Would a quantum computer give great answers? Or lie to cover its ignorance.

New research suggests you'd never be able to tell the difference


Is it possible to understand each separate detail of a project well enough to answer every question anyone could ask, but not know anything about the project as a whole?

Not in any normal universe, but researchers in quantum physics claim to have found evidence not only that knowing the details – or at least being able to prevent anyone proving you don't know the details -- does not mean you understand the whole.

" Let’s say that you study a book that has two pages: one and two," according to an interview co-author Stephanie Wehner of the Center for Quantum Technologies at the National University of Singapore did with Physorg.com.

"You are going to sit an exam in this class, and you only had a small amount of time to study. You don’t know everything that is on page one and page two. Can I point to a page that you don’t know, thereby exposing your ignorance? Classically, this is indeed true: For example, if you only know the information on page one, I can point to page two and expose your ignorance. I can catch you.”

In the quantum world that approach won't work because even a partly informed examinee can guess the answer to any question – almost perfectly. (Here's the abstract; here's a PDF of the press release; here's the title, which is the most complex thing you will read today: “Does Ignorance of the Whole Imply Ignorance of the Parts? Large Violations of Noncontextuality in Quantum Theory.”)

“Our example demonstrates that this effect can be rather dramatic. But does it always have to be this way? And, how do we tell," she asked.

Whether the effect has an impact depends on the type of memory involved – conventional or quantum. I think she's talking about computer memory, quantum states and quantum information, not human versions of those things.

Usually what happens at the quantum level has nothing to do with the behavior of anything big enough for us to perceive – which probably means we either don't understand what we're seeing on the quantum level or don't know what we're doing on the human one.

If states and conditions in the quantum world apply to the real (and biological one) that should mean tests wouldn't work to demonstrate whether a human had studied or not.

That doesn't mean it's not valid on the subatomic scale or that discovering a particle can lie so convincingly a human can't catch it won't have a big impact on the development of quantum computers, which may end up being as much more powerful than regular computers as the atom bomb was compared to the hundreds of thousands of bombs dropped over Europe during WWII.

It would be nice to know, though, if quantum computing lets us make breakthroughs in space travel or time travel or really, really good tax preparation that doesn't give us a headache, that the quantum computer we're relying on for an accurate course to Mars, for example, knows what it's talking about and isn't just making up the answers.

ITWorld DealPost: The best in tech deals and discounts.
You Might Like