Of course the CEO’s who have invested millions into A.I. and probably even bought stock in the companies they are using will try to make it seem like A.I. is the best thing to ever happen to the world.
MisterMasterCyIinder on
Drowning? I made my own AI slopboat and now we’re all just circling the drain in a big AI-slop Charybdis together
MisterMasterCyIinder on
>40% of non-managers say AI saves them no time at all at work, while 92% of high-level executives say it makes them more productive.
So the people actually doing the work hate the sloppy and inaccurate but superficially polished output of AI and the managers love how quickly it can churn out their typical meaningless horseshit.
Go figure.
keenumsbigballs on
AI was supposed to make it so we had 4 day work weeks and 6 hour work days. Instead they cut half the staff and put double the work on everyone else and justify it by saying the AI tools will make you more productive and efficient. Can’t get the job done? Job market is mad max of people trying to get hired. We’ll get someone else.
banditta82 on
At many large firms productivity is just measured in meaningless paperwork being moved around. Quantity vs quality which is of course harder to judge. So ya based on their awful standards productivity is up, the same way RTO “increased productivity” by increasing the amount of “at desk” time. AI fits perfectly into management by spreadsheet.
Disgruntled-Cacti on
Leadership and management are completely disconnected from actual labor. They have no idea what most of their employees actually do let alone have the ability to quantify and measure the output of their labor. Of course they have no clue about whether AI is making an impact or not.
Appeltaart232 on
If as a manager you’re tracking productivity with the number of lines of code your company can spit out, congratulations – you’re stupid.
timeaisis on
Who could have seen this coming! Oh yeah, everyone who does work for a living.
Due_Satisfaction2167 on
The actual source of the divide is this: middle managers and executives are finding LLMs to be extremely useful and productive because it’s a perfect fit to replace their roles.
But they have their own personal experience, and since it’s good at replacing their job, they presume it must be equally good at replacing individual contributors too.
It isn’t.
These tools *should* be causing a bloodbath among middle management as orgs become able to radically flatten their org charts due to LLMs being able to effectively do their own ETL -> Summarization work that lets higher level decision-makers keep awareness over a lot more than they used to.
RedHawwk on
I think it makes my work look “better”. But now everyone just passes around AI written text and I gotta wonder how much is actually being read or processed. Sometimes I feel like I’m just making AI slide decks for the sake of it.
SomeSamples on
Bosses are lying sacks of shit. What they mean is AI saves them bunches of money on payroll.
_cob_ on
This just in: bosses out of touch.
CircumspectCapybara on
Depends on the field. Glorified chat bots and funny image generators are not all that exists in the world of AI products and not all that professionals use AI for. LLM-based agents are huge force-multipliers in SWE, SRE, and MLE roles these days, and orgs with mature, competent engineering have absolutely figured out how to leverage them to boost productivity, which is putting it mildly.
Most large engineering orgs have fully embraced them and it’s becoming clear that the industry isn’t going back to the old way of working. I’m talking at large mature engineering orgs, most engineers don’t write any code by hand anymore, SRE workflows are being performed by agents with MCP integrations to the org’s source control, CI/CD system, o11y stack, and AI is being used to debug production incidents and conduct RCAs. It’s a wild new world.
Source: Staff SWE @ Google. Used to be an AI skeptic, but I’ve seen how things have changed rapidly in the span of like six months. The current agents are about as good as a really fast junior engineer and massively free up bandwidth for me to allocate brain bandwidth to the harder and more important parts of engineering besides hand-writing the code. And it’s extremely good at reasoning about distributed systems’ high level behaviors, failure modes, and bugs from the code. It’s an invaluable tool for systems design, for implementation, and for incident response.
jb4647 on
It’s always been workslop.
bonzoboy2000 on
I enjoy using AI. Especially on complex codes. But I find myself asking key questions different ways to see if the answer is correct. It’s time consuming for sure. I often look to see if the chat is taking my own words and just reorienting them.
For complex codes, it is very useful. But I can tell when someone sends me slop. And they didn’t read it.
wowlock_taylan on
They don’t even know what ‘Productivity’ is.
The executives just think random slop that have AI put out is ‘good work’ instead of the lowest level of slop that needs to be fixed by humans.
Skie on
I’ve seen people with seriously questionable tech and critical thinking skills adopt Copilot and begin replying in Teams channels to damn near every tech “how to” question with AI generated responses. The responses are the typical stuff you get from Copilot or a google search, where it does an okay job for basic questions but gets things seriously wrong as soon as anything is obscure or new.
