Gen AI as a peer learner - Stephen Chong

Summary

In a compilers class at the School of Engineering and Applied Sciences, Professor Stephen Chong discusses how students are allowed to use generative AI tools with specific restrictions. While high-level discussions and low-level troubleshooting with AI are encouraged, students must disable tools like Copilot to ensure they engage deeply with programming assignments. The goal is to foster intellectual effort in design and implementation, ensuring that despite advancements in AI, students still experience the critical learning process inherent to computer science education.

Takeaways

• 😀 Students in the CS153 compilers class primarily learn through intense programming assignments.
• 🤖 Generative AI tools are allowed in the course but with specific restrictions to ensure academic integrity.
• 🗣️ Students can discuss high-level concepts with peers and AI tools to enhance their understanding.
• 🔧 Low-level coding issues can also be discussed with peers and AI, but caution is needed with generative AI outputs.
• ⚠️ Students are advised to disable generative AI tools in their coding environments to avoid complete code implementations.
• 💡 The use of AI can improve programming productivity, but it should not replace the intellectual effort required in learning.
• 📚 Learning in computer science involves understanding design and implementation, which requires personal effort.
• 🔍 The course emphasizes critical thinking in coding beyond just using AI for solutions.
• 🛠️ AI tools should be used responsibly to aid in learning rather than do the work for students.
• 🚀 The future of programming education will likely include a balance between AI assistance and traditional learning methods.

Q & A

What course is Stephen Chong teaching this semester?

Stephen Chong is teaching CS153, a compilers class.

How is the CS153 course structured in terms of student assessment?

The course is very programming intensive, with students doing almost all their assessment through intense programming assignments.

What role does generative AI play in the CS153 course?

Students are allowed to use generative AI tools with restrictions to support their programming.

What high-level concepts can students discuss with generative AI tools?

Students can discuss high-level concepts and explanations of topics covered in lectures using generative AI tools like ChatGPT.

What kind of low-level issues can students consult generative AI tools about?

Students can use generative AI tools to seek low-level information about errors in their code or how to use specific libraries.

What is the primary restriction placed on the use of generative AI tools?

Students must disable generative AI tools, such as Copilot, in their development environments to prevent them from providing full implementations of assignments.

Why is it important for students to engage in the intellectual effort of programming?

Engaging in the intellectual effort helps students understand the design and implementation of code, which is essential for their learning in computer science courses.

How does Stephen Chong view the future role of AI in programming?

He anticipates that programmers will increasingly use AI as a tool to become more effective and productive, but acknowledges the necessity of learning through direct effort.

What does Stephen suggest about the ongoing need for traditional learning methods in computer science?

He suggests that even as AI improves, courses like the compilers course will still require students to turn off AI assistance and engage deeply with the material.

How does this approach to generative AI support student collaboration?

The approach encourages students to collaborate and discuss concepts, enhancing their understanding and learning through peer interaction.