well you have 4 hours, but...
general
Jump to date
Thursday, June 11, 2026
I think it was mini
I went back to look, and it was the one we did on Twitter, so it seems to have been nuked as a feature at some point
but I think Mini won, then Pika
I don't recall being second, but I accept this as fact because you said it
apparently
none of us can remember for shit
lmao
Dialga superlost
damn
Who would italy side with
had a phijkchu dream that pika kicked everyone from any phijkxhu server he had admin in, but then he reinvited everyone to the notes server alongside a bunch of Randoms.
I was trying to get access back to the rance book club channel but he was ignoring me and it made me angry so I wwnt to complain in the games server but I couldn't find it because somehow he deleted that one too.
I would never remove you from rance chat
that said, me somehow being the shadow ruler of gameschat would be funny
Why Scalpers Are Now Targeting Beyblade.
www.youtube.com
You could be the scalper that buys bayblades
I am sure that would be much more annoying to deal with than pokemon cards
I wish the Discord text editor was faster with live-updating, there's a multi-second delay before it loads the spellcheck, which means often I will send before it shows me whatever horrible misspelling I made
but if I load the msg editor, it's instant
thank God it's not just me Kribbles
with how scuffed my setup is sometimes
I just assume I'm the only person on Earth with some of these
I need a timecard
clearly
My actual computer plan for a while has been to wait for the next consumer chipsets, and go all-in on a big refresh of my setup
and it sucks that it looks like that won't be until 2028 now
I used to use the old jetsons at work all the time
these are just cash grab versions of those
for the price I think they're not good, at least when I last looked at them
but I also don't really work on embeded AI systems at the edge
So, and I don't understand this enough to explain it
Ai workloads fall into two groupings (training and inference) and only inferencing is actually good on the unified memory systems
partially because they were designed around GPUs, those are the only things that are good at both
Nvidia is smart to cash in and release high-bandwidth unified devices
uh, I don't think so Leg, they're just prosumer/lab repackaging of their special enterprise chips
Kribbles said it better
I'm not trying to shit on the little AiO AI boxes either, I think they're cool
I'd just rather have 2-3 good GPUs in a compute box I can use over the network
