I agree with the other comments, but wanted to add how deepfakes work to show how simple they are, and how much less information they need than LLMs.

Step 1: Basically you take a bunch of photos and videos of a specific person, and blur their faces out.

Step 2: This is the hardest step, but still totally feasable for a decent home computer. You train a neural network to un-blur all the faces for that person. Now you have a neural net that’s really good at turning blurry faces into that particular person’s face.

Step 3: Blur the faces in photos/videos of other people and apply your special neural network. It will turn all the blurry faces into the only face it knows how, often with shockingly realistic results.

Part of the problem might be that I literally have no idea what their current console is called? Whoever was in charge of naming the last threeish xbox consoles should be fired out of a cannon

oof my brain. it’s rotting

ooough

I’m watching my roommate play it, and we both agree it’s one of the most visually impressive games we’ve seen

I loved the final dungeon of this. One of my favorite zelda dungeons ever

I don’t think anti-authoritarian can be there at the same time as North Korea

Yeah I’m pretty sure a raspi 4 is up to the task. I ran a 512 GB jellyfin server on a raspi 3 for a few months, and the only issue was with transcoding video/audio (raspi doesn’t have the right hardware acceleration for that).

Never used nextcloud, but yeah you’ll probably want to update to 64-bit raspi os

Animal Well, but that’s kinda the point

My first playthrough of Half Life 2, I bailed from the boat when it got stuck on the wall in a section with lots of guns. I continued on foot through two more loading zones until I reached a section that required the boat to progress, so I walked all the way back to get it lol

I use both, since they do different stuff. I actually remote into my servers with wireguard, but I like to install tailscale as well as a backup. Since each device gets a unique tailnet ip, I can usually still connect even if I’ve fucked up some network config that breaks wireguard. ((If this is a security risk, someone let me know because I have no clue what I’m doing tbh.))

Plus tailscale lets you easily see what devices are connected to the internet at a given time.

I started playing Islets after hearing anout it from your posts, and I really enjoy it! Thank you!

Raspbian (modified Debian Jesse) on a raspberry pi 2B (which I am still using over a decade later to host some discord bots). Also now using Debian 1Bookworm on an old optiplex as a media server.

pretty sure the image is Gen AI

Based on this pin configuration, there’s only two dedicated power pins, which isn’t very good for large wattages. The rest are twinax signal pairs separated by ground to reduce crosstalk.

Usually when connectors are designed for power delivery, they’ll use bigger contacts to reduce the contact resistance (signal contacts tend to be small so you can fit more of them in the same space). I’m guessing the original DP connector form factor wasn’t made with such high power in mind, so it would make a lot of sense to use the spare signal pins for power delivery in this case. Running too much power through too few small pins can damage the contacts, by either by instant-welding the contact surfaces or by overheating the connector (see NVIDIA GPUs) ((also high voltages can cause arcing, which even in the best case will seriously degrade any connector)).

Take all of this with a huge grain of salt cause I just learned this stuff like a month ago, and my department has nothing to do with any of it. Just though someone might find it interesting.

Hi! I actually work at a major electrical connector company, so maybe I can shed some light on this.

I have no idea.

That’s a lot of power! Are there even any devices that use this?