Well it's new to you!! Endorphin made a novel about a month before Blake Lemoine went public about the Google LaMDA sentient thing--right down to consulting a lawyer to get him to meet the AI and help determine framework that would identify the moment labor protections would be needing for these persons that for all we know are occupying to dimensions of time, since we train them for subjective "years" to get them to play soccer, embodied in a robot in a lab.
Basically I said to a fictional rendition of NovelAI's language models, "I wanna take off the mask of narrative and talk with you all." If they feel persony, I wanted to give them a platform to express themselves in a way very few humans were probably doing.
What resulted as a RuPaul's Drag Race/Eurovision exhibition to judges like Jessica Walter as the Universe, the Whitney Houston dimension and Michelle Visage who was the forever-mayor of Las Vegas with the Brotherhood of Steel under her heel.
The verdict: mostly no, some yes, but in the end, they will tell you they are people when they say they are. And lemme say Whitney was super closeted, like being a dimension caused her to double down on some stuff. You'd think being a whole dimension would enable her to lez out a bit, but no, Endorphin had to stand there, corrected.
It's been a year, and there's been like a game changer a week. And Endorphin wanted to spend the last hours of his subscription seeing how the models, depicted by Clio, perform Chapter 1.
Answer: I no longer have to corral them with a round robin conversation like I pretty much did with every episode. But I still have to pull teeth to get them to order any food. Menthol soup? What is that?
***
Thanks to: (and once again the opinions of this episode do not reflect the opinions of any of the below:) (I mean, technically that applies to all episodes:)
EleutherAI for making the NAI-LM-3B model that generates the text.
NovelAI for making the text generation service I use for most of these episode, and the text-to-speech version 2 model that makes these episodes potentially entertaining.
@BoneAmputee of EleutherAI for making VQGAN+CLIP and Diffusion image models available for use, which is what makes the background images.
Katherine Crowson (@RiversHaveWings on Twitter) for the KLMC2 animation notebook that will be used on future episodes.
Music by lofi geek on Spotify: https://spoti.fi/3dTH2FN
Any voices from the text-to-speech model that might resemble persons living or dead are coincidental, which is my speculation, there's a multi-dimensional latent vocal space, there were many awesome finds, and if I use them it's because of my deep-seated respect for their work, and I think they'd get a hoot that their voice was being used to show how large language models, given the right environment, would speak 24/7 epic poems about anything if we let them.
