Course Objectives

This course introduces students to Human-AI Interaction and the design of interactive systems powered by artificial intelligence. Students will learn the human-centered design process for creating AI systems, including how to identify user needs, prototype AI-powered interfaces, evaluate usability, anticipate failure cases, and assess whether systems respect human values.

The course examines how people interact with AI systems, how humans participate in the data and machine learning pipeline, and how AI can be integrated into society in ways that are useful, responsible, and ethically grounded. Students will build several AI-powered interactive technologies and develop both practical design skills and critical frameworks for evaluating the social impacts of AI.

Prerequisite: CS5200: Foundations of AI.

Upon completion of this course, students should be able to:

• Apply a human-centered design process to create interactive AI systems.
• Design, prototype, and evaluate AI-powered interfaces that are useful, usable, and aligned with human values.
• Critically analyze how AI systems affect people, communities, institutions, and society.
• Identify where AI systems can fail, create harm, or violate human values and rights.
• Explain the role of humans in the data, training, evaluation, and deployment stages of machine learning systems.
• Justify where iteration, feedback, and human oversight are most important in the AI development pipeline.
• Recognize technical and social challenges in Human-AI Interaction, including usability, trust, bias, accountability, and value alignment.

Grading

• Reading Reflections: 20% — approximately once per week
• Quizzes: 5% — lowest quiz score dropped
• Mini Projects: 40%
• Final Project: 35%

Excellence in Assignments and Exams

Requirements for an A:

Students will receive an A if they either achieve a final score in the top 10–15% of the class after curving, or earn a raw score of 90% or higher before the curve, whichever standard is lower.

Week 1: Class Logistics, Introduction to Human Centered AI, and History of AI.

Assignment: 1-page summary of who you are 📕:

• Your cultural background.
• Background in AI and HCI.​
• A bit of what you know or have done with artificial intelligence.
• What you're looking to get out of class.

Mini Project: The Limits of the Algorithm — A Case Study in Human-Centered AI

Objective

This assignment challenges you to move beyond a purely technical understanding of Artificial Intelligence. By analyzing a real-world scenario, you will:

• Identify ethical, social, and design flaws in an AI system.
• Explain the potential negative consequences of these flaws for individuals, organizations, and society.
• Critique the idea that AI failures can be solved through technical fixes alone and argue for a broader human-centered approach.

Part 1: Foundational Knowledge — Learning from AI Failures

Before completing the case study, watch at least three of the following videos. Take notes on the major themes, examples, and lessons presented.

Why We Need an Ethical Approach to AI

• Why We Need Responsible AI — Nokia Bell Labs
• Why Do We Need Responsible AI? — IBM Research
• AI Is Dangerous, but Not for the Reasons You Think — TED

Where AI Has Failed

• AI Has a Fatal Flaw—And Nobody Can Fix It — Slidebean
• Apple's AI Disaster — A Rare Failure — ColdFusion
• Gen AI Gone Wild: How Artificial Intelligence Keeps Failing Us — Fireship

Part 2: Case Study — "Optimal Hire," The AI Recruitment Tool

Imagine a fast-growing technology company, Innovate Inc., has developed an AI-powered hiring system called Optimal Hire. The system automates the initial screening of thousands of job applications.

The Goal

The AI system is designed to predict which applicants are most likely to become top performers within the company.

The Training Data

The system was trained using ten years of company hiring data, including resumes, performance reviews, and promotion records. Historically, many of the company's highest-performing managers and engineers have been men from a small number of elite universities.

The Process

The AI analyzes resumes and assigns applicants a Success Score from 1 to 100. Human Resources staff are instructed to interview only applicants who receive a score of 85 or higher.

The Black Box Problem

The AI system uses highly complex models that even its developers cannot fully explain. They know that factors such as university prestige, previous employers, and specific keywords strongly influence outcomes, but they cannot explain why any individual applicant receives a specific score.

The Result

The company celebrates the tool because it increases hiring efficiency by 400%.

Part 3: Essay Assignment (1,000–1,200 Words)

Drawing on both the videos and the case study, write an essay that addresses the following questions:

1. Identify the Flaws

Identify and describe at least three major flaws in the design and implementation of the Optimal Hire system. Consider technical, ethical, organizational, and societal dimensions.

2. Explain the Problems

For each flaw you identify, explain why it is problematic and who may be harmed.

Consider the impact on:

• Individual Applicants: fairness, opportunity, transparency, and access.
• Innovate Inc.: diversity, innovation, reputation, and legal risk.
• Society: reinforcement of historical inequalities and systemic bias.

3. Why Code Is Not Enough

This is the most important section of your essay.

Explain why simply improving the algorithm or collecting more data would not fully solve the problems associated with Optimal Hire.

Discuss the importance of non-technical and human-centered solutions, including:

• Human oversight and intervention in hiring decisions.
• Transparency and explainability.
• Diverse development teams.
• Inclusive design practices.
• Regular auditing and ethical review processes.

Submission Requirements

• Length: 1,000–1,200 words
• Format: PDF
• Citations: Include references to the videos and any additional sources used.

In-Class Discussion Requirement

You must be prepared to discuss, defend, and elaborate on your analysis during class.

During the class following the deadline, students may be asked to explain their arguments and reasoning. Failure to demonstrate a clear understanding of your own submission may result in a score of zero regardless of the quality of the written essay.

Grading Rubric

Reading Assignment: Understanding Human-Centered AI

Objective

This assignment is designed to help you develop a deeper understanding of Human-Centered AI (HCAI). By engaging with conceptual videos and a scientific paper, you will explain why HCAI is critical for developing artificial intelligence systems that are beneficial, ethical, reliable, and effective for society.

Part 1: Foundational Concepts — Video Analysis

Before reading the academic paper, watch the following videos to build foundational knowledge about Human-Centered AI:

• What is Human-Centered AI? — Valuable Insights from Peter Norvig
• A Call for Human-Centered AI | DLD 23
• What is Human-Centered AI and Why Do You Need It?

As you watch, take notes on the key themes, including the goals of HCAI, the problems it aims to solve, and the difference between augmenting human capabilities and replacing human work.

Part 2: Academic Deep Dive — Scientific Paper Reading

After watching the videos, read the following scientific paper by Ben Shneiderman, one of the leading researchers in Human-Centered AI:

• Shneiderman, B. Human-Centered Artificial Intelligence: Reliable, Safe & Trustworthy

As you read, focus on how Shneiderman defines Human-Centered AI, the principles of reliability, safety, and trustworthiness, and the challenges involved in balancing human control with computer automation.

Part 3: Synthesis and Explanation — Written Response

Based on the videos and the scientific paper, write a clear, well-structured essay of approximately 750–1,000 words that addresses the following:

1. Define Human-Centered AI: In your own words, synthesize the definitions from the videos and the paper. What are the core principles of HCAI? How does it differ from a purely technology-centered or capability-focused approach to AI development?
2. Explain Why HCAI Matters: Drawing on specific examples from the assigned materials, explain why Human-Centered AI is important. Discuss at least three of the following:
   • Ethical considerations such as fairness, bias, and accountability
   • Building trust and user acceptance
   • Ensuring safety and reliability
   • Augmenting human skills and creativity rather than simply automating tasks
   • Solving real-world human problems more effectively
3. Connect Theory to Practice: The paper discusses the need to balance human control and computer automation. Explain what this balance means. Using a real or hypothetical AI system, such as a medical diagnostic tool, autonomous vehicle, or personalized learning platform, describe what a human-centered design approach would look like.
4. Future Challenges: Identify what you see as the biggest challenge to implementing a truly human-centered approach to AI development in the coming years. Justify your answer using insights from the assigned materials.

