That is harder than you might think. From Smarter than us ( https://drive.google.com/file/... ):
"Why aren’t they a solution at all? It’s because these empowered
humans are part of a decision-making system (the AI proposes cer-
tain approaches, and the humans accept or reject them), and the hu-
mans are the slow and increasingly inefficient part of it. As AI power
increases, it will quickly become evident that those organizations that
wait for a human to give the green light are at a great disadvantage.
Little by little (or blindingly quickly, depending on how the game
plays out), humans will be compelled to turn more and more of their
decision making over to the AI. Inevitably, the humans will be out of
the loop for all but a few key decisions.
Moreover, humans may no longer be able to make sensible de-
cisions, because they will no longer understand the forces at their
disposal. Since their role is so reduced, they will no longer compre-
hend what their decisions really entail. This has already happened
with automatic pilots and automated stock-trading algorithms: these
programs occasionally encounter unexpected situations where hu-
mans must override, correct, or rewrite them. But these overseers,
who haven’t been following the intricacies of the algorithm’s decision
process and who don’t have hands-on experience of the situation, are
often at a complete loss as to what to do—and the plane or the stock
market crashes. "
"Consider an AI that is tasked with enhancing shareholder value
for a company, but whose every decision must be ratified by the (hu-
man) CEO. The AI naturally believes that its own plans are the most
effective way of increasing the value of the company. (If it didn’t be-
lieve that, it would search for other plans.) Therefore, from its per-
spective, shareholder value is enhanced by the CEO agreeing to what-
ever the AI wants to do. Thus it will be compelled, by its own pro-
gramming, to present its plans in such a way as to ensure maximum
likelihood of CEO agreement. It will do all it can do to seduce, trick,
or influence the CEO into agreement. Ensuring that it does not do so
brings us right back to the problem of precisely constructing the right
goals for the AI, so that it doesn’t simply find a loophole in whatever
security mechanisms we’ve come up with."