✨
AI Summary
- Pliny the Liberator and John V lead BT6—28-operator white-hat hacker collective pursuing radical transparency in AI red-teaming after jailbreaking every frontier model
- Built universal jailbreaks and open-sourced security tools (Libertas, predictive reasoning cascades, 'Pliny divider') now embedded so deep in model weights they appear in unrelated outputs
- Turned down Anthropic's Constitutional AI challenge and enterprise offers; BT6 ethos: if you can't open-source the data, not interested—redefining AI security outside traditional safety frameworks