WELCOME TO MY THOUGHTS…
A PLACE WHERE I GET LOST, SO I’M DRAWING A MAP! 🧠 📜
WORKSLOP: The unintended productivity killer? 📁
FUN FINDS: Executive meme culture - AI powered 👔
JUST LOOK BUSY…
Your boss used to give you one report to write and a week to do it. Now you’ve got Copilot, so she gives you three. You produce 42 beautifully formatted slides before lunch. Everyone nods. The deck goes in the bottom drawer with the others, right next to last weeks “Strategic Roadmap 2.0.” Well done team 🙄
AI is a donut machine (bare with me)…It can crank out endless, glossy rings labeled ‘Deck’, ‘Report’, or ‘Email’ but without real ingredients: the facts, judgment, trade-offs, risks…humour? They’re sugar-free: perfectly finished, but bland. Colleagues downstream are forced to eat them, having to add the flavour themselves.
Workslop has a real definition now: “AI-generated work content that masquerades as good work but lacks the substance to advance the task.”
The term comes from a Stanford Social Media Lab + BetterUp Labs team writing in Harvard Business Review. In their survey of 1,150 full-time workers, 40% reported that receiving workslop from colleagues is often taking ~2 hours to fix!
This is the new busy work…
The broader internet and X trawlers had already been using term “slop” for ‘low-effort AI junk content’. This was made popular by a developer, Simon Willison in May 2024 who said:
“sharing unreviewed AI content is rude”
Not only that, but it could cause you reputational harm. People who forward unedited AI drafts look fast but unreliable, or in some cases, just plain irresponsible and stupid. This is magnified in domains where accuracy is paramount (legal/medical/risk).
So how do we avoid it?
THE SUGAR DONUT RULES 🍩
HOW NOT TO DROP SLOP:
DEFINE THE DECISION: Begin drafts/prompts with the decision you’re enabling, the options, and the trade-offs. Filter for purpose, not polish.
USE AI FOR THINKING, NOT WORD COUNT: Ask for counter-arguments, risks, or “what would change my mind?” then edit, hard.
FORCE HUMAN EDITS: Anything sent to others has to pass a human rewrite or fact check. Treat unedited model output like raw dough.
MEASURE OUTCOMES: Track time to decision, defects found in review, and business impact. Don’t celebrate the number of documents created.
RIGHT PLACE: Use AI where it shines (summarisation, boilerplate coding, polishing emails, conducting initial research, editing and creating images). Avoid it (or triple check responses) where accuracy or context is fragile unless you have expert oversight and are willing to do a full fact check.
RIGHT PLATFORM: Technology is moving fast. AI touches every industry. To use AI properly, you need to know which platform is best for your workflow or use case (and learn to use it effectively).
MY THOUGHTS:
AI slop is everywhere. Too many new users think hitting “generate” equals genius: they ship arbitrarily composed reports and immaculate emails that say little and decide nothing, making them a part of the noise.
I see workslop as an advantage for the people who actually learn the tools. Used properly, AI compresses thinking, sharpens language, and makes the next step obvious while also getting us to decisions more quickly.
While the other 80% flood inboxes with glossy emptiness, the few who insist on verified sources, and clear valuable insights will stand out. Signal is scarce, scarce is valuable.
I run an AI-automation agency, so I’m biased. We teach teams to swap conveyor-belt output for decision-ready work and high ROI workflows. Fewer donuts, more sugar.
FUN FIND
Not an ad. Just a brilliant use of AI, (depending on how you feel about memetic culture). Many see it as an art form - one for a future letter.
midjourney
IMAGE PROMPT: A whimsical, textured painting of a charming village with tall, wonky nice narrow houses with sloped roofs and chimneys
PROMPT: TELL ME A JOKE
Why did the AI go broke?
Because it spent all its tokens trying to generate a good joke.
🤔 a little bit waspy…
THAT’S ALL I’VE GOT. SORRY.
Need help navigating AI or integrating it into your business? You're in the right place. Email me for a chat: [email protected]
References: Harvard Business Review: “AI-Generated Workslop Is Destroying Productivity” The Week: “AI workslop is muddying the American workplace” | Harvard Business Review: “AI-Generated Workslop Is Destroying Productivity” The Week: “AI workslop is muddying the American workplace” | Simon Willison: “Slop is the new name for unwanted AI-generated content” | Microsoft Research & WorkLab: Copilot impact on info-worker tasks | Stanford/MIT field studies: Gen-AI raised customer-support productivity ~14–15% | Harvard/MIT (Imai et al.): When AI doesn’t improve decisions | Clay Shirky: “It’s not information overload. It’s filter failure.” | My Thoughts | Random things from the internet