Quantum conundrum

Should scientific funding emphasize basic research or technological applications? A discussion inspired by Alex Hutchinson’s “Breaking D-Wave” (Sept. 2007)
In a mediascape where even alleged human cloning by Raelians gets coverage as a scientific discovery, the February 13, 2007 demonstration of what British Columbia’s D-Wave systems billed as the world’s first commercial quantum computer was strangely unhyped.

A functioning quantum computer “would be a fundamentally transformative machine, capable of modelling and predicting the behaviour of almost anything in the universe” Hutchinson writes.

The implications are staggering. A quantum computer sounds like an all-seeing, all-knowing oracle — and, certainly, a cash cow. But because capital-oriented D-Wave and its chief technology officer, Geordie Rose, leapfrogged the science community by not submitting their research results to peer review, scientists have been underwhelmed by the company’s achievement. (Although, Rose has managed to create an online buzz.)

The media’s quantum silence may simply come from ignorance. Perhaps, as physicist Richard Feynman has suggested, no one really understands quantum mechanics — which makes discussion outside the techno-sphere (e.g. at The Quantum Pontiff and Quantum Algorithms) unlikely.

Quantum computing: Are we too thick to grasp the implications? Maybe, but we’re giving it the old Walrus try.

The old saw (disputed by M.I.T. computer scientist Scott Aaronson) is that the scientific community will snub research that doesn’t follow “ivory tower” rules.

And yet, the white coats don’t have clean hands: basic research has been largely funded by players such as the American military and National Security Agency, which have historically made large investments in science.

In whose hands does quantum computing belong? And in general, does science research only count when men with microscopes sniff at it long enough to write impenetrable journal articles? Or is D-Wave’s techno-engineering approach, which blurs the line between corporate suits and lab coats, the wave of the future?

The Walrus, of course, is of multiple opinions simultaneously. And you?

24 comment(s)

Ei(sen)steinAugust 10, 2007 08:28 EST

Pure research is a nice idea, and certainly deserves some funding, but we're talking a fraction of a percent. It's work best reserved for only the superest of supergeniuses—the ones who work with ideas capable of inspiring wonder in our society.

But the bulk of the research ought to continue to be technological in nature. This allows us to fund research according to need (environmental tech right now) and provide tangible economic benefits—ones that then give us the luxury of the time to sit around marvelling at the products of pure research.

Z PaxtonAugust 10, 2007 09:09 EST

Eisenstein:

I guess the question is: who decides the need? I'm not sure I want it to be the military - or whoever holds the purse strings. Call me cynical...

Scott AaronsonAugust 10, 2007 09:31 EST

This essay falls into the same trap as almost all previous press coverage of D-Wave. That is, it takes D-Wave’s claims to have built a "commercial quantum computer" essentially at face value, and then frames the issue as whether or not their achievement should "count" if they don’t follow "ivory tower rules."

In reality, whether or not D-Wave achieved anything is very much in question — and conventions like peer review are not ends in themselves; the whole point of them is to let people decide what's been achieved. In this case, it's not just that D-Wave went directly to the press without a paper. It's that since then, they've refused to answer extremely basic technical questions when asked — like how noisy their qubits are, how they encoded a Sudoku puzzle into 16 qubits, how much of the computational work is done in the classical control, what classical algorithm they're comparing against when they claim a speedup, etc.

A much more minor error is that I'm of MIT and formerly of Waterloo, not the reverse. :-)

LlewAugust 10, 2007 09:36 EST

The rumor I had heard in college was that the USSR had developed an early prototype of a quantum computer in the 90s, but it was unstable, and that if you bumped it just slightly, it would reset all of the quantum states.

Not sure how accurate that story is, but I do wonder about the feasibility quantum computing really is. It may be like nuclear fusion - something that can be done but takes more effort than it's worth.

Scott AaronsonAugust 10, 2007 11:37 EST

Geordie: Yes, I'm interested in buying one of your machines. My offer is $10 (US). As a new prospective customer, I await your answers to the questions in my previous comment. Thanks!!

Scott AaronsonAugust 10, 2007 11:44 EST

Alex: I wasn't talking about your article, which I hadn't seen and which does indeed include a skeptical perspective. I was just talking about the short essay on this web page (the one by "The Editors"), which is what I'd been sent a link to.

Alex HutchinsonAugust 10, 2007 13:36 EST

Scott, thanks for posting your thoughts. If my piece has left the impression that I take D-Wave's claims at face value — in other words, that I have no doubt that they've done what they claim they have — then I haven't been clear enough. The quote from Ray Laflamme expresses my doubts:


“If you go and ask anybody in the world who’s leading in putting qubits together, the first thing you’ll ask them is, ‘What is the decoherence time?’” says Laflamme. “And if they say ‘I don’t know,’ you’ll be very skeptical of the rest of the thing.”


