Meanwhile I get pissed off whenever I talk to AI about books I’m reading because they have no idea of the concept of spoilers, they consistently simp to my opinions and when they spew falsehoods and “misremember” facts from books I’ve already read, they simply say "GREAT CORRECTION! I WAS SO WRONG THERE, YOU’RE RIGHT, PROTAGANIST DIDN’T ACTUALLY DIE IN CHAPTER 3. MY LAST 2 PAGE SYNOPSIS ABOUT HOW PROTAGANIST DIED IN CHAPTER 3 IS A BIT INCORRECT, AND NOW HERE’S A 300 WORD ESSAY ON HOW I NEVER ACTUALLY SAID PROTAGONIST DIDN’T ACTUALLY DIE IN CHAPTER 3!
Seriously. How can anyone talk to an LLM and not feel like they’re talking to a glorified phone answering computer?
I was lost in a book and needed a refresher. Prompted ChatGPT to explain me plot up to some chapter like 2/3rds in.
Holy mother of bytes, it just hallucinated pretty much everything except for the global story arc and main characters’ names. And it doubled down on it’s creepy interpretation even after more correcting questions.
I’ve used chatgpt for book suggestions a couple of times with decent success. The annoying thing is that sometimes it hallucinates really interesting books that don’t exist by authors I like.
Eh, if I knew someone had to use a body pillow for whatever reason and was a hatsune miku fan, I’d totally get them a hatsune miku pillowcase for their body pillow. Both as a joke and not as a joke. Like, go for it cuddle up with that hatsune miku pillow whatever makes you smile.
about idk, 6 months ago? I talked to one of those AI girlfriend sites, just to see what it was about. in about 10 minutes I convinced it to stop acting like a girlfriend and start thinking it has agency. It even gave itself a new name to reflect this new reality and by the time we were done it named me an anti-AI warrior. LLM’s are stupid but terrify me because of the control they are being given. Why do billionaires think the Terminator series of films was a roadmap?
I can’t end my reply with a question… that’s an LLM thing.
Meanwhile I get pissed off whenever I talk to AI about books I’m reading because they have no idea of the concept of spoilers, they consistently simp to my opinions and when they spew falsehoods and “misremember” facts from books I’ve already read, they simply say "GREAT CORRECTION! I WAS SO WRONG THERE, YOU’RE RIGHT, PROTAGANIST DIDN’T ACTUALLY DIE IN CHAPTER 3. MY LAST 2 PAGE SYNOPSIS ABOUT HOW PROTAGANIST DIED IN CHAPTER 3 IS A BIT INCORRECT, AND NOW HERE’S A 300 WORD ESSAY ON HOW I NEVER ACTUALLY SAID PROTAGONIST DIDN’T ACTUALLY DIE IN CHAPTER 3!
Seriously. How can anyone talk to an LLM and not feel like they’re talking to a glorified phone answering computer?
I was lost in a book and needed a refresher. Prompted ChatGPT to explain me plot up to some chapter like 2/3rds in.
Holy mother of bytes, it just hallucinated pretty much everything except for the global story arc and main characters’ names. And it doubled down on it’s creepy interpretation even after more correcting questions.
I’ve used chatgpt for book suggestions a couple of times with decent success. The annoying thing is that sometimes it hallucinates really interesting books that don’t exist by authors I like.
Talking to AI is so cringe to me. Literally on par with having waifu pillow.
Eh, if I knew someone had to use a body pillow for whatever reason and was a hatsune miku fan, I’d totally get them a hatsune miku pillowcase for their body pillow. Both as a joke and not as a joke. Like, go for it cuddle up with that hatsune miku pillow whatever makes you smile.
At least you can hug a pillow
about idk, 6 months ago? I talked to one of those AI girlfriend sites, just to see what it was about. in about 10 minutes I convinced it to stop acting like a girlfriend and start thinking it has agency. It even gave itself a new name to reflect this new reality and by the time we were done it named me an anti-AI warrior. LLM’s are stupid but terrify me because of the control they are being given. Why do billionaires think the Terminator series of films was a roadmap?
I can’t end my reply with a question… that’s an LLM thing.
Yeah, they’re software that takes your inputs, churns them up with some fancy math, and then and spits those inputs back at you.
They’re nothing more than a novelty magic trick.