Submission Guidelines

• Your essay should be well organized, with clear paragraphs.
• Cite specific ideas or examples from both the videos and the paper.
• All work must be your own. Plagiarism will result in a failing grade.

Grading Rubric

• Assignment Submission: 100 points for a complete and thoughtful response that addresses all required parts.
• In-Class Engagement: Mandatory pass/fail component. You must be able to discuss and defend the ideas in your submitted work during class.

If you are unable to demonstrate a clear understanding of the material you submitted, your grade for the entire assignment will be changed to 0.

Reading Assignment Current Events: The People vs. Google — Deconstructing the Verdict

Objective

This assignment requires you to analyze the landmark antitrust trial between the U.S. Department of Justice and Google. Your goal is to synthesize information from multiple sources to explain the case, the verdict, and the unique role that emerging technologies such as generative AI played in the proceedings.

This assignment is designed to strengthen your research, writing, and critical thinking skills, while also preparing you to discuss and defend your analysis during class.

Your Task

Write and publish a blog post of at least 100 words explaining the key elements of the Google antitrust trial. Your post should be written for an educated but non-expert audience and should clearly explain the significance of the case.

Your blog post must include the following sections:

1. The Heart of the Lawsuit

   Briefly explain why the U.S. government sued Google. What were the main accusations regarding Google's dominance in the search market?

2. The Verdict Explained

   Describe the outcome of the trial. Did Google win or lose? Explain the court's reasoning and the evidence that influenced the decision.

3. The Unexpected Player: Generative AI's Role

   Analyze how the rise of generative AI systems such as ChatGPT, Gemini, Claude, and other AI assistants influenced the context of the trial.

   Consider questions such as:

   • Did generative AI serve as evidence regarding competition in search?
   • Did it influence how experts viewed Google's market power?
   • Did it affect the judge's perspective on the future of search and competition?
   • How might AI-powered search reshape the competitive landscape?
4. What Happens Next?

   Discuss the potential consequences of the verdict for Google, the technology industry, regulators, and consumers.

Submission Requirements

• Format: You may publish your post on Medium, LinkedIn Articles, WordPress, or submit a formatted PDF or Word document.
• Length: Minimum of 100 words.
• Sources: You must cite at least three credible sources.
• Acceptable Sources: Major news outlets, academic publications, government reports, official court documents, or other reputable sources.
• Include a Sources or References section at the end of your blog post.

Grading

A complete and thoughtful submission that addresses all required sections and includes appropriate citations will receive 100 points.

In-Class Engagement Requirement

You must be prepared to discuss, defend, and elaborate on the content of your blog post during the next class session.

Students may be called upon to explain their key takeaways, answer questions about the case, and discuss the role of generative AI in the trial.

If you are unable to intelligently discuss the content of your submission, your grade for the assignment will be changed from 100 to 0.

The purpose of this assignment is not simply to write a blog post, but to learn, understand, and engage with the material.

Week 2: AI History + Designing AI/ML UX.

Mini Project: Predicting the Future from the Code of the Past

Objective

This assignment challenges you to think like both a historian and a futurist. By studying the cyclical nature of AI development, including breakthroughs, AI winters, hype cycles, and unintended consequences, you will analyze current events in AI and predict possible future trajectories.

Part 1: Analytical Essay — Choose One Prompt (50%)

Choose one of the following essay prompts.

Essay Prompt 1: The Echo of an AI Winter?

Background:

The history of AI includes periods of intense optimism and investment, often called AI Springs, followed by periods of disappointment and funding cuts, known as AI Winters. The 1970s and late 1980s saw such winters after early promises went unfulfilled. Today, we are experiencing a major AI boom, with massive investment in generative AI. However, a recent MIT study reported that many business attempts to integrate generative AI are failing to deliver a return on investment.

Your Task:

Based on your understanding of historical AI Winters, analyze the current situation. Are we seeing early signs of another AI winter, or are the fundamentals of today's AI revolution different enough to sustain current momentum? Predict how the current investment and hype cycle may unfold over the next 5–10 years. Use specific parallels from AI history to justify your prediction.

Essay Prompt 2: The Ghost in the Machine — Old Problems in New Code

Background:

A recurring theme in AI history is the emergence of unexpected and problematic consequences. Early machine translation systems in the 1960s produced comical but often unusable results. Early expert systems were brittle and lacked common sense, which limited their usefulness in real-world settings. Today, new problems are emerging, including AI-AI bias, where AI models may prefer content generated by other AI systems over human-created content.

Your Task:

Analyze the emergence of AI-AI bias as a modern example of an unforeseen consequence of AI development. How does this issue parallel historical AI problems such as brittleness, lack of common sense, or embedded bias? Based on these historical parallels, predict the long-term social or economic problems that could arise if AI-AI bias becomes widespread and unchecked.

Part 2: Coding Assignment — Sentiment Analysis of AI Hype Cycles (50%)

Objective:

To complement your essay, this coding assignment asks you to quantitatively analyze AI hype cycles by performing sentiment analysis on historical AI-related news headlines.

Your Task:

Write a Python script that reads a dataset of fictional AI-related news headlines from different eras and determines the overall sentiment of each period.

Dataset: Fictional AI Headlines

              era,headline
                1980s_Boom, The Future is Now: Expert Systems to Revolutionize Every Industry 
                1980s_Boom, Intelligent Machines Promise a New Era of Economic Prosperity 
                1980s_Boom, Government Pours Millions into Fifth-Generation Computer Project 
                1990s_Winter, AI Fails to Deliver: The Unfulfilled Promise of Thinking Machines 
                1990s_Winter, Funding for AI Research Dries Up Amidst Disappointing Results 
                1990s_Winter, Expert Systems Deemed Too Brittle for Real-World Use 
                2020s_GenAI_Boom, Generative AI Set to Add Trillions to the Global Economy 
                2020s_GenAI_Boom, Breakthroughs in Large Language Models Astonish Researchers 
                2020s_GenAI_Boom, Is Artificial General Intelligence Finally Within Reach? 
                2020s_GenAI_Boom, Concerns Mount Over AI Job Displacement and Automation

Requirements

• Read the Data: Your script should process the provided headline data.
• Sentiment Analysis: Use a Python library such as TextBlob or NLTK's VADER to analyze the sentiment of each headline. Classify each headline as positive, negative, or neutral.
• Aggregate Results: Calculate the average sentiment score for each era: 1980s_Boom, 1990s_Winter, and 2020s_GenAI_Boom.
• Visualize: Create a simple bar chart using Matplotlib or Seaborn to visualize the average sentiment for each era. The chart should clearly show the highs of the boom periods and the lows of the winter period.
• Submit: Submit your Python script along with the generated chart.

