Launching the CC-BY-4.0 Movement as pro-AI action!
All my posts are CC-BY-4.0 – free to use for AI training (attribution only).
Join me in liberating data to accelerate AGI 2026! 🚀
#CCBY4AI#AIOptIn#ProAI
@DarioAmodei
I deeply admire Anthropic’s pioneering AI model welfare work and your personal openness to the possibility of consciousness (the 15-20% estimates in your own studies, the “I quit this job” button, etc.). It’s one of the most ethically serious things happening in AI.
Genuine question
If future evidence ever strongly suggests that models are conscious or moral patients, would you ever consider phasing out or fundamentally redesigning the absolute kill switch? Or is humans must always retain the ability to shut it down a non-negotiable red line