Saving time, or moving time from one category to another

This is a review of a few articles that came to my attention with similar themes around the use of AI and whether or not it saves time or consumes more time. From my experience, I think it does both as well as it compresses time. It makes the fast things go faster, but sometimes, it makes them take longer, too. Some of that can be attributed to user error (bad prompts, improper RAG source files, etc.) and some of that comes back to ambiguity. When you start from fuzzy and out of focus, don’t expect a few prompts to bring you clarity.

AI fatigue is real and nobody talks about it

An article from Siddhant Khare, February 7 2026 is written from one engineer to another. It opens with a brutal truth:

I shipped more code last quarter than any quarter in my career. I also felt more drained than any quarter in my career. These two facts are not unrelated.

I feel that. Being productive comes with a mental cost. Bring productive means holding more in your working memory. Start a task over here, leave the agent or AI or whatever to do its thing, start something else somewhere else. We don’t give ourselves a break and wait, staring into the abyss, letting our subconscious churn. No, we tokenmax. We start another thing. This leads to cognitive overload and context switching from one AI task to another.

Did that process finish? Were the outputs any good? How about this other process? Oh, not done yet, check back in a few. Go to this other thing. Decide whether or not to start a third process. Oh, shoot, I forgot about this other thing. I need to review the research plan before it goes off and does its thing. Oh and now I have 2000 words to read that I asked for.

Siddhant captures my experience perfectly:

Before AI, I might spend a full day on one design problem. I’d sketch on paper, think in the shower, go for a walk, come back with clarity. The pace was slow but the cognitive load was manageable. […] Now? I might touch six different problems in a day. Each one “only takes an hour with AI.” But context-switching between six problems is brutally expensive for the human brain. The AI doesn’t get tired between problems. I do. […] AI reduces the cost of production but increases the cost of coordination, review, and decision-making. And those costs fall entirely on the human.

He goes further into the psychology, which I appreciate, because it supports the phenomena I have been feeling:

Creating is energizing. Reviewing is draining. There’s research on this — the psychological difference between generative tasks and evaluative tasks. Generative work gives you flow states. Evaluative work gives you decision fatigue.

What if AI just makes us work harder?

An article from Tim Harford, April 2 2026 has a similar theme.

He takes a different tact, citing less his personal experience and instead leaning into a recent study from ethnographic researchers from the UC Berkeley Haas School of Business. And, surprise! They found that tech workers were more productive but were also voluntarily doing more work. Longer hours, more tasks, and feeding their own burnout.

When you pull back and look at what is at play, it makes sense:

Consider a freelance programmer, paid by results, who used to work 10 hours a day and suddenly finds that they can achieve the same results in two. Common sense might suggest that the coder will start to enjoy the pleasures of a two-hour workday, but economic theory is more ambiguous: the “income effect” suggests that the worker should work fewer hours because they can achieve so much by working so little. The “substitution effect” says that workers should work longer hours, as each extra hour yields bountiful rewards.

Familiar pattern, new technology

There we are. This is what always happens. New technology never saves us work, even though when it first hits the scene that is all anyone promises. In fact, this is the first technology that promised early in its lifecycle to eliminate work all together. (And jobs, by design! That’s the whole point! Automate the economy! But I digress…)

The internet promised the “paperless office” — predicted as early as 1964 but not a reality for most people until the past decade. The rise of smartphones starting in 2008 led to predictions like voice becoming the dominant input method, or tablets becoming more popular than PCs or laptops, or that passwords would become obsolete. We are still working on most of these.

Almost every new technology promises to reduce work, but instead they accelerate it. Email didn’t reduce work or meetings. Neither did instant messaging. Working from home, in many ways, blurred the lines between work and non-work time. Always having a smartphone on your person means you are always available to do something.

AI is no different in how it was promised to save time but is actually creating more available space for us to fill with more work. “But now I can use AI to create and act upon all the great idea I have had!” Sure, in what time? When? While watching your kid’s baseball game? While standing in line at the grocery store?

How about this instead. Next time you prompt something from AI, wait for it to finish. Don’t start a new task in another tab. Wait. Watch it, think about what you want to output to be. Stay in the task with it. Give your brain a mini-rest and keep it in the context you need to evaluate the output. I bet you will still be more productive and less burnt out on chasing the next output.