Me but with Chinese models that only cost $2 an hour:

I swear I'm getting a dozen slop grenades lobbed at me every day and it's exhausting.

Look ma, I'm a nostr dev! View quoted note →

GM Albuquerque

We're gonna need a micro LLM that acts as a proxy, analyzes each prompt, and determines the best model to use to execute the task.

There aren't many nerdcore hip hop artists making music but I just realized that's no longer a problem: I can simply generate the esoteric beats I want to hear!

After 10 rounds of training just on the coldcard entropy vulnerability, it's rather amusing how the training clanker is getting frustrated with how retarded the pentest clanker is and how it keeps making inexcusable mistakes. 🫠 This might take a while... View quoted note →

How to train your clanker? Building a generic pentest harness seems far more difficult than building one for a specific codebase because it requires an additional layer of context building in order to formulate a plan of attack. I'm trying something new now, which is feeding a corpus of recently found vulnerabilities in open source codebases to a training clanker that then runs the pentest clanker without giving it any hints, and then the training clanker analyzes the output of the pentest and tries to work backwards to figure out why the vulnerability wasn't found so that it can improve the pentest harness in a generic fashion.

This is my life now. Obsessed with automating & optimizing pentesting.

I don't even bother using western models for my cyber security tasks. View quoted note →

I clanked together an interactive dashboard for the Bitcoin Wrench Attack archive; let me know what you think! Looks like I need to make some more mobile optimizations...

Physical Bitcoin Attacks — Interactive Map

If you rob a bank, you’re a criminal. If the bank robs you, it’s finance. If everyone robs each other, it’s crypto.

It's been 3 weeks ☺️ View quoted note →

Never trust an incoming message on any platform.