AI on the Brink: Risk evaluations and the nuclear safety blueprint.

Dr Mark van Rijmenam, CSP -

AI experts call for early risk evaluation in AI development, akin to nuclear safety protocols. They suggest identifying 'extreme' risks and, if too high, halting development until mitigated. However, the lack of systematic risk assessment and understanding of AI algorithms' workings could lead to larger unforeseen harms.

Read the full article on MIT Technology Review.

----