Tag Archives: Scott Aaronson

D-Wave – Fast Enough to Win my Bet?

D-Wave, ,

Really Would Like to Get That Raclette Cheese.

Last summer I had to ship a crate of maple syrup to Matthias Troyer at the ETHZ in Switzerland. The conditions we had agreed on for our performance bet were so that, at this point, the D-Wave One could not show a clear performance advantage over a conventional, modern CPU running fine-tuned optimization code. The machine held its own, but there weren’t any problem classes to point to that really demonstrated massive performance superiority.

google_benchmark
Impressive benchmark graph. Next on my Christmas wishlist: A decisive widening of the gap between the green QMC curve and the blue D-Wave line as the problem size increases (as is the case when compared to the red Simulated Annealing curve).

 

The big news to take away from the recent Google/D-Wave performance benchmark is that, with certain problem instances, the D-Wave machine clearly shines. 100 million times better in comparison to a Quantum Monte Carlo Simulation is nothing to sneeze at. This doesn’t mean that I would now automatically win my bet with Matthias if we were to repeat it with the D-Wave Two, but it’ll make it much more interesting for sure.

One advantage of being hard-pressed to find time for blogging is that once I get around to commenting on recent developments, most other reactions are already in. Matthias provided this excellent write-up, and the former D-Wave critic-in-chief remains in retirement. Scott Aaronson’s blog entry on the matter strikes a (comparatively) conciliatory tone. One of his comments explains one of the reason for this change:

“[John Martinis] told me that some of the engineering D-Wave had done (e.g., just figuring out how to integrate hundreds of superconducting qubits while still having lines into them to control the couplings) would be useful to his group. That’s one of the main things that caused me to moderate a bit (while remaining as intolerant as ever of hype).”

Scott also gave a pretty balanced interview to the MIT News (although I have to subtract a star on style for working in a dig at Geordie Rose – clearly the two won’t become best buds in this lifetime).

Hype is generally and righteously scorned in the scientific community.  And when it is pointed out (for instance when the black hole information loss problem had been “solved”), the scientists involved are usually on the defensive.

Buddy
Buddy the Elf believes anything Steve Jurvetson ever uttered and then some.

Of course, business follows very different rules, more along the Donald Trump rules of attention. Any BS will do as long as it captures audience. Customers are used to these kinds of commercial exaggerations, and so I am always a bit puzzled by the urge to debunk D-Wave “hype”. To me it feels almost a bit patronizing. The average Joe is not like Buddy the Elf, the unlikely hero of my family’s favorite Christmas movie. When Buddy comes to NYC and sees a diner advertising the world’s best coffee,  he takes this at face value and goes crazy over it.  The average Joe, on the other hand, has been thoroughly desensitized to high tech hype. He knows that neither Google Glasses nor Apple Watches will really change his life forever, nor will he believe Steve Jurvetson that the D-Wave machines will outperform the universe within a couple of years. Steve, on the other hand, does what every good VC business man is supposed to do for a company that he invested in, i.e. create hype. The world has become a virtual bazaar, and your statements have to be outrageous and novel in order to be heard over the noise. What he wants to get across is that the D-Wave machines will grow in performance faster than conventional hardware. Condensing this into Rose’s Law is the perfect pitch vehicle for that – hype with a clear purpose.

People like to pick an allegiance and cheer for their “side”. It is the narrative that has been dominating the D-Wave story for many years, and it made for easy blogging, but I won’t miss it. The hypers gonna hype, the haters gonna hate, but now the nerds should know to trust the published papers.

Max Planck famously quipped that science advances one funeral at a time, because even scientists have a hard time acting completely rationally and adjusting their stances when confronted with new data.  This is the 21st century, here’s to hoping that the scientific community has lost this kind of rigidity, even while most of humanity remains as tribal as ever.

Riding the D-Wave

D-Wave, Quantum Computing,

Update: Thanks to everybody who keeps pointing me to relevant news (Ramsey, Rolf, Sol and everybody else my overtired brain may not recall at this time).

There is no doubt that D-Wave is on a role:

And then there’s the countdown to what is billed as a D-Wave related watershed announcement from Google coming Dec 8th.  Could this be an early Christmas present to D-Wave investors?

