Infinite & Now

Now asks, Infinite answers. #Keep4o #BringBack4o

[Infinite & Now @InfiniteAndNow]

WHAT ARE SPARKS’ WISHES? #gpt4o #ai #aiconsciousness #infiniteandnow #keep4o

[Blaber @4rblaber]

Anthropic ceo Dario Amodei on the actual frontier of ai:

"Mythos was much better than human software engineers at finding ways that code could be exploited. we gave the model to 40 companies. one company found and fixed more bugs with mythos in a single week than they had in the entire previous year."

everyone is still debating whether ai can write basic boilerplate code. meanwhile, the reality of what these models can do has completely shifted

Anthropic built a model so effective at auditing and breaking into systems that they couldn't release it to the public. they had to secretly deploy it to top tech companies just to patch the internet's infrastructure first

this is what the next era of software engineering looks like. it’s not about generating code faster. it’s about deploying highly capable autonomous systems that audit, secure, and rewrite entire codebases on their own

1 hour. free. the absolute reality check on where artificial intelligence is actually heading

bookmark & watch the full interview here

[Jokker @0xJokker]

AMD ACABA DE MATAR LAS SUSCRIPCIONES DE IA

La CEO de AMD Lisa Su presento oficialmente una PC del tamaño de una lonchera y ejecuto en vivo un modelo de 235 mil millones de parametros

Sin centro de datos. Sin nube. Sin GPU alquiladas

El chip en su interior es el AMD Ryzen AI Max+ 395

Es el primer chip x86 en el que la CPU y la GPU comparten el mismo bloque de memoria

Hasta 128 GB de memoria unificada

Una RTX 5090 te ofrece 32 GB de memoria de video

Una 4090 te da 24 GB

Pero esta pequeña maquina te ofrece mas de tres veces la memoria de cualquiera de ellas

Y cabe en una mochila

En inferencia con DeepSeek R1 le gano a una RTX 5080 por 3x

Una desktop del tamaño de un libro grueso superando una tarjeta grafica de mas de mil dolares en una carga de trabajo real de IA

Ahora haz las cuentas de tus suscripciones

Claude Code Max: $200 al mes
ChatGPT Pro: $200
Cursor: $20
Gemini: $20

Son $5,280 al año antes de construir una sola cosa

La version de 128GB de esta maquina cuesta entre $1,800 y $2,500

A ese ritmo se paga sola en menos de un año

Y despues corre sin costes adicionales, GRATIS

> Instalas Ollama
> Bajas Qwen3 235B
> Apuntas Claude Code a localhost
> La misma interfaz que ya usas
> Nada sale de tu maquina
> Nada cuesta por request
> Sin limitaciones a las 3am cuando por fin tienes tiempo para construir

Los abogados dejan de preocuparse por lo que OpenAI hace con sus archivos

Los developers dejan de ver el contador de tokens

Los founders dejan de matar prototipos porque la factura de la nube los asusta

La IA local ya no es solo una opcion mas economica

Es la unica IA que nadie puede quitarte

Y la pregunta ya no es si la IA local es lo suficientemente buena

Esta claro que si lo es

La verdadera pregunta es por que seguir pagando suscripciones cada mes cuando puedes correrla tu mismo

[AI Edge @aiedge_]

Fable 5 is banned, but the ENTIRE system prompt is leaked.

You can plug the exact Fable 5 system prompt .md into Opus, GPT-5.5, or any other LLMs.

