Cake
  • Log In
  • Sign Up
    • Mostly good things, especially running simulations that are computationally intractable today--protein folding, drug discovery, cosmology, climate, etc. One seriously bad one, which is that it will break current encryption methods. Maybe eventually it will produce new, more robust methods, but that's a bit of an unknown. I don't think it makes much of a difference whether the US or China gets there first, as the other will inevitably catch up soon enough. Unlike at least one scenario of super-intelligence, it's not a first-across-the-line competition. Think of it more like the nuclear arms race, in which technological advantages are not lasting.

      Can anyone recommend an introduction on how these machines are programmed? Do they actually use stored instructions? Are they Turing machines or something else? I have a vague understanding of superposition, but don't have a clue as to how one can use it solve problems--how do you know when you've got the right answer?

    • If you want to get your mind blown in a fascinating way, watch this. However, she doesn’t explain how it works

    • Here’s how to program them:

      However, I feel it helps to watch Veritasium’s excellent short explainer first:

    • Thanks @Chris. I'm old enough (and started learning about computers when I was young enough) that I remember the old IBM unit record accounting machines, which took punch card input and produced punch card output under the control of removable control panels of jumper wires, which expressed the program. These were not abstract machines--you could only do a limited number of operations and you needed to understand the hardware to get anything done. It sounds like we're sort of at an equivalent stage in the evolution of quantum computing, though the future potential is vastly greater.

    • Google is really promoting its latest achievements while IBM is saying wait, we already commercialized this, you have been able to program it in our cloud for awhile.

      Google is saying they solved a problem with it that would take the most powerful supercomputer 10,000 years to solve. IBM is saying wait, we have computers today that can solve that problem in two days.