I’ve raised it with people as a bit concerning, because they’re clearly wasting a lot of other peoples time on slop answers, but nobody wants to do anything about it because “they’re being helpful”.
niknacks on
The amount of times “leaders” at my company dump some truly aggregious slop on me, without also noting the information within is complete and utter garbage (because they either don’t understand our industry or business well enough to recognize it OR worse didn’t even bother to read it) is getting insane.
The only reason I believe ai can replace upper and middle management is because it already has at my company and yet we still keep paying for these people to show up and give an agent terrible prompts
gccumber on
I find it strange that companies aren’t talking about how AI can allow them to do more with current headcount. Instead they want to reduce… that doesn’t really make sense.
SomeGalNamedAshley on
And what do the actual workers say?
Pjpjpjpjpj on
The amount of “make work” AI allows middle managers to do is absolutely mind boggling.
My manager sent me a meeting request. It was scheduled for 30 minutes. The invite included a 4 page AI written agenda. Tons and tons of bullet points and topics to cover.
It would have taken me many hours to fully prepare. I thought screw it, and rolled the die.
I showed up (video call) and basically said “do you want to go through the agenda point by point or do you just want to discuss project X.” I don’t think he even remembered his AI agenda.
Then after the meeting, I get an email asking me to review a clearly-AI generated summary of our meeting that must have been done from a recorded transcript of our discussion. He wanted me to review 8 pages of notes and reply with detailed edits. That I’m sure he would have AI review and reply.
I just ignored the email and he never followed up.
There is a coming AI war – it’s going to be between managers creating AI-generated useless make-work demands and employees responding with AI-generated checkbox answers.
-The_Blazer- on
> While initial drafts were a breeze to create, Ken and his co-workers had to spend more time rewriting, correcting errors and resolving disagreements between each other’s chatbots than if they had never used AI at all.
This is my experience too. AI nowadays is impressively good at spitting out something and automatically testing and reviewing it so it becomes functional enough to meet stated requirements (mostly).
The problem comes when anyone else has to do anything with it. Half of the job of a knowledge worker is that whole *knowledge* part, not the literal typing of text or code. AI deprives you of that, so good luck maintaining or extending or generally doing anything with an AI output. And good luck if you need to explain it to anyone. And god help you if you need to provide guarantees about what it does or is to someone important.
That’s my problem fundamentally: if you told me to just generate slop with AI and then exempted me from responsibility, I’d be all for it. I’d make slop for one hour and then twiddle my thumbs the remaining seven. But companies want the speed of slop with the quality and knowledge of human work, and the two things are in direct opposition.
As it turns out, no sane manager *actually* wants to take the responsibility of truly AI-centric work.
Lower_Ad_1317 on
The only boost they’re seeing is the perceived future lay offs they’re hoping to throw at everyone.
Ai has very limited use scenarios right now. Everything else is smoke and mirrors combined with possible irresponsible ‘let’s see what happens’ scenarios.
Without actual self awareness it is a tool that must be monitored. And if it has to be monitored then it isn’t going to produce the results the AI bros are pushing for it
It is hand waving and shiny sparkly promises while the money flows. Wait until the bubble bursts and you will see how many are thrown under the bus.
srahsrah101 on
Every CEO I’ve had since chatGPT came out has used it to “discover” some new training method or model that magically works. Takes me precious time to unravel and explain “no, you can’t test on your training data” for the thousandth time.
APlannedBadIdea on
Not up professional standards. Bosses often love it. Speaks more to the level of review and specialized knowledge of middle management which is understandable given the role supervision provides. Doesn’t change that my heart sinks when the grammar of my colleagues’ work reads like unedited ChatGPT output.
Derpykins666 on
AI would be a potentially great boon if it weren’t for greed. The companies want to have the cake and eat it to. They all fired a lot of their work force and doubled the work of the next employee while making them train the AIs to eventually replace them as well and paying them next to nothing.
Not sure what the endgame of this looks like when nobody has jobs because everything is automated, who will have any money for any sort of products?
Rsubs33 on
My leadership is all about AI. I think it is useful to help check language and it’s decent at putting slides together though they all need to be edited but none of our leadership take the time to do that
i_am_13th_panic on
We tried using ai at my job. It takes longer to check results then just doing it ourselves most of the time.
icepick3383 on
all that’s happening is a bunch of new documentation is being made from conversations and turned into ai-written memos – which are then circulated and subsequently ignored because nobody can stay on top of the AI slop being passed around – and do their day to day stuff. I’ve got c-suite and svp’s doing ‘vibe coding’ all over the place and passing it over to eng saying ‘see, it’s easy’ – and that’s considered ‘being productive’. I’m so over it.
