According to IEEE, university professors are changing the way they teach students how to code. The emphasis is reportedly moving away from syntax and toward higher-level skills such as testing and debugging code so that AI-generated code can be verified as correct.

Besides testing and debugging code, lecturers also focus more on problem decomposition where you break down a big problem into smaller problems that an LLM might be able to solve more easily. Discussing this, Leo Porter, associate professor of computer science teaching at the University of California, San Diego, said:

“It’s a skill you should know early on, because you have to break down a big problem into smaller pieces that an LLM can solve. It’s hard to find where in the curriculum they teach – maybe in an algorithms or software engineering class, but these are advanced classes. Now, it’s becoming a priority in introductory classes.” .

While AI can certainly help with coding, professors still warn students to be careful about using it because it can be misleading and also infringe copyright. They also warn that AI should only be used as a co-pilot, not an autopilot.

Since generative artificial intelligence is fairly new, many professors are just adapting their courses. The forms these courses now take are unlikely to be the end result as they need to see how well they actually work.

Another issue that teachers will need to think about is how much the shift away from syntax actually helps people when they enter the workforce. Many companies limit the use of generative AI because they are working on secret code that they don’t want to leak out through AI. If students don’t learn as much syntax, they may find themselves struggling in the real world.

However, generative AI has some beneficial effects. One educator who spoke to IEEE said that they routinely ask students to write code and then grade them based on whether they pass some pre-written tests. The educator has now realized that this way of teaching is too restrictive and has expanded their teaching method, which is definitely positive.

source: IEEE Spectrum | Photo via Depositphotos.com