Group Actions
Engagement
  • Author Author: AngelSoto
  • Date Created: Date Created
  • Views 464 views
  • Likes 5 likes
  • Comments 3 comments
Related
Recommended

Reflections on Using AI in Software Development

AngelSoto

Context: A Practical Project

Over the past days, I have been developing a Python Qt application to generate weekly meal plans for my baby and family.
The project itself is not overly complex, but it has provided a valuable opportunity to reflect on something broader: the role of
artificial intelligence in our everyday engineering workflow.

My Approach to AI Assistance

My background is in C programming, with solid experience in embedded systems and software development. When shifting
into Python, however, I still consider myself at a more basic level. This mix has shaped how I approach learning: I rely on my
technical foundation, but I also use AI as a complementary resource.

The process is intentional. I first explore documentation and form my own reasoning. If questions remain, I consult ChatGPT to
clarify an approach or compare perspectives. In this sense, AI becomes more of a sparring partner than a code generator.
It helps me validate ideas, overcome mental blocks, and accelerate learning without replacing my own judgment.

Benefits Observed

Working in this way has changed how I experience the project. Instead of focusing only on syntax or debugging details, I
can step back and act more like a software lead: setting direction, defining structure, and designing tests. This perspective
is motivating. It allows me to keep learning by doing, while also practicing the skills needed to think at a higher project level.

Risks and Critical Perspective

That said, I am also aware of the risks. Misused, AI can easily foster intellectual laziness. It can discourage deeper exploration,
making it too tempting to accept quick answers instead of working through problems. The danger is subtle: what feels like
efficiency in the short term may lead to dependency and a weakening of critical skills in the long term.

Final Thoughts

This tension—between growth and dependency—is, to me, the most important reflection. On one hand, AI can help us break
barriers, learn faster, and concentrate on the bigger picture. On the other, it can erode the discipline and rigor that define our profession.

So the real question is not whether to use AI, but how to use it responsibly.
Done well, it becomes a powerful ally in both
learning and leadership.
Done poorly, it risks undermining the very skills we aim to build.

Parents

  • Hi,

    Interesting points.

    Regarding the AI laziness issue, I think in the future that could become a bigger possibility, but today, save for a basic level of software projects, AI isn't going to get everything right, and so one will be forced to examine and learn from what the AI has done, so for engineers at least, perhaps that's less of an issue currently. Currently, I more often than not find myself learning each time I use AI.

    Regarding dependency on AI other than laziness, there are some negatives about that (especially corporate and non-technology reasons), but the part of dependency that I don't mind particularly is that we are often dependent on acquiring information from other sources anyway, for instance, having access to the Internet in general when developing projects.

    I find AI is also highly useful when experimenting on your own, at low cost, with no colleagues around. Also, there are some topics we will just never wish to learn deeply for ourselves, especially if they are not all that strategic, and we might have offloaded them onto others. We would still be dependent on other sources, so it can be useful if AI can help with the mundane and non-strategic stuff too - I think that's particularly useful early on in a project, and one could always pay for professional help when it's more likely to turn into a commercial project.

Comment

  • Hi,

    Interesting points.

    Regarding the AI laziness issue, I think in the future that could become a bigger possibility, but today, save for a basic level of software projects, AI isn't going to get everything right, and so one will be forced to examine and learn from what the AI has done, so for engineers at least, perhaps that's less of an issue currently. Currently, I more often than not find myself learning each time I use AI.

    Regarding dependency on AI other than laziness, there are some negatives about that (especially corporate and non-technology reasons), but the part of dependency that I don't mind particularly is that we are often dependent on acquiring information from other sources anyway, for instance, having access to the Internet in general when developing projects.

    I find AI is also highly useful when experimenting on your own, at low cost, with no colleagues around. Also, there are some topics we will just never wish to learn deeply for ourselves, especially if they are not all that strategic, and we might have offloaded them onto others. We would still be dependent on other sources, so it can be useful if AI can help with the mundane and non-strategic stuff too - I think that's particularly useful early on in a project, and one could always pay for professional help when it's more likely to turn into a commercial project.

Children
  • in reply to shabaz

    Thanks, shabaz. Very good points, especially about how AI still forces us to review and learn, which makes the “laziness risk” less of an issue for now. I also find it interesting what you mention about relying on AI for the non-strategic parts — it gives a sense of confidence, helps achieve goals, and speeds up learning. Used wisely, I agree with you, it’s definitely worthwhile.