AI governance is in its HR phase

AI governance docs : Yeah. Looks like a HR manual. Forcing tech workers to relearn blackbox management.

Alternative situations:

1. Dependence on non-mechanistic interpretation implies blackbox management. Bots and meat are equally untrustable people. Same framework to govern.

2. Someone did their homework and has formal guarantees on behaviour. Then bots have significant differences from meat. "Old style IT security." 

The whole epoch is currently skewed to 1. for now.