Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep yeah and especially true when you are building AI features and products. I took a break in late Spring when I changed jobs and stopped working so much trying to understand LLMs and it was a cognitive rest.
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
The office work equivalent of the Cowan paradox: When “labor saving” devices like vacuum cleaners, dishwashers, etc. were introduced, the amount of time spent on domestic work did not significantly decrease over the years.
Sociologist Ruth Schwartz Cowan highlighted this phenomenon, suggesting that increased expectations and social pressures kept domestic workloads high.
“Spring cleaning” was a once a year activity that involved the entire household. Vacuums meant vacuuming multiple times a week.
Many middle class households would send clothing to commercial laundries, the washing machine meant doing the laundry at home.
-
The office work equivalent of the Cowan paradox: When “labor saving” devices like vacuum cleaners, dishwashers, etc. were introduced, the amount of time spent on domestic work did not significantly decrease over the years.
Sociologist Ruth Schwartz Cowan highlighted this phenomenon, suggesting that increased expectations and social pressures kept domestic workloads high.
“Spring cleaning” was a once a year activity that involved the entire household. Vacuums meant vacuuming multiple times a week.
Many middle class households would send clothing to commercial laundries, the washing machine meant doing the laundry at home.
@lain_7 I’ve heard of this phenomenon before—I think I stumbled across it in Crichton’s Jurassic Park as a teen, maybe?—but I’d never heard of Cowan before, or the paradox named after her. Thank you!
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep I mean it is Harvard Business Review so it's gonna be followed by "... and that is why you should deploy them at scale".
-
@Ashedryden nobody’s gonna render myself unhireable if i don’t do it
@beep @Ashedryden I'm looking for easier ways to render myself "unhireable"
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep I really don’t think people recognize what they’re giving up at the micro and macro scale.
For individuals, I think, depending too much on these things really hurts your ability to internalize and learn. I saw this when I was learning a new language and realized that three years ago I would’ve internalized things I was asking about repeatedly.
At the medium scale, letting it independently write code for you seems absolutely insane based on knowing what the limitations of LLMs are.
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep the cool thing about technology as a job is that you are always getting to learn and make decisions. I don’t know why anyone would want to give those up. It sounds miserable.
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep The human mind so readily steered away from the holistic view by tasty carrots. AI blindsighting many to the big picture context ...
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep it makes creative and exciting work boring.
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep
> While this sense of having a “partner” enabled a feeling of momentum, the reality was a continual switching of attention, frequent checking of AI outputs, and a growing number of open tasks. This created cognitive load and a sense of always juggling, even as the work felt productive.I suspect this almost throwaway paragraph explains A LOT about the disconnect between people's self-reported productivity and the actual outcomes of delegating work to chatbots
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
These findings have strong "reverse centaur" vibes.
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep
> They reduced dependence on othersWhy is this seen as a positive? (rhetorical)
-
I deeply hate to quote myself, but here’s me back in September: “…the technology’s real value isn’t improving productivity, or even in improving products. Rather, [“artificial intelligence” is] a social mechanism employed to ensure compliance in the workplace, and to weaken worker power.” https://ethanmarcotte.com/wrote/against-stocking-frames/
These platforms are not for you and I, and never were.
@beep oh neat to see you on here! "You deserve a tech union" was very good, I was surprised at how straightforward and readable it was
https://bookwyrm.social/user/samfirke/review/7446482/s/good-read-for-anyone-in-techit#anchor-7446482 -
@beep I think this is still WAY too optimistic about AI, but I guess what would you expect from an article written by managers who see humans as infinitely interchangeable and replaceable resources...
The problem is way more than burnout. It's shiting roles to people who complete the work faster *because* they don't have the training or experience to know if it's actually done well. So they don't do it well, they only do it fast. Designers start vibe coding, turning the engineers into testers just trying to cobble that slop together, then the testers are doing more dev and sysadmin work, and pretty soon everyone is doing every job EXCEPT the one they actually have the skills for. And then in a few years all your code is an unreadable, unmaintainable mess that nobody can work with, including the AI.
@admin @beep that's sort of buried in the article, in the bit where they say that some of the additional work includes Engineers reviewing 'partially complete' work begun by colleagues using AI... it's just that it's *buried* in the article because it absolutely doesn't want to dwell on the implications
-
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
"On their own initiative workers did more because AI made “doing more” feel possible, accessible, and in many cases intrinsically rewarding."
That's terrible, we definitely should go back to the days where workers feel they are accomplishing nothing, hard to get things done and unrewarding.
Those were the good old, pre #AIslop days.
-
@beep
Indeed!"In the study, 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. That may sound like a win..."
HA! Yes, that totally sounds like a "win" if your goal is to exploit workers.@kennypeanuts
Absolutely, in that AI is just the newest fad in neoliberal employee abuse.
@beep -
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep
Burnout in toxic tech companies is nothing new.
It’s about boundary setting and self respect.
Interestingly, people tend to respect you more when you draw a line.
I think the most concerning point of this article is people “reinventing” themselves while trusting the output of a text generator.
That’s the insidious long term risk for companies.
And by the time the adverse effects are visible it will take an enormous amount of effort to fix it. -
Whatever the output gains promised by LLMs, their initial productivity surge is erased over time, and replaced by heavier workloads—and that leads to workers experiencing “cognitive fatigue, burnout, and weakened decision-making.”
All this from research out of the notoriously pro-worker rag [checks notes] Harvard Business Review: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it
@beep @magichopi That article bugs me deeply because it still signals to bosses that they can get “efficiencies” even if it burns people out. But it doesn’t define what productivity means, or how they measure it, or how accurate any of the outputs were, or even what work these people are having it do. What jobs were these? What responsibilities? But toss some care-crumbs to your burnouts and it’ll be fine.
Feels like a head-fake. -
@beep @magichopi That article bugs me deeply because it still signals to bosses that they can get “efficiencies” even if it burns people out. But it doesn’t define what productivity means, or how they measure it, or how accurate any of the outputs were, or even what work these people are having it do. What jobs were these? What responsibilities? But toss some care-crumbs to your burnouts and it’ll be fine.
Feels like a head-fake.@andrewhinton Yeah, extremely fair point. I read it as having some limitations to how far the critique would go, given the source.
-
@andrewhinton Yeah, extremely fair point. I read it as having some limitations to how far the critique would go, given the source.
@beep Oh absolutely and the point you highlight in your post is still true! My anger about it is more about how sneaky an article like this is. Feels really dishonest about its supposed pro-worker positioning
