ChatGPT and the Future of Undergraduate Computer Science:
Challenges, Opportunities and Recommendations

Research Question

What are the strengths and weaknesses of ChatGPT when answering questions related to logic and theory in Computer Science?

To what degree can undergraduate computer science students rely on ChatGPT for solving their take-home exams and assignments?

What factors should instructors take into account while designing take-home exams and assignments for students who may potentially use ChatGPT?

How can ChatGPT be constructively used by students and instructors to enhance their learning and teaching experience respectively?

Methodology

Undergraduate computer science subjects

Graduate Aptitude Test in Engineering

Data Structures and Algorithms

Operating Systems

Database Management Systems (DBMS)

Machine Learning

Programming Questions from LeetCode

Future direction of CS pedagogy

Students

Instructors

Use ChatGPT to generate guide questions

Use ChatGPT for initial ideation and write-ups and then build upon these through their own skills

Formulate their own timeline and space

Train students in prompt engineering

Design. questions that require critical thinking and higher cognitive skills that cannot be easily solved by providing prompt

For take home assessments, questions that are precise and objective in nature might be a better way to test a student’s knowledge compared to subjective questions

New evaluation styles like giving students a topic where they will design a problem, then they will generate a solution and test cases

Future Work

Enhancing the answer of AI tools to make them more relevant and accurate

More inclusive and expansive data

Challenges

Variability in ChatGPT’s accuracy and the need for prompt
contextualization

Higher accuracy for subjective and theoretical questions

Bias in ChatGPT’s underlying language model