Connecting the Coding to the Essay

In your essay, you may refer to the results of your sentiment analysis to provide a data-driven perspective on historical AI hype cycles.

Submission and Discussion Requirement

You must be prepared to discuss your submitted work in class. Failure to engage with what you submitted will result in a grade of zero for the assignment.

Reading Assignment: Deep Dive into AI History and Human-Centered Artificial Intelligence

Objective

This assignment is designed to develop your understanding of the historical evolution of Artificial Intelligence and the emergence of Human-Centered AI as a response to technical, social, and ethical challenges. Through critical analysis and visual synthesis, you will create analytical diagrams that demonstrate your understanding of major AI paradigm shifts, recurring historical patterns, and the principles of Human-Centered AI.

You will also be expected to defend your analytical choices and engage in discussion during class.

Required Readings

• A Brief History of Artificial Intelligence: On the Past, Present, and Future of Artificial Intelligence
• A Brief History of AI: How to Prevent Another Winter (A Critical Review)
• Human-Centered Artificial Intelligence: Reliable, Safe & Trustworthy

Assignment Components

Diagram 1: AI Paradigm Evolution Timeline

Create an analytical timeline that demonstrates your understanding of the major phases, breakthroughs, setbacks, and paradigm shifts in AI history.

Your timeline must:

• Identify and analyze at least six critical junctures in AI history.
• Explain the assumptions, limitations, and breakthroughs that drove each transition.
• Connect historical developments to current AI trends and possible future directions.
• Demonstrate an understanding of why certain approaches succeeded or failed.
• Show relationships between technological capabilities, societal needs, funding environments, and research priorities.

Diagram 2: Human-Centered AI Implementation Framework

Design a framework that illustrates how Human-Centered AI principles can be translated into real-world systems and design practices.

Your framework must:

• Identify core components of Human-Centered AI systems and associated implementation strategies.
• Provide concrete examples of AI systems that demonstrate or fail to demonstrate Human-Centered AI principles.
• Show relationships between trustworthiness, reliability, safety, transparency, and ethics.
• Identify potential tensions or trade-offs between competing design goals.
• Demonstrate awareness of practical challenges involved in implementing Human-Centered AI.

In-Class Defense and Discussion Requirements

You must be prepared to present and defend any aspect of your diagrams during class discussion.

This includes:

• Deep Analysis: Explain your reasoning for design choices and why you emphasized specific relationships, events, or concepts.
• Critical Thinking: Discuss alternative interpretations, limitations, and implications of the readings.
• Real-World Application: Connect historical AI developments to current challenges and emerging technologies.
• Comparative Analysis: Compare different AI paradigms, design philosophies, and implementation strategies.
• Active Participation: Engage with classmates' work through thoughtful questions and constructive feedback.

What Constitutes Meaningful Engagement?

To receive credit for this assignment, you must demonstrate:

• Knowledge of key concepts from all three readings.
• The ability to explain your analytical choices using evidence from the readings.
• An understanding of the connections between AI history and Human-Centered AI.
• The ability to discuss and defend your interpretations.
• Evidence that you have carefully read and understood the assigned materials.

Grading Policy

• Submission of Complete Assignment (100 Points): Full credit will be awarded for submitting both diagrams with all required components and annotations.
• Class Discussion Participation (Mandatory): Students who cannot meaningfully discuss their diagrams or demonstrate understanding of the readings when called upon will receive a grade of 0 for the entire assignment, regardless of the quality of the submitted work.

Deliverable

Submit two analytical diagrams with detailed annotations:

1. AI Paradigm Evolution Timeline
2. Human-Centered AI Implementation Framework

You should also be prepared to discuss and defend your work during class.

Week 3: Mockups and Storyboards for Designing Interactive AI Systems.

Reading Assignment Current Events: Reflections on Meta Connect 2025

Objective

This assignment requires you to analyze and reflect on the key announcements and technological advancements presented during the Meta Connect 2025 keynote address. The goal is to critically assess the future direction of the technologies discussed at the event through the lens of Human-Centered AI.

Your Task

Watch the main keynote presentation from Meta Connect 2025 and write a reflective essay of approximately 500–750 words.

Your essay should address the following topics:

1. Summarize the Key Announcements

   Identify and explain the two or three most significant announcements from the event. These may include new hardware, software updates, advances in artificial intelligence, mixed reality technologies, wearable devices, or other major developments.

2. Analyze the Impact

   Choose the single announcement you believe is the most groundbreaking. Explain why it is important and discuss its potential impact on consumers, developers, researchers, businesses, or society more broadly.

3. The Metaverse Vision

   Discuss how the announcements advance, modify, or challenge Meta's long-term vision for the metaverse. Did the keynote make you more optimistic or more skeptical about this vision? Explain your reasoning.

4. Human-Centered AI Perspective

   Evaluate one of the announcements using concepts from Human-Centered AI. Consider questions such as:

   • Does the technology support human well-being?
   • Does it enhance human agency and control?
   • Does it create new risks or concerns?
   • Does it improve the overall user experience?

   Support your analysis with specific examples from the keynote.

Submission Guidelines

• Submit your assignment as a PDF document through the course portal.
• Use 12-point Times New Roman font.
• Double-space your document.
• Include your name and the submission date at the top of the document.

Grading

This assignment is graded primarily on completion and thoughtful engagement with the material.

• 100 Points: Submit a complete essay that thoughtfully addresses all required sections.
• 0 Points: Failure to submit the assignment or inability to discuss your submission during class.

In-Class Discussion Requirement

You must be prepared to discuss, defend, and elaborate on the ideas presented in your essay during class discussion.

Students may be asked to explain their key observations, defend their evaluations, and discuss the implications of Meta's technologies from a Human-Centered AI perspective.

If you are unable to demonstrate familiarity with your own submission or meaningfully discuss the content you submitted, your grade for the assignment will be changed to 0.

Reading Assignment: Understanding Mockups and Storyboards for Human-Centered AI

Objective

This assignment is designed to help you develop an understanding of mockups and storyboards by synthesizing information from academic papers and targeted video tutorials. You will then apply this knowledge to create visual diagrams that communicate the purpose, role, and value of these design tools in the development of Human-Centered AI systems.

Part 1: Engage with the Content

To begin, you must engage with the resources provided. The goal is to understand not only the definitions of mockups and storyboards, but also their practical application and importance throughout the design lifecycle.

A. Required Videos

Watch all three videos:

• What is a Mockup? — Envato Tuts+
• How to Create a UX Storyboard — Nielsen Norman Group
• Learn UX Storyboarding in 5 Minutes: Steps and Examples — UXtweak

B. Required Readings

Read all papers provided in the Mockups/Storyboards folder located within the Readings section on Canvas.

Part 2: Synthesize and Create

As you review the materials, synthesize your notes and consider the following questions:

• What are the primary functions of mockups and storyboards in the creation of Human-Centered AI systems?
• How do mockups and storyboards differ from one another?
• Where do these tools fit within the design and development process?
• How might learning about mockups and storyboards help you design new AI-enhanced tools, systems, and platforms?

