Rise of the 'slop zombies'
Corporate America and the wider world of work are suffering from a pandemic of AI-pilled slop zombies infected with brain rot and unearned confidence.
SILICON VALLEY, CALIFORNIA (JULY 10, 2026) — If you have a job in 2026, you know the type.
I’m talking about people who use AI chatbots for everything. They bludgeon you with massively long, detailed, mediocre reports, messages, emails, and slide decks they didn’t even read (but expect you to read). They illustrate everything with AI slop and marvel at their own creativity.
Churning out long, complicated missives with AI forces others in the company to use AI to keep up. The slop zombie virus is contagious.
Because of the low time and cognitive cost of producing AI-generated communications, and because AI chatbots by default tend to be extremely wordy and detailed, they incentivize you to use AI to boil down the slop zombies’ bloated content. People end up functioning as mere couriers of messages between AI chatbots.
As the habit sets in to use AI for everything, slop zombies get dumber and lazier while their confidence rises. They believe the ideas and work of the AI chatbots is their own. As chatbots churn out often false or hallucinatory junk with confidence, so do the slop zombies.
Developers using Claude Code or some other tool are filled with certainty about the quality of their work, bickering with other developer slop zombies who are equally confident and equally unfamiliar with the code they’re arguing about.
“Imposter mode,” whereby someone who is actually qualified feels unqualified, has been replaced by “slop zombie imposter mode,” whereby someone actually unqualified feels overqualified. Slop zombies feel that way because they mistake chatbot flattery and sycophancy for earned praise, and also because of the delusion that AI output is their own work.
A disproportionate number of bosses have become AI-pilled slop zombies, using AI to run companies or departments and even mandating that everyone over-rely on AI.
One slop zombie trait is to devalue feedback from people to the point of being dismissive of any human-generated ideas, and overvalue chatbot output, thus spreading the contagion.
AI is supposed to boost productivity. But slop zombies are killing it. When you work with a slop zombie, you waste all your time pouring over mountains of chatbot output in the form of long-winded reports or vibe-coded software packed with errors and hallucinations and off-topic ideas that are cognitively draining to find and expose. It’s a slop tax on productivity.
Slop zombies offload not only cognitive tasks to AI, but social skills and emotional intelligence. People are consoling co-workers and even firing people using words generated by AI chatbots. They use a machine to sound more human.
Slop zombies are the new office menace. But what is to be done?
READ: The Attachment Economy, Computerworld, Gastronomad newsletter, Gastronomad book
LISTEN: Superintelligent, TWiT, Gastronomad podcast
JOIN: The Gastronomad Experience
FOLLOW: Gastronomad on Surf Social, Machine Society on Surf Social, Bluesky, Reddit, Notes, Mastodon, Threads, X, Instagram, Facebook, Linkedin!
LOOK AT: Mike Elgan Photography
The trouble with slop zombies
Like regular zombies, slop zombies feed on human brains (in the form of LLM training data), yet suffer from personal brain rot.
They don’t realize that they’re not fully participating in their own lives anymore.
The worst part is that they believe themselves to be the opposite of what they are:
Obsessive over-dependence on AI chatbots has made them far dumber and far less productive, but they believe themselves to be super smart and highly productive.
They believe themselves to be full of great ideas, but they’ve outsource idea-making to the machines.
They believe themselves to be super valuable to the company, but in fact have already replaced themselves with AI and are therefore in fact of little value.
To accuse a boss or co-worker of being a slop zombie is hard to do. They’ll deny it because it strikes at the core of their professional identity, and because in any event, they don’t believe it.
Unlike in academia, where professors can test students with oral presentations and zero-AI test settings, it’s hard to prove that a co-worker didn’t do the work.
Also: The difference between a smart user of AI who uses it to become smarter on the one hand, and a slop zombie on the other, is actually hard to spot.
Is someone using AI to offer options to consider, or are they just copying and pasting ideas from chatbots? Superficially, it’s a fine line. In reality, the distinction is everything. Here are the differences between a smart user of AI and a slop zombie:
The smart user can explain why they chose this output over the alternatives the AI offered. The slop zombie can’t, because they never considered alternatives.
The smart user modifies, rejects, or overrides AI output based on domain knowledge the AI doesn’t have. The slop zombie presents it verbatim (or near-verbatim with superficial edits).
The smart user’s work improves over time because the AI interaction helps sharpen their thinking. The slop zombie’s work degrades over time because they’re outsourcing the thinking itself, which weakens cognition.
The smart user gets less dependent on AI as they learn more. The slop zombie gets more dependent.
What to do about slop zombies
Slop zombies have a defensive advantage, which is that accusing them of not doing their own jobs is socially awkward at least and an actionable offense at most. Still, even though you have to play along, there are ways to fight back:
