Chris Garrett has been publishing online long enough to have watched several waves of automation arrive with the same promise. His objection to the current one is not that the output is bad. It is narrower and more interesting than that.
The Connection AI Cannot Make
“But a human that read scientific advertising 30 years ago and reads like Seth Godin book this year is going to make a connection and is going to say something that even the best AI tool isn’t going to be able to find a connection between because it’s a human brain making random firing connections, possibly over caffeinated.”
Chris Garrett, ChrisG.com
The claim is precise. Not that machines cannot write. That a specific kind of connection — between two things nobody has previously connected, held in one idiosyncratic head across thirty years — is not in the training data by definition, because it has not been said yet.
That is the thing worth protecting in a content operation, and it is usually the first thing optimised away.
Where Are The People In All Of This
“Where are the people in all of this?”
Chris Garrett
Asked as a genuine question about the direction of the industry rather than a rhetorical flourish. As more of the pipeline becomes machine-mediated — generation, distribution, and increasingly consumption — it becomes possible to build a content operation with no identifiable human anywhere in it.
Garrett’s position is that this is a commercial problem before it is a philosophical one. An audience forms around people. It does not form around output.
Hallucinated Podcasts And Virtual Voices
The conversation moves into synthetic audio and the odd territory of shows that were never recorded by anyone. It is treated less as a novelty than as a test case for the previous point: if the format’s entire value came from a person having said something, what is left when nobody did?
A related discussion on being displaced by machine reproductions of your own work runs through Holly Homer’s episode on AI slop and being replaced by your own words.
What Garrett Still Bets On
Idiosyncrasy, accumulated reading, and a named person willing to be wrong in public. None of those scale, and that is the entire argument — the things that do not scale are the only things left that a competitor cannot copy by buying more compute.
More from the content side of the show in content strategy and marketing deep insights, and the machine side in AI and the future of search.
