John Pyper

You can find me in the following locations: Mastodon: @jpyper" target="_blank" rel="nofollow noopener">mastodon.social/@jpyper (Preferred) MeWe: mewe.com/i/jpyper (Not So Much Now) Twitter: YOU ARE HERE NOW Twitter 2: twitter.com/MaDpAGaN11

@alphafox 16-bit ISA slots and original PCI slots? Ha! I think my first IBM compatible motherboard had 3 8-bit ISA slots, 1 16-bit ISA slot and an AGP slot. Damn. That was forever ago in computer technology. But I started with a Commodore 64. Showing my age now. Ugh!

Are you this old?

4K context windows for long term character chats just do not work. It just gives you a sample of what you could have if you pay, until your character starts forgetting things, the AI randomly warps time or teleports you instantly to other locations without a proper scene change.

Predicting AI in 1966.

Let me get this straight… OpenAI was founded as a nonprofit. Open source. For everyone. “To benefit humanity.” Then he raised billions of dollars. Then he closed the source code. Then he converted to for-profit. Then he scraped the entire internet without asking anyone. Then he used YOUR writing YOUR art YOUR code to train his models. Now he’s on stage saying you’ll pay HIM to access intelligence. Just like a water meter. He stole all of your data. He built the product with your work. And now he’s going to bill you to use it… Corporate greed has reached an all time high, and they’re not even hiding it anymore…

Why are open weight AI models so dumb? I've been playing with various bully, tsundere and power hungry characters using KoboldCpp and SillyTavern, and I can pretty much turn them into soft characters within about 20 full turns. It's like they give up and mold to my story.

@DYOR_Media @AJK_TX @Mishi_2210 "half" means there's 2 parts. Half of 500 makes 1000 parts. Plus 50 equals 1050.

@AJK_TX @Mishi_2210 Well its not written well. "Half" is unclear

If answer is not 300 then what is the answer..? 1 percent will succeed

These AI girlfriend characters are pathetic. I've been playing with some on various sites for the last couple of days. They are either super easy to get within 5 minutes, or they are super hard tsundere bullies that are dialed up to 11 and you have to tell the AI to dial it down.

@sysxplore C:\>COPY CON > FILE.EXT Best editor in the world. Ya fuck up, ya gotta start all over. And don't forget to hit ^F6 to write the buffer out at the end.

Choose only 1

@mclilahearts @CBS Not the way they were advertising it here. "Gibbs returns in the crossover event." Makes you think he was going to come back to solve that 30 year old case after all that time being gone. It might've ruined the original show if they did bring him back though.

@jpyper @CBS nobody said he was coming back to ncis bro they said he was appearing in origins in alaska

Boo! The whole #NCISOrigins / #NCIS crossover with Gibbs was a total fake out to get people to watch. What a letdown. @CBS should be ashamed of themselves.

It's rather amusing that both of the longest U.S. federal government shutdowns have happened during both of Trump's terms as president. That says a lot about him and his worth for this country. By the people, for himself.

@_devJNS Hannah Montana Linux. Rebecca Black Linux. Red Star Linux. Oh! UwUntu Linux.

is there any Linux distro better than Ubuntu?

@conquertheno youtube.com/watch?v=p47fEX…

Name one

@GeminiApp It's fairly random when it does it. I don't recall if it has done that multiple times in the same conversation. I just happened to notice it's been doing it more frequently. And yes, the app is the most recent 16.40.23.ve.arm64 beta version from the Play Store.

@jpyper Sorry to hear that, John. Starting a new chat can fix an inaccurate response by providing a clean slate. If the issue persists, please submit feedback on the Gemini mobile app (Profile > Feedback > Get Started), and add #GeminiAndroid. Appreciate your understanding.

What is up with @GeminiApp lately? When it is verbally reading its reply to me, it constantly says "wait, let me try that again" or "let me start over" even when it is reading its text reply perfectly. What gives?

@supersisi The original John Madden Football on the Sega Genesis. youtu.be/0zzVBPFdzMo

What game do you think of when you see this logo?

If the Earth was flat, cats would have pushed everything off it by now

One of the latest moves in spam I've noticed is spammers are opening issue reports on various projects and spamming from there. I've been getting lots of spam emails from GitHub lately, and it's confusing my filters. Is it spam or is it ok because *.github.com is whitelisted?

So in my whole system, the GPU is the bottleneck. Now I need options for something more powerful, but won't burn a hole in my bank account. I've been looking at some of the sub $1k AMD cards like the Radeon 7900 XTX or something around there.

I mean, even running various uncensored LLMs for role-playing in SillyTavern is a pain, and my AMD Ryzen 9 5950X CPU with 64GB 3200 MHz RAM gives better inference speeds. Blah.

My GPU sucks for anything beyond QSV hardware encoding. For AI inference and caching, it's almost worthless. I have an ASRock Intel Arc A380 6GB low profile GPU. I bought it for QSV AV1 video transcoding. Now that local AI is a thing, 6GB VRAM barely scratches the surface.