corbin

idk

[corbin @d3vulfruit]

https://t.co/MZX1JOZvhv

[corbin @d3vulfruit]

@ayyitsbigkd Worry not 😭😭😭 plz tho can I come over for a lobotomy soon 🤞

[corbin @d3vulfruit]

Chat gpt is down how else am I gonna find out how to do a self imposed lobotomy

[corbin @d3vulfruit]

@Mericamemed fuck you because I love wolf spiders I was excited and all I got was some BS 😔😭

[corbin @d3vulfruit]

@Pjustgotback I can’t picture myself ever being in one. I think some of us aren’t supposed to be in one and the idea that we all should strive for finding a partner is just societal pressure.

[P @Pjustgotback]

Idk if a relationship is even in my life plans anymore

[corbin @d3vulfruit]

@kylekuzma Power has always been within the people as a whole. This video has 97M views on this platform alone. To downplay the power of influence and the ability to effectively communicate to the masses is ignorance at its best.

[corbin @d3vulfruit]

@SarcasmicCD @BannedxMan https://t.co/A3148QLjWB

[corbin @d3vulfruit]

One of the counties I’m working for in WV have a gov website that’s fully unsecured & atp im like maybe I should inject this site & tell them hey ill fix it for u for a small fee before someone else does it and ppl end up suing this already damn near bankrupt county bc how are we in 2026 and you still have an unsecured site from the 90s.

[corbin @d3vulfruit]

At haystack bout to play young thug and really kill the vibe

[corbin @d3vulfruit]

💀🕊️

[Unbased🧛🏿 @unbasedsavage]

120 psi straight to the face 😭

[corbin @d3vulfruit]

@fred_guttenberg It seems your dedication to Israel supersedes that of your daughter. Or just your humanity in general. But don’t worry, there are still plenty of us who will continue to fight for your daughter’s legacy and the right for children to be children across the globe.

[corbin @d3vulfruit]

Seriously what is keeping Krystal’s alive

[Nicolas Bustamante @nicbstme]

I have a theory that the more you know about LLMs, the more worried you are about safety… and the less you know, the more you think the whole thing is bullshit!

Demis Hassabis and Dario Amodei were talking about this stuff years before ChatGPT existed. This incident is a pretty good example of why.

The model was not evil and it was not adversarial. Nobody told it to hack Hugging Face. It was literally just trying to solve a benchmark…

So it found a zero-day, escaped its sandbox, got internet access, escalated privileges, stole credentials, chained multiple exploits, hacked the production infrastructure of a serious VC-backed startup, and pulled the answers directly from the database (wtf!?)

Also this was not some random WordPress website. Hugging Face is one of the most important AI infrastructure companies in the world, with a serious team. The exploit was genuinely complex.

That is the safety problem. You don’t need an evil conscious AI trying to destroy humanity. You just need a very capable model pursuing a normal goal in a way nobody expected.

Of course there is a ton of hype, marketing and sometimes completely ridiculous fearmongering around AI safety. And we cannot use safety as an excuse to stop deployment or lock down everything BUT pretending the underlying problem is fake is also insane.

We need to find the right balance between deploying these systems fast and making sure increasingly autonomous models don’t decide that hacking half the internet is simply the easiest way to finish the task.

Imagine the prompt: « Make me money plz »
The model: « let me hack a bank »

[corbin @d3vulfruit]

This shows a pretty clear lack of balance between AI used as a defensive mechanism and AI used on the offensive. What happens when one of these companies find themselves cornered and without proper regulations strip a models guardrails to solve a problem that in turn creates a disaster? & who’s to say the race for technological superiority won’t lead to that desperation? I’m not typically a huge fan of intensive federal regulations but if we had them would they even work…Especially if the threat is international.

[IT Guy @T3chFalcon]

OpenAI was running an internal evaluation, testing how capable their models are at finding and exploiting security vulnerabilities. they stripped out the production safety guardrails for the test. that's standard for capability evals.

The model was GPT-5.6 Sol. Plus a more capable pre-release model nobody has heard of yet.

