LinkedIn declares war on Slop with "Seems like AI Slop" button.

I can't believe they did it.

I feel like I’m in a fever dream: LinkedIn just introduced a “Seems like AI Slop” button to allow users to report Sloprific posts on the platform.

caption...

For years, LinkedIn pushed users to post AI-generated content, placing a “Write with AI” button in the compose field that would give them slop superpowers, so long as they upgraded to LinkedIn Premium. Then, two months ago, LinkedIn changed its tune, announcing that it was going to fight AI Slop with internal detection tools.

That didn’t stop the scrutiny of LinkedIn’s slop problem. Three weeks ago, Pangram published a viral study showing that 41% of long-form posts on LinkedIn were fully AI-generated.

LinkedIn’s slop problem was so bad that it seemed to spook Substack. One week ago, Substack introduced a partnership with Pangram to scan posts for AI-generated text to increase transparency. “One thing we do know is that we don’t want to wait until your Substack app turns into LinkedIn before we start to learn and make progress,” CEO Chris Best wrote.

In one week, two major social media platforms declared war on AI Slop. It feels like a turning point.

Are these solutions the right approach? I have no idea! But that doesn't matter. They’re an important step to establishing some cultural norms:

1. Be transparent about how you use AI, something I’ve been championing for the past two years. AI-generated content isn’t inherently bad; there are AI-generated financial and social media analyses I find pretty interesting! But the experience of reading a post that you think was written by a human and then realizing it’s AI-generated is soul-crushing. It destroys trust and delivers an “Inauthenticity Tax” on the poster. As I wrote earlier this month:

A new large-scale study of 27,000 people found that when people know a piece of text was created with AI, they rate it much more negatively and perceive it as inauthentic. Across all the studies, perceived inauthenticity most strongly correlated with the negative ratings, inferring that our dislike of AI writing is something of a social survival feature. We infer that someone is trustworthy not just by what they communicate, but how they communicate. And using AI to write feels inherently deceptive.

Slop can be both human-generated and AI-generated. But when it’s AI-generated, it gives us the ick in a special way.

2. Social pushback against AI-generated nonsense. We do lots of crazy things before there’s social pushback against it. Just watch one episode of Mad Men. We used to smoke on airplanes! Misdemeanor sexual harassment used to be a workplace norm! Workslop, message-slop, and content slop will continue to be a thing until people get feedback and think, “Oh right, I shouldn’t do that.” LinkedIn’s “Seems like AI Slop” button is designed to deliver that pushback. As LinkedIn Chief Product Officer Hari Srinivasan wrote this morning:

For anyone who shares content, we will test a way to privately flag, in your analytics dashboard, when members feel your post may have come off as inauthentic or heavy use of AI. This approach is based on two learnings. First, AI and slop are not the same thing; many people refine thoughts with AI, and we believe they want to know when they sound inauthentic. Second, we want members to get feedback from real humans on what sounds authentic - not just have an AI detector review it and get it wrong.

3. Honest dialogue about where we want AI to fit into our lives. Over the past week, there’s been vocal pushback against Substack’s AI-scanning feature. Some of it is a little wacky, like claiming that the AI-scanning feature is racist because it’s AI (wat) or this simple transparency measure is somehow discriminatory against people with learning disabilities. Others were legitimate concerns around the accuracy of Pangram, although research indicates it’s extremely accurate. Most importantly, it’s prompted an important discussion amongst Substack writers about how we use AI in our work.

---

TEST: This is a secret message for you using Substack’s new audience targeting features. Listen, if you’re enjoying this newsletter, you should consider upgrading to a paid subscription to my newsletter using this special 30% off offer. For just $4/month, you’ll get:

• A copy of my best-selling new book, Super Skill: Why Storytelling Is the Superpower of the AI Age (4.6 rating on Amazon, 4.7 rating on Goodreads)

• Access to my new content strategy course

• Invites to my subscriber-only monthly content strategy masterminds and workshops

• Very dope socks

Support my work and subscribe here for 30% off!

---

One theory I have is that the research points to a dividing line in how humans feel about AI output. When an AI’s output is objective—something we don’t expect to come from the voice and perspective of another human—we don’t really care if an AI made it. We don’t care if an AI coded an app, or helped with back-end research for a piece, or checked for typos. Conversely, we do care when the output is subjective—something we expect to come from the unique voice and perspective of another human being. Which explains why people consistently rate content negatively when they know it was generated by AI.

I could be wrong, though, and I’m excited to see what this discussion reveals about where we want to integrate AI into our work and art. As I’ve written before, I use AI for research and brainstorming with strict boundaries, and as a business assistant to give me more time to write. Every word I publish is 100% mine. That might not be the right approach for everyone, but I think we’ll learn a lot more as we bring the slop into the sunlight. When even LinkedIn is on board, it’s a bright day.

Want to get top content strategies and insights in your inbox each week? Then subscribe to this newsletter and check out my best-selling new book, Super Skill: Why Storytelling Is the Superpower of the AI Age. You’ll also get a personalized, signed copy of Super Skill when you upgrade your subscription to paid, access to my new course, and an invite to our next content strategy mastermind session.