Flashback

I am a master’s graduate in AI and a software engineer. I started my career 3-4 years ago when it was the right time to start IT because the entire world was desperate to seek developers and engineers in case there is a lack of the skilled IT specialists.

I personally started IT because I found it very interesting. When I and my girlfriend were living in Frankfurt, I saw my girlfriend learning Python and trying to solve problems from a web dev educating web with curriculums for IT beginners, if I remember right.

This was where we used to live. Frankfurt Süd.

She was on Pycharm terminal. I went, Let me try it too. I was writing a code, an extremely simple code like print(1+1) or print(2×4). The terminal showed exactly what I expected, I had goose bumps thinking, how on earth there is a system, like listening to my commands and show the answer which I expected it to be? Such an amazing experience it was. Then I learned algorithms, systems and network, eventually I became a developer.

Question

The question arises every day nowadays whenever I see news or Youtube videos speaking of “AI will substitute humans in dev”. Seriously the LLM models got bigger in IT market and everyone can use it. The AI tools such as Cursor, ChatGPT, V0 and others with only EUR 10-20 had us tremendously convenient experience of service development. Everyone without even a big knowledge could challenge themselves to build an app and services. I have seen many on SNS that they are not even familiar with IT, but they have launched services.

Now, the question is, Is it still worth studying software skills even though AI could assist on developing, debugging and all that?

Current Status of IT Employment

According to CompTIA review in February 2025, tech occupations in the US increased by approximately 228,000 for the month, however also unemployment rate increased to 2.9%.

What about the employment in South Korea? Software Policy & Research Institute(in short, SPRi) in January 2025 announced that software developer employment rate decreased due to economy downturn, just like the entire world is suffering from the severe difficulties in economy growth. Some big tech companies, for example Kakao group in Korea, even cut down the number of hiring developers up to 50% compared to 2021-2023.

Yet, it is still too early to assume that generative AI is substituting developer’s job. It seems unclear that the downfall in the number of jobs is caused by the AI, which has no clear statistics because the generative AI has reached the certain level(very human-like with texts, I might say) within the short period of time.

The sheer repetitive tasks will be handled by AI, resulting in shrinking the opportunities for junior level developers, but in the mean time it is giving more opportunities to enhance the software skills of the juniors.

Learn Software Skills?

It is said that the generative AI in general will not have enormous effects on the software developer’s ability, but it will challenge them by giving tasks to make them determine which information given by AI to make use of in their field.

Generative AI can do many things now. It may write codes and it may think which information is good and which is not. However it cannot, maybe for now, build the sophisticated logic or setup the environment of a service to build codes on top of it. The generative AI at the moment is an effective assistant when developing services. We may ask for advice as it has been trained with humongous information on the internet with the high dimensional tensor in the core.

Humans make mistakes and so do AI. We need to know the skills of the software, libraries and the framework to know exactly what is causing problems when there is an error. AI will not be able to answer to those questions that have not been used for training, but we can with our common sense or our experience, which may not be very related to software, but still be helpful to solve problems.

Developers say, and I also agree with it, that build and release, testing, doing QA tasks is still on human’s hand. What does it mean then? It means we need to understand what skills were used in the product or in the codes. It is something that cannot be subbed by AI like ChatGPT, Copilot and so on.

Conclusion

Well, thanks to AI we became familiar with IT and softwares. It is maybe the huge advantage for our generation. It is our choice to make us of AI or be used by AI in our software field.

I believe it gave a big chance now to us to think bigger and more creative as we do not need to think about the repetitive tasks. We only need to focus on what we would like to build. But what we also mush consider is that when it comes with maintenance of the codes, it is solely our task to understand the business logic and the flow and fix the bugs in accordance with the business logic. The ability to communicate between the development, design, DevOps, and sales teams, as well as with customers, is entirely up to us. It is something that cannot be substituted by AI.

So, would you make good use of AI or be overwhelmed by AI?