CBS Faculty Live: Empowering Classroom Discussion with AI with Professor Dan Wang

Key Concepts

• Casey: A voice-to-voice AI discussion partner designed to simulate one-on-one "tutorship" models for classroom preparation.
• Generative AI (GenAI): Large Language Models (LLMs) used to facilitate adaptive, interactive, and personalized learning.
• Pedagogical Shift: Moving from passive, one-way content consumption to active, synchronous, and critical debate.
• Super-linear/Exponential Growth: The observation that technological progress often feels stagnant before hitting an inflection point where it accelerates rapidly, causing societal stress.
• Psychological Safety: The benefit of using AI as a non-judgmental partner, allowing students to experiment with arguments without fear of social repercussions.

1. Main Topics and Key Points

• The "Socrates" Parallel: Professor Dan Wong draws a parallel between modern fears of AI eroding critical thinking and Socrates’ historical opposition to writing, which he feared would destroy human memory.
• The Technology Adoption Model: Breakthrough technologies (like electricity or AI) often face resistance not because they are ineffective, but due to organizational and managerial "frictions."
• The "Tutorship" Model: Inspired by Oxford University’s one-on-one tutorial system, Casey aims to provide personalized, high-pressure, and interactive debate for students at scale.
• Data-Driven Insights: Research on 1,300+ conversations revealed that voice-based AI interactions lead to higher verbosity and greater creative concept generation compared to text-based interactions.

2. Real-World Applications

• Business Strategy: Used in MBA courses (e.g., Netflix content strategy cases) to help students refine their positions before class.
• Accounting: Applied in accounting courses to help students navigate subjective judgment calls (e.g., revenue recognition) through debate.
• Inclusivity: The tool serves as an equalizer for students with diverse learning styles, such as those with dyslexia, who may struggle with traditional reading/writing-heavy assignments.

3. Methodology: The "Casey" Framework

• Preparation: Students read a case and input their initial stance into the Casey portal.
• Interaction: Casey acts as a "friendly discussion partner," challenging the student’s logic, asking for counterarguments, and providing business context (e.g., cost-per-episode data).
• Feedback Loop: After the 5–22 minute conversation, Casey provides a summary of the discussion, highlights strengths, and identifies areas for improvement.
• Classroom Integration: Students bring their Casey-generated notes to class, leading to more sophisticated, high-level discussions.

4. Key Arguments and Evidence

• Argument: AI does not replace the instructor; it enhances the classroom experience by ensuring students arrive better prepared and more engaged.
• Evidence: Students who changed their minds during an AI debate were 30% more likely to engage in longer, more frequent AI discussions.
• Argument: Voice-based interaction is superior to text for creativity.
• Evidence: Quantitative analysis showed that voice conversations, despite containing filler words, consistently covered a wider range of creative concepts than text-based chats of equal length.

5. Notable Quotes

• Professor Dan Wong: "The reason why you might see the uneven adoption of AI technologies today is not because AI technology isn't good... but rather because we face organizational, managerial and personal frictions."
• Professor Dan Wong (on hallucinations): "We don't [prevent hallucinations] because humans hallucinate all the time... hallucination in many ways is more of a feature than a bug [in a learning context]."

6. Technical Terms

• Large Language Model (LLM): A type of AI trained on vast amounts of text to predict and generate natural language responses.
• Latency: The delay in response time; Casey utilizes OpenAI’s real-time voice model to minimize this, creating a near-instantaneous conversational flow.
• Hallucination: When an AI generates factually incorrect information; in this context, it is treated as a prompt for students to practice critical verification.

7. Synthesis and Conclusion

The transition to AI-integrated education is inevitable, mirroring the historical shift from steam power to electricity. Professor Wong’s "Casey" demonstrates that when AI is used as an interactive, voice-based partner rather than a shortcut, it fosters deeper critical thinking, inclusivity, and student motivation. The primary takeaway is that the value of an MBA—and education in general—is shifting from the mere acquisition of information to the ability to engage in high-stakes, adaptive, and creative discourse. The future of learning lies in embracing these tools to overcome the "friction" of traditional, static teaching methods.