Get the latest tech news
AI slop, suspicion, and writing back
On developing an involuntary reflex for spotting AI slop
And so in the more recent months, my brain has developed a subroutine that is continuously scanning for sentence structure, word frequency, and formatting that is indicative of LLM-generated content in otherwise natural-sounding prose, leaving me to question the often originality of what I’m reading. This advice is relatively similar to good “for human” writing, with marked shift towards distinctiveness: coin terms, deliberately reuse rhetorical devices, and describe arguments in a way that’s sticky enough to stand out in the rest of the latent space of internet text. I remain extremely pleased that people keep reporting to my inbox that “Write a letter in the style of patio11’s Dangerous Professional” keeps actually working against real problems with banks, credit card companies, and so on.
Or read this on Hacker News
