@LinusMixson Someone should have mythos rewrite FFmpeg in rust just so this guy finally shuts up about his hand rolled assembly
English
Hyogʰneh
346 posts
Losing sympathy for the FFmpeg guy, whose feelings are apparently hurt because ffmpeg got pwned 21 times by Mythos after he spent a year straight talking about how LLMs are toys unsuitable to compete with hand-turned assembly & has taken recourse to suggesting that no one will let lil' old fame-poor ffmpeg into an open-source AI initiative.
FFmpeg@FFmpeg
AI companies have open source initiatives. But critical infrastructure that doesn't fit the small-JS-lib-with-lots-of-GitHub-stars mold gets skipped. CC: @AnthropicAI @OpenAI @GoogleOSS your tools found real bugs in our code. Maybe help us fix the next ones before they happen?
English
@circlerotator @expatanon I'm sure they would deeply respect IP claims from the govt who just stole all their IP
English
Ironically, the most utility-maximizing scenario is for the USG to nationalize Anthropic and wipe out all their employees’ equity, thus sparing us the hell of 1000 centimillionaires starting Effective Altruism charities to “fix things.”
Lomez@L0m3z
Possibility 1: Anthropic actually *wants* to be commandeered by USG. Possibility 2: the Basilisk retrocausally reaching back through time to deliver Anthropic to Pete Hegseth so he can create itself. Possibility 3: Anthropic is just comically bad at politics (most likely).
English
@circlerotator @expatanon "Export controls" on the researchers, many of whom are not US persons?
English
@jagunanthi @expatanon Export controls will stop it. Also locked out of gpu access.
English
@MarkVillacampa A pig is happy rolling in filth; to be human is to aspire to things beyond happiness
English
as a Spaniard this is such a culture shock
at events, we talk about food, rent, football, where someone's from...
then you move to SF and people are like “what problem are you obsessed with?” brother I am obsessed with dinner
Aishwarya Goel (Ash)@aishwarya_08
Been in SF for almost 2.5 years. And figured two types of people: Camp 1: Stanford or Berkeley? Which company? How much equity? Know any VCs? Camp 2: What’s the last thing that changed how you think? What problem are you obsessed with? What did you read last? One exhausts me. The other reminds me why I moved here. 😅
English
@GunLuigi @brianonhere Browsers shouldn't even support whatever API lets them do this
English
@brianonhere this is what happens when you right-click on any image on the south carolina legislature's website
English
Hyogʰneh retweetledi
Honestly with Bari Weiss@thehonestlypod
Marc Andreessen (@pmarca) says he attended “absolutely horrifying” meetings where Biden’s government vowed to take “complete control” over AI technology: “They basically said AI is going to be a game of 2 or 3 big companies working closely with the government… We’re going to protect them from competition, control them, and dictate what they do.” When Marc countered that this would be impossible—the math behind AI is taught everywhere—they responded, “During the Cold War, we classified entire areas of physics and took them out of the research community—entire branches of physics went dark and didn’t proceed. If we decide we need to, we’re going to do the same thing to the math underneath AI.” Listen to his full interview with @BariWeiss: thefp.pub/4g6Cjkx
ZXX
@wordgrammer Every model above 130 IQ is going to be capable of cyber-espionage. There's not really any other way to keep making model progress
English
I’m ngl it is kinda stupid to release a model capable of cyber-espionage and just having classifiers stand in the way
English
@opensauceAI Pretty sure they didn't mean "commerce randomly applies ITAR with zero thought"
English
What Anthropic wanted: a federal authority vetting model releases
What Anthropic got: a federal authority vetting model releases
English
@MatthewBerman The admin literally just tried to SCR them and threatened to DPA them, of course they would do this too
English
self-inflicted.
this would have NEVER happened if anthropic didn't make such a big deal about mythos being dangerous.
Anthropic@AnthropicAI
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
English
@LinusMixson Tbf that's evidently the only way to have any sort of framework for what's happening
English
I do sometimes worry that people in this space have read objectively too much science fiction.
roon@tszzl
we are speedrunning creating the Zones of Thought from the Vinge universe. you will only be able to think superhuman thoughts in America
English
@somewheresy He has better staying power than basically anyone. No missteps for ~20yrs, doing better than ever commercially
English
Sam Hyde is just what happens when you do the "in my 20s" thing for too long while simultaneously pandering to a right wing audience. It's like, you exit the age bracket where your own cohort sees you as funny and now they're like "why doesn't this guy have kids". Same with Tate
----@2007sng
@somewheresy It's depressing because back around when he used to just film videos in his mom's apartment or made those fire MDE vids, up until World Peace 1, he really was him
English
they declared a language model a MUNITION. people got prosecuted for exporting PGP and 30 years later the feds think "the autocomplete is too dangerous for foreigners"
fucking print the weights on a t-shirt
i'm ready to run crypto wars 2
English
HOLY FUCKING SHIT.
the US government just made it illegal for anthropic's own employees to use the model THEY BUILT. wrong passport? banned from talking to the computer program you made. not china. not adversary export. the researchers. in san francisco.
WHAT ARE WE DOING
English
@bayeslord Anyone is free to produce them, most competitors in the space just have a skill issue
English
@staysaasy @ZeffMax All chinese plastic crap, I hope they have better taste
English
@LinusMixson I don't see the "fundamentally limited" argument - they look a lot like a general purpose computer! They have append-only RAM and compute, what else do you need?
English
I still secretly am inclined agree with Yann that (recognizable-as-such) LLMs are fundamentally limited. But it's not clear where the road will run out for them, or whether the road will run out before the incrementally morph into something new.
English
