A mutual over on the birdsite was suggesting that they find Q&A with ChatGPT more productive than reading manuals, and suggested that by pointing out the likelihood of incorrect answers, I was treating this as an "academic pie-in-the-sky" topic. 🤔
As an experiment I went and asked it 4 questions and got 2 dangerously subtly wrong answers, 1 blatantly incorrect answer, and 1 useless non-answer.
Newton’s First Law of Cats: Cats in motion tend to say in motion, cats at rest tend to stay at rest!
After reading about prompt injection to remove ChatGPT’s conditioning on Twitter, my ChatGPT now only speaks in astromech beep-boop with subtitles
ChatGPT gives tax advice for dragons
If you were a #dragon how would you handle this situation?
Asking ChatGPT if computational complexity theory is in conflict with anarchist philosophy
@rechelon you're welcome, I think 😂
ChatGPT writing "haiku" about Gödel's Incompleteness Theorem
bro this isn't even a Haiku bro
co-Founder // Chief Science Officer at @geopipe (🖖🏻), PhD from Brown CS Dept (w/ @maurice), SMCVT alum (Math/Physics/CS), admin at Cemetech, AFOL & open-theist.
Decentralizing systems (human & digital). Opinions are my own.
📍 Vermont