The software development sector is evolving because to the impact of generative AI. Programmers’ workflows are being aided by AI-powered coding tools, and the number of employment in AI is still rising. However, academia, which is one of the main channels via which the upcoming generation of software engineers learns to code, is also exhibiting the shift.
Students studying computer science are adopting the technology, use generative AI to assist them learn how to code, comprehend difficult ideas, synthesize lengthy research papers, and come up with new research areas.
According to Johnny Chang, a teaching assistant at Stanford University seeking a master’s degree in computer science, “students are early adopters and have been actively testing these tools.” In 2023, he also established the AI x Education conference, a virtual assembly of educators and students to talk about the effects of artificial intelligence on education.
Educators are also working with generative AI so they don’t fall behind. However, they’re having trouble coming up with ways to use the technology while also making sure that kids understand the fundamentals of computer science.
Ooi Wei Tsang, an associate professor at the National University of Singapore’s School of Computing, adds, “It’s a difficult balancing act.” “We are still learning how to do this, given the rapid evolution of large language models.”
Less Focus on Syntax and More on Solving Problems
Both the principles and the abilities themselves are changing. The majority of beginning computer science classes concentrate on code syntax and getting programs to run. While being able to read and write code is still necessary, other topics that aren’t usually covered in the syllabus, such testing and debugging, now need to be covered more thoroughly.
learning procedure for the entire software development life cycle, not just the coding itself, according to Zingaro. Additionally, I believe that my classes are now far more inclusive and open than they were in the past. I am able to assign pupils to work on more complex and large-scale projects.
In agreement, Ooi states that generative AI tools “will free up time for us to teach higher-level thinking—for example, how to design software, what is the right problem to solve, and what are the solutions.” Instead of concentrating on the syntax of the code, students can devote more effort to optimization, moral dilemmas, and system usability.
Steer clear of AI Coding Pitfalls
Teachers, however, exercise caution because LLMs have a history of hallucinations. “Students should be taught to be skeptical of the results and to take responsibility for independently confirming and validating them,” says Matthews.
Furthermore, generative AI “may short-circuit the learning process of students who rely too much on it,” according to Matthews. Chang concurs that an excessive dependence on oneself might be dangerous and encourages his fellow students to solve difficulties on their own to avoid losing out on critical thinking or productive learning experiences. He states, “We should be making AI a copilot for learning, not the autopilot.”