Saide effects

AI assistants do not affect everyone on a team in the same way. Some people find them motivating, while others distrust them or lose interest in their work. None of these reactions is a flaw in the person who feels it. They depend on each person's personality and on what they enjoyed most about their work before these tools arrived.

The anxiety of not stopping

An AI assistant does not get tired, has no working hours and is always willing to try one more thing. For someone prone to anxiety, that availability is a trap. Every answer opens a new idea, every finished change invites the next one, and closing the session feels like wasting an advantage. The day stretches without you noticing, and the feeling of moving fast hides the tiredness until it is too late.

The tiredness does not come only from the hours, either. In March 2026, BCG research published in Harvard Business Review gave a name, AI brain fry, to the mental exhaustion of overseeing several AI tools or agents at once. The more agents work in parallel, the more decisions the person directing them has to review.

Working longer hours on a given day is not serious. The trouble starts when the tool's pace begins to set the person's, when it should be the other way around.

When programming was the fun part

For many people, programming is a game. What motivates them is precisely wrestling with a hard problem until they solve it. When the assistant solves that part in seconds, that person is left with the least enjoyable part of the job, which is reviewing, correcting and understanding what someone else did. It is not unusual for boredom to set in, along with a vague feeling of losing something, even when the results are good.

That boredom matters more than it seems. Someone who stops enjoying their work loses attention, and in a code review attention is exactly what keeps an error from reaching production.

In my case the opposite happened. Before, many ideas got left behind because I did not know how to build them. Today I can, because AI helps me with the part I do not know. What I need now is the creativity to imagine what to build. Sometimes it is something that already exists but does not fit what I need, and with AI I can make my own version, always reviewing what it does.

That is why I think boredom has a way out. What changes is the enjoyable part of the work, which stops being solving the problem and becomes deciding which problem is worth solving.

Distrust

Other people react with distrust. They know the assistant writes with the same confidence when it is right as when it is wrong, so they review every line, doubt every answer and end up working more than before.

It is not an isolated reaction. In Stack Overflow's 2026 survey, almost half of developers said they trust AI only when they can easily verify its output, and just 6.6% trust it for important work decisions. That distrust has a reasonable basis and, handled well, it is healthy. The problem appears when it becomes permanent, because reviewing everything with the same suspicion is exhausting and leaves no time for what really needs attention.

Frustration

Frustration appears when AI almost gets there. The answer looks right, the code seems complete and, when you test it, it fails on the detail that mattered most. Fixing something that is almost done can take longer than doing it from scratch, and after the third correction patience runs out.

There is another frustration, harder to name, that comes from the team. While one person already works with AI on almost everything, another is just trying it and a third is still not convinced. For me, the hardest part of this whole process is coordinating with others on progress and adoption, when everyone moves at a different speed.

Uncertainty

Finally there is uncertainty, which reaches almost all of us at some point. Every week brings a new tool, a more capable model or a headline announcing the end of some job. It is hard not to wonder what will happen to your own role in two or three years, and when that question has no answer, it keeps going round and wears you down. On top of that comes the constant comparison with people who seem to get more out of AI, and on social media there always seem to be many of them.

Even so, uncertainty does not come only from AI. Today almost nothing is certain, in technology or outside it. What is clear is that AI gives us tools to make better use of our creativity and productivity, as long as we learn to use them.

In a team, everyone moves at their own pace

Each person adopts AI at their own pace, and that pace cannot be forced. You cannot impose motivation, or the way the tool is used, or that everyone immediately understands what it is really for. It is something all of us have had to face, and not everyone knows how.

What I did was put together a shared way of working. That is how GATE was born, a framework that brings order to work with AI and that each company uses to build its own SDLC, which is the complete software development life cycle, from the moment someone raises a need until the change works in production. With a shared system, each person can move at their own pace without the team losing order, because everyone works with the same rules and the same checkpoints.

Some ways to handle it

There is no single recipe, because each reaction calls for something different. Even so, some practices help in almost every case.

• Start with a goal and stop when you reach it. It is what I do. I decide what I want to get done, finish it and close the session, even if the assistant offers to keep going. I am not afraid to close it, because my assistant keeps a long-term memory and picks up the next day where we left off.
• Keep part of the work for yourself. Every now and then pick a problem and solve it by hand, for the pleasure of it and to keep the judgement you later need to review what AI does.
• Turn distrust into method. Instead of reviewing everything with the same suspicion, rely on written rules, automated tests and review by another person, and put your attention where the risk is greatest.
• Talk about it in the team. Acknowledging that it affects each person differently takes pressure off. People who lead teams can help a lot simply by asking how each person is experiencing it, and not only how much faster they are moving.

The person sets the pace

AI assistants are here to stay and will keep improving. What does not change is that the work is done by people, with their own character, motivation and limits.

For me, the key is learning to use AI with clear boundaries. That means building and testing fast, getting things wrong early so you can correct them in time, and moving on. And instead of spending energy on small problems, focusing on big solutions for the problems that really matter.

Each person will find their own balance, and it is fine if it is not the same for everyone.