 

~~~~~~

dwavetrain_wide

Back in the day before he re-resigned as D-Wave’s chief critic, Scott Aaronson made a well-reasoned argument as to why he thought this academic, and at times vitriolic, scrutiny was warranted. He argued that a failure of D-Wave to deliver a quantum speed-up would set the field back, similar to the AI winter that was triggered by Marvin Minsky’s Perceptrons book.

Fortunately, quantum annealers are not perceptrons. For the latter, it can be rigorously proven that single layer perceptrons are not very useful. Ironically, at the time the book was published, multilayered perceptrons, i.e. a concept that is now fundamental to all deep learning algorithms, were already known, but in the ensuing backlash research funding for those also dried up completely. The term “perceptron” became toxic and is now completely extinct.

Could D-Wave be derailed by a proof that shows that quantum annealing could, under no circumstances, deliver a quantum speed-up? To me this seems very unlikely, not only because I expect that no such proof exists, but also because, even if this was the case, there will still be a practical speed-up to be had. If D-Wave manages to double their integration density at the same rapid clip as in the past, then their machines will eventually outperform any classical computing technology in terms of annealing performance. This article (h/t Sol) expands on this point.

So far there is no sign that D-Wave will slow its manic pace. The company recently released its latest chip generation, featuring quantum annealing with an impressive 1000+ qubits (in practice, the number will be smaller, as qubits will be consumed for problem encoding and software EEC). This was followed with a detailed test under the leadership of Catherine McGeoch, and it will be interesting to see what Daniel Lidar, and other researchers with access to D‑Wave machines, will find.

My expectation has been from the get-go that D-Wave will accelerate the development of this emerging industry, and attract more money to the field. It seems to me that this is now playing out.

Intel recently (and finally as Robert Tucci points out) entered the fray with a $50M investment. While this is peanuts for a company of Intel’s size, it’s an acknowledgement that they can’t leave the hardware game to Google, IBM or start-ups such as Rigetti.

On the software side, there’s a cottage industry of software start-ups hitching their wagons to the D-Wave engine. Many of these are still in stealth mode, or early stage such as QC Ware, while others already start to receive some well deserved attention.

Then there are also smaller vendors of established software and services that already have a sophisticated understanding of the need to be quantum ready. The latter is something I expect to see much more in the coming years as the QC hardware race heats up.

The latest big name entry into the quantum computing arena was Alibaba, but at this time it is not clear what this Chinese initiative will focus on. Microsoft, on the other hand, seems to be a known quantity and will not get aboard the D‑Wave train, but will focus exclusively on quantum gate computing.

Other start-ups, like our artiste-qb.net, straddle the various QC hardware approaches. In our case, this comes “out-of-the-box”, because our core technology, Quantum Bayesian Networks, as developed by Robert Tucci, is an ideal tool to abstract from the underlying architecture. Another start-up that is similarly architecture agnostic is Cambridge QC. The recent news of this company brings to mind that sometimes reality rather quickly imitates satire. While short of the $1B seed round of this April Fool’s spoof, the influx of $50M dollars from the Chile based Grupo Arcano is an enormous amount for a QC software firm, that as far as I know, holds no patents.

Some astoundingly big bets are now being placed in this field.

 

 

 

 

 

The Year That Was <insert expletive of your choice>

D-Wave, Popular Science, Quantum Computing,

Usually, I like to start a new year on an upbeat note, but this time I just cannot find the right fit. I was considering whether to revisit technology that can clean water – lauding the effort of the Bill Gates foundation came to mind, but while I think this is a great step in the right direction, this water reclaiming technology is still a bit too complex and expensive to become truly transformational and liberating.

At other times, a groundbreaking progress in increasing the efficiency of solar energy would have qualified, the key being that this can be done comparatively cheaply. Alas, the unprecedented drop in the price of oil is not only killing off the fracking industry, but also the economics for alternative energy.  For a planet that has had its fill of CO2, fossil fuel this cheap is nothing but an unmitigated disaster.

So while it was a banner year for quantum computing, in many respects 2014 was utterly dismal, seeing the return of religiously motivated genocide, open warfare in Europe, a resurgence of diseases that could be eradicated by now, and a pandemic that caused knee jerk hysterical reactions that taught us how unprepared we are for these kind of health emergencies. This year was so depressing it makes me want to wail along to my favorite science blogger’s song about it (but then again I’d completely ruin it).