I do use it to help with documentation and other tasks, but fuck me, it’s just drastically increased the amount of stuff people consider ‘work’. whole cloth presentations being churned out and disseminated without review or insight. it’s fuckin ponderous, man.
runyonave on
Company I work for has invested millions in to AI. We had an all team meeting where they mandated that we all use AI to write code, and not write our own code. It’s ridiculous.
VNM0601 on
I hear AI can replace C-suites pretty efficiently.
zippopwnage on
I don’t know man, I don’t care about my work. I just care to get some money. The company I work for doesn’t care about me or whatever. If anything happens to my health tomorrow, they will instantly replace me. I’m gonna use AI as much as possible to deliver the most mediocre output or whatever it delivers.
It helps me finish my tasks faster and it actually work for what it is. I don’t want to consume my life working.
I get the hate of AI replacing people, but those companies that replaces you for some shitty AI, was already actively looking on how to get less people working for more.
But I hate this thing, we act here like ohh no we care about the companies we work for.
ClvrNickname on
Whatever productivity boost AI gives me in writing code is cancelled out by the extra time I spend reviewing the absolute slop code pushed up by some of my coworkers who are clearly just accepting the first thing that Claude spits out at them without a second glance.
JohrDinh on
I can close/ignore/downvote/etc AI right now, but I can’t imagine having to be at a job where you’re forced work with it and read/look at it all day. I watch some podcasts that are now using AI to write their pre topic breakdowns and my brain feels grossed out when they start reading what is clearly slop now. Compared to the person just riffing like they used to it’s far more structured, boring, and just unenjoyable to listen to…can’t imagine being exposed to it by force. (tho least you’re getting paid for it I guess)
anon-a-SqueekSqueek on
Our product managers just make all our stories with ai now – so it’s impossible to tell what is a requirement from the business and what ai hallucinated. Not that they ever knew what they wanted to start with, but it’s worse now.
There are many issues with ai code, but at least we have a rigorous process of quality checks, and have to ensure it works, and are on the hook to support it at all hours if it doesn’t work.
The middle managers and C-suite execs use ai, they just walk away from their work output, and their team has to absorb the slop and deal with it, all while being lectured about not using ai enough and not going fast enough.
flamaryu on
Ai pushed after my job has instead made more work. Built automation scripts that saved hours of man power, that allowed us to focus on things that couldn’t be automated. Higher ups turned those jobs over to Ai now I have a new twice a week work session to fix Ai errors, 🤔🙃
Pomme2 on
AI is the excuse to cut expenses. The work is often shifted to remaining employees.
diablol3 on
Hasn’t worker productivity already woefully outpaced pay in the last 5 decades?
AntiSnoringDevice on
I went back to writing on a board. It is way faster than having to set-up an effing “teams” each time I have to share.
Want to know where the project stands? Come take a look at the board…we’re forced to be in the office anyway after the phasing out of remote.
Realistic_Run_649 on
Workslop is the right word for it. The productivity gain is real but only if someone who understands the task also reviews the output. When AI gets handed to people without the context to judge quality, you end up with well-formatted work that nobody trusts, which actually slows the team down because now two people have to debug what one person should have done right. The tool multiplies whatever judgment you bring to it, including the lack of it.
felixisthecat on
AI is meth for corporations, once they’re hooked and have laid off employees to fund the habit, they’re cooked!
Time-Industry-1364 on
pure copium. They want some kind of return on their investment and they’re desperate for anything that resembles an improvement.
solusolu on
At my company they force us to use AI and track our usage of it. Whenever there’s a report up to C-Suite they always say that productivity is up because of AI. It’s a combination of the top down setting implementation goals and the bottom up pretending to be successful.
lazyhustlermusic on
Output is up 500%.
Don’t actually look at what the output is, but the fact that there is -something- as output is up!
havenoir on
This is absolutely true. I’ve been using Claude and Gemini to do background research for me and both of them frequently generate references that they infer are correct because they match patterns, but are in fact completely wrong. They look correct, they seem official, but they don’t exist. It makes it very, very difficult to use these systems for even the simplest things that require accountability. you have to review absolutely everything.
Tripilot2025 on
The workslop framing is accurate. Executives measure productivity in outputs per hour, workers experience it as cognitive load per hour. AI can inflate the former while destroying the latter. A tool that generates 10 mediocre drafts has not improved productivity — it has shifted the bottleneck from creation to curation. Curation at scale is exhausting in a way that is hard to quantify on a dashboard.