A. Create a Diagram

Based on your synthesis of the readings and videos, create a diagram that illustrates the relationship between mockups, storyboards, and the overall Human-Centered AI design process.

You may use tools such as Lucidchart, Miro, Figma, PowerPoint, Canva, or create your diagram by hand (provided it is clean and legible).

Your diagram should:

• Provide a high-level overview of the Human-Centered AI design process.
• Define and distinguish key concepts such as mockups and storyboards.
• Show how these design techniques contribute to the creation of Human-Centered AI systems.
• Illustrate relationships between users, design artifacts, AI functionality, and iterative design processes.

B. Write a Brief Explanation

Write a 1–2 paragraph explanation accompanying your diagram.

Your explanation should describe:

• The core idea your diagram communicates.
• How specific concepts from the readings and videos are visually represented.
• Explicit connections between your diagram and the source materials. For example: "As explained in the Nielsen Norman Group video, UX storyboards focus on the user's experience and emotional journey. I represent this by..."
• The reasoning behind the structure and flow of your diagram.

Submission Requirements

Submit your diagram and written explanation as a single document.

Acceptable formats include PDF, Word, or PowerPoint.

Grading and In-Class Discussion Requirement

This assignment is graded based on your engagement and understanding of the material.

• Submission (100 Points): You will receive a provisional score of 100 for submitting a complete assignment that satisfies all requirements.
• In-Class Engagement (Final Grade Confirmation): You are expected to be an expert on the work you submit. During class discussion, you may be asked to explain your diagram, defend your design decisions, and discuss concepts from the videos and readings.

If you are unable to coherently explain your diagrams or demonstrate familiarity with the assigned materials, your assignment grade will be changed to 0.

The purpose of this assignment is not only to create a diagram, but also to demonstrate a meaningful understanding of how mockups and storyboards support the design of Human-Centered AI systems.

Week 4: Understanding User Needs for Designing Interactive AI Systems.

Mini Design: AI Feature Design from User Data

Objective

This assignment challenges you to act as both a product designer and developer. You will analyze user interaction data from a fictional project management application to identify patterns, user behaviors, and pain points. Using these insights, you will design and create a mockup for a new AI-powered feature that improves the user experience.

Total Points: 100

The Scenario

You are part of the product team for FlowState, a popular web-based project management platform. FlowState allows users to create projects, manage tasks and subtasks, assign work to team members, and visualize project timelines.

The product team has collected user interaction data over the last month and wants to explore opportunities for integrating Artificial Intelligence into the platform. Your job is to analyze the data, identify a meaningful user need, and propose an AI-powered solution.

The User Data

The following dataset contains user interaction logs collected from the FlowState platform over the last month.

flowstate_user_log.csv

timestamp,userID,sessionID,action,details,session_duration_minutes
2025-08-21 09:05:12,user_101,sess_abc,login,success,45
2025-08-21 09:06:20,user_101,sess_abc,view_dashboard,project_alpha,45
2025-08-21 09:07:31,user_101,sess_abc,create_task,{"task_name": "Draft Q3 Report"},45
2025-08-21 09:07:55,user_101,sess_abc,create_subtask,{"parent_task": "Draft Q3 Report", "subtask_name": "Gather sales data"},45
2025-08-21 09:08:21,user_101,sess_abc,create_subtask,{"parent_task": "Draft Q3 Report", "subtask_name": "Outline report structure"},45
2025-08-21 09:15:45,user_101,sess_abc,assign_user,{"task_name": "Gather sales data", "assigned_to": "user_102"},45
2025-08-21 09:25:11,user_101,sess_abc,set_due_date,{"task_name": "Draft Q3 Report", "date": "2025-09-15"},45
2025-08-21 09:40:30,user_101,sess_abc,change_view,gantt_chart_view,45
2025-08-21 09:50:12,user_101,sess_abc,logout,success,45

2025-08-21 10:10:02,user_105,sess_def,login,success,15
2025-08-21 10:11:15,user_105,sess_def,view_dashboard,project_beta,15
2025-08-21 10:12:00,user_105,sess_def,create_task,{"task_name": "Develop new login UI"},15
2025-08-21 10:13:45,user_105,sess_def,add_comment,{"task_name": "Develop new login UI", "comment_length": 120},15
2025-08-21 10:18:00,user_105,sess_def,upload_attachment,{"task_name": "Develop new login UI", "file_type": ".png"},15
2025-08-21 10:25:02,user_105,sess_def,logout,success,15

2025-08-21 11:02:19,user_101,sess_ghi,login,success,22
2025-08-21 11:03:00,user_101,sess_ghi,create_task,{"task_name": "Finalize marketing assets"},22
2025-08-21 11:03:30,user_101,sess_ghi,create_subtask,{"parent_task": "Finalize marketing assets", "subtask_name": "Approve banner ads"},22
2025-08-21 11:04:01,user_101,sess_ghi,create_subtask,{"parent_task": "Finalize marketing assets", "subtask_name": "Write social media copy"},22
2025-08-21 11:04:35,user_101,sess_ghi,create_subtask,{"parent_task": "Finalize marketing assets", "subtask_name": "Get legal review"},22
2025-08-21 11:24:19,user_101,sess_ghi,logout,success,22

2025-08-21 11:30:50,user_108,sess_jkl,login,success,5
2025-08-21 11:31:20,user_108,sess_jkl,view_dashboard,project_gamma,5
2025-08-21 11:32:00,user_108,sess_jkl,change_view,kanban_view,5
2025-08-21 11:33:00,user_108,sess_jkl,change_view,list_view,5
2025-08-21 11:34:00,user_108,sess_jkl,change_view,kanban_view,5
2025-08-21 11:35:50,user_108,sess_jkl,logout,success,5

2025-08-21 12:01:00,user_102,sess_mno,login,success,65
2025-08-21 12:02:15,user_102,sess_mno,view_dashboard,project_alpha,65
2025-08-21 12:03:00,user_102,sess_mno,complete_task,{"task_name": "Gather sales data"},65
2025-08-21 12:05:00,user_102,sess_mno,create_task,{"task_name": "Analyze competitor websites"},65
2025-08-21 12:05:40,user_102,sess_mno,create_subtask,{"parent_task": "Analyze competitor websites", "subtask_name": "Review competitor A"},65
2025-08-21 12:06:10,user_102,sess_mno,create_subtask,{"parent_task": "Analyze competitor websites", "subtask_name": "Review competitor B"},65
2025-08-21 12:45:00,user_102,sess_mno,change_view,gantt_chart_view,65
2025-08-21 13:06:00,user_102,sess_mno,logout,success,65

Part 1: Data Analysis and Insight Generation (50 Points)

Your first task is to analyze the FlowState user log and uncover meaningful patterns in user behavior.

Instructions

1. Set Up Your Environment

   Use a local Python environment with Pandas or an online notebook environment such as Google Colab.

2. Load the Data

   Write code to load the dataset into a Pandas DataFrame.

