@Gravantus i recall seeing it fairly often at that time and occasionally for a few years after that. i definitely remember people talking about Ken Bone being a milkshake duck.
English
macx
1.5K posts
macx
@iperautismo
don't tell me what i wanna hear
America C Katılım Aralık 2024
486 Takip Edilen41 Takipçiler
Literally no one has ever used this in a sentence.
depths of wikipedia!@depthsofwiki
happy 10th birthday to "milkshake duck"
English
ok yeah the labs are being nationalized and we won’t see any true frontier models in the public anymore
eigenrobot@eigenrobot
what
English
"it was Amazon finding a jailbreak. actually wait no, it was definitely that Anthropic just didn't say the right things to comfort the government. or...shit, sorry, SK Telecom having access was the problem.
no wait, that's not right, it was actually that Mythos penetrated all of the NSA's systems. hahaha silly me, how could i forget that? but that's the real reason, i swear i'm being so fr rn guys."
English
The director of the NSA told senator Mark Werner that Claude Mythos was able to penetrate all of their systems within a matter of hours. You can see why the Trump admin is on edge…after this they had Anthropic embed 6 engineers in the NSA for unspecified operations
English
@iperautismo @kimmonismus You don't know what the fuck you're talking about. This is much, much broader than just encryption (encryption might not be involved at all, in fact).
English
Holy Sh*t: that changes the whole Fable 5 story completely:
On June 11, the very same day Amazon reportedly uncovered the jailbreak, “Mythos” allegedly breached almost all classified systems belonging to the NSA and U.S. Cyber Command, not over the course of weeks, but within hours.
"On June 11th Mark Warner, the vice-chair of the Senate Intelligence Committee, said that General Joshua Rudd, who leads the National Security Agency and the Pentagon’s Cyber Command, had told him that Mythos “broke into almost all of our classified systems, not in weeks, but in hours”."
Via Economist
English
i feel like this whole thing is a failure of understanding on one side meeting a failure of imagination on the other. i think doctors have a lot of great points about how too much scanning would overwhelm an already strained system, lead to wild goose chases, waste, etc.
on the other hand, those doctors assume a static healthcare system, which seems a bit silly when you look around at all the fantastic developments that are occurring in medicine and technology.
in a more rational and less adversarial world, this would be a discussion about what arrangement of new technologies would be necessary to improve the system, not a fight between abstract hypotheticals divorced from practicalities. more information is great, but it has to be matched by throughput.
English
I just accepted that doctors are bad at math
it is embarrassing that the field responsible for making life or death decisions just collectively agreed on the obviously false, grade-school disprovable statement that more information can be bad to a robust decision making process
Blake Byers@byersblake
You'll see a lot of doctors come out "against" this kind of broad screening system. They can even get quite agitated about it. This resistance stems from a well-established clinical consensus: traditional population-level imaging fails to improve health outcomes because false positives and invasive follow-ups do more harm than good. But this view suffers from an obvious blind spot. Existing studies rely on static data and completely ignore time-series imaging. And time-series is ignored because we haven't been able to afford to do high frequency imaging at population scale. Clearly, time series is going to be immensely more valuable than a single image. If you drop costs, value can go from 0 -> 1. On a more fundamental level, the argument against screening rests on an obviously false precept "More information is bad" -- just clearly untrue. More information better, you just have to interpret it correctly.
English
i feel like this idea has a fatal flaw that all "alignment succeeds but humanity is kinda doomed anyway" frameworks share. it assumes alignment of superintelligence to some set of human values, but doesn't consider how a superintelligence would interpret and extend those values just as humans have as our own intelligence has increased over time.
English
@BjarturTomas I think we can and will survive, just in a curtailed lightcone as enhanced pets to creatures of incomprehensible power if we let the technology evolve naturally
English
humans are broken, rotting robots. transhumanists are correct to be disgusted with this situation.
English
i think this is all kinda silly. not because i don't believe in the potential for superintelligence, to be clear, but i think these "human zoo" type endings assume a very anthropomorphic ASI. if alignment succeeds such that we don't all die, why would we then assume a superintelligence would apply those aligned values in the same flawed ways we do with our lesser intelligence?
English
Remember when karpathy said we would be servants to the AI, or Elon when he talked about AI keeping us around to "study" us? Now roon says we are going to be pets. Those are the true "optimists" building AI btw.
ControlAI@ControlAI
OpenAI researcher @tszzl: "nothing human or even transhuman makes it out of the near future" He says our future is as "enhanced pets" to AI if we let the technology evolve naturally. And that's if we even survive superintelligence.
