That fits witj my experiences. And i want to add an otjer layer. In ai times its somtimes even nice to see some typos. You Casn be pretty sure it was not written by ai.
That fits witj my experiences. And i want to add an otjer layer. In ai times its somtimes even nice to see some typos. You Casn be pretty sure it was not written by ai.
Wow, this guy must be important.
You can prompt an LLM to add typos, though
interestingly, you can’t do the same thing with queries like “no em dashes”. it’ll agree, then proceed to use them regardless.
could be related to how so-called negative prompts fail to work when asking, say, ChatGPT to generate an image without a crocodile
My theory is that sprinkling emdashes into the output is some intentional measure to "watermark" LLM output.
"sent from my iphone"
Positive (tryhard) signaling: having a well designed email footer with all your contact info
Neutral signaling: no footer at all
-1 signaling: sent from my iPhone
-2 signaling: sent from my Samsung AI Family Hub 4-Door Flex Fridge
I think "Sent from my iPhone" is now less of a status symbol than it is an excuse for short replies / bad grammar.
There needs to be a new sig like "Written by Claude" or something. I'd rather somebody just openly admits it.
*Please forgive any typos