I opened up LinkedIn today and saw three posts stacked up with the following headlines:
“Most users have never...”
“Most teams think...”
“Something that people aren’t seeing.”
Behold, the new em-dash for AI-assisted writing. These phrases are doing something click-baity. The meta is that you feel like you’re behind and the writer knows something you missed.
Other patterns that put me on high alert:
“Nobody is talking about this.”
“Most misread this.”
The word “most” is doing the work of data without being data. And then there’s the big-number flex:
“I filtered 21,000 papers.”
“I analyzed 1,000 applications.”
“I spent 200 hours on this.”
Percentages are the same fun game:
“95% of AI projects fail.”
“84% of companies need an overhaul.”
Some of these reference real studies but there are no links or attribution. So I’ve started treating any percentage without a link as an opinion.
The insider knowledge thing is another trick that may be true but we have no way to verify:
“What companies tell me privately.”
“Based on what I’m hearing from leaders.”
Then there’s the second-person accusation.
“Your team is doing this wrong.”
“Your agents keep failing.”
“Everyone misses these three things about choosing an embedding model.”
Shame! But some of these shame-baiting techniques reference concepts that didn’t even exist a month ago.
And the urgency.
“Last chance to catch up.”
“The career ladder collapsed.”
Career advice built on fear is still just advice, and I notice most of it comes with no evidence that anyone actually suffered the consequences being described.
Does this invalidate the content? Many times no. I see this technique used with content creators that I’ll continue to follow given the core of their content is still good. But if you see these patterns, read, and more importantly, think carefully.