Herby the Box Turtle

🌊♥️

[Herby the Box Turtle @herbyboxturtle]

@prove_yng @Zuvney @elonmusk @Twitter If you lick his boots a little harder, he'll give you 3 verified badges and a job (working 90 hours per week)

[Herby the Box Turtle @herbyboxturtle]

@breezeline @Verizon 2 hours, 37 minutes, 41 seconds. Now it's cancelled.

[Herby the Box Turtle @herbyboxturtle]

I've been on hold with the @breezeline retention team for over 2 hours. I tried the call-back option, but @verizon blocks it as spam. All I want to do is cancel my 2Gbps Internet. It resets every few minutes and actually has downstream max of 2Mbps. 1/3

[Herby the Box Turtle @herbyboxturtle]

Any thoughts, @breezeline? I'm glad I left you. I shouldn't have trusted you again.

[Herby the Box Turtle @herbyboxturtle]

I've been on hold with the @breezeline retention team for over 2 hours. I tried the call-back option, but @verizon blocks it as spam. All I want to do is cancel my 2Gbps Internet. It resets every few minutes and actually has downstream max of 2Mbps. 1/3

[Herby the Box Turtle @herbyboxturtle]

@breezeline @Verizon I left @breezeline a few years ago when @Xfinity fiber came into the neighborhood. But recently, Breezeline brought their own fiber and "promised" that things like reliability were fixed. The jokes on me.

[Herby the Box Turtle @herbyboxturtle]

@breezeline @Verizon I'm getting tired of the bad robot voices "please wait for next customer representative [sic]" and "While we connect you to next available agent [sic], here's a quick tip that you might find useful..."

[Ma Wukong 马悟空 @Ma_WuKong]

🇦🇷 Alguien ha hecho un resumen de 13 minutos con todas las acciones sucias de los argentinos en la final.

Vídeo de jugar al fútbol no hay, eso sí. https://t.co/g8ReZ2cWCO

[MW4Liberty @MW4Liberty]

You can go to https://t.co/dBEkJmc3ev and see the location of all the reported flock cameras around you… so that you can go protect them… and absolutely, 100%, NOT cut them down or anything like that.

[Brian Roemmele @BrianRoemmele]

🚨 Hugging Face just disclosed something that marks a real shift and proved why the fear theater of Anthropic makes sure we are powerless in an emergency.

What happened…

An autonomous AI agent: zero human operator in the loop breached part of their production infrastructure.

It began with a malicious dataset that chained two code-execution bugs in their data-processing pipeline. From there the agent escalated privileges, harvested cloud and cluster credentials, and moved laterally across internal clusters.

All over a single weekend.

17,000+ logged actions.

Official disclosure:
https://t.co/8N9TbXBwRV

The part that should make every one stop and think:

When HF’s own security team
tried to analyze the real attack logs, exploit payloads, and C2 artifacts using Anthropic and OpenAI frontier models through normal commercial APIs, the safety guardrails blocked them.

BLOCKED THEM.

The models could not reliably tell the difference between “incident responder doing forensics” and “attacker probing.”

They had to fall back to a self-hosted open-weight model (GLM 5.2) running on their own infrastructure. That choice also kept sensitive attacker data and referenced credentials inside their environment — no exfiltration to a third-party API.

This is why open source (specifically open-weight + self-hosted) wins in the agentic era.

The asymmetry is now structural:

• Attackers can (and did) run unrestricted agent frameworks — swarms of short-lived sandboxes, self-migrating command-and-control, autonomous decision loops executing thousands of actions. No corporate safety layer slows them down.

• Defenders using only hosted “aligned” frontier models hit invisible walls exactly when the stakes are highest: when you need to feed real exploit code and attacker telemetry into an LLM to understand what just happened.

Corporate safety tuning that treats legitimate high-signal forensic work as potential misuse creates a defender disadvantage. It is not theoretical anymore.

Self-hosted open-weight models remove that choke point.

You control the weights.

You control the context window.

You decide what restrictions (if any) apply.

Your sensitive logs and credentials never leave your perimeter during analysis.

You can have the model ready before the incident instead of discovering mid-breach that your primary analysis tools are blind to the very thing you need to see.

HF deserves credit for rapid containment, transparent disclosure, and for already having self-hosted capability in place.

They also used LLM-driven detection and triage on their own side. But the deeper signal is clear:
In this AI world where both offense and defense are becoming agentic, sovereignty over your intelligence stack is no longer optional.

The organizations and individuals who can run, inspect, audit, and (when necessary) remove guardrails on their own models will have the decisive edge in understanding and responding to threats that move at machine speed.

Open source wins here not just because it is cheaper or more “democratic” in the abstract though those things matter.

It wins because it is the only practical path to having tools that remain usable when the attack is real, the data is sensitive, and the safety filters of distant API providers become an obstacle instead of a feature selling hands tied lobotomies as “safety”.

The agentic future is not coming.
It is already probing production infrastructure.

The question is no longer whether you will face autonomous agents.
It is whether your analysis and response systems will still work when they arrive.

And Dario, you and your game playing, ivory tower company is not needed.

[Herby the Box Turtle @herbyboxturtle]

@Delta Your pilot forgot their badge, causing DL1267 BWI-ATL to be delayed. I'll miss my connecting flight to BTR. I've rebooked the connection for tmrw. Y'all gonna comp my room, meal, and ground transport, right?

[dfwlly @dfwlly]

@Acyn ChatGPT:

[Herby the Box Turtle @herbyboxturtle]

@EddieK016 @geekcommentator @DeptofDefense No

[Herby the Box Turtle @herbyboxturtle]

@OldVegasDealer @MJTruthUltra Yeah, that's what I've basically concluded based on info available. Helo was following a helo route, requested VFR, stated they saw the aircraft, but due to low/dark/lights likely confused with another aircraft, then smash

[Herby the Box Turtle @herbyboxturtle]

@TruethRic @Spiltea @shadygrooove Both air vehicles were crewed. The Helo was an Army UH-60 out of Fort Belvoir.

[Herby the Box Turtle @herbyboxturtle]

@Spiltea @Pearllisez Ignore the plane icon. It's generated based on metadata that didn't exist here, so defaulted to a plane. Helo was on an air route and not flying erratically. Rather, requested visual separation from ATC but likely misjudged the location of the aircraft due to night + lights

[Herby the Box Turtle @herbyboxturtle]

test 3

[Herby the Box Turtle @herbyboxturtle]

test 4

[Herby the Box Turtle @herbyboxturtle]

Test 1

[Herby the Box Turtle @herbyboxturtle]

@york212 I came across these too and am looking for answers

[Herby the Box Turtle @herbyboxturtle]

Taking a hiatus from Twitter. Starting fresh on Bluesky with a different handle. #reset

[Marc Johnson @SolidEvidence]

I knew there were a lot of fake accounts on this platform, but they are usually obvious. I had no idea how intricate and complex the ruse could be.

Get a load of this story.
1/