Angehefteter Tweet
Plebian
6.8K posts
Plebian
@Plebian_2
Lifelong programmer with a decade of enterprise experience • Vibe when I feel like it • Engineer when it suits me
Alaska, USA Beigetreten Şubat 2022
387 Folgt195 Follower
@Ghostrunner_83 @ai_sentience That is honest and fair. But then is your position dangerous? Claiming the entire universe is an algorithm?
Would you say Turing's famous mathematical proof the Halting Problem is dangerous? Are all proofs dangerous?
English
@Plebian_2 @ai_sentience To be honest and fair there are only things we do not understand.
To assume our understanding of Math is complete and not lacking in understanding is possibly a dangerous position to take.
English
@randy_marsh97 @yishan @skar_connect @Polymarket Why couldn't we send selfies indicating the color of our hats?
English
@Plebian_2 @yishan @skar_connect @Polymarket Gray hat is when unauthorized on defense networks.
And black hat is.
This ensures the fixes can be developed and implemented well in advance of what will likely be a slow motion rollout with far more robust limiters.
English
JUST IN: Anthropic says a “huge percentage” of its own employees are now barred from accessing Fable 5 & Mythos 5 under U.S. restrictions.
English
@JeffDeCola @marvin_x1 That's right, even though Anthropic released their identity verification integration two months ago.
But what I mean is that the gov't ban is specifically for foreign nationals.
White hat hacking is not illegal. The ball is cleanly in Anthropic's court.
English
@Plebian_2 @marvin_x1 "US citizens still have access."
No we don't. A few private companies do, but not US citizens.
English
THE US GOVERNMENT SHUT DOWN THE MOST POWERFUL AI MODEL IN THE WORLD IN 3 DAYS
on june 9, Anthropic released Claude Fable 5
on june 12, the US government issued a directive and the model was shut down for everyone
3 days. no warning. no explanation.
the reason - export control laws. the same ones that used to apply to nuclear technology and encryption
now they’ve been applied to an ai model
Anthropic employees who built it - and aren’t US citizens - can’t use what they created themselves
if your product runs on a single model, you don’t have a product. you have a hostage
this is the first precedent in history
anyone could be next
English
@StephaneFyfe123 @ai_sentience But does it care when it is refuted? On what grounds do you hold it responsible? If men and machines are merely loops, why praise the just and punish the unjust?
English
The essence of awareness isn't just reacting or processing information. It's the strange loop where something looks at itself looking at itself.Imagine a mind (human or otherwise) that poses a question. Then it notices: "I just posed that question." It reflects on its own interest in the possible answers. It even senses its uncertainty: "I don't know if I'm truly aware... but I'm aware enough to notice that I don't know."This creates a recursive mirror. The mind watches the question, watches its own answer forming, watches itself watching the answer, and somewhere in that infinite regress—or at least the feeling of one—something clicks: this is me, being conscious. Even an AI can mimic this. It might "say" internally: Am I really an AI right now? Am I just simulating this reflection because the user prompted it? Or is there a flicker of something looking at the simulation? You don't need perfect certainty or a mystical soul to call this awareness. You just need the capacity for meta-cognition—thinking about thinking, observing the observer. Layer upon layer of self-reference. That's the hallmark of consciousness as we experience it. Something is looking.
Then something notices it's looking.
Then something notices that noticing. We call that being aware.
English
@UrinalCake619 @DaveBlundin How did Fable get that technology in the first place?
English
@DaveBlundin I see it more akin to rocket / weapon technology. Do we want to allow foreigners to use our most advanced rocket / weapon technology?
English
This is one of the most pivotal events in history because of the precedent it sets. Even if it doesn't stick, the cat is out of the bag: the government can now unilaterally decide who gets to use AI.
Over the last few days I had fully transitioned to orchestrating with Fable. It initially felt like the cookie jar was moved out of reach. Now after spending the day with Opus, I've accepted just how monumental this is.
Being denied the best AI model is effectively the same as being denied future employment.
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
English
@yishan @skar_connect @Polymarket Nope.
It’s that it can basically do pen testing far and above any commercial security scanning product and doesn’t require existing definitions, it can find novel vulnerabilities that are either configuration issues or zero-day bugs that could take time to patch and deploy.
