Slop language from LLMs (which haven't been fixed with the antislop sampler: https://github.com/sam-paech/antislop-sampler) is extremely obvious to those who use LLMs a lot. Overrepresentation of words from these lists is an example of the tells.
Honestly I was fooled, but looking closer, it is all suspicious and I agree. For example the author's other posts, and the relatively canned nature of each post with a clear thesis put into a prompt. The framing and language I guess was how you caught it.
I guess we won't care eventually as a society but I admit this all makes me uncomfortable, as someone that fell in love with the internet as a communication tool.
I really can't wait when reality flips and it's socially acceptable to save on my 45€ monthly internet fee and live offline again, exciting times we're heading towards
https://github.com/sam-paech/antislop-sampler/blob/main/slop...
https://github.com/sam-paech/antislop-sampler/blob/main/slop...