We will follow the DCS AI policy in this course (adapted from Prof. Irfan’s website).

Main point: Always acknowledge what generative AI tools were used to help you complete assignments.

  1. What counts as Generative AI? Generative AI refers to all the AI systems that can create new content like text, code, images, audio, and other types of media. This includes but is not limited to:
    • Large language models (ChatGPT, Claude, Gemini, etc.), apps and agents derived from them, and aggregator interfaces (Amplify, LibreChat, etc.)
    • Code generation tools (GitHub Copilot, Google Colab, Cursor, etc.)
    • Image generators (DALL-E, Midjourney, etc.)
    • AI-powered writing assistants and autocomplete features
    • NotebookLM, Quizlet AI features, and other AI-enhanced study tool.
  2. How can generative AI help you to learn in this course?
    • Converting slides and notes to study guides and podcasts (e.g., NotebookLM)
    • Creating practice materials like flashcards and mock quizzes or exams (e.g., Quizlet)
    • Proposing a study plan
    • Providing alternative explanations or examples
    • Suggesting refinements to grammar and corrections to spelling
    • A guide to debugging code
  3. How can generative AI impair your learning in this course?
    • For coding projects, AI often misses design layers, making results hard to debug
    • Reduces critical thinking and problem-solving abilities
    • Can limit creativity
    • Creates a sense of false confidence
    • Creates dependency
    • Reduces retention
    • Limits research skills
    • Denies you the possibility of self-expression in a constructive environment
  4. Other costs of using AI
    • Environmental impact – large energy use
    • Labor exploitation – poorly paid workers for moderation and labeling
    • Digital colonialism – reinforces inequalities and Western perspectives
    • Digital divide – widens gaps in tech access
    • Aggregation – concentrated power in few companies
    • Privacy – inputs may be stored despite disclaimers
  5. When is generative AI not allowed?
    • To generate code. AI code lacks structure and context and is often too complicated to debug (Prof. Irfan can give you many examples). You must write the initial draft of code yourself.
    • To generate text; e.g., to flesh out an outline, leading to an initial draft that you intend to revise
    • To brainstorm – your background is richer than AI’s
    • To solve problems – designed to strengthen critical thinking
    • As a search engine for facts – verify independently
    • To analyze data or interpret results without your own reasoning
    • To generate ideas for creative projects or research questions
    • To complete exams or quizzes unless permitted
    • To translate assignments or materials without permission
  6. When can generative AI be used?
    • Polishing text by checking for grammar, spelling, and styling suggestions
    • Outlining, i.e., preparing an initial structure with AI to be expanded with your own perspective (note that the opposite is not allowed)
    • Debugging your code, including generating test cases
    • Getting help on a software tool (e.g., how do I do X?)
    • Finding resources to support research
    • Testing discussion questions for class
    • Checking your work and getting preliminary feedback
  7. Academic Integrity Expectations
    • Be transparent about AI use
    • When in doubt, acknowledge it
    • Submit your prompts with assignments
    • Understand that overreliance hinders learning
    • Violations will be treated as academic dishonesty

  8. Why do we give you assignments?

    Assignments are designed to help you develop essential skills and ways of thinking that will serve you throughout your education and beyond. Each assignment is an opportunity to practice critical thinking, develop your unique voice, learn from struggle, engage with material, demonstrate learning, and prepare for future challenges.