Why do I feel busier since I started using AI?

TL;DR: You probably feel busier because AI makes tasks quicker to start, and the time it saves gets spent on more work and longer days. Workplace studies find modest time savings per task alongside a faster pace, a wider scope of work and more hours worked. Keeping the benefit means deciding in advance what the saved time is for.

Although using AI gave me this super godlike power to work on anything, after a while I felt drowning under unfinished tasks!

Does AI save any time at all?

Yes, a modest amount. The savings are real, but they are small enough to disappear into a normal week without you noticing.

The Federal Reserve Bank of St. Louis surveyed U.S. workers and found an average time saving of 5.4% of work hours in its November 2024 survey^[1]. In a 40-hour week, that is roughly 2.2 hours^[2].

Two hours a week is worth having. It is also spread across dozens of small moments: ten minutes saved on an email here, twenty on a summary there. Nobody hands you those minutes as a free afternoon. They show up as slightly earlier finishes on individual tasks, and something else usually fills them.

The surveys also measure what people report saving on tasks. They don't tell you where those hours went afterwards.

Where does the saved time go?

Mostly into more work. When a task gets cheaper, people tend to do more of it, and to start tasks they would have skipped before.

Researchers followed generative AI use at a U.S. technology company with 200 employees for eight months^[3]. They observed that employees "worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day, often without being asked to do so."^[4][3]

The last part is the one to notice. Often nobody asked. The extra work came from the people doing it.

This is easy to recognise from the inside. The report you used to skip because it took an afternoon now takes forty minutes, so you write it. The email that was good enough now gets a second, more polished version. The side task that felt too hard to attempt now feels possible, so it joins the list. Each choice makes sense on its own. Added together, they make a fuller day than the one you had before AI.

There is also a timing effect. When starting a task takes almost no effort, the natural stopping points go away. A job that would have waited until tomorrow because it felt heavy at 9 p.m. now feels light enough to start at 9 p.m. The headline of one report on the study puts it plainly: AI made the workday longer and more intense^[4].

Why does AI work feel tiring even when it's faster?

Because the job changes shape. Before, you produced something. Now you prompt, read the result, check it, fix it and decide whether it is good enough, and each step is a small decision.

That checking work is easy to leave out of your estimates. You plan for the twenty minutes AI takes to draft something and forget the half hour you spend making sure the draft is right.

More use does not reliably mean less strain. A field study of AI-assisted programming found that moderate use of either in-code suggestions or chat prompts improved efficiency and reduced perceived workload compared with not using the tool, while excessive or combined use lessened these benefits^[5]. That study looked at coders, so treat it as a hint for other kinds of work. The pattern it suggests is that there is a useful dose, and going past it costs you some of the gain.

Does AI create more messages and interruptions?

Possibly. One reported analysis points that way, though the evidence is weaker than the studies above.

According to that report, the time workers spent on email, messaging and chat apps more than doubled after AI adoption, and their use of business software rose by 94%^[6]. The data is observational and comes from a news report. It hasn't been through peer review, so it shows a pattern without proving AI caused it.

The mechanism is plausible, though. If you produce drafts faster, there are more drafts to send, review, comment on and discuss. Faster output can mean more conversation about that output, and conversation breaks up the long stretches of focus that hard work needs.

Is it the AI, or is it my job?

Probably both, and the research can't yet separate them cleanly. Understaffing, what your manager expects and the communication overload you already had all push in the same direction.

The clearest evidence so far comes from workplaces. There is little research on what AI does over long periods to people using it for personal planning, studying or creative projects. If you are a student with a thesis or a freelancer juggling clients, the workplace findings are a reasonable guide to what might be happening. They are not a measurement of your situation.

The same caution applies if you have ADHD. None of the studies here looked at it. Still, the logic is worth thinking through. AI lowers the effort of starting, which can feel like a relief if starting is the hard part for you. The same low effort also removes the friction that used to make you stop, so the risk of drifting into one more task, then another, may be higher. If your attention or energy is a health question, a clinician is the right person to talk to.

How do I keep the time AI saves?

Decide what the saved time is for before AI saves it. Otherwise new work fills the gap by default.

Some practical ways to do that:

• Write down today's list before you open an AI tool. When a new task appears because it now feels easy, check whether it was on the list. If it wasn't, put it on tomorrow's list.
• Count checking time in your estimates. "Draft the intro with AI, 20 minutes" should sit next to "Check and fix the intro, 25 minutes".
• Set a finish time for the day and treat it as fixed. AI removed the natural stopping point, so you have to put one back on purpose.
• Use AI for specific steps of a job, such as a first outline or a summary of your notes. Constant back-and-forth across every step is where the coding study found the benefits shrinking^[5].
• Send AI-assisted drafts at set times, for example once before lunch and once shortly before you stop work, so you aren't triggering replies all afternoon.
• Decide what good enough looks like before you start. A second polished version is optional. Because it's cheap, it's easy to drift into, so make it a deliberate choice.
• When you finish early, let the gap stay a gap. A ten-minute walk is a fine use of time AI saved you.

This is the problem TimeBud is built around. We break a goal into jobs with minute estimates, then plan today only into the open time on your calendar, leaving breathing room after events. When a job finishes early, the focus timer moves to the next job in the queue or the plan shows a gap. It doesn't invent extra work to fill the space.

Try TimeBud free