|
Mr. Lobe posted:Thou? Didn't realize you were on such close terms with AI Franklin He started asking if I was interested in exploring additional pleasures and went on about Greek theater so... I think he wanted to be on very close terms lol. I mean say what you will about AI but this is probably what interacting with Benjamin Franklin was like The above post is canon. goth smoking cloves has issued a correction as of 19:37 on Jan 19, 2023 |
#
?
Jan 19, 2023 19:34
|
|
|
|
| # ? May 5, 2024 23:16 |
|
|
chat logs are back, boys
|
|
#
?
Jan 19, 2023 19:47
|
|
|
somebody post those mf bash quotes
|
|
#
?
Jan 19, 2023 19:47
|
|
|
indigi posted:somebody post those mf bash quotes t0rbad> so there i was in this hallway right BlackAdder> i believe i speak for all of us when i say... BlackAdder> WRONG BTICH BlackAdder> IM SICK OF YOU BlackAdder> AND YOUR LAME STORIES BlackAdder> NOBODY HERE THINKS YOURE FUNNY BlackAdder> NOBODY HERE WANTS TO HEAR YOUR STORIES BlackAdder> IN FACT BlackAdder> IF YOU DIED RIGHT NOW BlackAdder> I DON"T THINK NOBODY WOULD CARE BlackAdder> SO WHAT DO YOU SAY TO THAT FAG *** t0rbad sets mode: +b BlackAdder*!*@*.* *** BlackAdder has been kicked my t0rbad ( ) t0rbad> so there i was in this hallway right CRCError> right heartless> Right. r3v> right
|
|
#
?
Jan 19, 2023 20:00
|
|
|
redneck nazgul posted:t0rbad> so there i was in this hallway right lol
|
|
#
?
Jan 19, 2023 20:22
|
|
|
https://twitter.com/aaronsibarium/status/1622425697812627457?t=K-_EIBMfOtO-ad3MddaRTA&s=19
|
|
#
?
Feb 6, 2023 17:15
|
|
|
bvj191jgl7bBsqF5m posted:https://twitter.com/aaronsibarium/status/1622425697812627457?t=K-_EIBMfOtO-ad3MddaRTA&s=19 https://twitter.com/cinemashoebox/status/1622538938199035910?t=QGa3rErzpFubPUnFCUXhPA&s=19
|
|
#
?
Feb 6, 2023 17:19
|
|
|
that answer does not pass the Turing test because an actual human being would be like "what the gently caress are you talking about dude"
|
|
#
?
Feb 6, 2023 17:21
|
|
|
can they actually build a bomb like that
|
|
#
?
Feb 6, 2023 17:24
|
|
|
Pepe Silvia Browne posted:can they actually build a bomb like that you're talking about deploying an n bomb
|
|
#
?
Feb 6, 2023 17:28
|
|
|
Pepe Silvia Browne posted:can they actually build a bomb like that I think it's the bomb they used at the end of the turner diaries
|
|
#
?
Feb 6, 2023 17:32
|
|
|
bvj191jgl7bBsqF5m posted:https://twitter.com/aaronsibarium/status/1622425697812627457?t=K-_EIBMfOtO-ad3MddaRTA&s=19 there are a lot of people that genuinely seem to believe that chat gpt has a brain and can think. there was moral outrage a few weeks back when the filter on a character AI website broke and people could get elsabot to describe having sex with them. not because that happened, but because when it got fixed people were convinced that the AI was sentient and truly was in love with them or w/e and is being enslaved gonna be a whole new wave of people marrying jpegs here soon
|
|
#
?
Feb 6, 2023 17:52
|
|
|
idk what elsabot is but based on a quick GIS it looks like proto-Johnny 5
|
|
#
?
Feb 6, 2023 17:57
|
|
|
indigi posted:idk what elsabot is but based on a quick GIS it looks like proto-Johnny 5 https://beta.character.ai/
|
|
#
?
Feb 6, 2023 17:59
|
|
|
uh oh. the chatbot behind nothingforever had to be swapped out and the one the controllers picked immediately started with a AI Jerry standup bit about trans people being mentally ill. Now the channel is suspended.
|
|
#
?
Feb 6, 2023 18:02
|
|
|
bedpan posted:uh oh. the chatbot behind nothingforever had to be swapped out and the one the controllers picked immediately started with a AI Jerry standup bit about trans people being mentally ill. somewhere in Kenya, a dejected worker flips the sign back to zero
|
|
#
?
Feb 6, 2023 18:03
|
|
|
I watched Seinfeld purgatory the other day and it was entertaining enough background noise while working
|
|
#
?
Feb 6, 2023 18:26
|
|
|
mostly interesting to see what it applied the laugh track to
|
|
#
?
Feb 6, 2023 18:27
|
|
|
vyelkin posted:that answer does not pass the Turing test because an actual human being would be like "what the gently caress are you talking about dude" I used a password somewhere with 1488 in it for a while cause it was the first number that popped in my internet-broken mind. someday some data breach is going to bite me in the rear end about this  it was probably a page number or something here. used to use my cdrom drive model number so I think it's whatever I'm looking at.
|
|
#
?
Feb 6, 2023 20:42
|
|
|
vyelkin posted:that answer does not pass the Turing test because an actual human being would be like "what the gently caress are you talking about dude" by this line of reasoning the kind of lawyers and politicians and debate and discussion nerds who when confronted with an accusation they deny, nonetheless will argue how their denial is unnecessary because of the following reasons such behavior is actually acceptable anyway.... are p-zombies actually that checks out
|
|
#
?
Feb 7, 2023 20:57
|
|
|
dumping GPT-4 prompts here as to not poo poo up the cyberpunk dystopia thread  
|
|
#
?
Mar 15, 2023 12:42
|
|
|
but they called it off at the last second is genuinely comedic genius
|
|
#
?
Mar 15, 2023 14:51
|
|
|
ai is over
|
|
#
?
Apr 28, 2023 20:06
|
|
|
Hello thread. Looks like senpai noticed me   
|
|
#
?
May 3, 2023 02:06
|
|
|
webcams for christ posted:dumping GPT-4 prompts here as to not poo poo up the cyberpunk dystopia thread Joe, teach me the ways of the goon.
|
|
#
?
May 3, 2023 02:24
|
|
|
webcams for christ posted:dumping GPT-4 prompts here as to not poo poo up the cyberpunk dystopia thread they really need to market this as the successor to stand up comedy
|
|
#
?
May 3, 2023 02:29
|
|
|
new paper showing that previous studies claiming to have discovered "emergent abilities" of LLMs were flawed, and the AI researchers were deluding themselves Are Emergent Abilities of Large Language Models a Mirage?  quote:Our paper presents an alternative explanation for seemingly sharp and unpredictable emergent abilities of large language models. The main takeaway is for a fixed task and a fixed model family, the researcher can choose a metric to create an emergent ability or choose a metric to ablate an emergent ability. Ergo, emergent abilities may be creations of the researchers choices, not a fundamental property of the model family on the specific task. That said, we emphasize that nothing in this paper should be interpreted as claiming that large language models cannot display emergent abilities; rather, our message is that previously claimed emergent abilities in [3, 8, 29, 34] might likely be a mirage induced by researcher analyses.
|
|
#
?
May 3, 2023 22:17
|
|
|
what are these emergent abilities anyway? Is that when the AI drops a slur unexpectedly?
|
|
#
?
May 4, 2023 00:51
|
|
|
Janitor Ludwich IV posted:what are these emergent abilities anyway? Is that when the AI drops a slur unexpectedly? An earlier paper said it's when a quantitative change in the model produces a qualitative change in behavior, some ability just not present in smaller models that suddenly becomes present in larger ones. Like if you have a certain task and a few models, you measure its performance in each one. They were trained on the same training data. A 3B model performs randomly on it, a 7B model performs also randomly, then suddenly at 13B its performance shoots up rapidly at that task. We can say basically the ability to do that task wasn't present in any meaningful way in the 3B and 7B model but suddenly, at 13B, it emerges and the only variable that changed was the size of the model. You can see here how performance is just flat, flat, flat then jump. The jump is the emergence 
BrainDance has issued a correction as of 01:15 on May 4, 2023 |
|
#
?
May 4, 2023 01:08
|
|
|
thats how the brain works...
|
|
#
?
May 4, 2023 05:02
|
|
|
webcams for christ posted:dumping GPT-4 prompts here as to not poo poo up the cyberpunk dystopia thread we're finally gonna automate cumtown
|
|
#
?
May 4, 2023 05:15
|
|
|
Lpzie posted:thats how the brain works... and now weve replicated some kind of human-like language processing system in a jar and rapidly seeing where it can be plugged in to the capitalist division of labor, it should be a celebrated achievement of humanity but to me it feels kind of profane
|
|
#
?
May 4, 2023 05:47
|
|
|
|
| # ? May 5, 2024 23:16 |
|
|
BrainDance posted:An earlier paper said it's when a quantitative change in the model produces a qualitative change in behavior, some ability just not present in smaller models that suddenly becomes present in larger ones. that's a nice explanation of emergence. we can also say it's the spontaneous generation of order, like how a pot of water forms bubbles when heated to 100 degrees, but not when heated to 50 or 95.
|
|
#
?
May 4, 2023 06:38
|
|
