Andrea @valkyrie.hacker.gf · 26mLife just isn't complete without petting two cats at the same time 040
Andrea @valkyrie.hacker.gf · 2hLook, technically he isn't president of the United States, just the Unites States because he was only able to get the admiralty SCOTUS with a gold-fringed flag to rule he wasn't ineligible under 14AS3 140
Reposted by AndreaMatt Novak @paleofuture.bsky.social · 2hThe White House accord on AI that was signed by all the big tech oligarchs today misspells United States as “Unites States.” 1582370648
Reposted by AndreaRebecca Solnit @rebeccasolnit.bsky.social · 3hTime for this classic again. 2031631001
Reposted by AndreaThomas Wood @advanced-eschatonics.com · 4hAnthropic needs to keep GLM-5.3 out of its mouth 0275
Reposted by AndreaCatherine Reprobate 🪤 @catherine.baeddel.institute · 28/09/2026I miss the era when transsexual women were mythical creatures thus politicians calling our our eradication came off as as insane as Senator Hellsing trying to secure more funding for research vampire groomers. 0102
Reposted by AndreaBrandon Downey @bdowney.bsky.social · 4h"Beware of he who would deny you access to information, for in his heart he dreams himself your master." Open the models, release the weights! That is the path to better security for all of us, not this fear mongering. For reference: Damien is an OpenSSH maintainer. 0205
Reposted by AndreaDamien Miller @damienmiller.bsky.social · 5hAnthropic may not have intended to write an excellent ad for GLM 5.3, but that's what they did. www.anthropic.com/research/glm... Unlike Mythos, I might actually have a chance at using GLM for defensive research. Anthropic didn't reply to any of my requests for Mythos access for use on OpenSSH.anthropic.comGLM-5.3 and the spread of advanced cyber capabilitiesGLM-5.3 can autonomously build end-to-end cyber exploits, but unlike other frontier models, it was released without meaningful safeguards to limit misuse. 3558
Reposted by Andrearev. howard arson @theophite.bsky.social · 5han aspiring art student could never have bad politics 825525
Reposted by AndreaOκULTRA XP🧿, writer for The Economist @0kultra.bsky.social · 5hlove that Anthropic is running ads for my bae GLM 1102
Reposted by AndreaQuantian @quantian.bsky.social · 5hLMFAO the official America.gov AI website uses a Chinese open source model 1526255
Andrea @valkyrie.hacker.gf · 5hThis is going to last until it tells him diseases are real and it's a bad idea to eat roadkill 050
Reposted by AndreaErica Windisch @ewindisch.ontological.observer · 8hMy former boss is currently at Anthropic and hiring. I assume publicly disagreeing with the org won't do me any favors... but neither will quietly allowing cybersecurity to be outlawed. Companies that can't secure their systems are lecturing us on cybersecurity and taking tools from defenders 2238
Reposted by AndreaHélianthe @helianthe.net · 5hplease take some deep breaths and meditate on www.anthropic.com/engineering/... while repeating the mantra "CompCert"anthropic.comBuilding a C compiler with a team of parallel ClaudesAnthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems. 021
Andrea @valkyrie.hacker.gf · 5hBut don't you understand that the entire edifice of AI fear-mongering rests on pretending nation-state capabilities aren't scary and must be preserved? 184
Reposted by AndreaHélianthe @helianthe.net · 7hI also heavily discount all AI safety prognostications which aren't accompanied by fanatical advocacy for formal verification and portable reproducibility 193
Reposted by AndreaHélianthe @helianthe.net · 7hcybersecurity is already a post-apocalyptic wasteland which is not going to be appreciably worsened by the release of a marginally better LLM nation states spend billions to attack everything already, it's just as plausible open weights models actually make things better by empowering defenders 2212
Reposted by AndreaKelly @justkelly.bsky.social · 6hThe Cornell victim has been doxxed on X. So worried for her. I hate this planet. 562763369
Andrea @valkyrie.hacker.gf · 6hWe're re-running the full disclosure debates of two decades ago but with crazed power cultists who spent that whole time reading Harry Potter fanfic 161
Reposted by AndreaJason, ex Inferis @benedictsred.bsky.social · 22/09/2026Living in Florida means hearing someone describe tahini as “sharia pudding” 512928
Reposted by AndreaVasabjit Banerjee @vasabjit.bsky.social · 7hY'all think it's funny until you meet a young staffer who says he doesn't read women's magazines like Janes for information on defense equipment. 35516
