What Is AI Slop and What Does It Cost a Community?

Overview

Everyone has an opinion about AI slop, and almost nobody agrees on what it is. For community managers that is not an abstract debate. Members are already using AI to write their questions, answers, and replies. The same tools that help an expert finally say what they mean can also help someone sound like an expert when they aren’t.

In this episode of CommunityAssistant.ai, Chris Heuer and Marius Ciortea try to pin the term down. They argue about where AI-assisted writing helps a community and where it quietly wears one down, and they end on what a community manager can actually do about it.

They start with their own slop. In the pilot, a statistic surfaced with AI said people trust AI answers over search; the source said people prefer them. Chris opens this episode by owning it. Relying on the machine being right set off a cascade of rework across the pilot’s clips and copy, and the two chose to talk about it rather than quietly fix it. Marius draws the lesson that frames the rest of the conversation: sometimes it is the human who tells the AI to make the slop.

What slop is

Marius goes first, with a reader’s definition. Slop is when a simple question gets a long answer that repeats itself and leans on phrasing you’ve seen a hundred times. There is something useful in there, but it takes too long to find.

Chris offers the working test he has been using, three questions: is it bloated, is it off, is it wrong? A fourth sits underneath all three, which is whether someone reached for the easy button instead of doing the thinking.

That raises the obvious question, and Marius asks it: can AI writing ever not be slop? Both hosts say yes. Neither thinks using AI is what makes something slop. Chris points to Matt Mullenweg’s view that code is poetry, a reminder that some people value the craft of the process and others care about the outcome. That split is why slop is in the eye of the beholder. It is also why each community manager has to define slop for their own community and their own use of AI, and then hold that line.

A worked case, and a better process

The case on the table is an op-ed attributed to investor Stanley Druckenmiller that Chris read as clearly written by AI. The defense that followed is a familiar one: executives’ op-eds have long been drafted by PR firms.

Chris’s answer to the easy button is what he calls a personal intelligence system. It is a defined body of your own writing, reading, and research, the material you believe is true, and you point your AI at it instead of letting it roam the whole internet. The PR-firm comparison holds here too. You don’t accept a draft and walk away. You review it, make sure it is accurate and sounds like you, and often discover something new along the way. That iterative loop is what turns a potential piece of slop into something useful to the people reading it.

Where AI writing helps a community, and where it breaks

Asked whether AI-assisted writing can actually make a community better, Marius says yes, with a caveat. He knows deep subject-matter experts who have always struggled to express themselves, because they aren’t strong writers or English isn’t their first language, and he counts himself among them. For those people, a tool that turns a rough draft into what they meant to say is a real gain for everyone reading.

It breaks when someone asks the AI to make them look good on a subject they don’t actually know. At that point, Marius wonders why that person is answering at all. It is the same dynamic behind the backlash on LinkedIn, where everyone has suddenly become a thought leader.

Chris adds intent to the older low-effort, low-quality definition, and brings a case from the show’s own feed. A comment on one of the pilot’s clips read like slop, from someone who seemed to be trying to sound more knowledgeable than they were. He could no longer tell whether they knew what they were talking about, so he trusted them less. If a contribution adds to the conversation but lowers trust in the person who made it, what did the community actually gain?

Marius is more optimistic about how communities handle it. They police themselves. Members are motivated by being recognized for what they know. When they are told a post reads like slop, they either leave or rethink how they show up. He doesn’t expect a future of AI answering AI, because the humans won’t put up with it.

Provenance, process, and intent

Then Chris argues against his own interest. He failed English in junior high and has never thought of himself as a natural writer. So what’s wrong with someone using Grammarly to do better? His answer is that it can make people seem better than they are. To know whether to trust what he is reading, he needs provenance, process, and intent.

Marius describes a loop members already run: ask the community, then check the human answer with AI. What a community offers that a bare AI answer doesn’t is the evidence. You get the answer and the thread beneath it that shows where it came from. Chris ties that to showing your work, citing research with Team Flow Institute. Marius brings back the op-ed: the fifteen-minute conversation with the PR firm was, in effect, always the prompt. The failure mode is the reverse, an 8,000-word wall of text in place of a fifteen-minute conversation, which pushes all the work onto the reader.

Where AI belongs in a community

Midway through, Chris asks where AI is actually best deployed in a community right now. He names three places: accessibility, enablement, and discoverability. Discoverability connects back to the pilot’s argument about generative engine optimization.

Marius takes accessibility somewhere most people miss: it runs both ways. Support, product, and other teams increasingly read the community through AI, so whatever is in the community travels. If the content is bad, it spreads everywhere. They also acknowledge the community managers who refuse to use AI at all.

Then what is the community for?

The episode’s best story comes from a workshop Marius attended. A community manager stood up and said they wanted a bot that answers every question. Marius’s reply: “Then what do you have the community for?”

The better design, he argues, is hybrid. An AI search surfaces the community’s answer when one exists, and when it doesn’t, it tells the member to go ask the community. Built that way, AI feeds the community rather than replacing it.

From there the conversation turns to platforms giving people a button to flag slop, and to the idea of a slop leaderboard, which Marius would rather nobody built. Chris flips it. Instead of ranking the worst offenders, show how much of a conversation is likely human.

