High on the list of Office Menaces, right up there with fish in the microwave and meetings that could have been an email, we have … workslop.
If you haven’t had the pleasure of experiencing it yet, workslop refers to a clearly AI-generated something that is presented as completed work yet merely resembles actual completed work.
For example, when you get partway through a report and realize it’s a bunch of pretty words that are coherent on face value but entirely devoid of substance. I suspect this is infinitely worse on engineering teams dealing with vibe-coded whatevers.
Not surprisingly, after investing time in literally nothing, co-workers report feeling “annoyed (54%), frustrated (46%), confused (38%) and offended (22%).” These sentiments are also rampant in the special circle of hell where you don’t want to produce workslop, but you can’t get the AI to produce anything but. I digress, though.
Workslop is a direct result of the steady drumbeat of “use AI to improve your productivity … or else.” Eager to stay employed, folks grab at every opportunity to use the tools and, in the quest to demonstrate time saved, either forget, gloss over, or don’t realize there’s a very important part where you need to validate the output.
You know what doesn’t save time? Reviewing someone else’s workslop. You know what also doesn’t save time? Redoing workslop.
Between that and the push to insert AI into legacy systems it’s not designed for, it should come as no surprise that meaningful productivity gains have yet to take off.
Pressure … pushing down on me
Yes, yes, I know. The models will get better. Agents will do the fixing-up. Glorious efficiencies are ahead, etc., etc. All of which is likely to happen. But it doesn’t fix the dumpster fire of a report that’s currently living rent-free in your rage.
Faster is not always better. More is not always good. It used to be that the limits of manual effort could moderate this, but AI has obviously removed that barrier and magnified the downstream effects.
One of those effects is an interesting twist on burnout, which is when chronic work-related stress turns you into an exhausted, cranky zombie who can’t be bothered to care anymore. This phenomenon was already a problem in high-pressure workplaces, and it’s spreading at the pace of AI.
It starts with the pressure of automating all your tasks — even the ones that might not need it — because AI literacy is now a core skill that affects performance ratings and such. Your outputs and deliverables increase (or seem to increase), and the pressure intensifies because you have to maintain and likely continue improving your new AI-assisted baseline or face the consequences of “slacking.” Ironically, this can really screw over the excited early adopters who devoted extra time and off-hours time to experiment with what was possible. Their outputs are now artificially inflated at a pace they meant to be temporary.
Quantity trumps quality in this scenario, and the people whose job is to receive or test the results are drowning in panic-driven workslop. They’ve got the same performance pressure plus a fair amount of additional anger and frustration at having to clean up everyone else’s mess.
Then, since the all-AI, all-the-time mentality is unavoidable and in your face every time you open a digital screen, the constant specter of robots outright replacing you can cause another sort of paralyzing distress altogether. The latter even has a fancy name: AI Replacement Dysfunction (AIRD).
There you have it: Another circle of hell unlocked. It’s amazing people still show up for work. And what do the execs do? They double-down on the threats.
Sure. That’s one way to handle it.
… Pressing down on you
AI accelerated it, but we were on a trajectory for trouble regardless. Emboldened by a favorable economic climate, employers have been desperately trying to roll the clock back to 2019, and so now we have rampant layoffs and mandates of various sorts that are the cherry on top of run-of-the-mill workplace stress. Hiring has slowed (that’s a whole ‘nother hellscape), and more than half of employees report job-hugging — sticking with their current role because there’s no good alternative. This isn’t a recipe for productivity either, and it amps the burnout risk significantly.
When left unchecked, burnout always backfires on both the employee and employer. The health consequences alone can have lifelong and life-changing impacts, and the cynicism it breeds is a contagion that easily spreads across entire teams. Absences go up, deadlines are missed, demoralized workers get even further disengaged and stop caring, workslop runs rampant. And customers will eventually notice.
We’ve got all the makings here for a crisis — one more to add to the pile, apparently — that makes dealing with workslop alone look easy. I wish I could say otherwise, but the situation is unlikely to change until other macro factors change. We can’t control that, but we can try to temper the burnout.
If dealing with other people’s workslop is dragging you down, make sure management is aware your “slowness” is tied to other people’s sloppiness. Kick the work back if possible, and avoid the temptation to tag it with a snarky ai;dr. Those folks are feeling the stress, too.
If productivity pressure is turning you into the workslopper AND if it’s safe to tell your manager without raising a big ol’ performance-management or lay-me-off-first flag, then make sure they know you’re overwhelmed. Their hands might be tied, but there also might be a way to reset your baseline to something more sustainable.
If you’re the manager who hears or observes a workslop problem, your team needs processes and expectations that reinforce critical thinking and human judgment. For example, requiring a “how I validated AI-generated outputs” sentence or two at the bottom of every report. If you’re an engineering team, maybe you do this during a retrospective. It might mean a temporary dip in quantity, but should be offset over the long term by less time spent fixing garbage. Here’s the message to communicate to leadership when they push back on numbers: You’re optimizing for real results, not optics.
And if you’re in a position where it’s not safe to tell your manager you’re struggling or your leaders only care about the optics, you are in a very tough situation. As your friend, my best advice is that no job is worth sacrificing your health or your family. If you have to job-hug to pay the bills, take back some control by mapping out a plan for what might be next when circumstances change.
The flip side of AI changing everything so quickly is that the drivers of the vicious burnout cycle could change quickly, too. This, too, shall pass. We just can’t predict when.
Meanwhile, you’re in very good company in the pressure cooker.
Requisite footnote: If you’re struggling with burnout or other health consequences of chronic stress, please consult with a professional. If your company offers an Employee Assistance Program, that’s a good place to start.
All opinions here are my own. All text is my own, too, including the em dashes. I welcome constructive comments and discussion on LinkedIn and Bluesky.