Horizontal wisdom transfer May 11, 2007
Posted by Geordie in QC-Related Posts.trackback
There used to be a time when one person could know pretty much everything that was known. Those days are long gone. Nowadays in order to become an expert in anything you need to focus pretty much all of your faculties on that one thing.
This leads to a couple interesting effects. One is that this focus tends to color everything about how you view the world. People tend to approach problems and form opinions through the lens of their expertise. This happens all the time when disciplines are close-for example solid state physicists tend to have a way of thinking about chemistry problems that is predictable and different from the way chemists think of these things-but it also happens in wierder situations, where the area of expertise is entirely disjoint from the situation being analyzed-like when theoretical computer scientists have opinions about real computers for example.
I think this is inevitable. But understanding this phenomenon is very important in dealing with any highly multidisciplinary technology effort. One thing you can actively try to do is make sure that basic results known to one class of expert are understood by the others. It’s easy to assume that just because something is well-known in one sector of science that leading experts in something else will be aware of these results, but this is almost always not the case.
So why am I going on about this? Well it turns out that something I thought was foundational knowledge in quantum computation is not common knowledge for a lot of QC theorists (although I’ve yet to meet an experimental QC-er who didn’t know about these. You guys should talk more).
An entry point into these sets of experiments is here, which in my opinion is a QC classic and one of the most important experimental QC papers written to date. A good theoretical analysis of this type of system is here.
Now that I get that a lot of people aren’t aware of these I’m going to spread the good word.
Of course this is a nice paper. Enough to base an entire speculative company on? Doesn’t even come close to passing muster with me.
And don’t assume that we “quantum theorists” haven’t read things like this. Some of us do read widely in spite of being told that this world is only for “experts.”
[...] at Roseblog, Geordie gets on a high horse and tells us quantum theorists what we haven’t read enough because we [...]
its a good idea to stop posting things so that angry people cant rip it apart. concentrate on the completion of the computer and you will come out on top 100%. Your frustration with negative feedback seems very strong. Hope you have good success toward your goal of 64 qubits.
@bacon. for gods sake this is the attitude that makes acedemic interaction unbearable. have a touch of class. hes only trying to show a “physical” result related to orion that some physicists might like to know for the sake of gaining knowledge and insight.
Hi Geordie!
Quantum Annealing = Quantum Adiabatic Computation
Just different physical frameworks.
Geordie the nature if QI is such that I’m sure on average we are much more widely read than most other physicists. I highly doubt that you personally read more widely than I do. So your condecension is somewhat irritating.
At the time I thought the “quantum annealing is faster than classical” work was vaguely interesting because it may give a nice operational way to look at “tunneling times” - an area of controversy. I didn’t then (and still dont) think of it as a particularly likely route to scalable quantum computation. You are welcome to try and change my mind about this with actual science.
Sorry for entering in your discussion, as I am a total ignorant in the matter.
Just two points:
1) investors have put some USD40millions+ in D-Wave. They have access to all the information to which other people ( Dave, Scott ) have not had access to. They also have probably, some theoretical physicists working for them as consultant, in order to give green light or not to funding D-Wave. So, why should a venture capital company or firm give away tens of millions of USD, if they do not have the realistic hope to get big profits out of it?
We are talking about tens of millions of US dollars.
And, again, the people who put the money, had probably had access to all the information, to which other people, like Dave and Scott, have not had access to.
2) I do not understand why so many people are asking D-Wave to give some more information away about the project, in order to understand if they have reached any real result or not. Why shound they do that?
If I were the CEO of D-Wave, I would never, never do that.
D-Wave does not need money ( at the moment ).
It would be a mistake for them, to give any information away ( except, for what is needed to get patents ), invite other people to see the machine ( Orion ), to publish anything, to submit any paper for peer review.
They have the money.
They do not need any peer review.
If they have hope that they will make it, it is good for them to keep things secret as long as possible, in order to get a big head-start over competition.
If they will fail, what is the difference between flopping now, and flopping at the mid-end of 2008, when they have planned to have a 1KQbit machine working?
In any case, either if they make it or if they flop, it is good for them to keep things secret as much as possible.
And, as long as possible.
Just my 0.000002 cents worth
nextquant: Not exactly. The annealing strategy is explicitly heuristic while AQC usually isn’t viewed in this way. The same physical hardware can run both strategies of course but the philosophy behind their operation is quite different.
The Professors didn’t get there first and they are angry; all truth must first come from them.
“have a touch of class”
I’ll get a touch of class when you get a real name! Deal?
“The Professors didn’t get there first and they are angry; all truth must first come from them.”
BS, I’d certainly be happy to see truth come out of D-wave.
BTW, that theory paper always buged me since it was done for the planar Ising model with no logitudal magnetic field, right? But this problem is in P: L. Bieche, R. Maynard, R. Rammal, and J.P. Uhry. “On the ground states of the frustration model of a spin glass by a matching method of graph theory.” J. Phys. A, 13:2553–2576, 1980.
“Hope you have good success toward your goal of 64 qubits.”
Note that the theory paper sited already simulates a system much larger than this, so what would such a chip demonstrate? Would it be able to beat quantum monte carlo?
“Quantum Annealing = Quantum Adiabatic Computation.” No isn’t it more like Quantum Annealing = Quantum Adiabatic Computation + Annealing?
[...] so talking about D-wave is certainly not good for my Zen-Budha nature. Not good at all! So let’s talk about something related to D-wave but only tangentially so that I can get that [...]
Hey Geordie;
Thanks for sharing. I haven’t been able to find sites for very many theorists. Also, anytime you want me to back you up on a conversation in the Barrens…Go to Altar of Storms…Horde side….look for Tenderluv or Morsonius.
[...] more posts about D-Wave and its commercial quantum computer for a while. But will you look at the bait that D-Wave founder Geordie Rose has been dangling in front of me on his blog? People tend to [...]
I think the major reason for the criticisms for Dave, Scott, et. al., is that D-wave seems (as I hear the hype, at least) to be unable to stick to a straight story about what they’ve built. Is it a system with 16 coherent qubits (coherent long enough to perform nontrivial calculations like even 2-by-2 Sudoko puzzles on)? That would certainly get everyone’s attention if true. 16 coherent qubits would be a huge advance over everyone else. Or is it a system that relies on a hope that a combination of classical and quantum annealing will get to the right state? In that case, 16 qubits just aren’t that much. In fact, in that case, 16 qubits are at best a toy to show some principles, and sitting in the computer in front of me, which cost only a few thousand dollars, is a processor many orders of magnitude faster (actually two of them), with about 10^9 times as much storage, not counting the hard drive. So, which is it?