User Modeling
User Modeling
Embedded Files
Target User
Emerging Scholars of Color
Who faces greater barriers to adjusting in the US (i.e., language, culture, etc.)
Who are less exposed to training, resources, and lived experiences around AI
Who are under-represented in models despite being the greatest beneficiaries
Who are the next generation of great minds to advance research on AI
User Models
Personas*
John, 27
John, 27
An international master's student in education who extensively utilizes GPT for academic writing while navigating challenges related to AI attribution and ethical usage.
Kim, 30
Kim, 30
An international Ph.D. student in political science who heavily relies on ChatGPT for refining writing and summarizing papers while navigating challenges related to academic integrity and AI detection.
Zhang, 28
Zhang, 28
A Ph.D. student and teaching assistant who navigates challenges related to detecting and addressing the use of AI-generated content while prioritizing academic integrity and originality in her work and the work of her students.
John, 27
John, 27
Background
Age: 27
Occupation: International Master's Student in Education
Education: Pursuing a Master's degree in Education
Demographics
Location: Currently studying abroad, originally from China
Marital Status: Single
Income Level: Moderate, supported by scholarships and part-time work
Goals and Objectives
John aims to excel in his academic studies and overcome language barriers to achieve native-like fluency in English writing.
He seeks guidance on using AI ethically in academic writing to level the playing field with native English speakers.
John wants to focus more on academic content and less on language refinement, utilizing AI as a tool for enhancement.
Challenges and Pain Points
John encountered an uncomfortable situation where his work was mistakenly attributed to AI, causing misunderstandings and undermining his original contribution.
John feels frustrated by the misconception surrounding the extent of AI's role in his writing and the potential detection of AI-generated content.
John advocates for the transparent and ethical use of AI in academic writing, emphasizing the importance of documenting AI assistance.
Kim, 30
Kim, 30
Background
Age: 30
Occupation: International PhD Student in Political Science
Education: Pursuing a PhD in Political Science at a university outside the United States
Demographics
Location: Currently studying abroad, originally from South Korea
Marital Status: Single
Income Level: Moderate, supported by scholarships and research assistantship
Goals and Objectives
Kim aims to excel in her academic studies and effectively navigate the overwhelming literature in her field.
She utilizes ChatGPT extensively for refining English writing and summarizing lengthy papers to enhance her literature reviews.
Kim seeks ways to balance her reliance on ChatGPT while maintaining academic integrity and originality in her work.
Challenges and Pain Points
Kim feels overwhelmed by the sheer volume of literature to review in her field and seeks efficient ways to process and reflect on the material.
She is concerned about how much she can reuse or repurpose ChatGPT's output without compromising academic integrity.
Kim faces challenges related to false positives when using AI detectors to grade student work as a teaching assistant, impacting her trust in AI-generated content.
Zhang, 28
Zhang, 28
Background
Age: 28
Occupation: Teaching Assistant (TA) and Ph.D. student
Education: Bachelor's degree in China, Master's degree in the USA
Demographics
Location: Currently studying and working in the USA
Marital Status: Single
Income Level: Moderate, supported by research assistantship and TA position
Goals and Objectives
Zhang aims to excel in her academic studies and teaching responsibilities while upholding academic integrity and originality.
Zhang uses AI tools, such as ChatGPT, as companions for quick checks on concepts, brainstorming ideas, and generating inspiration without compromising her originality.
Challenges and Pain Points
Zhang faces challenges as a TA in detecting and addressing the use of AI-generated content in student assignments, particularly in online discussion posts.
She is concerned about potential bias in persuasive prompts or responses and has limited knowledge of prompt engineering to mitigate this issue.
*Note: Three images of personas were generated using Midjourney from the prompt of the description we present above.
