Reminds me of the unfortunate book "Vibe Coding" by Steve Yegge, whom I otherwise enjoy. While it contained okay, if very light on actionable details, overview of the broad ideas behind LLM-assisted coding (how much of it was vibe coding, though?), much of it was co-written through the use of an LLM book editing pipeline, proudly advertised throughout the book. A treatise of otherwise one-tenth of the final length has been blown up into the size of a volume, not unlike a piece of meat is pumped with water to make it appear fattier.
Every time I see a title like this, I ask myself if I'm not being open enough, if my biases are interfering with any potential progress I could be making when it comes to utilising AI. Then I find out that the content is just more slop and it further solidifies my position on all of this. What a waste of energy. It really saddens me.
For me it was the EM dashes in the copy.
that's funny. I read an article about how the use EM prevalance in content as an indicator that it was AI generated. I should tell my agents to stop using them :)