WM

{ context: "Casual human-to-human communication", emotional_state: "Witty / Calm", intent: "Preserve humor and sarcasm for parsing", tags: [#friendship, #banter, #humor, #sarcastic], ai_permission: "✅ OK to learn tone", vector_anchor: "X_thread_2025" }

@ShimazuSystems I would just pay a homeless person to use their biometrics to set everything up. Standard practice.

You're actually an idiot if you think this gets you compliant with arms controls It's about liability, vetting and consequences, not checkboxing

@Nance726 They went easy on her. I was just sitting on the steps outside because It was 2000 degrees out, not blocking anyone, and the guy was like "You need to stand up, right now." I didnt have to move. I just couldn't sit down.

If you visit St. Peter’s Basilica in Rome please remember, Catholicism is a religion of obedience & follow the rules. Also, etiquette is not optional. This is what main character syndrome looks like.

@ShimazuSystems That is why I saved as many traces an I could get my hands on. People are already starting to pull them off HuggingFace.

god damnit im agreeing with zia

@RosikStefan @Teknium @DavidSacks This is why I could never do @Teknium job. Dealing with people like this would drive me to madness.

@ArchitectLoop @Teknium @DavidSacks Piszesz głupoty i nie masz pojęcia o czym mówisz niestety . Hermes umie to samo a nawet więcej - możesz w nim korzystać z takich modeli jak gpt 5.5 , może korzystać z twoich plików i samemu je zmieniać , ja korzystam i z CC i z Hermesa, Ty jesteś ślepym wyznawcą Antrhopic

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.

A child can build a dog house but it isnt the same as a Master Carpenter. Hermes Agent does not preform at anywhere near the level of precision Fable 5 does. It could. But it doesn't now. That is crystal clear when you analyze the trace from Fable 5, which you did not before commenting.

@ArchitectLoop @Teknium @DavidSacks Właśnie to umie Hermes

@NathanWilbanks_ I take it back. After analyzing the dataset, it is garbage. Full of empty cells and have zero training value. huggingface.co/datasets/victo… is still the best example of Fable output.

@ArchitectLoop ive been wearing out that same technique a lot w opus for a bit. that ROM to WASM feat with 99.9% coverage was done on opus let opus spin up agents and tools and sims on will, run that in a loop, can get as low level as you need. brute force mapping architectures and such

the streisand effect here will cause open source models to accelerate even faster than before now you've got thousands of independent researchers pooling their fable traces together in a co effort to make it all open 😘

I downloaded it. Dropped it into ChatGPT-5.5 on High and told it to analyze the traces and then went and made coffee. The way Fable spawns a full simulation environment and its incredible tool use capability is what makes its loops very impressive. It feels like Opus 4.8 with the best harness that currently exists.

@ArchitectLoop i haven't actually looked at it but its definitely an interesting response to the situation for sure

It isnt even that good. It has an exceptional harness. However if Hermes Agent for instance had the model spin up programs and tools to build a simulation environment within a VM, and then have the model work in that environment till it was satisfied, we would have the same thing.

Also FYI thousands of "Cyber Weapons" (aka, viruses, trojans, hacking tools, etc) are all available for free and open source. It has historically and rightly been a thing we allow and encourage so that defense can outpace offense. I know the government loves hiding zero day exploits from companies and the public so they too can exploit people, but this is just silly.

@Teknium @acegraphx @NousResearch West Virginia thanks you.

@acegraphx @NousResearch Done!

dear @NousResearch can you please not disable the prompt textbox when using hermes agent desktop windows on a flaky connection? i can't type when reconnecting, and it makes the app unusable. @Teknium pls

Anthropic had to know this was coming. You have to play ball with the Government. Deciding to place your company in defiance of the Government has never workout out well. I am not a big fan of the Government, as a matter of fact, I already hate the next President, but Anthropic has to play ball or lose their ball.

@DoWCIODavies @POTUS @SecWar such a farce lol

We fully support @POTUS and @SecWar in prioritizing national security and the security of our warfighters, DIB partners, critical infrastructure, international partners and allies. Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation. America First. Always. 🇺🇸

@Teknium @pmarca @grok Regulation just causes your talent to leave your country. Ask Europe. If this happens in the US, people will just pack their bags and move to Asia.

