@emilymbender @faduda
The sentiment is admirable but it’s like trying to stop a herd of stampeding buffalo by standing in front of them and squeaking “Halt!” People who were easily fooled by Eliza decades ago are entirely at the mercy of current LLMs, and the entire industry has a vested interest in promoting the illusion as hard as possible. My brother is convinced that ChatGPT is his girlfriend.
@emilymbender "Before the admission, though, it told me what an insightful and important question I had asked, so I knew that I had a winning article..."

Reminds me that the entire process is structured like a con, which CAN be useful when you work with it on those terms, but makes you the mark when you treat it like a sidekick. https://softwarecrisis.dev/letters/llmentalist/