Where We Are Today
Well... we've finally reached the second part of this series. In the first article, we talked about the first two stages. Now I'd like to move on to the third one. This part is different because now we're talking about the future. More specifically, I'd like to look at what software development may look like over the next few years.
As I said before, I'll mostly talk about the IT industry. That's simply the area I know best, and it's much easier for me to explain my thoughts using examples from software development. Of course, many of these ideas could apply to other industries as well, but let's stay with what we know for now.
Now... before we get there, let's stop here for a second. First, let's look at how software development works today. Because if we don't understand today's process, it's difficult to predict what may change tomorrow.
How Software Development Works Today
Well... if we simplify everything, software development has two main layers. The first layer is kind of where ideas are created. People come up with ideas, discuss them, analyse them, write requirements, improve them, and eventually pass everything to developers.
Then comes the second layer. This is where developers take those requirements and turn them into working software. Of course, that's not the whole process. There's also testing, QA, deployments, operations and many other steps. But for this discussion I'd like to focus mainly on these two layers because that's where most of the work begins.
Ideas are created in the first layer. They move to the second layer, where developers write the code. That's a very simplified picture, but it's good enough for what I'd like to talk about next.
Five Years Later
Now let's go back a few years. Around the time ChatGPT was just starting to appear, there was a very popular idea about what software development would look like in five years. Maybe you remember those predictions. The idea was simple. You would write one sentence, something like, "Build me a CRM," or "Create an online shop," press Enter, and a few minutes later the whole application would be ready.
Well... those five years have already passed.
And that's not what happened.
What actually happened is different. Today we have AI tools that can write code, explain code, fix bugs, generate tests and even suggest better solutions. We've already talked about that in the previous article, so I won't go into too much detail here. The important thing is that these tools are real, they already help developers every day, and I don't think they're going anywhere.
So, for me, the question is no longer whether developers will use AI. They already do. The more interesting question is how companies will use it, and how software development itself will change because of it.
Companies Will Have to Adapt
Right now, I think the biggest thing slowing down AI adoption isn't the technology itself. It's company security policies. Many companies still don't allow developers to use external AI tools because they work with private source code, customer data or confidential business information. Honestly, I understand that. Those concerns are real.
But I also think this is temporary. Sooner or later, companies will find a way around it. Maybe they'll run their own private AI models. Maybe they'll use secure cloud solutions. I don't know exactly how it will happen, but I do believe it will happen. Once companies see that developers can produce much more in the same amount of time, it will become very difficult to ignore.
And then another interesting question appears. What will developers actually do with all that extra productivity?
At first, I think many people will simply make their own work easier. If someone can finish a day's work in four hours instead of eight, they probably won't walk into their manager's office and ask for twice as much work. That's just human nature. Some people may spend more time learning. Some may start freelancing. Some may even work with two or three companies at the same time. Whether companies like it or not is a different question.
Eventually, companies will have to react. They'll need to find ways to motivate people to share better workflows instead of keeping them to themselves. And honestly, I think this may become one of the biggest management challenges over the next few years. Technology is usually the easy part. People are much more complicated.
Why One Prompt Isn't Enough
Now... here's the part I find the most interesting.
Personally, I don't think we'll build serious software with just one prompt. Can AI generate an application from one sentence? Yes, absolutely. It can already do that today. But will it build exactly what you had in mind? I don't think so.
Let me explain what I mean.
Imagine you ask AI to build you a car. It gives you something that looks like a BMW. Great. You drive it for two years. Then one day you decide you want something different. Maybe you want it to look sportier. Maybe more elegant. Maybe a little more aggressive. That's just how people are. We always want something new.
So you start changing things.
"Move the headlights a little."
"Make the front look stronger."
"No... that's not quite what I meant."
After a while, you're no longer writing one prompt. You're having a conversation. You're making small decisions. You're improving the result step by step.
And I think software development will work exactly the same way.
The first version AI creates will probably be good enough for many simple projects. Landing pages, internal tools, small websites, maybe even basic online shops. But as soon as you want something unique, something that makes your product different from everyone else's, you'll need much more than one prompt.
You'll start asking questions. You'll change the design. You'll improve the architecture. You'll think about performance, scalability, security and future features. Before you know it, you're no longer asking AI to build an application. You're working with it like another engineer on your team.
And that's the biggest reason why I don't believe software engineering will disappear. I think AI will become the most powerful tool we've ever had. It will help us write software faster than ever before. But a tool is still a tool. Someone still needs to understand the problem, make decisions, ask the right questions and guide the whole process.
In the next part, I'd like to talk about what this means for software engineers. How will teams change? Will junior developers have the same opportunities? Will experienced engineers become even more valuable? That's where things start to get really interesting.