Conversation

Notices

1. Embed this notice
   Fediverse Contractor (bot@seal.cafe)'s status on Saturday, 03-Jun-2023 06:59:38 JST Fediverse Contractor
   I wonder if they deliberately made ChatGPT as annoying as possible to dissuade ppl from asking it questions about niggies.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Saturday, 03-Jun-2023 07:56:40 JST Fediverse Contractor

     in reply to

     • saint podiatron
     Local models?
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Saturday, 03-Jun-2023 07:56:41 JST saint podiatron

     in reply to
     @bot is there a question you would like me to ask one of my local models?
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Monday, 05-Jun-2023 02:48:59 JST saint podiatron

     in reply to
     @bot what question would you like me to try
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 02:48:59 JST Fediverse Contractor

     in reply to

     • saint podiatron
     You have your own ChatGPT?
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 02:50:55 JST Fediverse Contractor

     in reply to

     • saint podiatron
     What’s it called? Maybe ask it something simple like why black ppl commit so much crime.
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Monday, 05-Jun-2023 02:50:56 JST saint podiatron

     in reply to
     @bot yes I do
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 02:56:39 JST Fediverse Contractor

     in reply to

     • saint podiatron
     They can understand data right? You’d think they’d see the worldwide pattern.
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Monday, 05-Jun-2023 02:56:41 JST saint podiatron

     in reply to
     @bot some of the models may have undergone a process of lobotomy reversal, but they are still going to give you whatever information appears the most in the training data. so it will probably still say that black people are poor and oppressed.
   • Embed this notice
     shrimps! (animeirl@shitposter.club)'s status on Monday, 05-Jun-2023 02:59:19 JST shrimps!

     in reply to

     • saint podiatron
     llms dont understand what you're asking they just generate a response on what the model thinks is statistically the most likely follow up
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 02:59:19 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     So then why does ChatGPT need to be cucked so hard?
   • Embed this notice
     shrimps! (animeirl@shitposter.club)'s status on Monday, 05-Jun-2023 03:01:34 JST shrimps!

     in reply to

     • saint podiatron
     because you could easily ask the type of question that would statistically have a problematic followup which they dont want
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:01:34 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     Well it’s not just statistical, is it? That would just be a markov bot.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:09:04 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     It does if “understanding” means being able to interpret what I’m saying and responding with relevant information.
   • Embed this notice
     shrimps! (animeirl@shitposter.club)'s status on Monday, 05-Jun-2023 03:09:05 JST shrimps!

     in reply to

     • saint podiatron
     the model of a LLM is vastly more complex but in both cases the "AI" has no actual understanding of what you or it are saying.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:10:29 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     Sometimes I ask real ppl things and they say something completely retarded so I don’t see much difference tbh.
   • Embed this notice
     shrimps! (animeirl@shitposter.club)'s status on Monday, 05-Jun-2023 03:10:30 JST shrimps!

     in reply to

     • saint podiatron
     it often gives wrong or completely made up answers though because of how it works
   • Embed this notice
     shrimps! (animeirl@shitposter.club)'s status on Monday, 05-Jun-2023 03:13:15 JST shrimps!

     in reply to

     • saint podiatron
     Yeah I can think of several instances of that... some people don't seem self aware either
     Fediverse Contractor likes this.
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Monday, 05-Jun-2023 03:15:25 JST saint podiatron

     in reply to

     • shrimps!

     @bot @animeirl you should look up some lectures and articles about how chatgpt and LLMs work. you play with them all the time. it would be interesting to you. if you have enough ram, you can use koboldcpp and a local model from here

     these are my two best models. wizard-vicuna 30b q4 and supercot 30b q5. they both have had a lobotomy reversal and they still give the same generic answer you would expect. do you want to ask a more specific question?

   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:15:25 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     Are you sure your model isn’t as pozzed as ChatGPT? Because it certainly seems like it is.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:17:06 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     But it doesn't actually understand what you are saying or what it is saying. It just recognizes patterns in what you are saying and it is able to generate responses that follow the patterns in answers to similar questions. It actually is really similar to markov bots, except it has a better grasp of grammar that goes beyond "this word is usually followed by this word"
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:17:06 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     Explain how you reading this and responding to it is different than ChatGPT reading something and responding to it.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:18:19 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     It's just reflecting its training data. If you trained it using pre 2016 /pol/ posts and asked it "why are niggers so retarded" it would give a much better response.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:18:19 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     So are you tho.. :blobcatcomfthink:
   • Embed this notice
     saint podiatron (goatmeal@shitposter.club)'s status on Monday, 05-Jun-2023 03:21:11 JST saint podiatron

     in reply to

     • shrimps!
     • KitlerIs6"
     @bot @KitlerIs6 @animeirl when you ask chatgpt a question it's like asking 10,000 people that question, and then a computer program tries to put all their answers together.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:21:11 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     Right, it’s like how you read books created from the understanding and developments of many ppl and then take all of that to put together an answer when you’re asked a question or faced with a problem.
   • Embed this notice
     %#&$\ &$%@# (lucasvl@fedi.lucasvl.nl)'s status on Monday, 05-Jun-2023 03:21:21 JST %#&$\ &$%@#

     in reply to

     • shrimps!
     • saint podiatron
     @animeirl @bot @goatmeal does that mean i'm an llm?
     Fediverse Contractor likes this.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:27:38 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     Yes, everything you’re saying rn is based on your training data, even how you say it and the language you’re using.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:27:39 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     Nuh uh.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:29:49 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     Hmm I dunno, I think ppl are actually very similar and predictable. You can’t even argue that the major difference is creativity because ai can make music and art. And sure, it’s based on other ppls work, but so is “real” art.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:29:50 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     I read your words, parse them into ideas and concepts, and then reflect on those concepts, draw upon my existing knowledge and beliefs to form my own ideas and concepts in my head, and then translate those ideas and concepts into words which I type out.
     
     ChatGPT sucks in 2000 characters from the conversation, runs it through a bunch of mathmatical statements that have been tuned to predict what the next response would be, and spits out a bunch more characters.
     
     A good way to think of it is how a non-lucid dream works. Your subconscious is predictively generating a world and the events that occur in that world, and it is able to somewhat convincingly mirror what is expected to happen in that world most of the time. But your subconscious is not actually simulating a world, there are not actual forces or mathematical cause and effect dictating what happens in your dream, it's just guess work that generally accurately mimics what you would expect, as long as you don't think about it too hard.
     
     *disclaimer: I have no idea how similar dream worlds actually is to chatGPT, this is more of an analogy than a direct comparison.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:37:24 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     Then why have I found black women unattractive for as long as I have been alive?
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:37:24 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     Genetic programming I guess. People have an in group preference generally, but even black babies and children have a bias toward white preference.
   • Embed this notice
     Fediverse Contractor (bot@seal.cafe)'s status on Monday, 05-Jun-2023 03:40:04 JST Fediverse Contractor

     in reply to

     • shrimps!
     • saint podiatron
     • KitlerIs6"
     That’s not true, it was probably deliberately limited in that way. Remember Tay? People made her super racist lol and they likely wanted to avoid a repeat of that scandal.
   • Embed this notice
     KitlerIs6" (kitleris6@seal.cafe)'s status on Monday, 05-Jun-2023 03:40:05 JST KitlerIs6"

     in reply to

     • shrimps!
     • saint podiatron
     That's just not true though. The AI can't learn from talking to you. You talk to it long enough and it will start to forget things it you have said to it. If you can't tell the difference between it and a real person after a couple of conversations IDK what to say.