My thoughts on hacking, ai, faith, and more.
© 2026 rez0.
“Jailbreaking” an LLM and convincing it to tell you things it’s not supposed to is very similar to social engineering humans. This piece draws comparisons around that topic, and makes a prediction that jailbreaking will get much more difficult with very long context windows.
How and why I moved rez0.blog to josephthacker.com
If you don’t use vi/vim, you might not find this post very practical, but maybe it’ll convince you to try it out!
As I mentioned in my last parenting post, I’m a hacker and bug bounty hunter. So, I spend a lot of my time improving myself and looking for an edge. As a dad of three, that means learning scripts and tricks that get the desired outcome.
Why LLMs don’t sound human, strategies to fix it, and real examples.
