Can someone advocate for the concern of one harm while not competing with the existence of other harms, unifying concern about AI into one multifaceted topic
or must all different aspects of concerns regarding ML (and eventually AI) by definition compete as different topics?

I, for example clearly see three primary facets (and many subvariants thereof) of concern about (currently ML, but called AI) and it's future effect on society:

1) unemployment/disemployment/permanent unemployability

2) empowering ill intentioned people

3) systems causing harm by themselves

A confounding factor is that 1) and 2) often overlap in cases where "capitalism" (almost always income inequality) harms people and the primary way in which we empower individuals at present is through money.
Morally our society has a tendency to give wealthy people a free pass when it comes to actions that in familiar settings, or person to person relations are considered wrong.

Separately, it is considered unfair when someone spends 5-13 years of their lives investing in education & training (often going quite deep in debt to do so) only for their choice of career to be made nonviable (a considerable financial harm).

2) and 3) can be confounded in cases where an empowered individual or organization's actions (which may already be considered harmful in themselves) hypothetically result in the creation of a Machine Learning system (present technology) or Artificial General Intelligence (as yet undiscovered technology) which can independently, and outside of any person's (whether ill intentioned or not) control harm people.

Many hypotheticals have been written on the topic, but in short; harms resulting from 3) can include both 1) and 2)

Is it possible to unify these views and discuss the problem's various facets or must people choose a side and cannibalize attention directed at the other facets?