AI Workslop

Welcome to the Macro AI Podcast with Gary and Scott. In this episode, we dive into one of the newest and most important concepts hitting boardrooms and executive teams: AI Workslop. 

AI Workslop describes polished, AI-generated work that looks good on the surface but lacks the substance, accuracy, or context to drive real decisions. It’s the long memo with no action, the glossy slide deck without insight, the email that shifts the burden onto the reader. And it’s not just annoying — it’s expensive. 

Recent research from Harvard Business Review, BetterUp Labs, and Stanford found that: 

• 40% of desk workers encountered AI Workslop in the last month. 
• Each incident wasted nearly 2 hours. 
• The hidden cost adds up to $186 per employee per month — over $9M annually for a 10,000-person company. 
• Colleagues perceive Workslop senders as less creative, less capable, and less reliable. 

In this episode, Gary and Scott explore: 

• What AI Workslop is — and how it differs from AI hallucinations. 
• Why it happens (old habits, new tools, and cultural pressure). 
• How leaders can spot Workslop before it derails productivity. 
• Why prompting skill matters — and why it’s not the full cure. 
• The Anti-Workslop Playbook: leadership guardrails, workflow templates, training strategies, and metrics. 
• Real-world examples of slop vs. substance in sales, operations, and contact centers. 
• The single KPI executives should watch: time-to-decision. 

AI isn’t the problem. Workslop is. And leaders who build the right norms, culture, and skills will see ROI instead of sludge. 

🔗 Resources mentioned in this episode: 

• Harvard Business Review article introducing “AI Workslop” (Sept 2025): https://hbr.org/2025/09/ai-workslop 
• BetterUp Labs research and resources: https://www.betterup.com/resources/research/ai-workslop  
• Stanford Social Media Lab collaboration: https://sml.stanford.edu  

About your AI Guides

Gary Sloper

https://www.linkedin.com/in/gsloper/

Scott Bryan

https://www.linkedin.com/in/scottjbryan/

Macro AI Website:

https://www.macroaipodcast.com/

Macro AI LinkedIn Page:

https://www.linkedin.com/company/macro-ai-podcast/

Gary's Free AI Readiness Assessment:

https://macronetservices.com/events/the-comprehensive-guide-to-ai-readiness

Scott's Content & Blog

https://www.macronomics.ai/blog

Transcript

00:00
Welcome to the Macro AI Podcast,  where your expert guides Gary Sloper and Scott Bryan navigate the ever-evolving world of artificial intelligence.  Step into the future with us  as we uncover how AI is revolutionizing the global business landscape  from nimble startups to Fortune 500 giants.  Whether you're a seasoned executive,  an ambitious entrepreneur,

00:27
or simply eager to harness AI's potential,  we've got you covered.  Expect actionable insights,  conversations with industry trailblazers  and service providers,  and proven strategies to keep you ahead in a world being shaped rapidly by innovation.  Gary and Scott are here to decode the complexities of AI  and to bring forward ideas that can transform cutting-edge technology  into real-world business success.

00:57
So join us,  let's explore, learn  and lead together.

01:05
Welcome to the Macro AI podcast. I'm Gary Sloper here with my cohost, Scott Bryan. On this show, we explore how artificial intelligence is reshaping business, leadership, and the future of work. Our goal is to cut through the hype and get you the insights you can always use. Yes. And today we're diving into a  brand new term that's been making waves really in the past week or two, uh AI work slump. You may not have heard of it before, but trust me, you've seen it.

01:41
Exactly. You open a report, a memo, or an email that looks sharp on the surface, but after a few minutes,  you realize it's empty. That's work slop. And it's not just annoying, it's costing organizations millions in lost productivity and trust. Yep. So in this episode, we're going to break it down.  What is work slop? How we got here? How to spot it?  And how leaders can stop it before it takes over their organization. Yeah. So let's get into it.

02:10
Okay. Yeah, let's just start with a clean definition. ah WorkSlop is  AI generated output that looks polished,  but doesn't actually advance the task. It's,  you know, for example, that long email with lots of words, but no decision,  or the deck of really fancy charts, but really in reality, zero insight. So it uh shifts the burden to the person receiving it or shifts it downstream.

02:39
Yeah, and that's a critical point. This isn't the same as AI hallucinations. Hallucinations are factually wrong. know, work slop might be dramatically perfect and even factually okay, but it lacks context, uh direction, or substance. It really forces someone else to redo the work. That's what makes it dangerous. If you think about it  in this way, it really steals time, attention, and trust of AI.