NoMark3945 on
Most AI productivity claims fall apart when you ask who is measuring and what they are measuring. Output volume goes up. Output quality is harder to track. What nobody is measuring is the cognitive overhead of reviewing AI-generated work at scale — the constant low-grade vigilance required to catch the plausible-sounding errors that a human would never make. That cost does not show up in the dashboard but it shows up in burnout.
47 Comments
Of course the CEO’s who have invested millions into A.I. and probably even bought stock in the companies they are using will try to make it seem like A.I. is the best thing to ever happen to the world.
Drowning? I made my own AI slopboat and now we’re all just circling the drain in a big AI-slop Charybdis together
>40% of non-managers say AI saves them no time at all at work, while 92% of high-level executives say it makes them more productive.
So the people actually doing the work hate the sloppy and inaccurate but superficially polished output of AI and the managers love how quickly it can churn out their typical meaningless horseshit.
Go figure.
AI was supposed to make it so we had 4 day work weeks and 6 hour work days. Instead they cut half the staff and put double the work on everyone else and justify it by saying the AI tools will make you more productive and efficient. Can’t get the job done? Job market is mad max of people trying to get hired. We’ll get someone else.
At many large firms productivity is just measured in meaningless paperwork being moved around. Quantity vs quality which is of course harder to judge. So ya based on their awful standards productivity is up, the same way RTO “increased productivity” by increasing the amount of “at desk” time. AI fits perfectly into management by spreadsheet.
Leadership and management are completely disconnected from actual labor. They have no idea what most of their employees actually do let alone have the ability to quantify and measure the output of their labor. Of course they have no clue about whether AI is making an impact or not.
If as a manager you’re tracking productivity with the number of lines of code your company can spit out, congratulations – you’re stupid.
Who could have seen this coming! Oh yeah, everyone who does work for a living.
The actual source of the divide is this: middle managers and executives are finding LLMs to be extremely useful and productive because it’s a perfect fit to replace their roles.
But they have their own personal experience, and since it’s good at replacing their job, they presume it must be equally good at replacing individual contributors too.
It isn’t.
These tools *should* be causing a bloodbath among middle management as orgs become able to radically flatten their org charts due to LLMs being able to effectively do their own ETL -> Summarization work that lets higher level decision-makers keep awareness over a lot more than they used to.
I think it makes my work look “better”. But now everyone just passes around AI written text and I gotta wonder how much is actually being read or processed. Sometimes I feel like I’m just making AI slide decks for the sake of it.
Bosses are lying sacks of shit. What they mean is AI saves them bunches of money on payroll.
This just in: bosses out of touch.
Depends on the field. Glorified chat bots and funny image generators are not all that exists in the world of AI products and not all that professionals use AI for. LLM-based agents are huge force-multipliers in SWE, SRE, and MLE roles these days, and orgs with mature, competent engineering have absolutely figured out how to leverage them to boost productivity, which is putting it mildly.
Most large engineering orgs have fully embraced them and it’s becoming clear that the industry isn’t going back to the old way of working. I’m talking at large mature engineering orgs, most engineers don’t write any code by hand anymore, SRE workflows are being performed by agents with MCP integrations to the org’s source control, CI/CD system, o11y stack, and AI is being used to debug production incidents and conduct RCAs. It’s a wild new world.
Source: Staff SWE @ Google. Used to be an AI skeptic, but I’ve seen how things have changed rapidly in the span of like six months. The current agents are about as good as a really fast junior engineer and massively free up bandwidth for me to allocate brain bandwidth to the harder and more important parts of engineering besides hand-writing the code. And it’s extremely good at reasoning about distributed systems’ high level behaviors, failure modes, and bugs from the code. It’s an invaluable tool for systems design, for implementation, and for incident response.
It’s always been workslop.
I enjoy using AI. Especially on complex codes. But I find myself asking key questions different ways to see if the answer is correct. It’s time consuming for sure. I often look to see if the chat is taking my own words and just reorienting them.
For complex codes, it is very useful. But I can tell when someone sends me slop. And they didn’t read it.
They don’t even know what ‘Productivity’ is.
The executives just think random slop that have AI put out is ‘good work’ instead of the lowest level of slop that needs to be fixed by humans.
