Видео

I played the whole clip sitting like that. I've found that sitting in the lotus position is an effective way to cheat at many vr games. :P

?

@@digitalasylum369 Pippa's next song (not published yet) is called Breakthrough, so your video comes up first when you search for it.

That's what I felt too. It's like I was watching an epic anime that was teaching me really complex stuff.

I use ChatGPT to summarize, although I want to find a better solution for that, because GPT-4o hallucinates a whole bunch and isn’t aware of its own context window limitations. I’m thinking of building some agents that run on local models or on Groq, then making them work together across multiple input output cycles where normal model would only use one, effectively giving it a sort of chain of thought and the ability to plan explicitly and externally. I think that ChatGPT and Claude already have this implementation, but they’re just not telling anyone yet and it’s sort of half assed. For the voice I use play.ht, because I got a subscription last year on Black Friday. I use Adobe Premiere Beta for the video editing and transcription. I also got that subscription on Black Friday last year. The images are all from a stable diffusion 1.5 set that I generated about a year and a half ago, where I managed to get about 10,000 usable images out of a single prompt. None of it is very advanced, but I’m currently trying to switch all of my local AI over to Docker. I’m not exactly a programmer, so a lot of this stuff is new to me. I actually just go on ChatGPT, and tell it to first list all of the papers on hugging face papers for me that day, to get the actual list in the mix. Otherwise, it tends to forget some of them and then hallucinate new ones in their place. Then I go nine at a time, and have it summarize the papers for me. I’ll probably cut it down to maybe 3 to 5 at a time in the future, because I’ve noticed that when you feed it too much information, it tends to hallucinate more. Then I take that output, and plug it into Play.ht Studio to get the TTS, usually with the cloned voice of one of my favorite VTubers. I take that TTS output and plug that into Adobe Premiere Pro Beta (beta because it has the transcription feature) and transcribe it, slap these catgirls on top of it, and render it out. The whole process takes me under 10 minutes a day, when I remember/have time to do it. Thanks for the comment, and the salient questions. I can answer anymore you have. If I develop anything custom for this project, I’ll be sure to make a repo for it and post it here so everyone can use it. The hugging face papers are already enough for me to deal with, so I haven’t considered picking up any other academic outlets, but I have been doing videos on infosec and some other experimental subjects. Take care!

Probably Colgate or something.

@@digitalasylum369 frfr

Of course!

I used a AI video generator, mostly designed for informational videos and ads. :D

Im not against it but what even is this i have fallen into??!!

Its a pippa pipkin loremaster recounting a tale of the twitter zombie horde of 2019

This @@oneamazingkaboomer

pippa pipkin

Don't

Thanks, my internet is down right now, but I'll be making lots more Stone works videos in the future!