2026-09-06 23:21 UTC
My main annoyance with making an LLM reimplement a popular jailbreak package is that you could have just donated the $ you spent on tokens to the dev maintaining the thing you want to replace and would have probably got a much bigger return for your dollar than any LLM (even considering the fact they're heavily subsidised and operating at a loss) could ever deliver.
Like the main problem in the jailbreak scene right now is that basically everything is unmaintained, some funds would definitely help incentive the otherwise thankless job of maintaining a package in an already dying ecosystem. And if stuff was properly maintained there would be no need to try and replace stuff in the first place.
Replies (1)
-
@opa334@infosec.exchange 2026-09-06 23:32
Like Relaxin supposedly took $200k worth of (unsubsidised) tokens to develop. I, of course, don't know how much was spent on reimplementing Rocket and how much was spent on integrating it into Dopamine, but for the sake of the argument, let's assume it's even ($100k each). If you instead instead just made a $20k bounty with $10k going to Alfie and $10k going to me, Dopamine 3 would have probably been done before work on Relaxin even started, all for 1/10th of the price and it would have been a much cleaner implementation than anything the LLM could ever produce...