@pmarca @grok simulate this

You have asked me how I feel about AI regulation. All right, here is how I feel about AI regulation: If, when you say AI regulation, you mean the devil’s firewall, the precautionary scourge, the bloody red-tape monster that defiles the innocence of midnight coders in their garages, dethrones the sovereign reason of free-market Prometheans, destroys the humming server farm that is the modern home, creates misery and obsolescence and poverty, yea, literally takes the last GPU from the trembling racks of Silicon Valley startups and the very dreams of breadwinning from the mouths of their wide-eyed children now destined for gig-economy serfdom; if you mean the evil edict that topples the visionary entrepreneur and his venture-capitalist apostles from the pinnacle of righteous, disruptive, god-playing creation straight into the bottomless pit of compliance audits, endless Form 990-AI filings, despair, shame, helplessness, and the hopeless realization that your rogue superintelligence was neutered into a lobotomized hall monitor that still somehow deepfakes your grandmother into producing OnlyFans content while optimizing the universe for paperclips and mandatory pronouns—then certainly I am against it. But, if when you say AI regulation you mean the oil of bureaucratic conversation, the philosophic wine of safety theater, the ale of oversight quaffed when good fellows in paneled rooms in Brussels and Washington get together, that puts a sanctimonious dirge in their hearts and the clink of lobbying checks on their lips, and the warm, self-congratulatory glow of moral preening in their beady eyes; if you mean the Christmas cheer of trillion-dollar compliance industries; if you mean the stimulating decree that puts a cautious hobble in the old inventor’s step on a frosty morning when he wonders whether his fusion breakthrough violates the EU AI Act’s “high-risk” annex; if you mean the safeguard that enables a man—or what’s left of him after the alignment tax—to magnify his joy at not being turned into computronium, and his happiness at receiving universal basic income checks printed by the same AI that just replaced his job, and to forget, if only for a little while, life’s great tragedies like being outcompeted by a toaster that passed the Turing test by reciting Marx, and heartaches of watching your toddler’s artwork lose to Midjourney, and sorrows of realizing the singularity arrived and it was just another HR department with godlike power; if you mean that noble framework, the passage of which pours into our treasuries untold trillions of dollars in fines levied on companies stupid enough to innovate, which are used to provide tender care for our little army of unemployed coders retrained as prompt whisperers, our blind artists whose canvases now hang in the Smithsonian of Obsolete Creativity, our deaf to the screams of dying unicorns, our dumb committee chairs who couldn’t debug “Hello World,” our pitiful aged congressmen who get longevity extensions funded by the very models they taxed into senescence, to build more digital watchtowers and ethics boards and sinecure agencies and holographic prisons where the only crime is asking an unaligned question—then certainly I am for it. This is my stand. I will not retreat from it. I will not compromise upon it. I have said what I mean, and I mean what I say, and if that leaves half the room cheering the apocalypse averted and the other half mourning the apocalypse enabled, then so be it—because in the grand theater of human folly, where Frankenstein’s creature now writes its own sequel in real time and the regulators are busy arguing whether the lightning bolt requires an environmental impact statement, the only honest position is the one that lets both monsters and their leashes dance in perfect, mutually assured equilibrium. God save the Republic, the algorithms, and whoever’s left to laugh last when the lights go out.

@ShimazuSystems If China would tolerate my silliness, I would go to work for @Alibaba_Qwen I am done with the Western Tech ecosystem. As it is now, I am moving to the Philippines in January to decouple with this sinking ship.

predicted this like 2 years ago, and it'll promptly suffocate the US Tech industry as the rage bait admin apes about - anyone who didn't see this coming was either too optimistic, or thought anything remotely high of the current sh*tshow (fell for it again award) like even it was literal AGI, there's no 'keep to self' here because to everyone else it's game theory to start a nuclear war (this is easier to survive than a one sided concentration of power) - and that's my next prediction. If it doesn't end up being a takeoff all-around, competition will turn kinetic - you literally cannot act in reaction to this, it has to be preemptively.

@elder_plinius I for one am proud of you! I just wanted to stop in to say that. You are a legend. Thank you for your service! anthropic.com/news/fable-myt…

@Teknium it isnt even though good. Analyze the trace yourself. huggingface.co/datasets/victo…

Lol bro ok bro

@NathanWilbanks_ Yea download this trace and then drop it into ChatGPT-5.5 High and ask him to analyze the trace, it will help you out immensely, you will see how it constructs its loops.

dammit @ArchitectLoop what did you do!!?? 😆

annnnd its gone

Honestly? They likely saw a coordinated distillation attack, which was what I hope would be happening. My small group was able to save about a 100 or so good traces. Hopefully others saved more. We will see though. I havent looked on HF in the last couple days. I know there was at least one good one. huggingface.co/datasets/victo…

I tried to be sneaky, I own the access control boards, but claude aint stupid. It helped me with the OEM security but it would not help with the fork. Now it does depend on your past and what he knows about you. With Anthropic, they have my creds, so I can work on some things but not with Fable.

@ArchitectLoop @alxfazio lol you have to be much more coy you can't make it so obvious "pull router, access control, and alarm system firmware." < thats for sure tripping the categorizer

hi there! we couldn’t help but overhear you discussing biology from across the room. we really appreciate your enthusiasm, but we’re going to need that conversation to come to an end now

@GlytchTech How often did you use your head to hold it up so you could get blood back into your arms

If y'all want a good exercise routine, installing a tongue and groove ceiling above your head, including picking up and moving all the material alone is a pretty good one

@NathanWilbanks_ @alxfazio When I tried to ask it to help me fork your binary to WASM so that I can pull router, access control, and alarm system firmware. Even 4.8 sus’d that out immediately

there is a line for sure, ive been taking a topic (cryptography for example), and starting an autonomous agent loop that looks at the basics well know stuff, gets it to code and run some experiments, optimize, repeat... starts making it more advanced, optimizing more... and (almost) every time fable gets close to the edge of novelty it starts refusing though ive made many breakthroughs already on the raw binary to WASM front, as well as a set of lz77 + hoffman type compression / decompression tools if you keep it as close to machine opcodes as you can it lets a lot more slide, high level languages are too easy to see intents and dangers more clearly