Michael Green's appointment to replace Stephen Hawking as the Lucasian chair, has, quite predictably, brought back into the spotlight the ever simmering STRING WARS!!!OMG!!!STRINGTHEORYRLZ!!. Okay, maybe not the spotlight, per se, but I did find the article about Green in the Guardian interesting (via the so wrong it hurts fellow):
But that was one of their arguments, that the academy is so biased towards string theory - hiring mostly string theorists, crowning mostly string theorists - that it has driven out all other ways of seeing (Smolin compared it to deciding that there was only one way to fight cancer, and pouring all available resources into that one way). "People do what they feel is going to be productive," says Green. "It's all very well to say they should be doing something else. But there is nothing else."
Now, of course, this is all part of a long series of arguments about the validity of string theory as an approach to a physical theory merging gravity and the standard model. Yawn, that is *so* 00s.
What it did make me think, however, was what the equivalent argument would be in a different field. And because, while I posses my fair share of extralusionary intelligence, I thought, oh I'd better stick to my own field when I think about this. So what would the equivalent be in quantum computing?
I hereby declare that there are only two valid approaches to building a quantum computer: ion trap quantum computers and superconducting based quantum computers. It's all very well to say that we should be spending our time working on other "ideas" for quantum computers. But there is nothing else.
- Log in to post comments
I can't actually speak for quantum computing to know if there are actually other valid approaches, but for quantum gravity, at least, there definitely are. Lee Smolin is right to still argue of a funding bias in quantum gravity, because there are plenty of great ideas that have been coming out, especially in the last 10 years, from very different approaches, that could certainly use some of that string theory money. String theory has had its day, as far as many people are concerned, productivity is going to happen in other areas now.
This reminds me of what I say to people that ask me how long it will be until we have a working (and usefully large) quantum computer. Instead of saying 10 or 20 years, I say: "10 or 20 years after the international community can agree there is only one valid way to go!". Though when your building a device it is more pressing to get everyone to agree on a design than when you devising foundational theories with sparse experimental predictions.
Ion traps? Never heard of 'em ;)
The 00s have been officially renamed "The Noughties" - you might want to amend your blog accordingly....
I'll humbly put forth that your posited analog is quantitatively different than our newly Lucasian'd friend's.
i.e. He made a statement about the existence of alternate consistent theoretical description that is not accommodated by the framework of String Theory.
You made a statement about what is feasible from an engineering (or experimental) perspective.
See what I'm getting at? For Michael string theory and it's non-perturbative completion gets everything -- it just needs to be teased out. Saying there's nothing outside of "everything" isn't a very big, interesting, or particularly controversial statement.
Saying "I know what *everything* is," may be a controversial statement. But presumably for him, this is just like saying "QM is behind chemistry."
In a certain sense his statement is already in your field btw. Let me try to translate. He's claiming that string theory as a framework is in a computability class that can handle scattering of gravitons. More than that, he's claiming that it's complete in that class. Or something. Limits of my extralusionary intelligence. (If "computability class" is the wrong phrase please let me know!).
Hi JJMC: Yes definitely my statement is different! It's hard to find suitable analogies across the field, which was why it was fun to try.
But maybe I am saying something beyond just "engineering." Maybe I'm saying that there is no way that the laws of physics as applied to systems besides ion traps or JJ qubits can be used to build a quantum computer. And if, heaven forbid, someone comes along and shows how to get linear optics quantum computing to work, well then it's going to have to use ion traps of JJ qubits. Ion traps and JJ qubits are a universality class for quantum computer implementations. :)
@IanW: I had not heard the term "The Noughties" but perhaps it is appropriate given the *ahem* behavior that seems to have occurred during the decade.
It boils down to putting those poor Schr"odinger kittens in the ultra cold ion traps to die of those poisonous ion vapors while they are freezing to death [technically they are plasma so they are vapor also] or let a super current be conducted through their flesh while a magnetic field flips through them... brutal! Think of the Eddy currents that will boil their innards [it's halloween season so some blood and gore is ok!]
Also, I can only see dead cats in your choices, meaning there is too much decoherence.
There must be a humane way.
Okay Dave, I'm of course happy to collapse your "maybes" to "definitelies" and applaud the analogy well made! :)
I'd only considered the straight reading of your statement which sounded like a constraint on engineering possibilities which honestly I'm very cautious of.
And since the internet is "4ever", I want to go on record as almost *never* putting a "just" in front of engineering (with or without airquotes!)-- too much awe for people who actually go out and build stuff that, you know, works ;)
And so `universality class' was the term I was clumsily groping for? I clearly need to spend more time around information theoretic folk!
What about measurement-based versus adiabatic quantum computing?
Measurement is the great Satan who will be crucified at the altar of adiabatic evolution. :)
I think about Quantum Neural Net Computers. Two ways: mixed state superposition of two or more classical Neural Net computers; or neural net architecture where each element is a quantum computer of one of your preferred implementations.
In my view the conscious brain does not act according to classical physics. It doesnât even act according to conventional quantum mechanics.
http://discovermagazine.com/2009/sep/06-discover-interview-roger-penros…
I think part of research should devoted to extracting existing choerent microtubule devices from Penrose's brain. There's certainly new physics there in his skull so it should be easier than producing coherence artificially. Plus afterwards he would be even more open-minded, state he so greatly enjoys.
C. Oward, Penrose proved the Penrose singularity theorem. What did you prove?
Satan crucified at an altar? Now there's a mixed metaphor if ever I saw one! :)
String theory is definitely in the correct direction. However, before doing any more mathematics; theorists really needs to ponder the paradoxes and metaphors of physics (especially those that have been ignored or discarded as being unphysical). The unphysical Dirac infinite sea of negative energy particles sure sound like extra dimensions to me. Feynmann decription of antiparticles as particles going backwards in time, seems like a big clue to me. These are not unphysical ideas that just happen to yield useful calculational shortcuts and predictions; these are strange but physical descriptions that need to be part of the foundation of string theory.