mathstodon.xyz is one of the many independent Mastodon servers you can use to participate in the fediverse.
A Mastodon instance for maths people. We have LaTeX rendering in the web interface!

Server stats:

2.8K
active users

#dat

0 posts0 participants0 posts today

wacoca.com/news/2509015/ 中国ディープシーク、無断で個人情報とプロンプトを転送=韓国当局 | ロイター ##中華人民共和国 #AMERS #ARTI #Asia #ASXPAC #China #CIV #CMPNY #CN #DAT #DEST:NOJPBSM #DEST:NOJPTPM #DEST:NOJPZTM #dip #DLI #EASIA #EMRG #Gen #ITSE #ITSE08 #ITSE1 #JFOR #JLN #JPTOPIC:BUSINESSTECHAI #KR #NAMER #POL #PRIVT #REGS #RSBI:DATAPRIVACY #SCI #SOCI #SWIT #Tech #TECH08 #TMT #TOPCMB #TOPNWS #TRN #US #www

I just read that nps.gov erased Harriet Tubman from a page about the Underground Railroad. I want not only to #archive copies of websites, but to use #cryptography to prove that this was the content fetchable from the real page at a date in the past. #provenance Why aren't there FAQs about how to do this already? I've just seen sciop.net which appears to be using both #BitTorrent and #RDF to advantage; and #ArchiveTeam appears to be doing distributed fetching at scale with their ArchiveWarrior thingy. But how should anyone trust that I don't tamper with the pages I fetch and archive? Or what if I just want free backups, so I pass off my photo archive as part of #memoryhole 'd DEI policies? And where's #ipfs or #dat or #hypercore ?

Replied in thread

There isn't, but that's a nice idea -- I'm going to think about that! Here's a janky script version that doesn't preserve pwd's:

1. set up the 4 windows on desktop and:
> cp /ram/shared/windows.pod snapshot.pod

2. to restore (assuming all terminal):
dat = fetch"snapshot.pod"
for i=1,#dat do w=dat[i] create_process("/system/apps/terminal.lua", {window_attribs={x=w.x,y=w.y,width=w.width,height=w.height}}) end

Replied in thread

@mogwai_poet @zep yeah, the fix is in the main thread! there's actually another bug in the fixed version when you try to compress 0x8000+ values -- `while pos <= #dat` fails b/c #dat overflows. I fixed that too but never published it anywhere. I was thinking I might publish it as a separate thread, so that I could be the person who needs to do bugfixes instead of you (zep). but I'd be happy to send it if you'd like to keep updating the PX9 cart

okubrowser.github.io/

Beaker Browser was a Chromium-based Web browser that incorporated the Dat protocol with the primary aim of enabling peer-to-peer static site sharing. To this end, Oku is similar; in fact, Beaker Browser was part of the inspiration to create Oku. The two projects diverge in their long-term ambitions, however; (...)

Hyped 🔥

okubrowser.github.io Oku · Your new home on the Internet Your new home on the Internet

“Intersectional hallucinations”: The AI flaw that could lead to dangerous misinformation

when we use synthetic data, intersectional hallucinations are always going to happen because the synthetic data must be slightly different to the original, otherwise it would simply be a copy of the original data. Synthetic data therefore requires hallucinations

#ArtificialIntelligence #AI #dat #bigdata #syntheticdata #hallucinations #technology #tech

psypost.org/intersectional-hal

PsyPost · “Intersectional hallucinations”: The AI flaw that could lead to dangerous misinformationBy Ericka Johnson