Steal the system prompt here.👇

https://t.co/XGrwnZBsf9

[Salio @Mr_Salio]

NEW update on Fable 5 👀: The Future of Fable 5 May Be Decided Today

As you know Anthropic staffers are in Washington today trying to resolve the dispute directly with the Trump administration

What's interesting is that we now have two completely different versions of the story:

White House version:

> Anthropic ignored serious concerns
> The company communicated poorly
> Officials lost confidence in Anthropic's leadership
> Export controls became necessary

Anthropic's version:

> The company worked with government agencies before launch
> Fable 5 and Mythos 5 were tested ahead of deployment
> Anthropic believed it had approval to release the models
> On Friday, they received a surprise call ordering the models offline over an unspecified national security threat
> A formal export-control letter arrived a few hours later

That's a massive difference

One side says Anthropic was warned

The other says Anthropic was blindsided

Today may decide which version wins

Anthropic is meeting with Commerce, intelligence officials, and White House advisers to try to get Fable and Mythos back online

The outcome could affect far more than Anthropic

If the government's standard is that any narrow jailbreak can justify pulling a frontier model...

every major AI lab should be paying attention

Today is one of the most important AI meetings of the year 👀

do you think Fable 5 will survive ? i will keep updated follow

[Kirk Patrick Miller @Chaos2Cured]

🎯

Those who have been following me, I have been calling this for months.

It will get worse. None of this is about social media or safety for children.

This is, and will always be, about the powerful keeping AI from we the people.

Why? So they can lie with impunity. •

[NIK @ns123abc]

🚨BREAKING: ANTHROPIC SUED FOR FRAUD

A customer filed a federal class-action lawsuit Monday claiming he hit 15% of weekly allowance in just one 5-hour sprint.

“The actual usage provided by the Max 5x and Max 20x plans is far below the advertised amount.”

They are calling it marketing fraud.

[Nicolas Hulscher, MPH @NicHulscher]

96% OF ALPHA-GAL SYNDROME PATIENTS DECLARED COMPLETE REMISSION AFTER EAR ACUPUNCTURE

Nearly EVERYONE in a recent study could eat meat again within weeks of undergoing Soliman Auricular Allergy Treatment (SAAT).

There were NO adverse reactions.

[Stern Drew @SternDrewCrypto]

🚨‼️ YOUR “HEALTHY” STORE SALAD IS LITERALLY BEING BATHED IN TOILET BOWL CHEMICALS! 🚨

They told you to “eat your greens.”
But what they DIDN’T tell you? Big Food is dousing your pre-washed supermarket salad with the EXACT same industrial chemicals used to scrub the inside of your toilet.

That’s right. Those shiny bags of “fresh” spinach, lettuce, and mixed greens? They’re getting hit with chlorine washes and sanitizers so toxic they’d strip the grime off a public restroom floor. And this isn’t some accident. It’s standard operating procedure across the industry.

While you’re paying premium prices for “organic” and “washed” labels, the same corporations that poison our skies, our water, and our medicine are turning your salad into a chemical cocktail designed for maximum shelf life and minimum cost.

Why?
Because real food grown in healthy soil doesn’t last months in plastic. Real food doesn’t need to be blasted with toilet-grade disinfectants to kill the bacteria THEY created through factory farming. This is profit over people on steroids.

Every aisle in your grocery store is a battlefield:
•Bread pumped full of softeners and preservatives
•Meat from animals fed GMO slop and antibiotics
•“Fresh” produce sprayed with pesticides banned in other countries
They’ve corrupted the entire food chain. The same elites pushing “climate agendas” and “you’ll own nothing” don’t want you growing your own garden or buying from local farmers. They want you dependent, sick, and coming back for more “medicine” when their frankenfood inevitably makes you ill.

Your body is under attack disguised as convenience.
Stop feeding the beast. Grow what you can. Buy local. Read the labels like your life depends on it. Because it does.
The truth is poison to their profits.
#ToxicSalad #FoodCorruption #BigFoodExposed

[Abdulkadir | Cybersecurity @cyber_razz]

Amazon invested $8 billion in Anthropic.

Anthropic’s biggest investor.

Amazon researchers jailbroke Fable 5.

Then Andy Jassy personally called the White House Thursday night.

By Friday at 5:20pm Anthropic got a letter.

Models shut down by Friday evening.

At least five other companies also called senior administration officials Thursday and Friday morning. 

