There are a discussions on the Internet and LinkedIn if AI is going to replace developers. I don't think AI is going replacing developers, like we're still going to need developers and engineers to create software. That job and responsibility isn't going to go anywhere. But with the help of CoPilot, ChatGPT and other AI systems, the way I work as a software engineer is definitely changing how I do this job.
I vibe coded an entire console application that moves data from one system to the other. This was a pretty fun experience but I have to say that I couldn't rely on the machine to do 100% of the work load. When CoPilot got to a certain road block it got stuck. I tried different ways to prompting to get through the issue, but no matter what I tried the system kept on reverting to the last three iterations. When the agent hit the third iteration of what it thought was the correct fix, it would revert back to the first and go through the loop again. This is where being the human in the loop and being a good engineer is important. At this point I needed to become the engineer and utilize my actual software engineering skills to fix the problem. That's the problem with AI agents, they can't think, they only know what they were trained on and if they come across an issue that isn't something they were trained on or trained a lot on they're not going to be able to process your instructions.
In this case the AI was my entry-level engineer who know all the correct syntax to execute and create the program I was telling it to create. There are two key things that happened that makes me believe engineers are going to be around for a while yet.
Where AI Helps Most
- Boilerplate Code: Let Copilot write the repetitive stuff—controllers, DTOs, test scaffolding.
- Code Reviews: Use AI to spot obvious bugs, style issues, and missing edge cases before PRs go to humans.
- Prompt Engineering: I keep a library of prompts for Copilot and ChatGPT to generate regex, LINQ queries, and more.
- Documentation: Summarize code or generate README drafts automatically.
Tips for .NET Developers
- Always review AI-generated code—don’t trust, verify.
- Use AI for speed, but keep your standards high.
- Share prompt examples with your team to level up everyone’s workflow.
Want to see how AI can help your .NET team? Let’s talk about practical ways to boost productivity.