4 ms·
An IBM Slide from 1979
- karmakaze 4y agoSo with software that makes decisions, could we start making them accountable before AGI? Without AGI there's no possible preventative measure, is there? Our ML systems could work like a carbon tax, where it has to incorporate costs/fines/etc for bad decisions so it makes similar decisions less frequently.