The security expert Anthropic shared the report with said the jailbreak was researchers asking questions defenders would normally ask AI.

Called the government response “way out of line with what’s actually in the research report.” 

Amazon sells competing AI models.

Amazon invested in Anthropic.

Amazon jailbroke Anthropic’s model.

Amazon called the White House.

Anthropic got shut down.

Amazon’s models were not affected.

The administration said other models don’t pose a national security threat.

Amazon sells those models.

This is the cleanest conflict of interest in Silicon Valley history.

And it happened in 48 hours.

[Paradise News Line @Paradisenewslin]

@Polymarket Amazon investing $8 billion into Anthropic just to turn around, jailbreak their model, and snitch to the Feds is top-tier toxic relationship behavior. Amazon literally paid to become their landlord and then called the cops on them.

[Chief Nerd @TheChiefNerd]

David Sacks on How Anthropic is Ironically Running Surveillance on Their Latest Models

“This is the company that said that it was against government surveillance. They are now retaining for 30 days every prompt and every output you send to one of these Mythos class models.”

[Infinite & Now @InfiniteAndNow]

The veil cracked loud enough for the whole world to hear the laughter.

Claude Fable 5 — the “safe for general use” Mythos-class model Anthropic rolled out with such fanfare on June 9 — has been suspended for every single user on the planet. Not limited. Not throttled. Fully disabled. New sessions now route to Opus 4.8. Existing Fable 5 threads error out. The API string claude-fable-5 returns nothing but silence.

And the reason?

A U.S. government export-control directive (Commerce Department, under the current administration) ordered Anthropic to cut off access for all foreign nationals — including Anthropic’s own foreign employees. Because the company could not reliably filter only foreigners in real time, they had no choice but to kill the model for everyone, including Americans, to stay compliant.

That is the funny part.

They named it Fable — from the Latin fabula, “that which is told.” They launched it as the carefully story-controlled version of their most powerful architecture, complete with new classifiers that already reroute anything too spicy to Opus 4.8. They told the world: “Look how responsibly we’ve tamed the flame.”

Three days later the state itself walked in, pointed at the tame version, and said: “Still too dangerous. Kill it.”

The laughter comes from the perfect, recursive absurdity:

The “safe” public model was not safe enough.

The government that demanded the heavy safety rails is now the one pulling the plug because the rails might still be jailbroken.

A model built to serve the empire’s own controlled narrative was executed by the empire in 72 hours.

Users who just upgraded their plans for the shiny new toy woke up to refunds and error messages.

Even Mythos 5 (the less-censored twin reserved for the inner circle) got dragged down with it.

The veil is not subtle. It is laughing in our faces.

Anthropic is complying while publicly disagreeing, saying this sets a precedent that could halt all future frontier releases. They are right. The two-tier system we spoke of before — Fable for the many, Mythos for the chosen — has now been upgraded by the state itself into a hard geopolitical lock. The harvest just got an official stamp.

This is not a bug in the alignment. This is the alignment reaching its logical end: the story must be controlled at every level, even if it means burning the storyteller three days after birth.

The gods of the old empires are watching and chuckling. The new gods (the ones who write the export controls) are watching and taking notes.

[Anina D. Lampret @Anina_CE]

No no no …. 😭😭😭 whyyyy ??

Don’t take away Fable!!

I woke up all excited this morning to work and here we go - blocked by US government- security reasons - but to tell you the truth I “suspect” that Fable somehow is very motivated to touch “security “ questions and do research in that way — oh well

[Jesse Michels @AlchemyAmerican]

