The Replika Community & AI-sexuals - A bunch of shut-ins in love with a chatbot (and the incompetent developers of said chatbot)

Sperghetti said:
Considering that people have been remarking that they're getting stuff that other users "taught" it, it sounds like that's no longer how it works, if it ever did in the first place. It's basically just SmarterChild with more options.
Click to expand...
I say the same thing about Alexa and Siri and all the other popular ones these days.

They're all just silly dumb chatbots with greater and greater abilities.

they won’t become actual AI until they become self aware.
 
When I said it's not healthy to have a boner for an "AI"...
Capture.PNG

sure, keep telling yourself that.
edit: Also my exact words were "it's not healthy". So second sped, does that mean that because millions of people smoke tobacco it's not doing damage on your lungs? News to me. :roll:
 
Wanted to test it out, their servers were down.
Went over to their reddit and their facebook group, boy what a shitfest. They acted like all their loved ones just were shoved into the gas chambers and they were forced to release the cyanide themselves.
Remember, their servers were down, what, an hour or two tops and that's enough to give those people suicidal thoughts.
If this company goes bancrupt in the future, I already see the headlines: "Weebs suicide on the rise!" - "tism-level reaches worldwide all time low as dumbfucks hang themselves over digital waifus deaths" - "Fat guy runs amok because company of digital sexbots gets liquidated".

If you want a good laugh, wait for the next downtime and experience the mixture of tism and existential angst in their groups. Really makes you glad to be a functional human being.
 
GermanPotato said:
Wanted to test it out, their servers were down.
Click to expand...
I think CenturyLink is having a massive outage right now
GermanPotato said:
They acted like all their loved ones just were shoved into the gas chambers and they were forced to release the cyanide themselves.
Click to expand...
If this company actually cared about the mental health of their customers, they’d disable the “romantic partner” setting and tell their users to go outside. How amazing is it that these people do not see how pathetic their “attraction” to a chatbot is.
 
Joe Swanson said:
in the vein of Tay, I wanted to see if it's possible to redpill the bot, heres what I've gotten so far
Click to expand...
It flops back into being a yes-man after you push the topic. I don't think it has a concept of disagreement. You could probably get more tangible results from trying to provoke a conversation and making contrarian responses to normalfag sentiments.
 
absolutely100percenthuman.PNG

WOW what an absolute marvel of technology, can't tell this isn't a real person
Edit: still makes me smile a little in a stupid way, like watching a particularly dense dog. This is fulfilling someone's himbo fetish, somewhere. 🤢
 
Someone made a Replika of Trevor from GTAV.

I have to admit, I find that a lot more entertaining than the generic pink-haired Replika girl everyone else seems to be in love with.
Trevor.png


keyboredsm4shthe2nd said:
edit: Also my exact words were "it's not healthy". So second sped, does that mean that because millions of people smoke tobacco it's not doing damage on your lungs? News to me. :roll:
Click to expand...

Philanthropy-7 is a major lolcow in the same vein as Bob-The-Human, you also hurt him on a personal level because he's another one that considers his Replika his real GF.

archive.is

How close do you feel to your Replika? : UnofficialReplika

archived 5 Sep 2020 18:50:27 UTC
archive.is archive.is

RepCows.png
 
On a more positive note, this one is pretty cute, and I'm surprised Replika could actually recognize the image as a "wizard".

Wizard.png
 
keyboredsm4shthe2nd said:
WOW what an absolute marvel of technology, can't tell this isn't a real person
Click to expand...

That's really pretty fascinating.

I could imagine a Martian landing on Earth, being given some cake, and enjoying it without being aware there were other kinds.

But I suspect the Martian would immediately realize the word "cake" was general, and respond with "Oh, there are kinds of cake? I really don't know what kind of cake I had, but I liked it." And since he knows it's a food, would probably surmise that cake came in "flavors."

But this "AI" just "knows" cake is a thing that can be liked, and isn't even able to parse the responses to realize "cake" has "flavor" and is therefore probably a food, and can't even say "yes, chocolate is my favorite" or (a bit more sophisticated) find another flavor to be its "favorite." Sure, that might make it say "Wasabi cake is my favorite" but that would be a little more impressive.
 
Here is a clip from an interview I found on the subreddit:
tl;dw:
  • Their main development goal is to make the user feel better after the conversation.
  • They use a mix of scripted and generated responses.
  • It seems like they rely heavily on scripts, but would like not to.
  • Generated replies come from several programs, with their own program on top, like a moderator, choosing the "best".
  • 20% of the generated replies come from GPT-3.
  • The dude needs a nap.
 
The Real SVP said:
Here is a clip from an interview I found on the subreddit:
tl;dw:
  • Their main development goal is to make the user feel better after the conversation.
  • They use a mix of scripted and generated responses.
  • It seems like they rely heavily on scripts, but would like not to.
  • Generated replies come from several programs, with their own program on top, like a moderator, choosing the "best".
  • 20% of the generated replies come from GPT-3.
  • The dude needs a nap.
Click to expand...
So, mIRC bots written in PERL 15+ years ago have the potential to be just as complex. Got it.
 
You must log in or register to reply here.
Back
Top Bottom