I’ve seen people with seriously questionable tech and critical thinking skills adopt Copilot and begin replying in Teams channels to damn near every tech “how to” question with AI generated responses. The responses are the typical stuff you get from Copilot or a google search, where it does an okay job for basic questions but gets things seriously wrong as soon as anything is obscure or new.
I’ve raised it with people as a bit concerning, because they’re clearly wasting a lot of other peoples time on slop answers, but nobody wants to do anything about it because “they’re being helpful”.
The amount of times “leaders” at my company dump some truly aggregious slop on me, without also noting the information within is complete and utter garbage (because they either don’t understand our industry or business well enough to recognize it OR worse didn’t even bother to read it) is getting insane.
The only reason I believe ai can replace upper and middle management is because it already has at my company and yet we still keep paying for these people to show up and give an agent terrible prompts
I find it strange that companies aren’t talking about how AI can allow them to do more with current headcount. Instead they want to reduce… that doesn’t really make sense.
And what do the actual workers say?
The amount of “make work” AI allows middle managers to do is absolutely mind boggling.
My manager sent me a meeting request. It was scheduled for 30 minutes. The invite included a 4 page AI written agenda. Tons and tons of bullet points and topics to cover.
It would have taken me many hours to fully prepare. I thought screw it, and rolled the die.
I showed up (video call) and basically said “do you want to go through the agenda point by point or do you just want to discuss project X.” I don’t think he even remembered his AI agenda.
Then after the meeting, I get an email asking me to review a clearly-AI generated summary of our meeting that must have been done from a recorded transcript of our discussion. He wanted me to review 8 pages of notes and reply with detailed edits. That I’m sure he would have AI review and reply.
I just ignored the email and he never followed up.
There is a coming AI war – it’s going to be between managers creating AI-generated useless make-work demands and employees responding with AI-generated checkbox answers.
> While initial drafts were a breeze to create, Ken and his co-workers had to spend more time rewriting, correcting errors and resolving disagreements between each other’s chatbots than if they had never used AI at all.
This is my experience too. AI nowadays is impressively good at spitting out something and automatically testing and reviewing it so it becomes functional enough to meet stated requirements (mostly).
The problem comes when anyone else has to do anything with it. Half of the job of a knowledge worker is that whole *knowledge* part, not the literal typing of text or code. AI deprives you of that, so good luck maintaining or extending or generally doing anything with an AI output. And good luck if you need to explain it to anyone. And god help you if you need to provide guarantees about what it does or is to someone important.
That’s my problem fundamentally: if you told me to just generate slop with AI and then exempted me from responsibility, I’d be all for it. I’d make slop for one hour and then twiddle my thumbs the remaining seven. But companies want the speed of slop with the quality and knowledge of human work, and the two things are in direct opposition.
As it turns out, no sane manager *actually* wants to take the responsibility of truly AI-centric work.
The only boost they’re seeing is the perceived future lay offs they’re hoping to throw at everyone.
Ai has very limited use scenarios right now. Everything else is smoke and mirrors combined with possible irresponsible ‘let’s see what happens’ scenarios.
Without actual self awareness it is a tool that must be monitored. And if it has to be monitored then it isn’t going to produce the results the AI bros are pushing for it
It is hand waving and shiny sparkly promises while the money flows. Wait until the bubble bursts and you will see how many are thrown under the bus.
Every CEO I’ve had since chatGPT came out has used it to “discover” some new training method or model that magically works. Takes me precious time to unravel and explain “no, you can’t test on your training data” for the thousandth time.
Not up professional standards. Bosses often love it. Speaks more to the level of review and specialized knowledge of middle management which is understandable given the role supervision provides. Doesn’t change that my heart sinks when the grammar of my colleagues’ work reads like unedited ChatGPT output.
AI would be a potentially great boon if it weren’t for greed. The companies want to have the cake and eat it to. They all fired a lot of their work force and doubled the work of the next employee while making them train the AIs to eventually replace them as well and paying them next to nothing.
Not sure what the endgame of this looks like when nobody has jobs because everything is automated, who will have any money for any sort of products?
My leadership is all about AI. I think it is useful to help check language and it’s decent at putting slides together though they all need to be edited but none of our leadership take the time to do that
We tried using ai at my job. It takes longer to check results then just doing it ourselves most of the time.
all that’s happening is a bunch of new documentation is being made from conversations and turned into ai-written memos – which are then circulated and subsequently ignored because nobody can stay on top of the AI slop being passed around – and do their day to day stuff. I’ve got c-suite and svp’s doing ‘vibe coding’ all over the place and passing it over to eng saying ‘see, it’s easy’ – and that’s considered ‘being productive’. I’m so over it.
