You might have seen videos of OpenAI’s latest GPT-5 model and all the impressive things it can do when it comes to coding: creating a fluid dynamics simulation, a 3D racing game or even a basic working Photoshop clone.
That leaves the question on the mind of every software engineering student, and anyone considering studying software: “is AI going to take my job?!”
While it may be exciting to imagine a future where we have to fight AI in order to keep our jobs and prevent world domination, the short answer is no. Software engineers are not going anywhere. If you want to study software today, there’ll be jobs for you when you graduate and twenty years after that. There’s no need to drop out and pick another degree.
Thanks for reading. Go check out this great post by one of the other bloggers!
Not convinced? Have you seen how stupid AI can be sometimes? Really? Fine, if you insist, here’s my explanation on why AI will not be replacing software engineers:
“there goes my career in software”
It is true that LLM-based tools are very powerful for coding. Just by typing a few natural language prompts, in less than 30 seconds it can generate hundreds of lines of code that would have taken a human hours, if not days to write. And for the most part, this generated code actually works!
The thing is, software engineers do much more than just code. Coding is certainly part of the job, but the actual time spent typing code makes up a small percentage. Instead, software engineers spend their time researching, planning, debugging, and thinking about how to tackle problems. Knowing how to code is one thing, but knowing what to code is another. Too much time is wasted working on the wrong thing. An underrated skill is product sensibility—at the end of the day, we’re building software products for users. But if our users don’t like the products we build or we don’t meet the needs of the users, then what’s the point?
That’s why it doesn’t really matter how you write the code. You can write it all by hand, or use an AI tool to do it for you. Anyone can go online and learn to code, it’s not exclusive knowledge. But you don’t see everybody making the next viral app. Coding is simply a means to an end. You still need skills to figure out what to build, and AI can’t do that for you.
“we’re so cooked”
In their current state, AI tools are not fully accurate, and likely will never be. On a very basic level, LLM-based tools work like autocomplete, predicting the most likely outcome based on the data it was trained on and the inputs given. When it generates a piece of code, it does not know if it actually works, so a human is required to run and test that code. This makes AI tools like an improved StackOverflow: there may be lots of answers but you only know if you’ve found what you’re looking for once you try it out.
As always, integrating bits of code you didn’t write into the rest of your code takes time and understanding. What if something breaks? Sure, AI tools can be great for debugging. But when bugs are more complex or deeply rooted in the design of a system, the AI tools fall short and you’re left on your own. It’s crucial to know how to fit the pieces together. The best software engineers have developed an intuition for this which allows them to be faster debuggers.
“rip junior devs”
The power of these AI tools means good engineers can do their work more efficiently. Less time spent coding (or doing “grind work”), and more time making decisions on product design and things that have a bigger impact on the final outcome. As such, the work previously done by a team of five engineers can now be done by a team of four. In this tough economy, if you can get away with hiring one fewer engineer, that’s big savings for a company. This has unfortunately led to fewer jobs for the lesser-experienced engineers, which happens to include junior and entry-level engineers.
Still, software engineers aren’t going anywhere. As building software becomes easier, the demand for more software only increases. History has shown this before and things are no different today. AI is not an enemy; it’s an ally. But it still needs to be directed by us humans. And that’s much harder than it sounds.
Entry level engineers won’t go away because once all the senior engineers retire, who will be left to replace them? It may be harder for us to find jobs now, but it just means we have to work a little harder and prove that we’re still worth hiring. We need to show that we can bring exciting new ideas to the table while using AI tools to our advantage.
Don’t study software engineering because it’ll be an easy, stable job. Do it because you’re truly interested and passionate about building software. Engineering is all about working as a team to design and build solutions for humans. Your connection to other humans is what makes you special. No AI can ever replace that.