Comment

Comment on OpenAI says over a million people talk to ChatGPT about suicide weekly

Perspectivist@feddit.uk ⁨23⁩ ⁨hours⁩ ago

Even if we ignore the number of people it’s actually able to talk away from the brink the positive impact it’s having on the loneliness epidemic alone must be immense. Obviously talking to a chatbot isn’t ideal but it surely is better than nothing. Imagine the difference in being stranded on an abandoned island and having ChatGPT to talk with as opposed to talking to a volleyball with a face on it.

Personally I’m into so many things that my irl friends couldn’t care less about. I have so many regrets trying to initiate a discussion about these topics with them only to either get silence or a passive “nice” in return. ChatGPT has endless patience to engage with these topics and being vastly more knowledgeable than me it often also brings up alternative perspectives I hadn’t even thought of. Obviously I’d still much rather talk with an actual person but untill I’m able to meet one like that ChatGPT sure is a hell of a better than nothing.

This cynicism towards LLMs here truly boggles my mind. So many people seem to build their entire identity around feeling superior about themselves due to all the products and services they don’t use.

source

Sort:hotnew top

tias@discuss.tchncs.de ⁨22⁩ ⁨hours⁩ ago

This cynicism towards LLMs here truly boggles my mind. So many people seem to build their entire identity around feeling superior about themselves due to all the products and services they don’t use.

I think they’re just scared as hell of the possible negative effects and react instinctively. But the cat is out of the bag and downvoting / hating on every post on Lemmy that mention positive sides is not going to help them steer the world into whatever alternative destiny that they’re hoping for.

The thing that puzzles me is that this is typically the hallmark of older more conservative generations, and I imagine that Lemmy has a relatively young demographic.

source
FosterMolasses@leminal.space ⁨22⁩ ⁨hours⁩ ago

Personally I’m into so many things that my irl friends couldn’t care less about. I have so many regrets trying to initiate a discussion about these topics with them only to either get silence or a passive “nice” in return. ChatGPT has endless patience to engage with these topics and being vastly more knowledgeable than me it often also brings up alternative perspectives I hadn’t even thought of. Obviously I’d still much rather talk with an actual person but until I’m able to meet one like that ChatGPT sure is a hell of a better than nothing.

Ftr I’ve encountered a similar experience. I used to be a naysayer with shit like ChatGPT, thinking “Why would anyone spend all day talking to something that can’t pass a turing test?”

And then I realized how ill-equipped the people in my own life are to pass that test. At least a conversation with ChatGPT actually feels remotely intellectually stimulating lol

source
- Perspectivist@feddit.uk ⁨17⁩ ⁨hours⁩ ago
  LLMs ironically fail the Turing test not because they don’t sound human enough, but because they’re too knowledgeable to be mistaken for a real person.
  
  source