> “Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers. On social media, which is increasingly clogged with low-quality AI-generated posts, this content is often referred to as “AI slop.” In the context of work, we refer to this phenomenon as “workslop.” We define workslop as AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task.”

> “Based on participants’ estimates of time spent, as well as on their self-reported salary, we find that these workslop incidents carry an invisible tax of $186 per month. For an organization of 10,000 workers, given the estimated prevalence of workslop (41%), this yields over $9 million per year in lost productivity.”

https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity

AI-Generated “Workslop” Is Destroying Productivity

Despite a surge in generative AI use across workplaces, most companies are seeing little measurable ROI. One possible reason is because AI tools are being used to produce “workslop”—content that appears polished but lacks real substance, offloading cognitive labor onto coworkers. Research from BetterUp Labs and Stanford found that 41% of workers have encountered such AI-generated output, costing nearly two hours of rework per instance and creating downstream productivity, trust, and collaboration issues. Leaders need to consider how they may be encouraging indiscriminate organizational mandates and offering too little guidance on quality standards. To counteract workslop, leaders should model purposeful AI use, establish clear norms, and encourage a “pilot mindset” that combines high agency with optimism—promoting AI as a collaborative tool, not a shortcut.

Harvard Business Review

That article will probably not surprise any of you. Generative AI produces results that are either so unreliable or of such poor quality that it tends to make processes less efficient and productive than they would be without it.

Combine the idea of workslop with the general failure of private firms to turn profits off investments in AI—MIT estimates 95% of generative AI pilots are not generating profits—and we’re faced with the conclusion that generative AI is making firms *less* efficient and “productive,” and probably costing them money.

https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/

MIT report: 95% of generative AI pilots at companies are failing

There’s a stark difference in success rates between companies that purchase AI tools from vendors and those that build them internally.

Fortune

In the US, something like a third of GDP growth this year can be attributed to generative AI firms and, even more specifically, to the growth of data centers.

Setting aside the incredible damage that will happen to people’s lives when the AI speculative bubble bursts, the thing that interests me the most in all of this is the ratio of investment in AI—the astronomical sums of money, labor, electricity and water, land, construction materials, and metals and plastics—to the return on that investment—which we saw above is actually making life harder, less efficient, and less productive for the firms trying to use it as a profit-generating tool.

https://www.exponentialview.co/p/is-ai-a-bubble

🫧 Is AI a bubble?

A practical framework to test whether AI is building value - or inflating it.

Exponential View

I say this because I’m a big fan of Joseph Tainter’s deceptively simple proposal that societies become vulnerable to collapse when their returns on investment in additional complexity begin to experience declining marginal returns.

Complexity is a problem-solving tool and it carries energy costs. Every new specialization, every new layer of hierarchical control that a society might implement to solve a new problem adds complexity and creates new demands for energy. That is, every techbro who styles himself a “prompt engineer” and every “AI will kill us all” grifter is being fed by people who grow, process, ship, sell, and prepare food for them while they’re busy tinkering with AI. If the returns on that additional complexity are declining in value, or have even turned negative, then they serve as net drains on those primary food producers, who can only work so hard to feed people engaged in less-than-productive work.

https://www.exponentialview.co/p/is-ai-a-bubble

🫧 Is AI a bubble?

A practical framework to test whether AI is building value - or inflating it.

Exponential View
So I wonder if the AI catastrophes the are correct, just for the absolute wrong reason. Generative AI is not going to come to life and transform us all into paperclips. It might, however, be such a shitty product at such astronomical cost that it undermines the ability of the status quo to sustain itself, devouring more and more resources only to create more and not less work for everyone.
@HeavenlyPossum
In my opinion this is the best blog for understanding "AI".

www.wheresyoured.at

A favorite post of mine. Their goals are vapid or harmful, and they have no idea how to achieve them.

www.wheresyoured.at/make-fun-of-them/
Ed Zitron's Where's Your Ed At

The Words of Ed Zitron, a PR person and writer.

