What is the future for software engineers? In the age of Artificial Intelligence (AI), this question is asked with increasing curiosity…and fear.
As a software engineer, I can tell you that AI cannot do everything a human coder can do. AI’s coding creativity, left minimally guided and unreviewed, can produce expensive errors in your final product and unnecessary headaches in your workflow. It needs oversight.
However, adopting AI as a tool is crucial to your development as an engineer. If you try to ignore it, you’ll wake up one day to realize you’ve been left behind. Used correctly, AI can help you code more efficiently, freeing up your time for the tasks that really need a human touch.
The shiny newness of LLMs may tempt overzealous employers to fire their engineers, but they may soon realize why they need to hire them back. Here’s why.
AI needs an architect
LLMs need your direct input to produce an output. It does not provide unprompted solutions, or ponder business challenges long after interacting with you. With AI, the work stops and starts at your desk.
But sometimes, the best ideas come from the parallels drawn between our lived experiences and the problems we face–a thought while you’re in the shower or taking your dog for a walk.
Innovation is a human trait that businesses need and that LLMs have yet to replace. And innovation does not discriminate between the junior and senior software engineer.
Innovation is simply a functionally different way of solving a problem. Senior software engineers can innovate from experience. Junior developers can innovate through untrained imagination.
Software engineers must guard against AI slop
LLMs can produce code faster than any human, but is it reliable code or what we would term “AI slop”? When I produce code, I understand the intent, structure and ultimate purpose. This makes it easier to navigate and troubleshoot the code if there is an error.
When LLMs dump a chunk of code on your screen, you own the outcome of that code if you choose to use it. You cannot hold a chatbot accountable if the code it produced causes your system to crash or malfunction.
Being able to own your code is especially important when you are handling sensitive information. Software engineers in healthcare, government and banking for example must ensure that the code in their systems protects private information.
How to integrate AI in your career
Software engineers should embrace a more directive role to integrate AI in their workflow:
- Understand coding fundamentals to produce effective prompts. Learn more about vibe coding vs. software engineering
- Plan your workflows and integrations before prompting AI. Then prompt in sections to better manage and monitor the output
- Use AI to troubleshoot for minor syntax issues
- Train junior developers on your AI workflow to free up time for more projects!
The job of a software engineer is not going away, it is just evolving—fast.
Book a call with me for more information on how to integrate AI in your workflow.