Regarding this: https://social.vivaldi.net/users/lproven/statuses/117087977477686112
Putting aside "The Point Prover Has Entered The Chat" aspect of this response, if we are serious about whether or not people can fluently differentiate LLM text from human-written text, we're not going to understand how or whether we can do it if our basis is a series of KPIs like # of languages we speak, education level we've attained, this or that credential, and so on.(1) None of those things confer reading comprehension by themselves, and the mindset that leads one to conflate them with demonstrated ability is literally the problem. "I know 6 languages, and I'm not able to tell the difference between LLM-generated text and human-written text" is, if anything, a self own. I read it as a limitation of self awareness, not a demonstration of fact.
#AI #GenAI #GenerativeAI #LLM #reading #text #interpretation
(1) If you haven't yet met a person with a PhD in English who lacks basic reading comprehension, I recommend it as a one-time life experience.
Putting aside "The Point Prover Has Entered The Chat" aspect of this response, if we are serious about whether or not people can fluently differentiate LLM text from human-written text, we're not going to understand how or whether we can do it if our basis is a series of KPIs like # of languages we speak, education level we've attained, this or that credential, and so on.(1) None of those things confer reading comprehension by themselves, and the mindset that leads one to conflate them with demonstrated ability is literally the problem. "I know 6 languages, and I'm not able to tell the difference between LLM-generated text and human-written text" is, if anything, a self own. I read it as a limitation of self awareness, not a demonstration of fact.
#AI #GenAI #GenerativeAI #LLM #reading #text #interpretation
(1) If you haven't yet met a person with a PhD in English who lacks basic reading comprehension, I recommend it as a one-time life experience.