Students using AI: Cheating or smarter learning? | Deep Dive podcast

AI and Homework: A Deep Dive

Key Concepts:

• Generative AI in education
• Academic fraud/dishonesty
• AI detection in code and essays
• Formative vs. summative assessments
• Personalized learning
• Misuse vs. effective use of AI
• Fairness in grading
• Evolving role of teachers
• Motivation in learning
• AI skills gap
• Adapting assessment methods

NTU Academic Fraud Case

• NTU accused three students of academic fraud for using generative AI in assignments.
• Students received zero marks and are appealing, questioning the university's justification and the line between acceptable and unacceptable AI use.

AI Usage by Students

• Writing essays and homework: A standard use case, predating ChatGPT with models like GPT-3.
• Research: AI is used to gather, synthesize, and report information, reducing workload.
• Creative applications: Media school students use AI in art and editing, expanding beyond traditional academic reporting.

Detecting AI Use and Plagiarism

• Detecting plagiarism in code is challenging because good code often looks similar.
• Traditional plagiarism detection relies on identifying poorly written or incorrect code copied from others.
• AI can generate standard-looking code, making detection harder.

The Purpose of Homework and Assessments

• Formative assessments: Designed to facilitate learning through struggle and engagement. AI circumvents this process.
• Summative assessments: Used to grade students for job market purposes. AI use can skew grading and hinder employers' ability to differentiate candidates.
• The learning journey is crucial for developing problem-solving skills applicable to novel real-world situations.

Personalized Learning and the Role of Teachers

• AI can personalize learning, adapting to individual learning styles.
• This raises questions about the evolving role of teachers, potentially shifting towards individualized tutoring.
• Effective AI use requires specific prompts and instructions, demanding critical thinking and problem-solving skills.

Misuse vs. Effective Use of AI

• The key question is how AI is used: to enhance learning and understanding or to avoid effort.
• Fairness is compromised if some students use AI to gain an unfair advantage over those who don't, especially when AI use is prohibited.
• Rich students may have access to better AI tools, exacerbating existing inequalities (similar to tuition).

The Learning Gap

• A small group of students are using AI effectively, accelerating their learning.
• A larger group may be becoming "dumber" by relying too heavily on AI and losing critical thinking skills.
• It's crucial to understand how students are using AI to address this gap.

Evolving Assessment Methods

• Traditional essays may be a poor measure of understanding, as AI can generate them algorithmically.
• There's a need to evolve assessment methods, potentially incorporating face-to-face discussions.
• Humanities disciplines face greater challenges in adapting to AI due to reliance on essay-based grading.

Clarity and Anxiety Among Students

• Students lack clarity on the acceptable use of AI, leading to anxiety and avoidance.
• Schools may not be adapting quickly enough to the changing landscape.

The Importance of Foundational Knowledge

• Students need to reach a certain level of competency before effectively using AI.
• Without a solid foundation, they may never cross the "AI chasm" to become experts.
• While some argue that certain foundational knowledge (e.g., simultaneous equations) is no longer necessary, others maintain its importance for developing learning skills.

Managing Motivation

• The core business of education is managing motivation, not just communicating knowledge.
• Large language models may negatively impact motivation if students become overly reliant on them.
• Educational institutions need to adapt their methodologies to maintain student motivation in the age of AI.

Acing Exams with AI

• Focus on outputs and replicating successful results.
• AI can condense information and provide high-level overviews, increasing motivation.
• Universities need to set clear guidelines on AI use and address abuse of the system.

The Role of Instructors and Institutions

• Instructors have the right to set rules regarding AI use in their courses.
• The key question is how to respond when those rules are broken.
• Institutions are still learning how to navigate the AI landscape and are taking the issue seriously.
• Students should ask professors about AI usage policies for specific projects.

Conclusion

• AI is here to stay and will continue to evolve.
• There is no one-size-fits-all solution for integrating AI into education.
• Institutions, instructors, and students must work together to find the best ways to leverage AI for learning while maintaining fairness and academic integrity.
• Open communication and clear guidelines are essential.