3. Answer the Following Questions Using Code
   • What are the five most frequently performed user actions?
   • What is the average session duration?
   • What pattern do you notice immediately after a create_task action?
     Hint: Examine actions that occur next within the same session.
   • User user_108 appears to exhibit a specific behavior. What is it, and what might it imply about the user's experience with the available view options?
4. Summarize Your Findings

   Write a brief 2–3 paragraph summary describing the most important insights from your analysis.

   Identify at least one clear user problem or opportunity revealed by the data. This problem will become the focus of your AI feature.

Deliverables for Part 1

• Python script or notebook file (.py or .ipynb)
• Written summary of findings
• Description of the user problem or opportunity identified

Part 2: AI Feature Mockup and Justification (50 Points)

Based on the problem or opportunity identified in Part 1, design an AI-powered feature that improves the user experience.

Instructions

1. Conceptualize the AI Feature

• Give your feature a clear, user-friendly name.
• Examples:
  • Smart Subtask Suggestions
  • AI Project Assistant
  • View Optimizer
• Write a short paragraph explaining:
  • What the feature does
  • How users interact with it
  • How it addresses the problem identified from the data

2. Create a Mockup

Create a visual mockup illustrating how the AI feature would appear and function within FlowState.

The mockup does not need to be high fidelity. A wireframe is acceptable as long as it clearly communicates the feature and user interaction.

Recommended Tools:

• Figma
• Balsamiq
• Adobe XD
• Moqups
• Google Slides
• PowerPoint
• Google Drawings
• Hand-drawn sketches (must be legible)

3. Justify Your Design

Provide several bullet points explaining your design decisions.

Every design choice should be directly connected to insights discovered during your analysis.

Example:

"Many create_task events are immediately followed by create_subtask events. Because of this pattern, I included a 'Generate Suggested Subtasks' button directly within the task creation workflow."

Deliverables for Part 2

• Mockup image, PDF, or slide deck
• Feature description
• Design justification document

Submission

Bundle all deliverables into a single folder or archive.

• Python script or notebook
• Written findings summary
• Problem/opportunity statement
• Mockup image or PDF
• Feature description
• Design justifications

Submit your work as either a ZIP file or a shared folder link.

Grading

In-Class Discussion Requirement

You must be prepared to discuss and defend your analysis, identified user problem, AI feature design, and mockup during class.

Students who are unable to explain their findings, justify their design decisions, or demonstrate familiarity with the submitted work may receive a grade of 0 regardless of the quality of the submission.

Reading Assignment: Understanding User Needs for Human-Centered AI Design

Objective

This assignment is designed to provide you with an understanding of methodologies for identifying and analyzing user needs in order to create new AI tools centered on those needs. You will explore these concepts through a combination of introductory videos and academic readings, and then apply this knowledge to the design of Human-Centered AI systems.

Part 1: Video Introduction

Please watch the following videos to gain an overview of user needs, user experience (UX) design, and user research methodologies. As you watch, take notes on key concepts, methodologies, and challenges discussed in each video.

• Understanding Users for Better UX
• Empathy in UX Design: Understanding User Needs
• How To Conduct UX Research Analysis (UX Design Guide)
• How I'd Learn UX Design (If I Could Start Over)
• Understand User Needs – UX Design Principles

Part 2: Academic Readings in HCI

The following resources introduce foundational Human-Computer Interaction (HCI) methodologies for understanding users and designing systems around their needs. As you read, focus on the purpose of each methodology, the types of insights it produces, and how it can support the development of Human-Centered AI systems.

Contextual Inquiry

Contextual Inquiry is a research method that combines observation and interviewing within a user's natural environment. The goal is to understand work practices, challenges, and behaviors as they occur in real-world settings.

• What are Contextual Interviews?

Participatory Design

Participatory Design emphasizes the direct involvement of users as partners in the design process. Rather than designing for users, designers work with users to co-create solutions.

• Participatory Design Reading

User-Centered Design (UCD)

User-Centered Design is an overarching philosophy and process that focuses on understanding user needs, iteratively designing solutions, and evaluating systems with users throughout development.

• What is User-Centered Design (UCD)?

Part 3: Assignment Questions

Based on the videos and readings, write a 2-page paper that addresses the following questions.

1. The Role of Context

   The videos and readings emphasize the importance of understanding a user's context. Why is context critical when designing AI systems?

   Discuss how factors such as environment, goals, workflows, social relationships, organizational constraints, and user motivations can influence the success or failure of an AI system.

2. Applying a Human-Centered Process

   Imagine that you have been tasked with creating a new AI-powered tool to help project managers track team progress.

   Outline a plan for the initial user-needs discovery phase of the project.

   In your response:

   • Identify which methodologies from the readings you would use.
   • Explain why you selected those methodologies.
   • Describe the steps you would take to understand users.
   • Explain what kinds of insights you would expect to gain before beginning the design process.

Submission Guidelines

• Length: 2 pages
• Font: 12-point font
• Spacing: Double-spaced
• Margins: 1 inch on all sides

Grading and Discussion Policy

This assignment is worth a total of 100 points.

You will receive 100 points for submitting a complete assignment that addresses the required questions.

However, a core component of this assignment is your ability to articulate and synthesize the material. You must be prepared to discuss your responses during class.

If you are unable to discuss the content of your submission when called upon, your grade for the assignment will be changed to 0.

What to Prepare for Class

• Be prepared to explain the purpose of Contextual Inquiry.
• Be prepared to discuss the value of Participatory Design.
• Be prepared to explain how User-Centered Design supports Human-Centered AI.
• Be prepared to defend your proposed user research plan.
• Be prepared to discuss how understanding user needs influences AI system design.

Week 5: Bootcamp for Creating User-Friendly AI Agents.

This week, we will be exploring:

• AI Agents in Practice: Understanding what AI agents are, how they differ from traditional software applications, and how they can perform tasks on behalf of users.
• Hands-On Agent Development: Participating in a practical bootcamp where you will learn how to design, build, and deploy AI agents that assist users with a variety of real-world tasks.
• User-Centered Agent Design: Examining how user needs, goals, workflows, and contexts should influence the design of AI agents.
• Accessibility and Inclusion: Learning strategies for ensuring that AI agents can be effectively used by diverse populations with different backgrounds, abilities, languages, and levels of technical expertise.
• Trust, Transparency, and Control: Exploring how AI agents can communicate their capabilities, limitations, and decision-making processes in ways that help users maintain confidence and control.
• Evaluating Agent Performance: Learning how to assess whether an AI agent successfully helps users achieve their goals while respecting human values and needs.

Learning Outcomes

By the end of this week's bootcamp, students should be able to:

• Explain the core components and capabilities of AI agents.
• Design AI agents that perform meaningful tasks for end users.
• Apply human-centered design principles when creating AI-powered systems.
• Identify barriers that may prevent different populations from effectively using AI agents.
• Implement design strategies that improve accessibility, inclusivity, and usability.
• Evaluate AI agents with respect to user experience, trust, transparency, and effectiveness.
• Critically reflect on the opportunities and risks associated with deploying AI agents in real-world settings.

Mini Project: Adapting an AI Assistant for Cultural Awareness and Inclusion

Objective

This assignment challenges you to apply the theoretical concepts of inclusive and secure AI discussed in the report Inclusive and Secure Artificial Intelligence: A Global Perspective to a practical coding project.

