I would agree. We already know that a great part of our systems, such as communication, is basically monitored by machines. Humans have very little part in that or even the diagnostics of the problem. The computer simply spits out what the issue is and, if it cant correct it, notifies humans to get someone to take care of the problem.

This will only expand as machines get more "intelligent". Even if we are not going the AGI route, the machines will get better in their niche meaning they will be able to better oversee what they were designed for.

We are in the early stages of this but it is shaping up quickly. They will have to have a different set of rules than what humans operate by.