uncertainty
- Final recommendation (choose from allowed list)
- Reasons
- Connection to Assignment 1 reflection
- Questions for later course units
Allowed recommendations
- Proceed
- Proceed with safeguards
- Pilot or limit
- Redesign
- Use a non‑AI alternative
- Obtain specialist review
- Defer
- Do not proceed
Deliverable: Submit a written report (approximately 1,250–1,500 words or equivalent structured submission).
- Use only synthetic, public, or authorised information.
- Do not use real personal, confidential, or proprietary information.
- Do not use a use case that involves a high‑impact decision unless you have the appropriate expertise and authority.
Show self‑review checklist
- ☐ All 39 sections are addressed.
- ☐ The reasoning is clear and justified.
- ☐ Risks are identified and mitigated with proportionate safeguards.
- ☐ Human authority and oversight are meaningful.
- ☐ Transparency and accountability are explicit.
- ☐ Escalation and stop conditions are defined.
- ☐ The recommendation is justified and chosen from the allowed list.
- ☐ Residual uncertainty is documented.
- ☐ The report is well‑structured and professional.
Show solution architecture
Structure for a strong report:
- Title and overview – brief summary of the use case.
- Problem and purpose – define the human problem and legitimate purpose.
- Stakeholders and outcomes – identify who is affected and what success looks like.
- Alternatives and AI role – compare non‑AI alternatives and specify the AI's role.
- Risk assessment – use the AI Paradox and ORBIT, assess consequences, detectability, reversibility, scale, vulnerability, and autonomy.
- Risk classification and safeguards – classify risk and design proportionate safeguards.
- Authority and oversight – assign human authority and design meaningful oversight.
- Transparency and accountability – draft statements.
- Escalation and stop – define triggers.
- Capability and relationship – consider human skills and relationships.
- Residual uncertainty and recommendation – document unknowns and state final decision.
- Reflection – connect to Assignment 1 and future units.
Show illustrative outline (different use case)
Use case: AI‑generated project status reports for a construction company.
- Problem: Project managers spend hours writing status reports, delaying other work.
- Purpose: To produce accurate, timely reports that keep stakeholders informed.
- AI role: Generate draft reports from site data and notes.
- Risk: Inaccurate reports could mislead stakeholders and affect project decisions.
- Safeguards: Mandatory project manager review, source verification, and a correction process.
- Recommendation: Proceed with safeguards, pilot on one project, and monitor accuracy.
Note: This is a partial outline. Your report should be more detailed and include all 39 sections.
D. Module 1 Synthesis Activity
- Purpose: Link the three units of Module 1 into a coherent understanding.
- Estimated time: 30–45 minutes.
In Unit 1, you learned what generative AI is and how its capabilities create opportunities and risks. In Unit 2, you learned how AI produces and extends outputs, including its technical limitations. In Unit 3, you learned how to evaluate whether and under what conditions AI should be used.
Your task
For the following use case, answer the nine synthesis questions below. This activity integrates all three units.
Use case: A local government wants to use generative AI to draft responses to routine citizen inquiries about recycling schedules, permit applications, and service requests. The responses would be reviewed by a staff member before being sent.
- What capability is involved? (Unit 1)
- How does the system produce or obtain the output? (Unit 2)
- What technical limitations matter? (Unit 2)
- What opportunity is intended? (Unit 1)
- What corresponding risk exists? (Unit 1)
- Is the use contextually suitable? (Unit 3)
- What evidence and oversight are required? (Unit 3)
- Who remains accountable? (Unit 3)
- What is the final recommendation? (Unit 3)
Show solution design
- 1. Capability: Text generation, natural language understanding, information retrieval from a knowledge base.
- 2. Output production: The system takes the citizen's query, retrieves relevant information from a database, and generates a fluent, context‑appropriate response using a language model.
- 3. Limitations: May misinterpret queries, hallucinate facts, or fail to handle nuanced cases. Context window limits may exclude relevant information.
- 4. Opportunity: Faster response times, reduced staff workload, consistent information delivery.