You will analyze the report, adapt the AI assistant you created during the bootcamp, and reflect on the challenges of building culturally aware and inclusive AI systems.

Total Points: 100

Required Materials

• Report: Inclusive and Secure Artificial Intelligence: A Global Perspective
• Podcast: Inclusive and Secure Artificial Intelligence Podcast Discussion

Part 1: Reading and Analysis (30 Points)

Before beginning the coding portion, read the report and focus on:

• Section 2: Inclusive AI
• Section 3: Advancing Secure AI
• Section 4: Policy Recommendations

Answer the following questions in 2–4 sentences each.

1. AI Is Not Neutral

   The report argues that AI is not neutral. Explain what this means using the example of facial recognition technology discussed in the report.

2. Digital Colonialism

   What is digital colonialism as described in the report, and how does it relate to the dominance of the Global North in AI development?

3. Content Moderation Challenges

   According to Section 3.1, why do AI content moderation systems often fail when dealing with non-Western languages and dialects? Provide one example from the report.

4. Cultural Impact Assessments

   What is the purpose of a cultural impact assessment for high-risk AI systems?

5. Participatory Design

   Why does the report advocate for participatory design and community involvement as a strategy for creating fairer AI systems?

Part 2: Coding Challenge — Adapting Your Bootcamp AI Assistant (40 Points)

In our recent Build Your Own AI Assistant bootcamp, you created a working prototype of an AI assistant. Your task is now to adapt that assistant using the principles of inclusion, cultural awareness, and fairness discussed in the report.

Even if your original assistant served a completely different purpose (e.g., homework support, gaming assistance, productivity coaching), you must add a layer of cultural intelligence to its behavior.

Required Adaptations and Enhancements

1. Culturally Aware Greetings

Modify your greeting system so that it can provide culturally appropriate greetings based on user preferences such as language, country, or region.

Connection to the Report:

This adaptation addresses linguistic diversity and reduces reliance on Western-centric defaults.

2. Diversify the Knowledge Base

Review the resources, examples, recommendations, or content generated by your assistant.

Adapt these materials so that they include perspectives and examples from diverse cultures.

Example:

• Diwali Festival of Lights
• International Film Night
• Lunar New Year Celebrations
• Local Community Festivals

Connection to the Report:

This addresses concerns about cultural homogenization and promotes cultural pluralism.

3. Audit for and Mitigate Bias

Review your assistant's outputs and identify potential biases, particularly gender stereotypes.

Adapt the assistant's language and responses to be gender-neutral and inclusive.

Examples:

• Avoid associating professions with specific genders.
• Avoid assumptions about interests, hobbies, or capabilities.
• Use inclusive language whenever possible.

Connection to the Report:

This directly addresses concerns regarding gender bias and discriminatory AI behavior.

4. Handle Ambiguity and Cultural Context

Improve your assistant's ability to recognize culturally dependent concepts and ask clarifying questions instead of making assumptions.

Examples:

• Holiday
• Family celebration
• Traditional food
• Community event

Connection to the Report:

This demonstrates a shift from a one-size-fits-all approach toward context-aware AI behavior.

Deliverable for Part 2

• Link to your updated AI assistant
• Description of the adaptations implemented
• Examples demonstrating each enhancement

Part 3: Reflection Essay (30 Points)

Write a reflection essay of approximately 400–600 words.

Your essay should address the following prompts:

1. Challenges of Retrofitting AI

   Describe the biggest challenge you faced while adapting your assistant to become more culturally aware.

   How does this challenge reflect the broader problems of retrofitting existing technologies discussed in the report?

2. Applying Policy Recommendations

   Choose two policy recommendations from Section 4 of the report.

   Examples include:

   • Integrating perspectives from diverse communities
   • Promoting transparency and accountability
   • Supporting participatory design
   • Strengthening cultural representation

   Explain how your modifications to the AI assistant attempt to implement these recommendations.

3. Future Improvements

   If you continued developing your assistant, what additional feature would you add to make it more inclusive or secure?

   Explain how your proposed feature connects to ideas presented in the report.

Submission Requirements

Submit a single PDF document containing:

• Your answers to Part 1
• A link to your adapted AI assistant for Part 2
• Your reflection essay for Part 3

Grading Breakdown

In-Class Discussion Requirement

You must be prepared to discuss your assistant, explain the modifications you made, and connect your design decisions to concepts from the report.

Students who cannot demonstrate familiarity with their own submission or explain how their assistant incorporates principles of inclusion and security may receive a grade of 0 regardless of submission quality.

Week 6: Problematic AI (AI Risks)

Mini Project: Adapting an AI Assistant for Cultural Awareness and Inclusion

Objective

This assignment challenges you to apply the theoretical concepts of inclusive and secure AI discussed in the report Inclusive and Secure Artificial Intelligence: A Global Perspective to a practical coding project.

You will analyze the report, adapt the AI assistant you created during the bootcamp, and reflect on the challenges of building culturally aware and inclusive AI systems.

Total Points: 100

Required Materials

• Report: Inclusive and Secure Artificial Intelligence: A Global Perspective
• Podcast: Inclusive and Secure Artificial Intelligence Podcast Discussion

Part 1: Reading and Analysis (30 Points)

Before beginning the coding portion, read the report and focus on:

• Section 2: Inclusive AI
• Section 3: Advancing Secure AI
• Section 4: Policy Recommendations

Answer the following questions in 2–4 sentences each.

1. AI Is Not Neutral

   The report argues that AI is not neutral. Explain what this means using the example of facial recognition technology discussed in the report.

2. Digital Colonialism

   What is digital colonialism as described in the report, and how does it relate to the dominance of the Global North in AI development?

3. Content Moderation Challenges

   According to Section 3.1, why do AI content moderation systems often fail when dealing with non-Western languages and dialects? Provide one example from the report.

4. Cultural Impact Assessments

   What is the purpose of a cultural impact assessment for high-risk AI systems?

5. Participatory Design

   Why does the report advocate for participatory design and community involvement as a strategy for creating fairer AI systems?

Part 2: Coding Challenge — Adapting Your Bootcamp AI Assistant (40 Points)

In our recent Build Your Own AI Assistant bootcamp, you created a working prototype of an AI assistant. Your task is now to adapt that assistant using the principles of inclusion, cultural awareness, and fairness discussed in the report.

Even if your original assistant served a completely different purpose (e.g., homework support, gaming assistance, productivity coaching), you must add a layer of cultural intelligence to its behavior.

Required Adaptations and Enhancements

1. Culturally Aware Greetings

Modify your greeting system so that it can provide culturally appropriate greetings based on user preferences such as language, country, or region.

Connection to the Report:

This adaptation addresses linguistic diversity and reduces reliance on Western-centric defaults.

2. Diversify the Knowledge Base

Review the resources, examples, recommendations, or content generated by your assistant.

Adapt these materials so that they include perspectives and examples from diverse cultures.

Example:

• Diwali Festival of Lights
• International Film Night
• Lunar New Year Celebrations
• Local Community Festivals

Connection to the Report:

This addresses concerns about cultural homogenization and promotes cultural pluralism.