Ed Zitron's Where's Your Ed At
@HeavenlyPossum
I think your conclusion is pretty much spot on too.
@walnut @HeavenlyPossum Seconded, Zitron's economic analysis is very good. I would also add https://pivot-to-ai.com/ for short snarky articles about AI nonsense.
Pivot to AI

It can't be that stupid, you must be prompting it wrong

Pivot to AI
@HeavenlyPossum Thank you for making the article accessible for anyone.
I wonder if factoring in the climate crisis might change the conclusion.
Devouring resources as resources get more and more scarce seems a risky business…
@HeavenlyPossum The proliferation of slop has the potential to destroy all public culture and communications mediums just by drowning every forum in so much slop it becomes impossible to find anything of actual worth.

@HeavenlyPossum

only $186/month? I guess I should read the paper, but some people seem to say that they spend all of their time vibe coding. Their entire salary would be negative value.

@richpuchalsky

Even if that number is accurate—and I assume it is a grotesque under-estimate—it still means that all of that investment in generative AI is almost entirely a loss for the economy, money and labor and physical resources flushed down the drain.

@HeavenlyPossum

One of these days I'm going to have to write that bit I want to write about workerism and how society would actually be better off if all of the telemarketers lost their jobs

@richpuchalsky @HeavenlyPossum Can we just abolish the entire advertising sector? It is a linchpin of all wasteful economics, because without advertising to convince people they need crap they don't even want, consumer spending would plummet. And we would all be much happier, no longer bombarded with messages telling us of our crippling inadequacy which can only be addressed by buying goods to fill the emptiness in our lives.

@richpuchalsky @HeavenlyPossum

This is the only study I know of that looks at the question of coding specifically.

https://time.com/7302351/ai-software-coding-study/

---
METR measured the speed of 16 developers working on complex software projects, both with and without AI assistance. After finishing their tasks, the developers estimated that access to AI had accelerated their work by 20% on average. In fact, the measurements showed that AI had slowed them down by about 20%.
---

I think there's an interesting analysis to be done looking at programming from the perspective of increasing complexity. Programs have to get more complex in order to keep reselling it (think Win 95 vs Window 11), there's a whole industry of tooling to help manage this complexity, both markets seem to be pretty saturated and yet complexity *must* increase. Enter AI...

In the Loop: AI Promised Faster Coding. This Study Disagrees

A new AI coding study yields a result that few expected.

Time

@HeavenlyPossum whoever could have possibly seen this coming, except for every large language model and machine learning domain expert without financial incentives in the wide scale adoption of this flash in the pan card trick.

Ok, maybe I'm being a bit harsh, but it's felt like watching someone push on a door clearly labeled "pull" for the past 5+ years while any critique is derided as not understanding the full potential of door technology.

@HeavenlyPossum note that that article is mostly marketing for a different AI product – although the term and concept are good. Pivot to AI has covered this today. https://pivot-to-ai.com/ https://youtu.be/xnI-Yr6plXI
Pivot to AI

It can't be that stupid, you must be prompting it wrong

Pivot to AI

@HeavenlyPossum

Thanks for giving me a word to put on this. The concept makes me extremely annoyed.

There was a translation gig I refused that was basically to proof-read chatgpt's work. The mere idea was already very insulting, but the reality of the job was even worse. In essence would have to 1) probably redo the entire thing, 2) with a worse pay and untenable timeline, 3) and let the chatbot get all the credit.

@HeavenlyPossum @Iris „If you have ever experienced this, you might recall the feeling of confusion after opening such a document, followed by frustration—Wait, what is this exactly?—before you begin to wonder if the sender simply used AI to generate large blocks of text instead of thinking it through. If this sounds familiar, you have been workslopped.“

thx for giving me the right vocabulary to be adequately annoyed now ;)

@HeavenlyPossum Now that we have demonstrated that artificial intelligence costs $186 per month in lost productivity, we will turn our study on the hidden costs of natural stupidity. 😇

@HeavenlyPossum

Not sure if that's the same article but I've been told it's not reliable.

https://www.youtube.com/watch?v=xnI-Yr6plXI

Workslop: bad ‘study’, but an excellent word

YouTube

@PandaCab @HeavenlyPossum That's the one. We've been had! It's some fake science written to sell another AI productivity tool using a deliberately terrible survey.

"Your workers are costing you money because they don't have the right AI tools and make work slop. But coincidentally we happen to sell a better office assistant AI that will boost your workers productivity the right way, and our research shows our product solves all these problems. So give us money. Lots of it please."