5 ms·

Legal disclaimer: I know nothing at all about physics, prepare eye roll. Isn't the fact that certain quantum particles only "exist" when measured validation tha

Legal disclaimer: I know nothing at all about physics, prepare eye roll. Isn't the fact that certain quantum particles only "exist" when measured validation that they get "computed" when needed?

euyyn 11y ago
Depends on what you mean by <<certain quantum particles only "exist" when measured>> :)
gus_massa 11y ago
It's very difficult to explain the quantum mechanics without math, and most of the "explanations" without math are misleading or plain wrong. This is a usual misconception that I saw a few times before. I'll repot an old comment that I made in a similar article (with some minor changes): Quite the opposite, the quantum version of the word needs more process power instead of less process power. When you use Classic Mechanics, you only need to "compute" one trajectory. With Quantum Mechanics the system may have a many choices, and you must keep the details of all the possibilities. The easy case is a qbit that only has two choices. But if you have a few qbits nearby, they may interact and now you must "compute" the 2^n possibilities. This is why if we can ever build a quantum computer, it would be exponentially more powerful than a classic computer.
- gizmo686 11y ago
  >This is why if we can ever build a quantum computer, it would be exponentially more powerful than a classic computer. The truth is a little more complicated than that. If we could precisely inspect the quantum state of a system, then quantum computers trivially give us an exponential speed up. However, once we measure the result of the computation, we collapse the state down to a pure state[0], and all future measurements of the system will give us the same result. This means that we would need to repeat the computation an exponential number of times to get the full result with a bounded probability of error. As it turns out, there are some tricks we can do with interference to improve efficiency, and for some problems achieve an exponential speed up. [0] Or rather, according to the many worlds interpenetration, we become entangled with the system; however this is one of those cases where the Copenhagen interpretation is actually easier to use.

Legal disclaimer: I know nothing at all about physics, prepare eye roll. Isn't the fact that certain quantum particles only "exist" when measured validation that they get "computed" when needed?

euyyn 11y ago

Depends on what you mean by <<certain quantum particles only "exist" when measured>> :)

gus_massa 11y ago

It's very difficult to explain the quantum mechanics without math, and most of the "explanations" without math are misleading or plain wrong. This is a usual misconception that I saw a few times before. I'll repot an old comment that I made in a similar article (with some minor changes): Quite the opposite, the quantum version of the word needs more process power instead of less process power. When you use Classic Mechanics, you only need to "compute" one trajectory. With Quantum Mechanics the system may have a many choices, and you must keep the details of all the possibilities. The easy case is a qbit that only has two choices. But if you have a few qbits nearby, they may interact and now you must "compute" the 2^n possibilities. This is why if we can ever build a quantum computer, it would be exponentially more powerful than a classic computer.

gizmo686 11y ago
>This is why if we can ever build a quantum computer, it would be exponentially more powerful than a classic computer. The truth is a little more complicated than that. If we could precisely inspect the quantum state of a system, then quantum computers trivially give us an exponential speed up. However, once we measure the result of the computation, we collapse the state down to a pure state[0], and all future measurements of the system will give us the same result. This means that we would need to repeat the computation an exponential number of times to get the full result with a bounded probability of error. As it turns out, there are some tricks we can do with interference to improve efficiency, and for some problems achieve an exponential speed up. [0] Or rather, according to the many worlds interpenetration, we become entangled with the system; however this is one of those cases where the Copenhagen interpretation is actually easier to use.