With ChatGPT and other similar large-language processors becoming more
common, so too are tools designed to /detect/ verbiage created by
these algorithms. Tools like ZeroGPT (
https://www.zerogpt.com/), or
GPTZero (
https://gptzero.me/).
So, of course I had to try them.
On what? You ask. On myself, of course. Am I an AI? Because - reading
back some of the drivel I've posted over the years - I can only assume
I must be at least partly driven by some imperfectly coded AI that
thinks extraneous amounts of verbiage and sesquipedalianism is a
convincing writing methodology.
But, no. "Your Text is Human written", says ZeroGPT. 0% AI generated
text, claims GPTZero.
Why am I suddenly feeling disappointed... and perhaps a little
insulted? What, is my writing is not good (bad?) enough to be
something created by an AI? Is it not shallow enough? Is it that I I
doesn't claim enough facts without backing them up? Is that what
you're saying, AI-detectors? 'Cause that sounds like a challenge to
me!
I reassure myself that these AI detection tools are still very new,
and imperfect themselves. They've been known to make egregious
mistakes (even insisting that the entirity of the US Constitution is
completely AI written*, which I'm sure is almost entirely incorrect!).
So they could be wrong about me too.
But until their coding errors are fixed, I shall gleefully pick up the
gauntlet they tossed at my feet. Expect more long-winded, fact-free
nonsense from me in the future; I'll claim the AI crown yet!
*
https://arstechnica.com/information-technology/2023/07/why-ai-detectors-think-the-us-constitution-was-written-by-ai/
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)