The Role of AI in Software Development

There is a growing buzz in the software development community that we foresee that AI will soon put us out of work. We will be replaced by thousands of conscious bots that will hold meetings and discuss with each other the best ways to make lists of requirements and discuss them again and then discuss more about how to implement them.

That is, they will do what we do (arguing it seems), but orders of magnitude faster. However, that time has not yet come and right now their role is being the best it could be: they are helping us to be more productive.

The future of what's coming

I want to put on record before I start that I am writing this article on May 14, 2024. Just yesterday OpenAI announced its real-time voice assistant and demonstrated what many of us already saw coming: the way we interact with technology has changed forever.

And I want to put the date on record because the last year has shown me that we are making great strides and that, surely, this article and everything I am going to comment on here will be obsolete by the time it passes all the phases of revision and I have published it.

I have to admit that before the first steps of a baby copilot by GitHub I remained skeptical about what it could and couldn't do, of all the incredible promises it made and that I expected to be the smoke and mirrors that I'm used to in the startup ecosystem. That was one of the reasons why I didn't decide to try it out and why I let it grow and grow in the shadows while I looked away.

However, zapping around the web (does Gen Z know what zapping is?) I found that they were giving GitHub Copilot free for students. And it turns out that I pay a PhD student tuition every year that allowed me to try it for free. So I decided to give it a try.

Copilot

Half a year later, I can't go back.

Using copilot is one of those things that you experience in life that marks a before and after in how you see and understand something. In just half an hour, I went from skepticism to excitement about the future ahead.

Copilot works. And it works great.

And by "working" I don't mean that it makes correct code or that it understands what I'm saying. By working, I mean that it's useful, tremendously useful, and that it offers me value that I didn't think it could give me.

The fact that it increases your productivity is totally true, it autocompletes the code as if it were reading your mind. It fills functions for you, writes tests, applies refactors that you have in your head. Just by showing the intention of what you want to do, it already knows where you are going and it gets ahead of you. Something magical.

A few articles ago I talked about how cognitive load affects us in development and that it takes us away from focusing on what really matters, providing value to our users.

And boom, copilot takes that cognitive load away from you in one autocomplete. You no longer need to implement the tedious methods of a collection. You no longer need to remember how to use the http request library. Copilot knows everything and starts stuffing code like a champ. Offering you the opportunity to focus on what's important: the business logic, the needs you're covering, your users.

The role of AI

Therefore, and after so much hype, I have to say that we have led AI (at least the one that refers to programming) to the best terrain for everyone. Code assistants are not replacing us by any means, they are a tool. And one that gives us super powers.

Thinking about all this AI and the danger to our work continuity, I have realized that what is actually happening is something that has been happening for centuries. Something that every society has feared more than once and from which it has always emerged unscathed. What we are seeing with AI is the inevitable advance of progress.

I can easily imagine the farmer of the mid-1950s seeing how mechanization threatens his work. If machines do everything for us, what will happen to our jobs? What will we do?

Well, we'll work on other things, new jobs will be created as has always happened and progress will allow us to continue advancing. Mechanization in agriculture saved millions of lives and ended hunger in many places. And in the same way, AI will democratize software to levels that we are not yet able to imagine.

And if there's one thing we have to fear, it's ourselves. Because it won't be AI that replaces you in your work, it will be someone who knows how to use it better than you.

So my advice is clear. If you don't have a copilot (or alternatives) yet, get it now. Train in tools that bring AI closer to our day-to-day lives and keep learning so you don't become obsolete. The future is optimistic, let's take advantage of it.

TL;DR

Artificial Intelligence is already transforming the way we develop software. Tools like GitHub Copilot prove that AI isn't here to replace us, but to boost our productivity and allow us to focus on the most complex and valuable aspects of software development.

Although, we should not be complacent. Those who know how to best leverage the capabilities of AI will have a competitive advantage. Therefore, it is crucial to keep up with the latest tools and techniques, continuously training yourself so as not to become obsolete.