Monday

For the practical move, Chris points to the Work Map’s entry on moderating content for policy compliance. Marius lays out the policy work in order:

Decide what’s acceptable, and cover two separate things: bots presenting as humans, and members using AI to write.
Whiteboard it with your team and your executives.
Run the draft past your community before it becomes terms.
When you remove something, use a template that says so plainly, says you value the member’s view, and invites them to rewrite and repost.

Underneath the process is a view of the role: community managers are enablers.

Human-written, AI-backed

Marius offers the line both hosts land on. Community content should be human-written and AI-backed, not AI-written with a little human backing.

Chris then asks the harder question. Do people come to a community for answers or for connection? And should they come for answers at all, if an AI can give a better one? Marius’s answer is that AI only knows what it knew at a point in time. People come to communities for answers that are uniquely human and highly specialized, built on experience too particular for a model to pick up. It isn’t how to fix a pipe. It is: here’s my situation and my view, do you agree? AI can’t replicate different points of view.

Chris’s advice is to revisit your community’s mission and values and take a stand. Marius’s is to get in front of a whiteboard with your team and executives, walk through the what-if scenarios, and decide what you want the garden to look like. Chris closes where the Work Map begins. Don’t turn on a technology because it can do something. Redesign the workflow on purpose: decide what can be offloaded, what is uniquely human and should stay that way, and what those choices are worth.

A correction

Correction: the pilot episode described a study as showing people trust AI answers over search. The study said prefer. The error was reproduced across every asset made from that episode. It’s discussed in the cold open, and it’s the reason this episode’s subject is a live one for us. Episode 0 →

Chapters

Time	Chapter
00:00	Cold open: human slop, and why we’re talking about it
00:25	Welcome
01:37	The pilot correction
02:18	Preferring AI answers, not trusting them
02:35	What’s at stake is trust
03:19	What is AI slop? Too long to find the useful part
04:07	Bloated, off, or wrong, and the easy button
05:08	Can AI writing not be slop?
05:16	Code is poetry; slop is in the eye of the beholder
06:39	The Druckenmiller op-ed: clearly written by AI
08:55	The PR-firm defense
09:17	Your personal intelligence system
12:22	Where AI writing helps a community: the expert who can’t write
14:04	Where it breaks: “make me look good”
14:42	Low effort, low quality, and intent: the LinkedIn problem
16:10	Communities self-police
17:51	“I failed English”: provenance, process, intent
19:17	The verification loop
20:44	The evidence under the answer
21:43	Show your work
22:05	The 15-minute conversation was the prompt
22:50	The 8,000-word wall of text
23:40	Where is AI actually best deployed in a community?
24:00	Higher Logic
24:47	Accessibility, enablement, discoverability
25:53	Accessibility runs both ways
28:38	Bad answers proliferate
28:52	The community managers who refuse AI
29:33	“I want a bot that answers all questions.” Then what do you have the community for?
31:41	LinkedIn’s slop button
33:08	The leaderboard, flipped
33:59	Monday: automated moderation and the Work Map
34:49	Write your community’s AI policy, with your community
37:36	Community managers are enablers
39:20	Human-written, AI-backed
40:31	What is a community for? Answers or connection?
41:16	Uniquely human: built on experience
42:28	Revisit your mission, your values, and take your stand
44:30	Set what the garden should look like
44:42	Not because the technology can: the Work Map thesis
45:56	Close
46:23	Give us your feedback on the Work Map report

Key takeaways

Slop is not “written with AI.” It is bloated, off, or wrong, and usually reached for as an easy button. Define it for your community and say so publicly.
AI writing helps the member who is an expert and a poor writer, or writing in a second language. It breaks the moment the prompt is “make me look good” on a question the poster can’t answer.
Communities self-police, because recognition is why people contribute. Negative feedback on slop either drives the poster out or makes them rewrite to be heard.
Trust now requires provenance, process, and intent. Readers can’t tell from the text alone, so show the evidence under the answer.
If an AI can answer, let it, and send the member into the community when it can’t. That enriches the community with new questions instead of repeats.
Accessibility runs both ways: AI reading the community feeds support, product, and sales. A wrong answer in the community proliferates through all of them.
Human-written and AI-backed keeps trust. AI-written with human backing loses it.
Monday: write the AI policy with your team and your executive, run the draft past the community before it becomes terms, and template the removal message so it’s kind and fast.
Communities exist for the questions AI can’t answer: “I’m in this situation, do you agree?” Experience is not replicable.
The same failure produces human slop: a person stopped deciding. Our own pilot is the worked example.

About the show

CommunityAssistant.ai is the podcast and open resource library exploring the human work of community management: what AI can take off your plate, what only you can do, and how to grow into the role that’s coming. Hosted by Chris Heuer and Marius Ciortea. Presented by Higher Logic.

Our sponsor. Community Assistant is made possible by Higher Logic, supporting the practitioners who build the communities that matter.

Our independence. Sponsors make this show possible. They don’t make this show. Marius works at Higher Logic; he appears here as a practitioner speaking for himself, not for his employer.