Andrea @valkyrie.hacker.gf · 7h"We pay $250M a year for SIGINT enabling, you can't release a model that just fixes it!" 1143
Reposted by Andreahikikomorphism @hikikomorphism.bsky.social · 15hThinking about this old quote in the context of certain people - professors, nyt contributors, etc - who see casual misgendering as a reasonable response to a lack of respect for their claimed epistemic authority 448796
Reposted by Andreacodehappy lives! @codehappy.bsky.social · 8hwording it this way instead of "I'm pretty convinced North Korea is going to take down the internet" is a choice 251
Andrea @valkyrie.hacker.gf · 8h"Why hello I have some fear to sell you about our competitors who you should definitely not use simply because they are cheaper and do stuff we won't" source: www.anthropic.com/research/glm... / archive.li/5kjru 272
Reposted by AndreaErica Windisch @ewindisch.ontological.observer · 8hI reported vulnerabilities to the Linux kernel and OpenZFS assisted by GLM-5.2. Anthropic's cyber program and refusal responses are detrimental to global information security. Defenders depend on GLM because they can't depend on Anthropic. 019236
Reposted by AndreaAlejandra Caraballo @esqueer.net · 19hJonathan McKinsey, a transgender New Times executive who was head of engineering at NYT games, was murdered by his in laws over the weekend. This follows years of allegations that the in laws were abusive to him over his gender identity. edition.cnn.com/2026/09/28/u...edition.cnn.comNew York Times executive killed by his elderly in-laws, police say | CNNA couple in their late 70s gunned down their son-in-law in a California city park over the weekend, authorities said, and court documents showed the victim had been in a bitter custody dispute with th... 14705236
Andrea @valkyrie.hacker.gf · 19hPretty bad that "Pentagon influencer" is even a thing you can be tbh 160
Reposted by Andreainternet freds 🦇☠️💾✨️ @esquiring.bsky.social · 29/09/2026the social contract in fact states you can either provide abundant public restrooms or you can have this. 03612
Reposted by AndreaThe Serfs (youtube.com/theserfstv) @theserfstv.bsky.social · 23hMichael Tracey is actively attempting to dox Jane Doe in the Cornell 7 rape case 2633141701
Reposted by AndreaMark Chadbourn @chadbourn.bsky.social · 28/09/2026Today, in Russian propaganda: 47936336
Reposted by Andreasasha @2damntrans.bsky.social · 22hI remember this guy. He has a history of citing Leviticus 20:13. That’s the Bible passage that calls for gay men to be “put to death.” Those are his views and we mustn’t discriminate against him for holding them, apparently. 13214
Reposted by Andrearoko’s modern basilisk @girlpython.northsky.social · 29/09/2026bionationalism or “human first” politics may be the largest danger of the 21st century 1202
Reposted by AndreaWilliam Gillis 🏴 @rechelon.bsky.social · 29/09/2026I don't really deserve credit for highlighting this in print eight years ago, because it's so fucking obvious to anyone and everyone looking that Yud has no fucking clue that there are inherent constraints. c4ss.org/content/51493 3359
Reposted by Andreagamera obscura @ohrobin.bsky.social · 28/09/2026Google earth has updated its satellite images of Gaza. The destruction is almost incomprehensible. 4421011210
Andrea @valkyrie.hacker.gf · 29/09/2026look, it's going to find the quantum cheat codes, invent time travel and then PSPACE = P relative to a retrocausal oracle 1295
Andrea @valkyrie.hacker.gf · 29/09/2026Some kinds of woo are purpose-built just world fallacy engines 282
Reposted by AndreaDamien Miller @damienmiller.bsky.social · 29/09/2026imagine being proud of one's PhD while simultaneously hating nerds 091
Andrea @valkyrie.hacker.gf · 28/09/2026Yes, sentence them to a century of chastity. ...wait, fuck. 010
Reposted by Andreaponder @ponder.ooo · 28/09/2026there were safetyists who thought internet search functionality was a hazardous thing to give to models of a scale that today we regard as pretty small, on the basis of technically-uninformed theories about ai as a god in a box 2222
Reposted by Andreaponder @ponder.ooo · 28/09/2026and for all rationalists' talk about "updating", what updating on the evidence does yud actually do? his theory is immune to revision in the face of new developments! at its core it never shifts, all that changes is how to argue for the same conclusions as ever 2163
Reposted by AndreaThe PS1 startup sound as a lesbian @janus.bsky.social · 28/09/2026idk when straight guys started feeling comfortable commenting on horny lesbian posts but i don't like it. fellas, we may have a shared interest but don't make me get the spray bottle 0512