03:07
Yeah. So how big is this problem really? A survey of more than a thousand desk workers found that about 40 % had run into AI work slop in just the past month. And each time it costs them about two hours to sort it all out, make it better, make it actionable. Yeah. And then when you translate that into dollars, it's about $186 per employee per month. So if you take a 10,000 person company, that's $9 million a year.

03:37
just in hidden costs. And that's just the productivity side. The perception cost is really huge as well. Colleagues reported feeling annoyed, confused, even offended that they received the work slop to begin with. Yeah, thanks for that. Appreciate it. Yeah, exactly. And once someone gets labeled as the person who sends empty fluff, their credibility really goes down fast. And that helps explain why so many executives tell us

04:06
You know, we're, investing in generative AI, but we, but we don't see the return on investment. So the AI is powerful, but if it's generating work slop, the business  never sees the value or just doesn't see it quickly enough. Yeah. And, and, know, really if you break it down, you know, the question that we're starting to see is, you know, why work slop happens.  So, you know, maybe we discussed some of the root causes.  Um, you know, first.

04:36
People lean on AI as a crutch. I think we've all seen that, family, friends. ah In the workplace,  folks are, they'll fire off ah one prompt  and copy and paste the result and hit send. So they're not editing, no fact checking.  I've seen this firsthand, so it's imperative to check for accuracy.  Second, companies push AI use without setting clear guardrails.

05:04
So leaders might say use AI everywhere, but they never define what a good AI deliverable looks like or where, and I'm using my air quotes, everywhere actually is. And third, context gets lost. So outputs don't explain the inputs ah or even the assumptions ah or the intended audience. So the next person down the chain is left guessing on the context that was delivered. Right.

05:30
Yeah, and then outside of your company walls, you see this everywhere. uh Blogs,  LinkedIn posts,  even news articles that they all kind of sound the same, the same cookie cutter phrasing, paragraphs that feel polished, but they're starting to feel generic and you recognize the uh word patterns. And then the slop culture bleeds into the enterprise if you're not careful.

05:55
So on those posts,  everybody now is pretty familiar. see the little rockets or the other icons. That's kind of a dead giveaway.  Maybe the author changed a few things. Maybe they didn't, but now people are starting to be able to identify just pure content generated by AI that really hasn't been tweaked. Yeah, I agree. see it as well.  All  right. So I think maybe there is a uh bigger question. How do we even get here?

06:23
know, bad work didn't start with AI. People have been making, you know, fancy decks with little or no substance for decades. So, so what changed? Gary, you want to take a shot at that? Yeah. I was thinking about this before as we were getting ready for the show. a way, AI, in my opinion, is just poured fuel on old fires. So corporate culture has long rewarded activity over outcomes. We've celebrated long memos, big, thick decks.

06:53
volume of output. Now, artificial intelligence makes it 10 times easier to crank out that instantaneously. Companies like Amazon, who remove PowerPoint decks  from their meetings, thrive  in this culture. uh So left unchanged, the same bad incentives go supercharged. Right. And add to that the pressure from the top. So  over the last two years, every boardroom has been saying, we need to show we're doing AI. So managers

07:23
They push their teams,  use it, show results, but managers are not defining what good AI output is. So that mandate without guardrails  and without inspection and coaching, I think that's a recipe for, for workslop. Yeah. And I feel like it's Groundhog Day. We saw this when cloud first came out. Everyone said, I have to be in the cloud. And they didn't really put good thought into this. So we really need to learn from that mistake.

07:50
There's also the psychology side. AI gives you  something that looks finished right away. So, you know, you get that dopamine hit and it feels done, but editing the fact checking, the tailoring, that's really the hard work with AI. And that's  the exact, you know, uh part that gets skipped through this whole process.  At least  what I've seen more, especially most recently. Agreed. Yeah. And then finally there's the noise problem.

08:17
So once  AI generated content starts  flooding, know, Slack channels, inboxes, SharePoint,  you don't just get bad deliverables, you get uh a signal to noise crisis that's,  I think people are really getting a feel for. And leaders can't tell what's real insight versus AI sludge. And then that overwhelms and erodes trust pretty quickly in an organization. Yeah, I fully agree. And so the answer to how did we get here?

