The real risk in playing it safe

Breakthroughs rarely emerge from cautious committees. For agencies testing AI, alignment, autonomy and acceptance of failure outweigh top-down declarations.

by Sameer Sankhe