I find it bizarre how often these points are repeated. They were both obviously wrong in 1993, and obviously wrong now.
1) A nitpick I've had since grad school: the answer to "can we create a machine equivalent to a human mind [assuming arbitrary resources]?" is "yes, of course." The atoms in a human body can be described by a hideously ugly system of Schrödinger equations and a Turing machine can solve that to arbitrary numerical precision. Even Penrose's loopy stuff about consciousness doesn't change this. QED.
2) The more serious issue: I sincerely have no idea why people believe so strongly that a human-level AI can build a superhuman AI. It is bizarre that this claim is accepted with "little doubt" when there are very good reasons to doubt it: how on earth would such an AI even know it succeeded? How would it define the goal? This idea makes sense for improving Steven Tyler-level AI to Thelonious Monk-level; it makes no sense for a transition like chimp->human. Yet that is precisely the magnitude of transition envisioned with these singularity stories.
You might defend the first point by emphasizing "can we create a human-level AI?" i.e. not whether it's theoretically possible, but humanly feasible. This just makes the second point even more incoherent! If humans are too stoopid to build a human-level AI, why would a human-level AI be...smarter than us?
I just don't understand how anyone can rationally accept this stuff! It's so dumb! Tech folks (and too many philosophers) are hopped up on science fiction: the reason these things are accepted with "little doubt" is that this is religious faith dressed up in the language of science.