08:48
It's really old habits, new tools and cultural pressure. It's not one person being lazy. It's systematic. you know, leaders have to understand that, or they'll misdiagnose the problem at the end of the day. So just keep that in mind within your organization. Yep. Yeah. So let's talk a little bit about how, how you spot works, what works a lot quickly. I think for, for me,

09:15
Uh, the biggest red flag is when,  uh,  sources are missing, you know, no specific actions,  uh, no internal facts, internal specific facts,  uh, really no way to verify anything. I think that's kind of a dead giveaway. Obviously not a lot of sources listed in a lot of internal writings, but you know, the lack of internal facts or specific actions is the giveaway. Yeah, that's, that's a good point. I think another one, it looks.

09:45
overpolished but empty, lots of words,  no decision, or you see cookie cutter phrasing, the kind of generic corporate speak you've read a hundred times.  And maybe the most dangerous one, context gaps. ah You read it and you realize the purpose,  the decision, or the next step just really isn't there.  And that's pretty easy to spot and pretty easy to follow, you

10:10
into that workstream. Again, getting back to that dopamine effect, I  cranked something out, but you really didn't  proofread what you delivered. Yeah. And like we talked about before the show, there are some tests. A question would be, before you send it out, does this help me make good decisions specific to our business?  If not,  think twice. Is that something you really want to forward off? Yeah, I like that. We should come up with some sort of uh

10:39
acronym equation for that. With formula. Yeah, little formula. So, at the end of the day, how do you go about this and trying everything you can do to avoiding the trap? How do you keep work slop out of your business? And I've always said, with anything, it really starts with leadership. It's not just work slop. It's really how you define the organization. So set clear rules and expectations for output.

11:08
Spell out where AI can draft and where it can't. Require people to label AI assisted work and list their sources. You were just talking about sources  a few minutes ago.  And absolutely prohibit one click shipping. Nothing leaves a team ah without a human edit before it gets sent off. Right.  And  then you need to make sure that you're enforcing that your teams  bake quality into the workflow. So perhaps

11:38
uh set standards or templates that force people to state the purpose, the inputs, the method, uh the decision and the next steps. that's not, I  don't think that's too difficult to do as a leader of a team. You're just really enforcing that method.  And I think managers should continuously coach their people to focus on substance and not just bulk content and grammar that you can pull right out of a...

12:07
a generative AI tool. Yeah. And as a leader, raise your hand, say, I'm also going to commit to this as well. Uh, that's how you lead by, by example. yeah, good point up skill, up skill, just on the skills, up skill, your skills. Right. And, and, and to your point, don't forget skills, train your teams to be pilots, not passengers, uh, teach prompting. We recently posted an episode on AI prompts. We did a couple of weeks back. Check it out if you haven't,  um, error spotting.

12:37
and rewriting,  celebrate concise source linked work. That cultural reinforcement is what makes the difference in the long run for your team.  Yeah, and since you mentioned prompting, this is obviously a big one. We've got a question over, well, the previous weekend from Christine in St. Augustine. She says, does prompting reduce,  does better prompting reduce work slope?  And  I think the answer is yes.  And thank you, Christine.

13:06
So if you type in a vague prompt, you get vague, generic output. If you write in a precise, contextual prompt, the output is far more likely to be useful for you and for your team and everybody downstream. So  prompting well actually forces the  author, the person to think at the business level.  What is the purpose? Who's the audience?  What's the decision that we're trying to make?  And the data that we're considering. um

13:35
clarity upfront  always  cuts down on the downstream slop. Yeah, exactly.  But I would say, let's be clear though, prompting isn't the whole cure, right?  Even a great prompt can still churn out something that looks unfinished,  but isn't actionable. Or looks finished, but isn't actionable is what I should have said.  That's why you still need human editing. You still need the human review and the cultural guardrails we've been talking about. So prompting reduces the risk.

14:04
but it doesn't eliminate it. So that's a really important component that you need to really share across your teams. Yeah. And then you can add in uh metrics on top of everything else we talked about.  can measure it. can  track time to decision,  track how often work has to be rewritten.  And when you start seeing those metrics improve,  that's when AI starts to pay for itself. um

14:33
And Gary, I'm going to pull up our prompting checklist.  Okay. So it just an idea, you know, before hitting enter on your AI request, ask yourself, what's the purpose? So that's number one, what's the purpose? What decision, action or audience am I writing for? ah Two, what's the context? So have I given AI the background, the constraints or examples that it needs to generate a  more thoughtful and  concise output?