- 5. Risk: Inaccurate or misleading information could frustrate citizens, undermine trust, and lead to legal liability.
- 6. Suitability: Suitable – the task is bounded, the information is public/authorised, and human review provides a safeguard.
- 7. Evidence and oversight: Staff must have access to authoritative information sources and must review each response for accuracy. A feedback mechanism for citizens is needed.
- 8. Accountability: The staff member who reviews and sends the response is accountable. The government agency is accountable for the overall service.
- 9. Recommendation: Proceed with safeguards – pilot with low‑risk inquiries, ensure thorough staff training, and monitor performance.
E. Unit 3 Cumulative Self‑Check Quiz
This comprehensive quiz covers all the key concepts from Unit 3. Each question includes explanatory feedback.
Q1. What is the first step in the responsible‑use decision sequence?
Show answer and explanation
Correct answer: b) Define the human problem.
Explanation: The decision sequence begins with defining the human problem, not with the technology. This ensures that the use is grounded in a legitimate need.
Q2. Which of the following is a characteristic of effective AI use?
Show answer and explanation
Correct answer: c) It contributes to a legitimate human purpose with adequate quality.
Explanation: Effectiveness is about achieving the intended purpose with adequate quality, not just speed, volume, or effort reduction.
Q3. A non‑AI alternative should be considered because:
Show answer and explanation
Correct answer: b) It may be more reliable, transparent, or cost‑effective.
Explanation: Non‑AI alternatives are not inherently inferior. They may be preferable depending on the context.
Q4. Which of the following is not a factor in risk assessment?
Show answer and explanation
Correct answer: c) The AI's training data size.
Explanation: Training data size is a technical detail, not a risk factor. Risk factors include consequence, sensitivity, scale, vulnerability, detectability, reversibility, and autonomy.
Q5. Proportionality in safeguards means:
Show answer and explanation
Correct answer: b) Matching safeguards to the level of risk.
Explanation: Proportionality means that safeguards should be commensurate with the seriousness and nature of the risk.
Q6. A use is classified as unsuitable when:
Show answer and explanation
Correct answer: b) The purpose is illegitimate or authority is absent.
Explanation: Unsuitability arises from fundamental issues like illegitimate purpose, lack of authority, or unacceptable residual risk, not from minor practical concerns.
Q7. Meaningful human oversight requires all of the following except:
Show answer and explanation
Correct answer: c) Approval of the AI's recommendation by default.
Explanation: Meaningful oversight requires the authority to disagree and the ability to reject or revise. Automatic approval is a sign of nominal oversight.
Q8. Nominal oversight is characterised by:
Show answer and explanation
Correct answer: c) Rubber‑stamping without meaningful checks.
Explanation: Nominal oversight is superficial, often due to volume, time pressure, or lack of authority.
Q9. Which of the following is not a form of human authority?
Show answer and explanation
Correct answer: c) AI authority.
Explanation: Authority is a human concept. AI systems do not hold legitimate authority; they are tools used by humans.
Q10. Responsible refusal is:
Show answer and explanation
Correct answer: b) A competent decision to limit AI use when conditions are not right.
Explanation: Responsible refusal is a professional judgment that prioritises human welfare and accountability.
Q11. A stop condition is:
Show answer and explanation
Correct answer: b) A predefined trigger that halts the AI process.
Explanation: Stop conditions are safety mechanisms that stop a process when certain criteria are met.
Q12. The AI Paradox describes the relationship between:
Show answer and explanation
Correct answer: b) AI capability, opportunity, and risk.
Explanation: The AI Paradox highlights that the same capabilities that create opportunities also create corresponding risks.
Q13. ORBIT stands for:
Show answer and explanation
Correct answer: c) Opportunities, Risks, Beneficiaries and Burden Bearers, Intervention and Oversight, Transparency and Accountability.
Explanation: ORBIT is a framework for structuring a responsible‑use analysis.
Q14. Which of the following is not an allowed recommendation?
