The future is here:
There was an announcement I saw in a tiny blurb in the current events section of a journal that was utterly mind boggling. A more extensive treatment is now available. (Nicola Jones The Quantum Company NATURE vol. 489 no. 7454 June 20, 2013 page 286)
For many years there have been reports of arcane quantum phenomena. The first I ever saw was done with photons. Pictures were scanned with laser light that had been split into two beams. One beam scanned a picture of interest. The other beam scanned a gallery of pictures. The beams were recombined and, by means I could not quite fathom from the article, the best match was identified from the gallery. I thought then, “It’s over. Quantum computing is real and can be scaled up for best-match searches as much as you want.” It was not to be. The way they do it now is outsourcing. If you want a best match you line up a whole army of people with small computers and send them images over the internet. People can do it better than computers. How very odd.
A quantum object is one that is not fixed in its characteristics until it interacts with the real world. A photon coming from a distant galaxy, for instance, has a reality that is distributed along every path it could conceivable have followed; its energy is fixed but its polarity is not established until that interaction occurs. At that moment, the only reality of the photon is local and things like its polarity are retroactively immutable. So every time somebody demonstrates a new quantum effect it is tribal custom for the person to say, “And maybe this can be harnessed for quantum computing.”
And now something has been. The contraption, they say, is 3,600 times faster than anything an ordinary digital computer can do. “Moore’s Law is an empirical observation that computers double in their power every 18 months. That power sometimes gets redefined, but pretty much the law seems to hold. And the result is that the world changes far faster than my poor brain can keep pace.
Well we just jumped about 18 years into the future. The only rational response is to gibber and drool.
One might take comfort in thinking that what is now available is little more than a prototype. But things are more mind boggling than that. It seems that somebody has figured out a way to get it to check an ordinary computer program and see whether it has bugs.
And that, oh best beloved, is utterly and permanently beyond the capability of any normal computer. Alan Turing showed that to be true decades ago. I even kind of almost sort of followed the logic when I read about it, but don’t ask me now.
Well writing a program isn’t easy, but it is at last possible. Working the bugs out is far harder. And you know what? If you can tell if it can rule out bugs it can write the program from scratch. I mean that follows, doesn’t it? So all that obscure cyber work we live with and depend on but do not understand can now be analyzed. There are things we probably depend on that have been written in Cobol or some other language that is utterly obsolete and that nobody is eager to make a career of leaning about. And even if they could, figuring out somebody else’s program is just about impossible even if you know the language unless great pains have been taken to document just how it works.
If the contraption works as described, which seems beyond doubt since so many capable people have looked into it, and if it can be scaled up, which even the designer is not so sure of, computers will be doing things we cannot now even imagine.
It’s going to be interesting, anyway.
There have been 69 visitors over the past month.