-
The dilution is the right frame but the thing that goes underdiscussed is that the AI content problem isn’t uniform across platforms. On LinkedIn and Twitter, the visible AI patterns (em dashes, “it’s not about X, it’s about Y” constructions, rhetorical question hooks) are starting to actively hurt engagement because audiences have learned to recognise them. On Instagram carousels and TikTok scripts the same patterns get less penalty because the format hides them better.
What’s working for my content specifically is leaning harder into what AI can’t do well: specific numbers from personal experience, named examples that aren’t in training data, contrarian takes where the argument is grounded in something measurable. The more context-specific the content, the less diluted it feels.
On distribution email lists and communities are the right answer but they take 6-12 months to build. The shorter-term move is being in comment sections of the people your audience already follows. You skip the feed entirely, land in threads people are already reading, and the visibility compounds faster than building your own audience from zero.