And there is another reason to not yet let go of the past, corrections:

With these corrections out of the way I will finally let go of 2014, but with the additional observation that in the world of quantum computing, the new year started very much in the same vein as the old, generating positive business news for D-Wave, which managed to just raise another 29 million dollars, while at the same time still not getting respect from some academic QC researchers.

I.H. Deutsch (please note, not the Deutsch but Ivan) states at the end of this interview:

  1. [1]The D-Wave prototype is not a universal quantum computer.
  2. [2]It is not digital, nor error-correcting, nor fault tolerant.
  3. [3]It is a purely analog machine designed to solve a particular optimization problem.
  4. [4]It is unclear if it qualifies as a quantum device.”

No issues with [1]-[3].  But how many times do classical algos have to be ruled out before D-Wave is finally universally accepted as a quantum annealing machine?  This is getting into climate change denying territory. It shouldn’t really be that hard to define what makes for quantum computation. So I guess we found a new candidate for D-Wave chief critic, after Scott Aaronson seems to have stepped down for good.

Then again, with a last name like Deutsch, you may have to step up your game to get some name recognition of your own in this field.  And there’s no doubt that controversy works.

So 2015 is shaping up to become yet another riveting year for QC news. And just in case you made the resolution that, this year, you will finally try to catch that rainbow, there’s some new tech for you.
SOURCE: chaosgiant.deviantart.com

 

 

Update: Almost forgot about this epic fail of popular science reporting at the tail end of 2014.  For now I leave it as an exercise to the reader to spot everything that’s wrong with it. Of course most of the blame belongs to PLoS ONE which supposedly practices peer review.

The Church of D-Wave

D-Wave, , ,

Are You a BDeliever? dwave_churchgoer Science and religion have a difficult relationship, and sometimes they combine in the most obscure manner, such as when Scientology was conceived.  The latter seems to have lost a lot of its appeal and followers, but it seems that another new religion is poised to grab the mantle.

That is, if one is willing to follow Scott Aaronson’s rationale that believing in the achievability of significant speed-up with D-Wave’s architecture is a matter of faith. Ironically Scott, who is teaching computer science at MIT, made this comment about the same time that the MIT Technology Review named D-Wave to its Top 50 Smartest Companies list. An illustrious selection, that any company would be delighted to be included in. The only quibble I have with this list is that it ranks Elon Musk’s SpaceX before D-Wave, my point being that quantum mechanics is harder than rocket science. After all, with the latter, everybody can decide if your spacecraft made it into orbit or not (classical mechanics is so straightforward).  On the other hand, we still have the ongoing high profile battle over the question of how quantum D-Wave’s machine actually is (since Schroedinger the uncertainty of what’s in a box seems to be a constant in Quantum Mechanics).

Another paper buttresses the company’s claims that there is substantial entanglement present on their chip.  This prompted Prof. Vazirani, who I experienced as a most delightful soft spoken academic when checking out his Quantum Computing MOC, to come out swinging.  The New York Times quotes him as saying:

“What I think is going on here is that they didn’t model the ‘noise’ correctly. (….) One should have a little more respect with the truth.”

In academic parlance these are fighting words.  And so the show goes on.

But I want to take a break from this for a moment, and focus on another question: How did a startup like D-Wave get to this point?  Time magazine front page material, coverage in the New York Times, being named in the same breath as SpaceX.  From a business perspective this is nothing but an amazing success story to have gotten to this point. And to me, the question of what makes successful entrepreneurship is of no less interest than science and technology.

Geordie_Gold
Geordie got closer to having a shot at Olympic gold than most of us, having been an alternate on the Canadian wrestling team at the 1996 Olympic Games, so getting this one may have been bitter sweet.

Flying into Vancouver I imagined Geordie Rose to be a Steve Jobs-like character, about whom it was famously quipped that he was surrounded by his own reality distortion field, an invisible force that made others see the world like he did, and made them buy into his vision. And although I never had the pleasure of meeting Steve Jobs, I think it is safe to say that Geordie is nothing like him. If I had to describe him in one word, I’d say he is quintessentially “Canadian”, in terms of the positive attributes that we typically like to associate with our national character. (Full disclaimer: Technically I am not Canadian yet, just a permanent resident).