15:03
uh What uh three, what is the output format? Do I want a one pager, a bulleted summary, a table  or a draft email?  And then lastly, number four, what's missing?  Did I specify my sources?  Did I specify the data? Did I specify the tone that should be included in the output? So if you can't answer all four, you're probably prompting the  AI into some level of  works love. Yeah, those

15:33
Those are spot on. And I think number one, what's the purpose? I for a lot of initial users, it's  to get my boss off my back, right? They are just churning work slop out because of other tasks that are upstream. So you as a leader, think about what you're asking of your teams as well, because they may not fully understand the purpose other than I just have to get this off my plate. Yeah. So it's like a constant, constant continuous management improvement process with your team.

16:02
Totally agree. And I know you deal a lot with this just at the, you know, introductory  baseline level. When you're, when you're doing some of your workshops, you see it often and it gets back into the numbers. We just talked about at the productivity levels, the amount of productivity waste that's going in because the purpose really isn't defined. Exactly. So real world examples, let's, let's make this real. before the show,  Scott and I talked through a few quick examples. So.

16:32
So we'll kind of go through this, I'll start. Okay, so here's a real world example. So imagine a sales email, the slop version is 400 words of buzzwords with no tie to the customer. The anti-slop version, kind of moving all that, 120 words, specific reference to the customer's last three support tickets, and concrete next steps, like a 20 minute review call. In the end,

16:59
the customer probably won't be impressed by buzzwords. They want clear, summarized responses to their three support tickets. Do feel bad that they had three open cases,  but if you think about it, if you're on the receiving end of that support requirement, you don't want to read through 400 words. You wanna know- Yeah, you wanna  get to a resolution. Yeah, you wanna get to the resolution. Where are you at with my three issues? Period, end of discussion.  Now, one of those cases may require an in-depth write-up because-

17:29
be very technical, it could be laborious, to make that fix, that's okay. But if you start talking about industry standards and how you were named in the Gartner Magic Quadrant because you provided work slop, that's just going to anger the customer even more. I'll shift gears into operations. um So operations, obviously you're trying to uh get to  solutions.  So in an operations  update,

17:57
The slop version is  lots of busy graphs, say five busy graphs. The anti-slop version is  one chart, one variance number, the root cause, ah and the owner with an ETA. Something that is clear, concise, actionable. Yeah, that's a great one. And  I'd say the last one I kind of thought of even in the contact center, we had  a lot of conversations around contact center and you might see.

18:26
an AI generated summary that just says, quote unquote, agents handled calls well. That's completely useless. The real version is top three drivers of calls, two fixed, one outstanding, cost impact, and the change in operating procedure. That's real insight that the call center operations  lead or team  can use. Right. Yep. That's a good one. So, uh

18:55
Just moving over to some of the general questions that business leaders might ask. ah So if, if AI creates  slop, should we stop using it? So obviously the answer is no.  AI  is an incredible first draft engine and a summarizer when used responsibly and like we talked about when, when prompted  with a Yep. Totally agree. Another one. uh What's the one KPI I should watch?

19:25
It's an often asked question. I'd say time to decision. If AI speeds that up, you're winning. If it slows you down because you're cleaning up slop, you've got a problem.  Yep. Yep. Just spreading that problem around your team and various other downstream teams. m And then um some specific industries are even more exposed. you know, tech, healthcare, professional services, they're just

19:53
obviously, you know, swimming and information and client deliverables. So when they see work slop, you know, they, they see it. And that means that you need even more stringent guard rails in certain industries  and quickly. Yes.  Good point. So if we're starting to wind down the episode, um, at the end of the day,  AI work slop isn't an AI problem. It's a management problem.

20:21
If you set norms,  measure outcomes,  train your people to be pilots, not passengers,  as we mentioned earlier, you'll see the return on investment. Yeah, I like that.  Pilots, not passengers. uh If  you don't, you're going to pay the  work slop tax. That's wasted time, frustrated teams,  and  a credibility hit for you and your organization. uh

20:50
Don't let some of this polished nonsense drag your business down. Focus on improvements, continuous improvements like we talked about. Yes. Before we go, we will leave a link to the Harvard Business Review article, The Better Up Labs Research. Hit us up on LinkedIn anytime. And it was a great episode,

21:13
Yeah, that's yeah, I appreciate that. That's a good one, Gary. That's it for today on the Macro AI podcast. So thanks, everyone, for listening. Please keep sending in the questions. Please like and subscribe and share with your network. Talk to you next time.