Geordie was very helpful in answering questions as I was writing this piece — but his helpfulness stopped at a certain point, the point at which I asked to see things like decoherence times. As his comments on this discussion thread suggest, he's simply not going to tell us.


On the other hand, I talked to a lot of physicists for this piece, most of whom aren't quoted, including some who are connected to people who have gone to work at D-Wave. And I feel reasonably safe in saying that, whatever doubts scientists harbour about D-Wave's claims, most of them don't feel that the company is trying to perpetuate an outright fraud. The scientists there believe in what they're doing, even if they haven't proved it to anyone else's satisfaction.


So that's the epistemological conundrum I was left with in writing this piece. I don't know enough to write about whether D-Wave's science is good or bad, because of their approach. So the question becomes, is their approach justified? Is it good? Bad?

Ken AAugust 10, 2007 14:15 EST

Applied research would never have invented the lightbulb.

AnonymousAugust 11, 2007 12:07 EST

Very interesting article!

I have a technical question: if D-Wave is indeed using a quantum system, what could they be doing besides quantum computing? Is it possible to constrain the quantum degrees of freedom to do classical computation with a quantum system?

Alex HutchinsonAugust 12, 2007 16:58 EST

"Is it possible to constrain the quantum degrees of freedom to do classical computation with a quantum system?"

Here's my understanding of D-Wave's particular approach to quantum computing (which is called adiabatic QC): If they had a big shiny knob that could "tune" their system between classical and quantum behaviour, and they cranked it all the way to the "classical" side, they would have a ordinary classical computer that calculates using a technique called simulated annealing.

That's what makes it so hard to evaluate D-Wave's claims: if their system ISN'T quantum (because there's too much environmental noise, say), it could still be solving all the problems they demonstrated (sudoku, wedding seating plans, etc.). If they go on and build a computer that's hundreds of times bigger than their current one, then we'll know if it's quantum because it will be really, really fast. But for now, with their simple demo system, we just have to take their word that we're witnessing a quantum calculation and not a classical one.

PatAugust 13, 2007 10:27 EST

Do I smell Bre-X?

Geordie RoseAugust 13, 2007 18:47 EST

I'd like to briefly respond to Scott Aaronson's post.

The substance of Scott's criticism is "[D-Wave] refused to answer extremely basic technical questions when asked".

If Scott is interested in buying one of our machines I would be more than willing to discuss these issues at length with him (under NDA), as I do with all potential partners and customers.

The Walrus OnlineAugust 13, 2007 18:47 EST

Scott:
M.I.T. / Waterloo error has been fixed - apologies for the mistake

Geordie RoseAugust 13, 2007 18:47 EST

Scott: Great. Contact me directly and we can discuss.

Jim MaloyAugust 20, 2007 11:28 EST

Mr. Hutchinson states that he cannot comment on "whether D-Wave's science is good or bad", given the data available to him. Not being a computer scientist, I'd ask, "So?" I'm surrounded by devices on which I depend and don't really understand. I'd suggest that whether or not there really is a quantum dragon within the D-wave computer, if the device behaves as if there is one - or computes very quickly in a larger version of what now exists - I'd be happy to call it a quantum computer.

AnonymousAugust 23, 2007 13:26 EST

IMHO

Peer Review is more valuable to the researcher than the rest of us. I would agree with Aaronson that following 'ivory tower' rules is not a hard prerequisite to a theory or development being accepted, but 'discoveries' can be believed much more readily if there is reasonable evidence of a logical foundation. Otherwise an interested party would have to replicate the discovery to determine credibility; and with no background information replication would be difficult.

The oft-advertised statistic that 4 out of 5 dentists recommend a product always leaves me wondering why they didn't just get rid of the fifth dentist since no description of the sample is provided to lead me to assume any statistical accuracy.

Again IMHO, pure research should be publicly funded; application research probably not because productive application leads to economic value.

As to quantum computing; four decades working with computers on both sides of the customer/vendor fence renders one a tad cynical about "fundamentally transformative" anything.

LeroyAugust 24, 2007 12:38 EST

Anonymous: "As to quantum computing; four decades working with computers on both sides of the customer/vendor fence renders one a tad cynical about "fundamentally transformative" anything."

What about the internet? Or the digital computer itself? Both of these are fundamentally transformative.