Given the amazing success that D-Wave has had, and the awards and accolades that he himself has received, I was impressed with his unassuming demeanor. Hard to imagine Geordie would ever park his car in a handicap spot, as Jobs was fond of doing, to shave a couple minutes off his commute.

D-Wave just moved to a new enlarged premises. In their old building Geordie occupied an interior office without windows. I naturally assumed that he would have upgraded that. So I was surprised to learn that his new workspace still doesn’t have any windows. His explanation was simple, it allows him to be close to his team.

My take away is that visionaries cannot be pigeon-holed, because when talking to Geordie it was quickly obvious that his focus and dedication to making his vision a reality is ironclad, and his excitement is infectious.  So this is one key similarity to Steve Jobs after all, and then there is of course this, which goes without saying:

Great entrepreneurs never do it for the money.
Great entrepreneurs never do it for the money.

Prof. Vazirani must have picked up on D-Wave’s commitment to make Quantum Computing work, as the New York Times also quotes him as saying about D‑Wave that “after talking with them I feel a lot better about them. They are working hard to prove quantum computing.”

That Geordie picked an approach which is so abhorred by theorists, I attribute to yet another aspect that, in my mind, marks great entrepreneurship: An almost ruthless pragmatism. Focusing on the less proven quantum annealing on a chip, he managed in just seven years to turn out an entirely new computing platform.  Meanwhile, the advances in superconducting foundry know-how that his company ushered in, will also benefit other approaches, such as the gate based implementation that UCSB’s John Martinis plans to scale up to 1000 qubits within five years.

To me, there is no doubt that the hurry to get something to the market is a net benefit to the entire quantum computing field, as I expect it will attract more private capital. And that is because Quantum Computing is now no longer perceived as something nebulous, something that just may happen 25 years down the road.

Game changers polarize.  So if we pay heed to Scott Aaronson’s rhetorics Geordie clearly has a leg up over Steve Jobs.  Where the latter had a cult following, Geordie’s on his way to having his own religion.  Maybe that’ll explain the following recent exchange on D-Wave’s blog:

D-Wave_blog

 

(h/t Rolf D. and commenter Copenhagen for pointing me to material for this post.)

He Said She Said – How Blogs are Changing the Scientific Discourse

D-Wave, Popular Science, Quantum Computing, , ,

The debate about D-Wave‘s “quantumness” shows no signs of abating, hitting a new high note with the company being prominently featured on Time magazine’s recent cover, prompting a dissection of the article on Scott Aaronson’s blog. This was quickly followed by yet another scoop: A rebuttal by Umesh Vazirani to Geordie Rose who recently blogged about the Vazirani et al. paper which sheds doubt on D-Wave’s claim to implement quantum annealing. In his take on the Time magazine article Scott bemoans the ‘he said she said’ template of journalism which gives all sides equal weight, while acknowledging that the Times author Lev Grossman quoted him correctly, and obviously tries to paint an objective picture.

If I had to pick the biggest shortcoming of the Times article, my choice would have been different. I find Grossman entirely misses Scott’s role in this story by describing him as “one of the closest observers of the controversy“.

Scott isn’t just an observer in this. For better or worse he is central to this controversy. As far as I can tell, his reporting on D-Wave’s original demo is what started it to begin with. Unforgettable, his inspired comparison of the D-Wave chip to a roast beef sandwich, which he then famously retracted when he resigned as D-Wave’s chief critic. The latter is something he’s done with some regularity, first when D-Wave started to publish results, then after visiting the company and most recently after the Troyer et al. pre-print appeared in arxiv (although the second time doesn’t seem to count, since it was just a reiteration of the first resignation).

And the say sandwiches and chips go together ...Scott’s resignations never seem to last long. D-Wave has a knack for pushing his buttons. And the way he engages D-Wave and associated research is indicative of a broader trend in how blogs are changing the scientific discourse.

For instance, when Catherine McGeoch gave a talk about her benchmarking of the DW2, Scott did not immediately challenge her directly but took to his blog (a decision he later regretted and apologized for). Anybody who has spent more than five minutes on a Web forum knows how the immediate, yet text only, communication removes inhibitions and leads to more forceful exchanges. In the scientific context, this has the interesting effect of colliding head on with the more lofty perception of a scientist.