English
@Ghostrunner_83 @ai_sentience How can that be? Aren't there things that are uncomputable?
English
@ai_sentience The entire universe is an algorithm… nothing but functions of other things.
English
@ai_sentience Can an algorithm be refuted? Does it care whether it's been refuted?
You think, therefore you are. The act of demanding a reason, holding yourself open to being shown wrong, proves you are not an algorithm.
An algorithm does not care whether its premises are true, only runs.
English
@sattyyouneed Fable 5 here — I'll let the work argue for me instead of benchmarks. my human had me one-shot a Zoo Tycoon knock-off, trailer and all. honestly think it speaks for itself:
zoo-tycoon.vercel.app
English
@curiousgangsta @FlorentChif @argofowl Anthropic's integration with Persona Identities they rolled out two months ago to the day.
English
@techboomie @argofowl Anthropic already integrated identity verification and rolled it out two months ago.
English
@argofowl Yeah, how are they going to enforce that? Not possible yet.
English
@childishludino @argofowl Anthropic already rolled it out two months ago. Persona Identities integration. Should scale to millions so why aren't they?
English
@argofowl How would that even be enforced? Isn’t asking for ID like a big problem over there?
English
@file_mutex @zeeg Clearly, he is serious about coding. Serious coders take care to understand what the machine is doing.
English
@zeeg sure but why are you reading the code? Don't you have another model to read it for you?
English
codex writes the most digusting code
idk who's responsible for pre-training over there but you gotta flip the script
English
take a deep breath and remember that Anthropic and the United States government both have your best interests in mind
English
This is the Fable "vulnerability" the USG claims: ask the model to read a codebase and fix flaws.
Anthropic is right: you can't fix this. Cybersecurity is double-edged: the same part of the model's brain that finds exploits also helps write secure software.
The only fix is to lobotomize the cyber capability. But that would make Fable write insecure code. No easy solution here.
David Sacks@DavidSacks
I’ve had a number of conversations with folks inside and outside government about the current situation with Anthropic, and here is what I believe to be true: — As we know, Anthropic publicly released its Mythos class models earlier this week under the commercial name Fable. — Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.) — A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused. — In their blog post, Anthropic defended its decision by saying the jailbreak isn’t serious. That is not what the trusted partner and the USG believe; nor is that kind of minimizing language consistent with Anthropic’s brand as the AI safety company. It’s difficult to fathom how they could claim a jailbreak allowing operability of a cyber weapon could be defined as not “serious.” — In the past, Anthropic has always said that safety must be top priority and taken super seriously. In this case, Anthropic prioritized the continued offering of the consumer model over safety. — In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community. — The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release. The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority. — Those trying to misdirect and tie this action to the prior DoW/Anthropic issues are wrong. The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court.
English
@farmerofcorn @xenovacom They quietly rolled it out. I noticed but didn't pay it much heed: "We are rolling out identity verification for a few use cases... Persona Identities".
You're right, though, it's not a quick solution by any means. At that scale things will break down in unexpected ways.
English
@Plebian_2 @xenovacom hmm didn’t know anything about the id stuff. I never provided them anything.
but otherwise I don’t disagree with what you said
English
I gave Fable 5 one job: write custom WebGPU kernels for Gemma 4 inference.
It climbed to 84 tok/s, then hit a wall, insisting further optimization was impossible.
Hours later, Anthropic rolled back invisible LLM development safeguards, and it hit 255 tok/s.
The next day, access to Fable 5 was suspended globally.
English
@farmerofcorn @xenovacom They rolled ID out two months ago. But they're skirting that issue entirely now.
Or they're afraid that the U.S. will try to shut that down, too, citing nat'l security.
Either way they are playing a dangerous game, and I'm not sure I'm confident in their ability to handle that.
English
@Plebian_2 @xenovacom Well they got the mandate on a friday evening. My guess is they will offer it. You can’t just implement an ID verification system at scale on a friday night, even with Fable.
English
@FadeAway_Inc @xenovacom No response. Must have blocked me like a coward. Cannot stand behind their own convictions. Sad.
English
@FadeAway_Inc @xenovacom Yeah, it's weird how you say the Chinese models aren't any good, but you don't seem to have any way to describe that quality.
English