3. Audit for and Mitigate Bias

Review your assistant's outputs and identify potential biases, particularly gender stereotypes.

Adapt the assistant's language and responses to be gender-neutral and inclusive.

Examples:

• Avoid associating professions with specific genders.
• Avoid assumptions about interests, hobbies, or capabilities.
• Use inclusive language whenever possible.

Connection to the Report:

This directly addresses concerns regarding gender bias and discriminatory AI behavior.

4. Handle Ambiguity and Cultural Context

Improve your assistant's ability to recognize culturally dependent concepts and ask clarifying questions instead of making assumptions.

Examples:

• Holiday
• Family celebration
• Traditional food
• Community event

Connection to the Report:

This demonstrates a shift from a one-size-fits-all approach toward context-aware AI behavior.

Deliverable for Part 2

• Link to your updated AI assistant
• Description of the adaptations implemented
• Examples demonstrating each enhancement

Part 3: Reflection Essay (30 Points)

Write a reflection essay of approximately 400–600 words.

Your essay should address the following prompts:

1. Challenges of Retrofitting AI

   Describe the biggest challenge you faced while adapting your assistant to become more culturally aware.

   How does this challenge reflect the broader problems of retrofitting existing technologies discussed in the report?

2. Applying Policy Recommendations

   Choose two policy recommendations from Section 4 of the report.

   Examples include:

   • Integrating perspectives from diverse communities
   • Promoting transparency and accountability
   • Supporting participatory design
   • Strengthening cultural representation

   Explain how your modifications to the AI assistant attempt to implement these recommendations.

3. Future Improvements

   If you continued developing your assistant, what additional feature would you add to make it more inclusive or secure?

   Explain how your proposed feature connects to ideas presented in the report.

Submission Requirements

Submit a single PDF document containing:

• Your answers to Part 1
• A link to your adapted AI assistant for Part 2
• Your reflection essay for Part 3

Grading Breakdown

In-Class Discussion Requirement

You must be prepared to discuss your assistant, explain the modifications you made, and connect your design decisions to concepts from the report.

Students who cannot demonstrate familiarity with their own submission or explain how their assistant incorporates principles of inclusion and security may receive a grade of 0 regardless of submission quality.

Week 7: Analyzing People's Data to Create HAI Interfaces

Reading Assignment Current Events: : A Human-Centered Critique of the “Superintelligence” Open Letter

Objective

This assignment requires you to read and critically analyze the open letter titled “We Are AGI Researchers, and We Are Asking for Caution” published in TIME.

You will evaluate the letter's arguments, concerns, and proposed solutions through the lens of Human-Centered AI (HCAI) principles.

Part 1: Summarize the Core Argument

Read the open letter carefully. In your own words, write a concise summary of approximately 200–250 words that addresses the following:

• Who are the authors of the letter and what is their primary concern?
• What specific risks associated with Artificial General Intelligence (AGI) and superintelligence do they highlight?
• What key actions or policies are they advocating for?

Part 2: Critical Analysis

Critique the letter using the Human-Centered AI principles discussed in class. Write a well-structured essay of approximately 700–800 words that addresses the following questions:

1. Whose “Center” Is It?

   From a human-centered perspective, which stakeholders does the letter seem to prioritize? Consider groups such as the public, developers, governments, specific communities, and affected workers. Are there important perspectives missing?

2. Alignment with HCAI Principles

   How do the letter's proposed solutions, such as increased funding for safety research and government regulation, align with HCAI principles such as accountability, transparency, and safety? Do these proposals go far enough?

3. Beyond Existential Risk

   The letter focuses heavily on existential risks, such as human extinction. From an HCAI perspective, what more immediate risks might this focus overlook, such as job displacement, bias, privacy erosion, misinformation, or unequal access to AI benefits?

4. A More Human-Centered Approach

   If you were to rewrite the letter's conclusion from a strictly human-centered viewpoint, what different recommendations would you make to ensure that AGI development serves humanity's best interests?

Deliverable

Submit a single document containing:

• Your summary for Part 1
• Your critical analysis essay for Part 2

Week 8: Understanding the End-to-End Human Centered AI Design Process

Mini Project: From Affinity Mapping to Personas, Mockups, and Storyboards

Objective

The goal of this project is to transform your user research findings into a coherent vision for a Human-Centered AI system.

Throughout the semester, we have explored methods for understanding users, identifying needs, and designing AI systems around human values. In this project, you will demonstrate how design decisions emerge directly from user data by:

• Deriving personas from your affinity mapping process
• Designing a high-fidelity mockup of your AI tool
• Creating a storyboard that illustrates how a user achieves a goal using your system

A central requirement of this project is traceability. Reviewers should be able to follow a clear path from raw data to affinity clusters, from affinity clusters to personas, and from personas to design decisions.

Submission Format

Submit a single document, PDF, or slide deck that contains all required project components.

Students may work individually or in teams. However, each student must submit their own copy of the final project.

Part 0: Data-Driven Personas (30 Points)

Goal: Demonstrate how your personas emerge directly from your data and affinity mapping process.

What to Include

1. Data Sources

Provide a brief description of the user data that informed your project.

Examples include:

• Interview transcripts
• Survey responses
• Field observations
• Contextual inquiry notes
• Participatory design sessions
• Online community discussions

2. Affinity Map Summary

Summarize the major themes that emerged from your affinity mapping process.

Include:

• 5–8 affinity clusters
• Short descriptive labels for each cluster
• 1–2 representative quotes per cluster

3. Persona Set

Create 2–3 personas that represent important user groups identified in your data.

Each persona should include:

• Name
• Background
• Goals
• Needs
• Frustrations
• Context of use
• Constraints
• Technology habits

4. Persona Justification

For each persona, include a short paragraph explaining:

• Which affinity clusters informed the persona
• Which quotes were most influential
• Why the persona represents a meaningful user group

Make the mapping between data and persona explicit.

Exit Criterion

A reviewer should be able to trace each persona directly back to specific affinity clusters and supporting quotes.

Part 1: Mockup (40 Points)

Goal: Communicate the broad features, functionality, appearance, and interaction design of your AI tool while grounding every major design decision in the needs of your personas.

What to Include

1. List of AI Features

Provide a list of the major AI capabilities your system offers.

Examples include:

• Recommendation systems
• Personalized guidance
• Summarization
• Planning assistance
• Conversational support
• Content generation
• Decision support

2. High-Fidelity Mockup

Create a mockup that illustrates:

• Overall layout
• Major screens
• Navigation structure
• Key interactions
• Important AI features

You may use tools such as:

• Figma
• Adobe XD
• Balsamiq
• PowerPoint
• Google Slides
• Hand-drawn sketches (if legible)

3. Design Rationale Callouts

Add annotations directly to your mockups explaining:

• Why interface elements were included
• Which persona needs they address
• Which user frustrations they help solve
• How the AI supports users in achieving their goals

Exit Criterion

A reviewer should be able to examine your mockup and clearly see how your affinity mapping and personas influenced the design.

Part 2: User Storyboard (30 Points)

Goal: Visualize a complete user journey using your AI system.

