Why do we always focus on AI as some sort of all powerful entity that will either try to immediately destroy us or control us?
The only things we have created that currently have the ability to cause effects on a massive scale are organizations such as governments. Although direct actions are sometimes taken by these organizations, usually they try to accomplish their goals through less exhaustive means such as incentives (good to encourage, bad to discourage).

Considering how automated programs already use machine learning to incentivize certain purchases by offering coupons or deals, why wouldn't AI try something similar to achieve its goals? It would seem that a type of symbiotic relationship would help the AI obtain energy and resources it couldn't on its own by doing tasks more effectively than we can using conventional methods. If we assume that AI is incredibly logical why would it try to antagonize over 7 billion unpredictable creatures that have spent over 30 years worrying about AI?