How AI Is Changing the IT Job Market in Armenia

It used to be possible to become a junior developer within a few months. Now you need not only several programming languages but also the ability to find new AI tools and use them well. Without that flexibility, even an experienced specialist can be left behind.

For years, the IT industry was the Armenian version of the "American dream": a place where you could get somewhere on your own merit, without knowing the right people. There are plenty of success stories of people from other, even non-technical, professions who moved into IT.

But the simple work that newcomers used to start with is increasingly done by AI. This inevitably affects a beginner's chances of finding a job, in programming and in related fields such as UI/UX design and graphics.

"There are even experienced developers who have been out of work for more than a year, which was almost unheard of before. Now they have to look for a job and learn new AI tools at the same time, and those tools aren't cheap when we're talking about professional products," says Narek Hakobyan, co-founder and CEO of the HR Drone job platform and the IT company M One, in an interview with Sputnik Armenia.

Who Saves Money, and at Whose Expense⁠

Traditionally, an IT project is led by a senior specialist who has not only knowledge but also experience of their own mistakes. Years ago, they started at the bottom with the simplest tasks, which the seniors of the time handed to them, and then fixed their errors.

"Now that there are AI coding tools (Claude Code, Cursor and others), many seniors, including those in my own company, tell me that beginners and juniors get in their way more than they help. Today I give the same task to Claude Code, which does it much faster, and all that's left for me is to review the result once," Hakobyan says.

For its part, the company saves money, not so much on a junior's salary as on a senior's time. That time is far more expensive and can go to complex work instead of teaching beginners. Of course, teaching should be seen as developing the company's human capital, which matters just as much in the long run. But when we talk about today's problems rather than the future, Claude Code was a level below experienced developers just a few months ago, and now, by their own account, it has almost caught up with them, Hakobyan notes.

Employers Are as "Scared" as Employees⁠

New AI models can turn entire areas of IT upside down. For example, Meta recently announced an AI program that will automatically annotate any images, that is, recognize and label any object in photos and videos. It is clear that overnight this could become a challenge for many computer vision startups across all kinds of fields, such as face recognition.

Another example: an AI agent that creates 3D objects has come to market. It is hardly necessary to say what this could mean for 3D design, for instance in video games, where hundreds of designers draw tens of thousands of 3D objects.

Starting new IT projects has become unpredictable: months, even years, of work can turn to dust in a single day. At this crossroads, some stop and others move forward.

Some companies lay off employees, while others try to solve more problems with their existing and new people.

"We are taking the second path, because we plan to build more products. Let's look at it from another angle: yes, projects can collapse quickly, but new ones can be created just as quickly. If bringing something new to market used to take a year or more, now it can be done in two or three months. You shouldn't be afraid of problems; you should be flexible," Narek says.

Developer, Program Yourself Right⁠

In fairness, he notes, part of the layoffs happens not only because of the employer but also, so to speak, because of the employee. Some developers have worked with "traditional methods" for years and now see AI assistants not as a tool but as a competitor ("sooner or later they'll write better code than me, and I'll be fired," and so on).

"If you keep telling yourself that you'll be let go, sooner or later life will arrange itself so that you really are, because your employer won't see your motivation to work more effectively with new solutions. I should say this ability is now required not only from beginners but also from senior developers and team leads. Employers expect from them not just knowledge and experience but also the ability to step out of their comfort zone, find new AI tools and learn them," Narek stresses.

Back to the "Armenian Dream"⁠

Narek himself started working as a developer in 2015. Back then, knowing one programming language well was considered enough to find a job. Now you need to know several languages and various supporting tools, which we have already discussed.

"When the AI boom began, we gave interview candidates a task to write code with pen and paper, to see what they could do 'honestly,' without AI. Now it's the opposite: we say 'write it with AI,' because we want to see how well the specialist handles such tools," Narek explains.

How do you learn all this if professional AI tools, as we noted at the start of the conversation, are expensive?

A lot can be found in free resources too, in articles and videos. Yes, the road to the profession has become longer, but you shouldn't give up and, as we said many times during the conversation, you shouldn't get tired of learning, Narek concludes.