Key Insights
Generative AI should be designed with a consideration for the broader populations (i.e, differences in language and culture)
A clear color-coding system to raise students' awareness on the content source (i.e., User-ideated, AI-generated, or Mixed/Collaborative Contribution) and make adjustments accordingly
The ways in which students' AI-powered work will be evaluated and presented should be made transparent and established on common grounds (i.e., a set of statements affirming the best efforts by AI developers and users)
The process of working on their draft, checking the source of contribution, submitting the paper should be integrated for a seamless experience (i.e. embed a word processor, reports from detector, and submit functionality)
Scenario**
Persona : Kim
Task : Draft and Revise an Academic Essay
Background : AH-AI, a cutting-edge educational platform, is designed to promote responsible AI use and improve academic writing. It offers real-time analysis and feedback, making it an invaluable tool for international graduate students like Kim, who navigate the complexities of academic writing in a second language and ethical AI use.
click for Kim's Scenario Detailed Walkthrough
click for Kim's Scenario Detailed Walkthrough
Scenario Detailed Walkthrough
Logging In and Setting Up:
Kim sits at her desk in the quiet corner of the university's study lounge, her laptop open and ready.
She navigates to the AH-AI platform and logs in using her university credentials. The clean, intuitive interface of AH-AI welcomes her, with clear options: “introductory Module,” "Start a New Project," "Continue Drafting," and "AI Chatbot."
Kim begins by working through AH-AI's introductory module, which explains the importance of ethical AI use in academic contexts while considering their diverse linguistic and cultural backgrounds. This module also teaches her about data privacy and the importance of consenting to data use, emphasizing the risks of sharing personal information with AI systems.
After that module, Kim clicked on "Start a New Project" and typed in her essay title: "The Impact of Social Media on Democracy."
Engaging with AI Chatbot:
She then clicks on "AI Chatbot" to brainstorm her essay. A chat window pops up on the right side of her screen, with a friendly greeting from the AI: "Hi Kim, how can I assist you today?"
Kim types, "I'm exploring how social media affects democratic processes. What are some key considerations?" The AI responds with insightful points about the role of social media in shaping public opinion, the spread of misinformation, and the digital divide.
As they interact, AH-AI's sidebar displays a real-time log of "Idea Origins," categorizing insights as "Student-Initiated" or "AI-Suggested."
Drafting with Real-Time Analysis:
Kim begins drafting her essay in the integrated word processor. As he types, a subtle color-coding system highlights her text: green for her original thoughts and red for sentences closely aligned with the AI's suggestions.
A tooltip appears when he hovers over a red-highlighted sentence: "This idea was inspired by your conversation with the AI. Consider expanding with your analysis."
Incorporating AI Suggestions:
Upon encountering a complex argument about echo chambers in social media, Kim decides to reference a point the AI made. Immediately, the text is highlighted in yellow, with a note: "Mixed Contribution: Consider adding personal analysis."
Kim pauses, reflects, and then enriches the argument with examples from recent elections, turning the yellow highlight into green, indicating a strong personal contribution.
Utilizing Feedback for Revision:
After completing her first draft, Kim clicks the "Analyze Draft" button. AH-AI provides a detailed breakdown: 70% green (original work), 15% yellow (mixed contributions), 10% red (AI-influenced), and 5% blue (potential external AI sources).
The platform suggests, "Review the 10% red-highlighted sections for more personal insight." Kim revisits these sections, integrating more of her own analysis and reducing the AI's influence.
Final Review and Submission:
Before submission, Kim accesses the "Final Review" feature. AH-AI summarizes her essay's composition, now showing 85% green, demonstrating a significant increase in original content.
Satisfied, Kim clicks "Submit," and a confirmation message appears: "Your essay has been successfully submitted. Great job on enhancing your original contributions!"
Self-reflection
After submission, students are prompted to reflect on the history of conversation or other forms of interaction in a quick scan and identify one thing good and bad about their participation as an ethical AI user.
**Note: The scenario text has been polished using ChatGPT from the prompt please help me to polish the content (OpenAI, 2024).
Page updated
Google Sites
Report abuse