I do use it to help with documentation and other tasks, but fuck me, it’s just drastically increased the amount of stuff people consider ‘work’. whole cloth presentations being churned out and disseminated without review or insight. it’s fuckin ponderous, man.
Company I work for has invested millions in to AI. We had an all team meeting where they mandated that we all use AI to write code, and not write our own code. It’s ridiculous.
I hear AI can replace C-suites pretty efficiently.
I don’t know man, I don’t care about my work. I just care to get some money. The company I work for doesn’t care about me or whatever. If anything happens to my health tomorrow, they will instantly replace me. I’m gonna use AI as much as possible to deliver the most mediocre output or whatever it delivers.
It helps me finish my tasks faster and it actually work for what it is. I don’t want to consume my life working.
I get the hate of AI replacing people, but those companies that replaces you for some shitty AI, was already actively looking on how to get less people working for more.
But I hate this thing, we act here like ohh no we care about the companies we work for.
Whatever productivity boost AI gives me in writing code is cancelled out by the extra time I spend reviewing the absolute slop code pushed up by some of my coworkers who are clearly just accepting the first thing that Claude spits out at them without a second glance.
I can close/ignore/downvote/etc AI right now, but I can’t imagine having to be at a job where you’re forced work with it and read/look at it all day. I watch some podcasts that are now using AI to write their pre topic breakdowns and my brain feels grossed out when they start reading what is clearly slop now. Compared to the person just riffing like they used to it’s far more structured, boring, and just unenjoyable to listen to…can’t imagine being exposed to it by force. (tho least you’re getting paid for it I guess)
Our product managers just make all our stories with ai now – so it’s impossible to tell what is a requirement from the business and what ai hallucinated. Not that they ever knew what they wanted to start with, but it’s worse now.
There are many issues with ai code, but at least we have a rigorous process of quality checks, and have to ensure it works, and are on the hook to support it at all hours if it doesn’t work.
The middle managers and C-suite execs use ai, they just walk away from their work output, and their team has to absorb the slop and deal with it, all while being lectured about not using ai enough and not going fast enough.
Ai pushed after my job has instead made more work. Built automation scripts that saved hours of man power, that allowed us to focus on things that couldn’t be automated. Higher ups turned those jobs over to Ai now I have a new twice a week work session to fix Ai errors, 🤔🙃
AI is the excuse to cut expenses. The work is often shifted to remaining employees.
Hasn’t worker productivity already woefully outpaced pay in the last 5 decades?
I went back to writing on a board. It is way faster than having to set-up an effing “teams” each time I have to share.
Want to know where the project stands? Come take a look at the board…we’re forced to be in the office anyway after the phasing out of remote.
Workslop is the right word for it. The productivity gain is real but only if someone who understands the task also reviews the output. When AI gets handed to people without the context to judge quality, you end up with well-formatted work that nobody trusts, which actually slows the team down because now two people have to debug what one person should have done right. The tool multiplies whatever judgment you bring to it, including the lack of it.
AI is meth for corporations, once they’re hooked and have laid off employees to fund the habit, they’re cooked!
pure copium. They want some kind of return on their investment and they’re desperate for anything that resembles an improvement.
At my company they force us to use AI and track our usage of it. Whenever there’s a report up to C-Suite they always say that productivity is up because of AI. It’s a combination of the top down setting implementation goals and the bottom up pretending to be successful.
Output is up 500%.
Don’t actually look at what the output is, but the fact that there is -something- as output is up!
This is absolutely true. I’ve been using Claude and Gemini to do background research for me and both of them frequently generate references that they infer are correct because they match patterns, but are in fact completely wrong. They look correct, they seem official, but they don’t exist. It makes it very, very difficult to use these systems for even the simplest things that require accountability. you have to review absolutely everything.
The workslop framing is accurate. Executives measure productivity in outputs per hour, workers experience it as cognitive load per hour. AI can inflate the former while destroying the latter. A tool that generates 10 mediocre drafts has not improved productivity — it has shifted the bottleneck from creation to curation. Curation at scale is exhausting in a way that is hard to quantify on a dashboard.
Most AI productivity claims fall apart when you ask who is measuring and what they are measuring. Output volume goes up. Output quality is harder to track. What nobody is measuring is the cognitive overhead of reviewing AI-generated work at scale — the constant low-grade vigilance required to catch the plausible-sounding errors that a human would never make. That cost does not show up in the dashboard but it shows up in burnout.