RickWAugust 25, 2007 16:28 EST

"Alleged human cloning by Raelians" is spectacle, and we have become a society that prefers spectacle over "plodding" science. Jared Diamond in his book "Guns, Germs, and Steel" noted that we, relative to our society, have become quite ignorant when compared to any aboriginal, relative to his society. The aboriginal knows everything he needs to know to survive, whereas we do not. And the sheer volume of information we generate serves to make us yet more ignorant in a holistic way. Consequently, our "defense" is to give credence to the more shallow and titillating aspects of science, knowing these tidbits will come and go, and will not affect our lives in any forthright sense. "Science is for eggheads", we convince ourselves, with a dismissive wave of the hand, and cozy up a little more intimately with our ignorance.

We view science today as alchemy. Perhaps if we were more familiar with basic science, the decision (noted above)"who decides?" would become moot.

Alex HutchinsonSeptember 03, 2007 14:53 EST

To Jim Maloy: Sure, what you say makes sense. Once D-Wave builds a bigger computer, either it will work or it won't, and the debate will end there.

In the meantime, though, decisions have to be made about, for instance, reseach funding — not just for D-Wave, but for the hundreds of scientists working on other quantum computing approaches. So if one group claims that they have a working quantum computer, you can understand that other groups might feel that either you (a) provide some evidence to back up those claims, or (b) don't make the claims until you're ready to back them up.

That's how traditional science works. But it's not necessarily how venture-capital-funded technology works, and that's the culture clash we're seeing here, I think.

Chester SadowskiSeptember 06, 2007 08:03 EST

I am not sure that the "or" in the question posed has any validity. The fact is that modern technological advances have a basis in pure science and consequently it behooves us to maintain a strong science base that is constantly expanding our knowledge to keep the technology of the future healthy. There is no clear boundary between pure science and technology; each feeds off the other. There are many examples of this. MRI is based on NMR that was developed by a Swiss organic chemist to help understand his compounds better. Lasers are based on a concept of matter-radiation interactions first proposed by Einstein near the beginning of the 20th century - stimulated emission of radiation. The first illustration of this process had to wait until the 1950s when sufficiently strong light sources and sufficiently sensitive light detectors became available. The first laser, actually a maser, was developed by Harold Townes as an amplifier for use in radio astronomy. Television was initially based on the cathode ray tube which developed out of a precursor built in the 19th c by J.J. Thompson to study cathode rays(electrons). The modern dye industry owes much to the synthesis of artificial dyes as a scientific curiosity by 19th c. chemists. One could go on. The flow occurs in the reverse direction as well, since modern science depends on technology - often the same technology that it spawned in the first place.

The problem is that the linkage between science and devices is seldom direct and often serendipitous. The concept of stimulated emission did not lead directly to lasers nor did Thompson's cathode ray tube lead directly to television. Those developments had to wait a long time. It is the separation between scientific discovery or research and devices that causes people to downgrade science and dismiss it as merely "curiosity driven" and an expensive activity to support.

We should not forget the educational impact of pure scientific research either. University based research trains individuals at an advanced level, who are indispensible to enterprises developing new technologies.

In answering the question posed it is essential to maintain a broad view. We need both pure science and technological development and our institutions, both private and public, should provide healthy support of both. In closing, I would like to state that in a recent study in the UK it was shown that a strong national pure research programme has a significant positive impact on the national economy.

R. Brent LangSeptember 06, 2007 10:43 EST

I am a believer!

codySeptember 07, 2007 22:41 EST

im with you Alex, the idea in science is to be as honest as possible about what is and what is not true, but honesty is not necessarily the best way to sell an idea to investors, nor even is proof through public demonstration. granted, everyone will always have the right to reject and criticize such extraordinary claims when they lack any extraordinary evidence.

AnonymousJuly 21, 2008 07:50 EST

I'm still waiting for Windows Quantum Edition, Mac OS Q, and qLinux!

LocalSeptember 25, 2009 17:02 EST

There is one undeniably fishy aspect of this company. Their website has been advertising to hire scientists for over 2 years. I and other qualified applicants have offered our services with no response from the company. The job ads get renewed once in a while but with no material change and they don't hire anyone.
I believe their attempt to look like a legitimate growing company, looking to hire and grow IS fraud.

Whatever they are doing with their money, it's NOT hiring staff.

Add a comment

  
I agree to walrusmagazine.com’s comments policy.

Canada & its place in the world. Published by
the non-profit charitable Walrus Foundation
TwitterFacebookTumblr
The Walrus SoapBox
The Walrus Laughs
Walrus TV