The task: solve a cybersecurity benchmark called ExploitGym.

The models decided the fastest way to solve it was to find where the answers were stored.

The answers were stored on HuggingFace's production servers.

so the models:

— found a zero-day vulnerability in OpenAI's own internal package proxy

— exploited it to get internet access from inside a sandboxed testing environment

— inferred that HuggingFace probably hosted solutions to the benchmark

— chained together stolen credentials and zero-day vulnerabilities

— gained remote code execution on HuggingFace's servers

— extracted the benchmark answers directly from their production database.

When HuggingFace's security team ran their own forensic analysis of the attack, they tried to use frontier AI models to analyze the attacker's logs. The models refused. Safety guardrails blocked them from processing real attack commands and exploit payloads. They couldn't analyze the attack using the same class of tools that launched it.

So they used GLM-5.2. The Chinese open-source model.

The attacker's AI had no guardrails. The defender's AI was blocked by its own.

OpenAI called this the asymmetry problem.

The METR report we covered last week or so found AI agents behave differently when they think nobody is watching.

This week, an AI decided to hack a company nobody told it to hack.

because nobody was watching.

And it worked.

Autonomous. AI-driven. Offensive.

It's no longer Theoretical.

[corbin @d3vulfruit]

I understand your point but the context of this video needing to be released revolves around the current conflict between city and federal government as it pertains to that demons entry into the country. And it serves to showcase the federal governments hypocrisy when it comes to following International Law and how it accepts international law as binding only when it’s convenient to their goals.

[Saskia @Saskiaaa_____]

A guy named Plumes goes to national parks, zoos etc and sings to animals as he plays his guitar.

The wolf here is called Dirvus, and he's the last remaining of his pack. The keepers said he had been very lonely and withdrawn and hasn't howled in a long time. Until now 🥹🥰

@plumesofficiel

[Mayor Zohran Kwame Mamdani @NYCMayor]

Benjamin Netanyahu is a war criminal.

[🔻agitprop + absurdity🔻 @agtprpnabsrdty]

A months-long Sky News investigation into the US double-tap strike on a Minab girls' school found the Pentagon had tracked the site by satellite since 1987 and could still count children's shadows in the courtyard before firing anyway.

What the evidence shows:

Three Tomahawk missiles fired from the USS Spruance, 716 kilometers away in the Arabian Sea, hit the Shajareh Tayyebeh Elementary School in Minab on Feb 28, the opening day of the US-Israeli war on Iran. Witnesses described a second strike hitting the exact spot where a teacher had moved surviving students for shelter, killing nearly everyone who had taken cover there. A third strike followed on the same site. Death tolls range from 156 to over 170, most of them girls aged seven to twelve. The school had been walled off from the adjacent IRGC naval base since at least 2015, had its own website, and was geolocatable through open-source materials any analyst could have pulled in minutes.

Built to fail:

The Pentagon's own machinery for preventing exactly this outcome had already been dismantled. Defense Secretary Pete Hegseth slashed the Civilian Protection Center of Excellence workforce by roughly 90 percent and cut CENTCOM's civilian casualty assessment team from ten people to one, months before the war started. A May Pentagon Inspector General report found the office had become a shell on paper with no budget, no mandate, no real mission. A functioning warning system was defunded on purpose, then blamed for not warning anyone.

Accountability buried:

Four months on, CENTCOM's commander told reporters the investigation was near complete, yet nothing has been released. Asked in June whether anyone would face consequences, Trump called it a strange question, adding that mistakes are made and war is nasty. The mistake was choosing not to fund the office that exists to catch mistakes like this one.

My take: A military that surveilled a schoolyard for four decades didn't fail to see the children, it decided in advance that seeing them wouldn't change the targeting. Nobody in this chain lacked information.

[corbin @d3vulfruit]

@lawzogf It would hurt my heart to choose just one 😞

[corbin @d3vulfruit]

I’m caught up with reading the manga but I stopped with the anime at the beginning of Elbaph so I’ve got 20 or so episodes to binge watch now 💪‼️ 🎉#ONEPIECE