Yah - it lies outside of the narrowly technical (though technical systems come up a lot) and part of what I would talk about would have been: how much is this a trick and how much of this is real? Like, is software doing slight of hand and really Uber (or whoever) should be taxed on an externality / risk? Or does this electronic machine of software genuinely create a new arrangement of responsibility? My unhelpful understanding is "it depends" and even in the Uber case it's a bit mixed, though on balance I think Uber is more of a scam than a truly new thing (even though there's some new there).
My intuition is similar; software can distribute responsibility in ways no previous vehicle could, by virtue of speed, scale, and mode of interaction (in a McLuhan-ish "medium as message" sense). But it's still an evolution of long pre-existing social dynamics, not (or relatively rarely) an absolute novelty. Scamming is an ancient art which we did not invent, I suspect, but rather discovered ourselves doing already. Only painstakingly (and still only partially) have we dragged the concept into the light of consciousness.
I feel the same way about artificial intelligence: it's not new, it's all around us, digital computation merely crystallizes the concept. But shiny objects should not distract us from the much more general phenomenon.
