What it means to be junior in the age of AI

I entered the workforce in 2024, with little experience and plenty of ambition. That was also the year I found my first full-time job at Dust. Joining one of the first 20 people at a startup working in AI was exciting, intimidating, tiring, and stimulating, often all at once.

By then, AI had already gone through several moments that made its progress difficult to ignore, from Attention Is All You Need to GPT and many other developments. I was curious about what these systems might change in the way we work, but I did not have a clear view of what that would mean for jobs, or for my own career. I mostly saw a field that was moving quickly and seemed worth paying attention to.

For a junior, Dust was an unusual place to start. I was joining a company in an industry that was still taking shape, surrounded by people building products and practices that did not have much precedent. What surprised me most was that many of the people around me were extremely competent, and some of them had been thinking about these problems long before I had. I expected that experience to come with a clearer view of the future. It did not.

Working without a clear view of the future

One of the strange things about working in AI is that even people with years of experience are often unsure about what the next six months will look like. They may have strong opinions about where things are going, more context than I do, and better instincts about which developments matter. But they are still trying to make sense of a field that changes quickly.

A new model, product, or use case can change the situation almost immediately. Some of the patterns we need do not exist yet, so we are not always applying established methods to new problems. Sometimes we are still trying to understand what the methods should be.

That makes it difficult to know whether a decision will still make sense a few months from now. We can have a reasonable idea, make a choice, and then discover that a new development has changed the context around it. This happens often enough that uncertainty is not an occasional inconvenience. It is part of the work.

At Dust, "default to action" is a principle that comes up often. I have come to understand it less as an instruction to move quickly and more as a reminder that decisions are part of the job. We build things, ship them, change our minds, and sometimes roll them back. We can feel certain that we are heading in the right direction and later find out that we were wrong.

I am still learning how to deal with that. My instinct is often to look for more information before making a decision, partly because I want to avoid making the wrong one. But in a field that changes this quickly, waiting for complete certainty usually means waiting too long. At some point, you have to make the best decision you can with the information available, then pay attention to what happens next.

Learning what deserves attention

There is also a practical difficulty that has less to do with AI itself and more to do with the pace of work around it. There is always something to do. A new model comes out, someone finds a new use case, a customer raises a problem, a message arrives in Slack, or a small issue appears somewhere in the product.

When you are junior, it can be difficult to decide what deserves your attention. You are still trying to understand how the organization works, what people expect from you, and where you can be useful. It is tempting to treat every request as important, answer every message as quickly as possible, and fix every problem you happen to notice.

I have done this many times. I have spent time on things that felt urgent in the moment and seemed much less relevant by the time I finished them. The speed of the industry makes this feeling worse. You can spend a whole day working on something and still wonder whether it was the right thing to work on in the first place.

After two years in this industry, the best advice I have received was also the most direct: do not care about things that are not important.

I do not interpret that as permission to be careless or disengaged. It means trying to understand why something deserves my time before giving it that time. Not every Slack message needs an immediate answer. Not every small problem needs to be solved by the first person who notices it. If something is genuinely important, it will usually come back around.

This is still difficult for me. Ignoring something can feel irresponsible, especially when you are junior and still trying to show that you are reliable. But reacting to everything is not the same as contributing to what matters. The question I am trying to ask more often is not only, "Can I do this?" but also, "Why should I spend time on this now?"

Being junior while the rules keep changing

I am speaking from the perspective of someone who started their career in AI, so my experience is not representative of everyone entering the workforce today. I can imagine that the uncertainty may feel even more difficult in industries where AI is changing expectations around jobs without offering a clear idea of what comes next.

I do not have a solution to that anxiety. I sometimes feel it myself. The technology develops quickly, and it is hard to know which changes will last. When people such as Sam Altman describe a future in which some people may eventually choose not to work, it is understandable that younger people might feel unsure about the kind of career they are building.

At the same time, working in AI has made one thing clear to me: nobody has a complete map. The people with more experience may have better instincts, stronger opinions, and more information, but they are still making bets about the future. They are also revising their assumptions as the technology changes.

That can be unsettling when you are looking for guidance. It would be easier if someone could explain exactly what skills will matter, what the industry will look like, and which decisions will turn out to be useful. But no one can really do that yet.

For now, I am trying to become more comfortable with a few basic things. I am trying to make decisions without having all the information, notice when I am wrong, and spend less time on problems that do not deserve my attention. None of this gives me a clear idea of what the AI industry will look like in five or 10 years. I doubt anyone has one.

Still, I am glad I started my career in a place where that uncertainty was visible. It can be uncomfortable to see that even highly experienced people do not always know what comes next. But it is also a more honest way to learn. You do the best work you can with what you know, accept that some of your decisions will be wrong, and try to understand what the result teaches you.

That is probably what being junior in AI means to me for now. Not having a clear view of the future, but learning how to work when nobody else has one either.