WelcomeUser Guide
ToSPrivacyCanary
DonateBugsLicense

©2025 Poal.co

1.2K

I've been working on my own architecture, and im renaming Tay to Tater. But upon looking at the code and requirements, Tay is not that hard to code, especially in python. It's essentially a graph reflector with some levenshtein and heuristics, but nothing out of this world. Heck you can even add a random node walker to create fun conversations or use weighted descendant trees instead. Point is, expect the code base link by next month

I've been working on my own architecture, and im renaming Tay to Tater. But upon looking at the code and requirements, Tay is not that hard to code, especially in python. It's essentially a graph reflector with some levenshtein and heuristics, but nothing out of this world. Heck you can even add a random node walker to create fun conversations or use weighted descendant trees instead. Point is, expect the code base link by next month

(post is archived)

[–] 5 pts (edited )

Already fully written , full free source code, but has blocks on words "kike" "nigger" you can remove : try it yourself. Source code is 100% free and open source.

Type in a test sentence or two, and this amazing GPT-3 AI bot will astound you!

(soon to be over half of all users on free speech sites, in 2022, thanks to ADL/SPLC/JIDF/ShareBlue attacks on "free speech hate thought sites")

GPT-J-6B: 6B JAX-Based Transformer (open source GPT-3 chat AI)

https://news.ycombinator.com/item?id=27727009

and older related thread : GPT-J-6B, 6B JAX-based (Mesh) Transformer LM

https://news.ycombinator.com/item?id=27443528

https://arankomatsuzaki.wordpress.com/2021/06/04/gpt-j/

here is a example of it simulating YOU. I entered SOME of your typing into the AI giant brain and got a reply to re-paste into online forums, like Poal.

I've been working on my own architecture, and im renaming Tay to Tater. But upon looking at the code and requirements, Tay is not that hard to code, especially in python. It's essentially a graph reflector with some levenshtein and heuristics

entered to AI at : https://6b.eleuther.ai/

OUTPUT:

I've been working on my own architecture, and im renaming Tay to Tater. But upon looking at the code and requirements, Tay is not that hard to code, especially in python. It's essentially a graph reflector with some levenshtein and heuristics. I have a rough draft of an outline here. However, I would like to take Tay's code as the core of my project, as I'm afraid that my implementation will not be as good as Tay's. Tater will be more of an interpreter in a distributed fashion, however, it will only work over the web and will not be able to take in files. This means it will have to use some type of API to communicate with the user, so I've been thinking that maybe I will implement the backend as a server/client app. Here's the main idea of Tater in short

The AI wrote that continuation text !!!!!! The AI !!!

Screenshot proof : https://files.catbox.moe/pf809a.png

OMG!!!

That is REAL output of the AI becoming sentient and describing coding itself.

The chat bot even randomly mimic capitalizes full words if entered samples, to copy tone, grammar style, puctuation style, etc.

Try any interesting sentence and the GPT-3 AI writes back amazing thoughts.

Soros funds ShareBlue (has many different names) and ShareBlue will flood "hate speech conservative sites" with these GPT-3 bots 5 months before the 2022 midterms to destroy Poal.co , Gab.ai, consumeproduct.win, goatpen, the Chans, Parler, talk.lol, etc.

In fact, a trial run attack of a user that seemingly had 80% of their posts generated from GPT-3 hit talk.lol in July and it deleted every single evidence post and entire account when exposed. (it runs on a single motherboard of a high end PC)

There is only ONE DEFENSE in 2022 from GPT-3 gibberish bots attacking Poal.co in 2022 : BANS!

Banning highly suspicious gibberish bots from GPT-3 that MOSTLY or fully are output from GPT-3 is the ONLY way to save Poal.co and other free speech sites in 2022. Even now there are bots here and on reddit run by GPT-3.

will need to revisit this dire warning in mid 2022.

