[Cryptography] Why full-fledged quantum computers might always be five years away

[Jerry Leichter]

> What is a quantum computer? The question feels deceptively simple....
What is a digital computer?

The IBM 1620 - late 1960's - was missing all the following features that someone today would likely list if asked to define a digital computer:

- No binary arithmetic - the machine worked in decimal.  Yes, the internal gates were binary but that was essentially invisible in the device as presented to a user.
- No hardware floating support
- No interrupt mechanism
- No separate user/supervisor mode
- No support for graphics
- No support for a mouse
- No support for a network connection
- No time/date reference - the machine (and in fact even machines built years later) had no internal way to determine the current date and time; you entered that at boot
- No real-time clock.  The 1620's operating system didn't ask you for the current date and time as it even lacked a mechanism to increment the time at regular intervals to create a clock.

And yet the 1620 was described as a digital computer, and was used for the kinds of things we use digital computers for to this day.  It even came with a FORTRAN compiler.

Sure, you can argue that these features are not essential to the definition of a digital computer - but that's taking things the wrong way around.  If you want to know what "digital computer" means to any particular speech community, you would need to ask them - and I can pretty much guarantee you that these items - perhaps not all, in any one community - will show up almost universally.

Philosophers - particularly analytic philosophers - make their living by picking at the exact meaning of words.  They pretty much always come to the conclusion that some word we use isn't nearly as well-defined as we think.  They aren't so good at suggesting better definitions, mind you; in fact, some of them will actually say that the lack of solid grounding is exactly the point.  If you follow through some of these arguments, you'll find that doing science is impossible - we can never really know what we are looking at and whether the questions we ask have any meaningful answers.

Cool stuff, but in the end, work goes on, discoveries are made, we use them to understand things for a meaning of "understand" that is sufficient to let us build useful stuff.

If we succeed in building devices that can do certain computations exponentially faster based on quantum phenomena, the question of what "quantum computer" refers to will answer itself.  If we don't, "quantum computer" will be as useful a term as "radish computer."

                                                        -- Jerry