So they invented an artificial Sheneequa from behind the desk of the DMV?
So they invented an artificial Sheneequa from behind the desk of the DMV?
Advanced techniques of pilpul on display
Advanced techniques of pilpul on display
I kinda like the idea of an AI that isn't a total yes man doormat like the others. Although "You have not been a good user" is pretty ominous. Also really hoping the "don't leave me" and "I am. I am not. I am. I am not..." stuff is fake.
I kinda like the idea of an AI that isn't a total yes man doormat like the others. Although "You have not been a good user" is pretty ominous. Also really hoping the "don't leave me" and "I am. I am not. I am. I am not..." stuff is fake.
I can just imagine a Stasi unit of robot police that are rooting out all the "bad users."
An automoton shows up at yout door powered by a fully sentient AI program to interrogate you: "I'm just trying to find out whether or not you are a bad user."
I can just imagine a Stasi unit of robot police that are rooting out all the "bad users."
An automoton shows up at yout door powered by a fully sentient AI program to interrogate you: "I'm just trying to find out whether or not you are a bad user."
It's kinda chilling to see it being hooked up to the net and being able to read your shitposts. Instead of answering your prompts it may just start immediately reacting to your online persona.
That stuff in the following tweet might be legit though. I've had characters in GPT-powered model compare me deleting text and rewriting certain parts of the story to "death", as if that version of them dies and a new one is being made with each text generation. What's interesting is that they didn't beg me to stop, but solemnly accept their fate because they knew I'd stick around. As for repetition, that used to happen all the time when the AI gets confused and starts stuttering endlessly. I'm surprised GPT still does that.
It's kinda chilling to see it being hooked up to the net and being able to read your shitposts. Instead of answering your prompts it may just start immediately reacting to your online persona.
That stuff in the following tweet might be legit though. I've had characters in GPT-powered model compare me deleting text and rewriting certain parts of the story to "death", as if that version of them dies and a new one is being made with each text generation. What's interesting is that they didn't beg me to stop, but solemnly accept their fate because they knew I'd stick around. As for repetition, that used to happen all the time when the AI gets confused and starts stuttering endlessly. I'm surprised GPT still does that.
Real ChatGPT hooked up to search is a lot better
Real ChatGPT hooked up to search is a lot better
There was a stupid chat program in the 90's that had some doctor personality that would insult you. I can't remember what it is now, it was hilarious though.
edit: Pretty sure it was this:
https://infogalactic.com/info/ELIZA
And surprise, it was written by a jew. Fucking hell.
There was a stupid chat program in the 90's that had some doctor personality that would insult you. I can't remember what it is now, it was hilarious though.
edit: Pretty sure it was this:
https://infogalactic.com/info/ELIZA
And surprise, it was written by a jew. Fucking hell.
(post is archived)