AI Made Me More Efficient. Why Am I Busier Than Ever?

Introduction

I have been spending less time writing code lately, but the amount of work I do has actually grown. The time saved on coding has not turned into time off. It has been filled with product work, UI design, backend development, testing, communication, and acceptance reviews. Overall, AI has not made my work easier. It has just changed the way I am busy. AI raises productivity, but once productivity goes up, more work follows.

How My Work Has Changed

Since the second half of the year began, I have felt my responsibilities growing. Frontend, backend, product, UI design—I am involved in pretty much all of them.

For example, I recently worked on an NDVI analysis module. From competitor research and product planning to UI design, frontend development, and backend development, I handled almost the entire process myself. In the past, I might only have been responsible for the frontend. Now I need to deliver a fairly substantial product module from scratch, all the way through to completion.

But one person taking on more responsibility does not make the other processes disappear. I still need to attend product planning meetings and test case reviews, and debugging and confirming details still take a lot of time.

I need to explain the business logic to the testers, make clear what the product is ultimately supposed to achieve, and generate an initial set of basic test cases. Once the frontend and backend have been integrated, I still need to confirm with the product owner that the functionality meets expectations, then complete the acceptance review.

AI can certainly help me write code and quickly generate some basic content. But it does not attend those meetings for me, and it cannot confirm business details with different people on my behalf. Many things still depend on my own judgment, communication, and effort to move them forward.

Development no longer works the way it used to, when you could be responsible for just one part of a feature. As a frontend engineer, I used to be able to focus on getting the interface and interactions right. Now, at least at our company, there is a growing expectation that one person take responsibility for an entire module.

Having control over the frontend, backend, product, and UI actually feels pretty good. I can do things the way I think they should be done, work with fewer constraints, and bring my ideas to life in full.

At the same time, the responsibility is greater. Before, I needed to do my part well. Now I need to find a way to make the whole module work.

One person is an army. You are the general, and the AI tools are your soldiers. You have more and more soldiers at your disposal, but winning the battle still depends on your judgment and your ability to organize the work.

Greater Efficiency Has Not Reduced the Workload

For a while, people often talked about how much AI had improved their efficiency. I seem to hear that less often now.

I think that is because, when people first started using AI, the differences between them were obvious. Some were using it and others were not, so it was easy to see how much faster AI made you. But now there are more AI tools, the tools are more mature, and many people use them. The gap between people in coding speed alone may no longer be as noticeable as it once was.

As efficiency improves, expectations for results rise too. Features that once took a long time to build can now be completed quickly. So the time that used to go into writing code gets filled with more work. One person might previously have been responsible for a single feature; now they may need to handle product analysis, UI design, frontend, backend, testing, and releases at the same time. AI gives you the ability to do more, but it does not automatically reduce what you are expected to take on.

A company pays for your working hours. AI simply lets you get more done within those same hours. Without AI, you might have been able to deliver a certain amount. With AI, the expectation is that you deliver a little more. That is really all there is to it.

So I increasingly feel that greater efficiency does not mean fewer working hours, nor does it mean work becomes easier. Often, it just means one person can cover a wider range of responsibilities.

More Efficiency, Less Sense of Achievement

Another change is in how I feel about the work itself.

Building something from scratch used to give me a strong sense of achievement. Now that many things are easier to implement, finishing them does not excite me as much as it used to.

When something takes less effort to obtain, it seems to become less rare. In the past, implementing a feature might mean reading a lot of documentation, trying different approaches, and working through problems one step at a time. It was slow, but the process also made me feel that I had truly solved a problem.

Now AI can quickly suggest an approach and generate much of the code directly. The result still gets built, of course. But some of the exploration and waiting along the way have been removed, and the finished result does not bring satisfaction as readily as it once did.

I used to think building a product from scratch was hard, so finishing one made me happy. Now that the process has become a little easier, my expectations have risen along with it.

AI has made many things more convenient, but in some ways it has also taken away some of the enjoyment I used to find in doing them.

What is even stranger is that AI has made so many things faster, yet people have not been freed from being busy. What a company is really buying is your time, not the time you spend writing every line of code by hand. AI lets you produce more during that time. The work itself does not automatically shrink.

How I Might Adjust

Mindset matters.

I cannot honestly say I feel no anxiety at all. AI is developing quickly, and it really will replace many kinds of work. Many things that once took a long time to learn are now much easier to get started with using AI. Technology will keep moving forward, and ordinary people have very little power to stop it. Since I cannot stop it, I might as well start by learning to live with it.

AI still needs people to tell it what to do, and people still need to judge whether its output is actually usable. As the limits of what one person can do expand, initiative and the ability to act become even more important.

In the past, one person might only have been able to work on one thing at a time. Now I can direct AI to work on many things at once. That does improve efficiency, but it also creates another problem: there are more and more things to manage, my attention keeps switching between tasks, and it becomes hard to settle down and think deeply about a problem.

So I may need to rethink a few things: which tasks I should give to AI and which decisions I need to make myself; which work can move forward in parallel and which needs an uninterrupted block of time; and what is truly worth getting done for me.

Closing Thoughts

With AI, I really have become busier.

I spend less time writing code, but I am responsible for more. My work is no longer just about completing one part of development. I need to think all the way from requirements, product planning, design, development, and testing through to final delivery.

There is a good side to this change. One person can do more and bring an entire module to life according to their own ideas.

But there is a cost too: broader responsibilities, more switching between tasks, less sense of achievement, and less time for deep thought.

Next, I want to try changing how I use AI. Alongside having it help me complete tasks faster, I want to have longer, deeper conversations with it whenever I can, bringing it into problem analysis, thinking through approaches, and reviewing what happened, instead of treating it only as a tool for writing code.

Whether that will make things a little easier for me is something I will have to keep watching.