Show answer and explanation
Correct answer: c) Automate without human oversight.
Explanation: Automation without human oversight is not a responsible recommendation because it lacks human authority and accountability.
Q15. Personal AI boundaries should be:
Show answer and explanation
Correct answer: b) Reviewed and updated over time.
Explanation: Boundaries should evolve as you gain experience, as technology changes, and as your context shifts.
F. Assignment 1 Journal Guide
- Weight: 15%
- Completion period: Modules 1 through 6
This guide helps you prepare your journal entries for Unit 3 and Module 1.
Unit 3 Journal Topics
Reflect on the following topics, drawing on your work in Lessons 3.1–3.4:
- How do you distinguish effective from efficient AI use?
- Why should a human problem be defined before selecting AI?
- Describe one use for which a non‑AI alternative may be preferable.
- How do consequence, detectability, and reversibility affect risk?
- What does meaningful human oversight require?
- Why is human presence not equal to human control?
- What do transparency and accountability mean to you?
- Describe a situation requiring specialist authority.
- Describe a condition requiring escalation.
- Describe a condition requiring stopping.
- Give an example of responsible refusal.
- How have your preliminary personal AI boundaries changed?
- How has your understanding of generative AI changed from Unit 1 to Unit 3?
Module 1 Synthesis Prompts
- How do Units 1, 2, and 3 fit together to give you a complete picture of responsible AI use?
- What is the relationship between the AI Paradox, ORBIT, and the responsible‑use decision sequence?
- What is the most important insight you have gained from Module 1?
Suggested Organization
- For each entry, start with a specific example or scenario from your own experience or from the course.
- Apply the concepts from the unit (e.g., suitability, risk, oversight).
- Reflect on what you learned and how your thinking has changed.
- Identify any remaining questions or uncertainties.
Evidence to Retain
- Your responses to exercises, homework, and mini‑projects.
- Notes from your personal AI boundary charter.
- Your synthesis activity responses.
- Any AI interaction logs (if you used an AI system for activities).
AI‑Interaction Documentation Reminder
If you used an AI system for any course activities, document the interaction (prompts, outputs, and your review). This provides evidence for your journal and demonstrates your responsible use.
Academic‑Integrity Reminder
Your journal should reflect your own thinking and experience. Do not use AI to generate your journal entries. Using AI to produce your personal reflections would undermine the purpose of the assignment and violate academic integrity.
Privacy Reminder
Do not include personal, confidential, or proprietary information in your journal. Use synthetic, public, or generalised examples.
Show self‑review checklist for journal entries
- ☐ My entry is grounded in a specific example or scenario.
- ☐ I have applied course concepts (e.g., suitability, risk, oversight).
- ☐ I have reflected on my own learning and changes in my thinking.
- ☐ I have considered both opportunities and risks.
- ☐ I have identified uncertainties or questions.
- ☐ My entry is written in my own words.
Show reflection planning checklist
- ☐ I have reviewed all Unit 3 lesson materials.
- ☐ I have completed the consolidation exercise and project.
- ☐ I have drafted responses to at least 5 of the journal topics.
- ☐ I have connected my learning to my own context (academic, professional, or personal).
- ☐ I have documented any AI use for activities.
- ☐ I have reviewed the Assignment 1 rubric.
Looking Ahead
Congratulations on completing Module 1! You have built a strong foundation in responsible and effective AI use. You now understand:
- What generative AI is and how it works (Units 1–2).
- How to define problems, assess suitability, evaluate risks, and design oversight (Unit 3).
- How to make integrated responsible‑use decisions and establish personal boundaries (Unit 3).
You are now ready to move on to Module 2: Prompt Engineering and Human‑AI Collaboration, beginning with Unit 4: Prompt Design Fundamentals. There, you will learn to craft effective prompts that leverage AI's capabilities while maintaining control, quality, and responsibility.
- Complete the consolidation exercise and cumulative project.
- Draft your Unit 3 journal entries.
- Review your personal AI boundaries and update them.
- Begin reading Unit 4 materials.