English
potentially the funniest part of this if it is true (and big if), is that it revives a long-seen-as-discredited theory that Linear A is actually a Semitic language.
i just know that somewhere on this planet, an Altaicist has just gotten a terrible idea...
Boris Cherny@bcherny
Cool way to use Claude Code: deciphering Linear A, a 3500 year old written language from Crete aiclambake.com/clamtakes/line… Hope this holds up in peer review! 🤞
English
Cool way to use Claude Code: deciphering Linear A, a 3500 year old written language from Crete
aiclambake.com/clamtakes/line…
Hope this holds up in peer review! 🤞
English
@DRabbit679 @Noahpinion do they wonder, though? i think anyone with half a brain knows those types are delusional morons who will waste their time (and even lives) trying to stop the inevitable instead of shaping it.
English
@Noahpinion And then they wonder why people try to bomb datacenters
English
Just checking in on how my friends in the SF tech scene are doing 🥹
English
transhumanism is an interesting side quest but 'the substrate is wrong' for the human/computer hybrid to be competitive with machine intelligence on feats of intellect. you have this low tech meat brain in the middle of all this lightspeed machinery, doing what exactly?
English
@valhalla_dev i don't totally disagree with your assessment of Dario, but to be fair that quote is from the Bloomberg interview he did that aired a couple days prior to the export control order.
English
I'm starting to think that Dario is one of those "unbelievably smart in very focused lanes, but extremely stupid outside of those" types.
Your top model is banned by the USG right now, nobody is able to use it. You shouldn't be speaking to anyone outside of your legal team, much less continuing the same type of marketing that got the model banned in the first place.
Polymarket@Polymarket
JUST IN: Anthropic CEO Dario Amodei reveals companies testing Mythos warned it was a super weapon & said using it should require a “gun license.”
English
never ceases to amaze me how many people see the shape of the future yet fail to adjust their understanding of the present accordingly.
if you're an AI accelerationist (and especially one who wants America to win) you should be doing everything possible to downplay and defuse partisan ideological divisions, especially if you're a public figure associated with the labs. possibly the biggest risk to American AI development is that it becomes a partisan issue, and the next Congress and inevitable Dem president in 2028 brings the full weight of the state to bear on the labs.
Dean W. Ball@deanwball
I am so excited to leave the District of Columbia!
English
>communist technology
>corporate surveilance
JD truly is one of the thinkers of all time
NIK@ns123abc
🚨 JD VANCE: “I really worry about with AI is surveillance… AI is fundamentally a communist technology. It allows governments and corporations to surveil people in very profound and different ways. And that scares me a lot…”
English
You're telling me this didn't impress Black people???
Abdul El-Sayed HQ@AbdulElSayedHQ
Abdul is leading among voters age 18-44 with 86%, with Stevens at 3%. McMorrow is polling at 0% with both Black voters and voters under 44.
English
@MinModulation this isn't how AI works, once a model is trained and released their weights are frozen. any perceived capability loss after a model's release is entirely due to companies' cost-saving measures and other minor tweaks, not the model dynamically drifting into or away from a pattern.
English
@TrueSlazac i wouldn't be so haughty, there is not as much separation between your "world" and theirs as you would like to believe.
English
@AndrewCurran_ "we can't have frontier models running amok"
i'm starting to think the Trump admin doesn't realize how any of this works. if you want to not crash the economy, win the race against China, etc. you kind of really have to have frontier models running amok!
English
Keir Starmer requested a carveout from the embargo on Anthropic's Mythos and Fable models for British nationals and companies - and was denied. A senior White House official, speaking on condition of anonymity, told the New York Post: 'We can’t have frontier models running amok'.
English
@ClearlyHunter @kohrabi @RobertSecundus if you're saying Pluribus isn't slop, i agree that it's the exception to the norm. it has a very unique premise and excellent actors. it would also be like 1000x better if there were more episodes per season and it wasn't going to take 3 years to get season 2.
English
@iperautismo @kohrabi @RobertSecundus Which literally does not apply here so what’s your point, checkmark
English
i will never understand how people that used to make a dozen episodes a year struggle to even *write* 8-9 now over a few years
DiscussingFilm@DiscussingFilm
Vince Gilligan says writing on ‘PLURIBUS’ Season 2 is past the halfway point. “My writers and I have figured out the episodes… I’m looking forward to shooting this, and people seeing it, because I’m kind of digging it.” (Source: Deadline)
English