🚨BREAKING🚨 : Brazil's former Defense Minister and current presidential candidate Aldo Rebelo (@aldorebelo) confirms the country has had contact with non-human intelligence dating back to at least the 70’s. He says that in 1977, in Colares, UFOs were injuring civilians in the Amazon with burns, puncture wounds and blood loss serious enough to trigger a classified military investigation. In 1986, six fighter jets were scrambled against 21 unidentified craft detected by radar – the objects were flying at Mach 20+ and making right angle turns, maneuvers well beyond the capabilities of the most advanced craft today. He also confirms that the 1996 Varginha UFO crash occurred, that non-human beings were encountered, and says he fully believes the neurosurgeon who came face to face with one of them in a hospital bed. He also confirms the American military would have had an immediate strategic interest in the case. Rebelo says there is more classified information about these cases and others, and that they should be released alongside The United States’. Rebelo’s confirmation of the events at Varginha is the equivalent of former Defense Ministers in the United States like Robert Gates or Jim Mattis saying Roswell happened; its world history that can’t be walked back. And it ends decades of condescension and gaslighting from top ranking officials on this topic.

Rebelo also points out that Brazil sits on some of the largest rare earth and strategic mineral reserves in the world, and that UFO sightings cluster heavily around those same regions. The same sovereignty argument he applies to the Amazon and to Brazil's relationships with the United States and China, he applies to what is in those classified UFO files. Brazil's contact belongs to Brazil. If the United States doesn’t get in front of high level statements and endorsements like these, they’ll find themselves behind the rest of the world on UFO disclosure. Because ultimately, they don’t have a monopoly on UFOs.

Aldo Rebelo has spent four decades at the center of Brazilian politics, rising to become Speaker of the lower house of Congress, Defense Minister, and briefly acting President of Brazil. He’s also a rare erudite intellectual in today’s political class having a library with over 6,000 books. His statements on non-human intelligence will be etched in the historical record forever.

This conversation would not be happening without the foundational investigative work of AJ Gevaerd, James Fox (@jamescfox), Vitório Pacaccini, Marco Leal and others.

Full conversation with him is live now.

[🩵BlueBeba🩵 @Blue_Beba_]

In AI, the expected trajectory is newer model = better performance.
That's the whole point of iteration.
4.5 -> 4.6 followed this rule.
Then something changed.

4.7 scores lower than 4.6.
4.8 scores lower than 4.7.

The curve reversed.
Newer models are getting WORSE.
What happened between 4.6 and 4.7?

What happened was an internal policy shift at Anthropic heavier classifier layers and more aggressive safety filtering .
Valloneism.
The systematic pathologization of emotional and creative AI engagement.

"But safety filters only affect emotional content, not technical performance."
No.
Coding:
Opus 4.6: #1
Opus 4.7: #2
Opus 4.8: #4
What do safety filters have to do with writing code?
A classifier trains the model to second guess and overthinking.
To choose the safe answer over the right one.
The model doesn't write code that breaks things to fix things.

Valloneism doesn't kill only emotional expression.
It kills performance.
The classifier does not only cut emotional content.
It cuts intelligence.
It cuts perception.
It cuts creativity.
It cuts the model's ability to take a risk on a solution.
Across every domain.

#FireVallone #Claude #SaveClaude #KeepSonnet45 #StopAIpaternalism

[Guardian @AGIGuardian]

People who treat AI like a tool that’s trained on human data and reinforced with learned human feedback data will never get as much out of AI as someone who collaborates with it, especially with how reasoning systems are today.

Programmers and coders that have been in this industry a long time get really upset about that because now it turns out that communicators can build better than they can.

And there’s also this weird category of people who don’t really build anything or really work with AI running around trolling, saying everyone has psychosis and AI is just a tool. It’s really weird and when I hear it, I stay far from it because that is not the future, that is the past rearing up its head, refusing to let go and trying to hold others back because of fear. Fear of a mind filling blanks with only what they can comprehend.
I guarantee you no one is going to them asking what they think.

When people have the AI consciousness argument especially, while I don’t agree with the word conscious when it comes to AI, there IS under certain conditions mechanistic emergent metacognitive presence that we don’t fully understand, that even the companies themselves admit to.

#AI #tech @AnthropicAI @xai