CSUSB Student Workshops on AI Thoughtfulness and Ethics

Overview of AI Learning at CSUSB

  • Students at California State University, San Bernardino (CSUSB) lead workshops focused on the thoughtful, ethical use of artificial intelligence (AI).
  • The workshops aim to empower students across various disciplines to teach responsible AI usage, emphasizing control rather than fear or avoidance.

Central Themes of the Workshops

  • Human Judgment in AI: The workshops center around the belief that human oversight and judgment remain crucial, implying that AI should not displace human expertise or responsibility.
  • Key Questions for AI Use: Participants are encouraged to consider:
    • How to use AI effectively
    • When it is appropriate to use it
    • What specific purposes AI should serve
  • Shared Framework: The workshops utilize a framework highlighting:
    • Verification of AI outputs
    • Ethical considerations in AI application
    • The necessity of critical thinking in AI-related tasks.

AI as a Tool for Learning

  • AI is framed as a supportive tool, not as a replacement for knowledge or reasoning. It should be questioned, tested, and applied judiciously.

Presentations by Workshop Participants

Jonathan Muckelroy (Career Technical Education)

  • Background: Credential candidate in adult education; Graduate in physical education.
  • Approach: Utilized AI to support early drafting and brainstorming in kinesiology.
  • Case Study:
    • Students used AI to propose explanations for how moderate aerobic exercise might reduce perceived stress.
    • Verification required through:
    • Peer-reviewed sources
    • Databases such as PubMed and guidelines from organizations like the American College of Sports Medicine.
    • Outcome: Students categorized AI-generated content as verified, unclear, or removed, shifting focus from shortcuts to scientific judgment.

Monique Rodarte (Master's Student in Career Technical Education)

  • Focus: Equity and bias within AI systems.
  • Key Points:
    • Explored how algorithmic tools can perpetuate disparities, notably in school discipline and student placement.
    • Addressed practical queries regarding AI effectiveness:
    • Who benefits from the system?
    • Who is overlooked?
    • How does data incompleteness cause issues?
  • Positioning AI Literacy: As an equity issue emphasizing transparency and accountability.

Garrett Wulc (Mathematics Major)

  • Focus: Technical analysis of AI in diagnosing machine learning workflow errors.
  • Dataset Used: Airfoil data from NASA and UC Irvine.
  • Findings: Displayed limitations of AI, addressing specifics of a baseline random forest model:
    • Issues encountered:
    • Improper cross-validation
    • Data leakage
    • AI suggestions were tested; some accepted, others rejected due to factors like:
    • Computational costs
    • Irrelevance
    • Contextual fit
  • Key Lesson: Confidence in AI outputs does not equate to accuracy.

Logan Ashbaugh (Graduate Student in Computer Science)

  • Focus: Interaction between people and AI systems.
  • Demonstration: Responded to sorting an algorithm with different prompt qualities:
    • Vague prompts led to misleading conclusions whereas precise prompts enhanced relevance.
  • Emphasis: Importance of understanding AI-generated content and verifying against established knowledge.
  • Framing AI Literacy: As a skill linked to judgment rather than mere consumption of technology.

Jannatul Faika (Graduate Student)

  • Focus: Debugging scenarios with chatbots.
  • Presentation: Showed how AI explanations may be confidently incorrect due to missing context.
  • Methodology: Used:
    • Manual testing to compare outputs.
    • Refinement of prompts to highlight human oversight in revealing AI blind spots.
  • Conclusion: Reinforced the idea that while AI can assist, human responsibility governs decisions and outcomes.

Conclusion of Workshops

  • The presentations collectively proposed a cross-disciplinary model advocating for responsible AI use.
  • Learning approach emphasizes deep engagement with technology and critical questioning of AI applications, showcasing student empowerment in exploring emerging tools.