[dasmoth]

If one were being cynical and thinking that optimising for (maximum) headcount is a major driver here, one might want to avoid technologies with substantial barriers to entry, since that might clear the market of qualified people and make further recruitment very hard. The "ideal" technology would be one where almost anyone can pick it up quickly and at least look like they're being vaguely productive, but where even a 0.01-%ile expert will struggle to achieve really exceptional things.