If we actually had superintelligent AI, I might be concerned. But what we have instead is stochastic parrots with no innate volition. In and of themselves, they aren't dangerous at all—it's the humans backing them that we have to be wary of.