For my Capstone project my team is developing a React client for visualizing Fire and EMS call data. These visualizations include charts, heatmaps, and graphs which aim to assist first responder organizations in developing strategies to reduce response time and increase efficiency.
ChatGPT has been a large help for me during this project. I have not had to waste time writing as much boiler plate code, I can simple communicate to ChatGPT that I need a block of code which can accomplish some task, for example creating a particular JavaScript Fetch request or a generic React dropdown component. I can then ask clarifying questions about the generated code and modify it as needed, sometimes even with iterative help from ChatGPT (ie “I need x”, “OK, now I need x but with y”, “I want y to be more like z”). This has helped to significantly speed up development time and leave more time for me to make important design decisions.
I have also had success with using ChatGPT for debugging purposes the ability to have a conversation to work through a particularly tricky bug is extremely helpful. Prior to ChatGPT, if I ran into some odd behavior that I couldn’t figure out, I would have to scour the internet hoping to find some similar issue on Stack Overflow or some other forum. Often times what I would find would not be quit what I was looking for, the solution wouldn’t work in the context of my project, or I would simply have follow up questions which couldn’t be answered. ChatGPT on the other hand will give me tips on how to find the source of a bug in my code specifically, and then explain exactly what the issue is once the bug has been found.
In both these cases discussed above, I find myself with a greater understanding of the underlying tech stack after my conversations with ChatGPT. But in order to get the maximum learning benefit from this tool, it’s important to ask follow up questions. If I were to blindly copy and post code from ChatGPT into my codebase I would not only risk introducing bugs into the code, but I wouldn’t develop my skills as a Software Engineer. The same goes for debugging with ChatGPT. If I don’t fully understand what’s causing the bug that I’m using ChatGPT to fix, I will likely run into the same bug again in the future, requiring another lengthy debugging conversation with ChatGPT which would have been avoided from the beginning.
Overall, I have felt that using ChatGPT is nearly equivalent to having an on demand tutor or mentor at your finger tips. Blindly following the suggestions from a mentor can help you finish your project, but wont help you grow your skills. Following the suggestions and taking the time to truly understand them, will help you do both.
Leave a Reply