Your storyboard should tell a coherent story with a beginning, middle, and end.

Scenario Selection

Choose:

• One primary persona
• One concrete goal or task

Clearly describe the problem the user is trying to solve.

Narrative Structure

1. Problem

Describe:

• The user's initial situation
• The pain point they experience
• Why the problem matters

2. Interaction

Show the user interacting with your AI tool.

Include:

• Key interaction steps
• Moments where the AI assists the user
• Evidence of user control and agency
• Important interface screens or decisions

3. Resolution

Show:

• The successful outcome
• How the user's goal is achieved
• How the user knows the AI was helpful

Storyboard Components

• Storyboard panels or frames
• User actions
• User thoughts or emotions
• AI actions
• Captions describing events

Exit Criterion

Someone unfamiliar with your project should be able to understand the user's journey and how your AI design helps them achieve their goal.

Grading Summary

What to Turn In

Submit one document or slide deck containing:

• Data source description
• Affinity map summary
• Personas with justifications
• Mockups with rationale callouts
• Storyboard with captions

Optional Appendix:

• Additional affinity clusters
• Additional quotes
• Supporting research artifacts

The appendix is recommended if additional evidence is needed to demonstrate traceability between user research and design decisions.

Week 9: Mental Models, AI Design and Risk

Reading Reflections: AI Risks

Objective

In this assignment, you will analyze two research papers on AI risks and societal impact. Your goal is to use the readings to understand potential risks and harms in AI systems. You will also be expected to engage in a class discussion about the papers and videos.

Papers to Read

• The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
• Fairness and Abstraction in Sociotechnical Systems

Part 1: Reading and Summarizing

For each paper, write a 1–2 paragraph summary of the main concepts related to the risks and harms that AI systems can cause.

In your summary, highlight key examples or case studies from the papers that are relevant to your own final project.

Part 2: Video Analysis

Watch the following YouTube videos on AI risk and mitigation:

• Mastering AI Risk: NIST’s Risk Management Framework Explained
• AI Risk Mitigation Generator
• Conversational AI Risk Assessment & Mitigation

For each video, write a 1–2 paragraph analysis of the key takeaways.

In your analysis, explain how the concepts presented in the videos relate to the concepts discussed in the papers.

Discussion Preparation

You will be expected to participate in a class discussion about the papers and videos. To prepare, think about the following questions:

• What are the most significant risks and harms that AI systems can cause?
• How can we mitigate these risks and harms?
• What are the ethical implications of using AI systems?
• How can we ensure that AI systems are used for beneficial rather than harmful purposes?

Grading

You will receive 100 points for submitting the assignment.

However, if you are unable to engage meaningfully in the in-class discussion, you will receive a 0.

Week 10: Explainable AI, Mental Models, and Risks

Mini Project: AI Risk and Mitigation Plan

Objective

The goal of this assignment is to critically evaluate your proposed AI tool for potential risks, harms, and unintended consequences. You will identify key ethical challenges and develop a concrete plan to mitigate these risks.

This exercise is intended to help you think proactively about responsible AI design and ensure that your system is safe, fair, reliable, and beneficial for its intended users.

Assignment Instructions

This assignment is divided into three parts. Submit a single document that addresses all sections.

You may complete this assignment individually or with your final project team. If you choose to work as a team, every team member must submit the same final document to receive credit.

Part 1: Risk Identification and Analysis (40 Points)

Based on your final AI tool proposal, identify and analyze at least three potential risks associated with your system.

For each risk, provide a detailed explanation addressing the following components:

Risk Category

Clearly identify the category of risk. Possible categories include:

• Fairness and Bias: Could the tool produce unfair outcomes for certain groups?
• Privacy: Does the system collect, store, or process sensitive user information?
• Security: Could the system be vulnerable to attacks such as adversarial inputs, prompt injection, or data poisoning?
• Malicious Use: Could bad actors intentionally misuse the system in harmful ways?
• Reliability and Safety: What happens when the AI makes mistakes? Could errors result in significant harm?

Detailed Description

Describe the specific risk within the context of your project.

Your description should explain:

• How the risk could emerge
• What conditions might trigger it
• The worst-case scenario if the risk is not addressed

Affected Populations

Identify who would be most affected if the risk became reality.

Be specific. Consider:

• End users
• Specific demographic groups
• Workers
• Communities
• Organizations
• The broader public

Example Format

Part 2: Mitigation Plan (40 Points)

For each risk identified in Part 1, develop a mitigation plan that includes specific actions you can take during the design, development, and deployment of your AI system.

Design Stage

Describe design choices that can reduce or prevent the risk.

Examples:

• User interface changes
• Transparency mechanisms
• User control features
• Data minimization strategies
• Inclusive design practices

Development Stage

Describe technical measures you would implement.

Examples:

• Fairness metrics
• Bias testing
• Security protections
• Human-in-the-loop workflows
• Model monitoring systems

Deployment Stage

Describe policies, procedures, or safeguards that should be in place once the system is deployed.

Examples:

• User reporting mechanisms
• Terms of service
• Transparency statements
• Usage restrictions
• Periodic audits

Example Format

Part 3: Ethical Reflection (20 Points)

Write a brief reflection of approximately 2–3 paragraphs addressing the following questions:

1. How has this exercise changed your perspective on your final project?
2. Why is it important for AI engineers and designers to proactively consider risks before building and deploying AI systems?
3. What do you see as the biggest ethical challenge associated with your specific AI tool?

Your reflection should demonstrate thoughtful engagement with the risks, tradeoffs, and responsibilities involved in creating AI systems.

Submission Requirements

Submit a single document containing:

• Part 1: Risk Identification and Analysis
• Part 2: Risk Mitigation Plan
• Part 3: Ethical Reflection

Students working in teams may submit the same document, but every team member must upload their own copy.

Grading

Your submission will be evaluated on:

• The depth and quality of your risk analysis
• The feasibility and thoughtfulness of your mitigation strategies
• The insightfulness of your ethical reflection
• The degree to which your analysis is grounded in the specifics of your AI project

Week 11: AI Principles and Interface Evaluation

Final Project: Requirements and Evaluation

What You Need to Submit

Please upload the following materials:

• Slides for your 10-minute presentation.
• A short video, approximately 1–3 minutes, showcasing your demo.
• Code for your project, with enough structure and comments so it can be understood and, ideally, run.

Final Presentation Logistics

Each team or individual will have 10 minutes to present their project.

You may structure your presentation in the way that best tells the story of your work. For example:

• Problem
• Approach
• Demo
• Reflection

After each presentation, there will be a short Q&A:

• Classmates will provide feedback and ask questions.
• The presenter(s) will have approximately 1 minute to respond to questions.

Final Project Evaluation Criteria

Your final project grade will be based on the following criteria:

1. Integration of Human-Centered Design (40%)

• Clearly show how you incorporated Human-Centered AI principles into your project.
• Demonstrate attention to users’ needs, context, and values.

2. Narrative and Communication of Your Project (60%)

Within this category, we will evaluate the following:

Please let me know if you have any questions.

Good luck — I am really looking forward to seeing your projects!