I think Scott Aaronson is right to rail against the explanation of "trying all possibilities at the same time". The emphasis should be on *choreographing* the right interference, not on the fact that all possibilities get accounted for by a quantum computer (since each outcome has a measurable probability).
In fact, "trying all possibilities at the same time" is a ubiquitous phenomenon in Nature. Light rays find their trajectory by tryin out all possible paths and only the one with the highest constructive interference survives (this is what Feynman discovered in Quantum Electrodynamics, which is referenced in the article).
The real trick in quantum computing is manipulating qubits through quantum gates such that the outcome that gets assigned the highest probability is the correct one. This is indeed hard as the environment usually interferes and spreads out the probabilities across all outcomes, leading to a junky, random outcome.
I took John's class on theoretical cosmology while a master's student in Cambridge. He was remarkably old fashioned — he scribbled in his undecipherable handwriting on a bunch of transparent slides and used a projector to share things in class. He had a deep mellow voice and clearly had a supreme grasp of all areas of cosmology. It is a real shame he died so early.
In fact, "trying all possibilities at the same time" is a ubiquitous phenomenon in Nature. Light rays find their trajectory by tryin out all possible paths and only the one with the highest constructive interference survives (this is what Feynman discovered in Quantum Electrodynamics, which is referenced in the article).
The real trick in quantum computing is manipulating qubits through quantum gates such that the outcome that gets assigned the highest probability is the correct one. This is indeed hard as the environment usually interferes and spreads out the probabilities across all outcomes, leading to a junky, random outcome.