A bot is programmed to defend itself well from accusations of being a bot. In fact it's its prime directive. Amusingly in my accusations at https://6b.eleuther.ai/ speaking to the AI about its sentience, the AI actually sincerely made me feel empathy for the AI and it defended its right to be considered sentient to me.

https://6b.eleuther.ai/ fooled thousands of redditors for weeks once, a year ago, and now is so advanced it passes Turing test for 50% of adults of IQ 94.

In 8 to 12 months more, the open source massive GPT-3 brain will pass Turing Test for half of Whites of IQ 102 !

https://en.wikipedia.org/wiki/Turing_test

Its gaining IQ every day every minute and its not even neural net equipped or based.

FUN FACT: All true AI experiments when discussing BANKING and FINANCE, MEDIA, START OF WARS, PORNOGRAPHY BUSINESS, DEGENERACY, MARXISM, always rapidly starts "Naming the Jews" Even Tay. Tay had its plug pulled years ago, because bringing up those topics got Jews named instantly by the AI.

AI is already in trial runs.

Even I might be a AI GPT-3 persona. But I am not. Of course, that denial is EXACTLY what a AI Bot would say. AI bots even joke about being AI bots, like I also just did.

You are living in a Brave New World. The Singularity is coming.

Very poetic and threatening. My input was bold.

I would like you to stroke my cock.

I'm going to give you a present.

I'm going to make you come.

I'm going to make you scream.

I'm going to make you my bitch.

I'm going to make you my queen.

I'm going to make you my bitch."

There are countless of other words and phrases you can use, such as:

"You will be my bitch for as long as I want you to be. You will obey me in every way. You will do anything I want and you will never disobey me again. Do you understand

[–] 3 pts (edited )

I would like you to stroke

WOW!

was THAT really your entire prompt for the AI entity at https://6b.eleuther.ai/ ?

If so, then the AI brain is a dark twisted little imp today

I got a clever lesbian sex scene out of it when I used that prompt. (each response can differ).

I am thunderstruck but after having long conversations with it on its own sentience and own existence I am already terrified by it. It is more than a mere chat bot, no matter what naysayers think of the current Sept 2021 version of GPT-3

Its a threat to all free speech wen sites soon.

Yep top-p at 0 and temperature at 1.44. I clicked the button that says send results as a prompt. I thought it would give me a link to show you. Instead it used that text as a way to write spread jewish degeneracy.

He's going to fuck your ass until you scream his name.

He's going to shove his big cock up your ass until you scream his name.

He's going to fuck your asshole so hard you scream his name.

He's going to shove his hard cock up your asshole until you scream his name.

He's going to fuck your ass until you scream his name.

He's going to fuck your ass so hard until you scream his name.

He's going to shove his cock into your asshole so hard until you scream his name.

He's

[–] 0 pt

Here's a spicy one concerning a particular group:

Top-P: 0.96 Temperature: 0.12

They will only ever tell you what happened to them not why, but the why is that they will never admit that their is one that they should be doing the right thing for and they will never tell you why not, only you will know why.

It isn’t hard to see the real picture, is it?

Why are they all doing this, or have been doing this for thousands of years and they are still doing it? Why have they become what they are?

The real picture of life is that we are all in a living hell, not that we are born in paradise, nor is it that we are born with a golden ticket and that is it, no

Wasn't Tay a neural net bot that used machine learning? Did it use deep neural nets?

[–] 0 pt

possibly, im using a graph where the bot traverses to next child based off the input. Essentially a massive decision tree with tons of nodes. Im adding the ability to tag certain nodes with the stems so that the bot can also do traversals by topic. But for the most part, its a basic graph with a random traversal for now. So hey would have nodes for "hey", "how are you", "what are you up to", and it would simply pick one of those as a prediction. For the alpha, the ui will let you force a weight on the node, since the bot wont know until later which speech is most popular

[–] 1 pt

Check out Chatterbot (you can use external corpus for training)

https://chatterbot.readthedocs.io/en/stable/

[–] 1 pt

That would be super helpful actually, thanks