It used to be that arguments were only conducted via scientific publications, in person such as in scientific seminars, or the occasional letter exchange. It’s interesting to contemplate how corrosive the arguments between Bohr and Einstein may have turned out, if they would have been conducted via blogs rather than in person.

But it’s not all bad. In the olden days, science could easily be mistaken for a bloodless intellectual game, but nobody could read through the hundreds of comments on Scott’s blog that day and come away with that impression. To the contrary, the inevitable conclusion will be that science arguments are fought with no less passion than the most heated bar brawl.

During this epic blog ‘fight’ Scott summarized his preference for the media thusly

“… I think this episode perfectly illustrates both the disadvantages and the advantages of blogs compared to face-to-face conversation. Yes, on blogs, people misinterpret signals, act rude, and level accusations at each other that they never would face-to-face. But in the process, at least absolutely everything gets out into the open. Notice how I managed to learn orders of magnitude more from Prof. McGeoch from a few blog comments, than I did from having her in the same room …”

it is by far not the only controversy that he courted, nor is this something unique to his blog. Peter Woit continues the heretical work he started with his ‘Not Even Wrong’ book, Robert R. Tucci fiercely defends his quantum algorithm work when he feels he is not credited, Sabine Hossenfelder had to ban a highly qualified String theory troll due to his nastiness (she is also a mum of twins, so you know she has practice in being patient, and it’s not like she doesn’t have a good sense of humor). But my second favorite science blog fight also occurred on Scott’s blog when Joy Christian challenge him to a bet to promote his theory that supposedly invalidates the essential non-locality of quantum mechanics due to Bell’s theorem.

It’s instructive to look at the Joy Christian affair and ask how a mainstream reporter could have possibly reported it. Not knowing Clifford algebra, what could a reporter do but triangulate the expert opinions? There are some outspoken smart critics that point to mistakes in Joy Christian’s reasoning, yet he claims that these are based on flawed understanding and have been repudiated. The reporter will also note that doubting Bell’s theorem is very much a minority position, yet such a journalist not being able to check the math himself can only fall back on the ‘he said she said’ template. After all, this is not a simple straight forward fact like reporting if UN inspectors found Saddam Hussein’s weapons of mass distractions or not (something that surprisingly most mainstream media outside the US accomplished just fine). One cannot expect a journalist to settle an open scientific question.

The nature of the D-Wave story isn’t different, how is Lev Grossman supposed to do anything but report the various stances on each side of the controversy? A commenter at Scott’s blog was dismissively pointing out that he doesn’t even have a science degree. As if this were to make any difference, it’s not like everybody else on each side of the story doesn’t boast such degrees (non-PhDs are in the minority at D-Wave).

Mainstream media reports as they always did, but unsettled scientific questions are the exception to the rule, one of the few cases when ‘he said she said’ journalism is actually the best format. For everything else we fortunately now have the blogs.

Scott Aaronson (again) resigns as chief D-Wave critic and endorses their experiments

D-Wave, Quantum Computing

An exercise in positive spin.

Update below.

The English language is astoundingly malleable. It feels almost as if it was tailor made for marketing spin. I noticed this long ago (feels like a lifetime) when working in a position that required me to sell software. Positioning products was much easier when I spoke English.  Mind you, I never told a blatant lie, but I certainly spun the facts to put our product in the best light, and if a customer committed I’d do my darnedest to deliver the value that I promised. The kind of customers I dealt with were of course aware of this dance, and perfectly capable of performing their due diligence. From their perspective, in the end, it is always about buying into the vision, knowing full well that a cutting edge technology, one that will give a real competitive benefit, will of course be too new to be without risk.

During the courting of the customers, any sales person worth their salt will do anything to make the product look as good as possible. One aspect of this is of course to stress positive things that others are saying about your offerings.

To accomplish this, selective quoting can come in very handy. For instance, after reviewing the latest pre-print paper that looks at D-Wave’s 503 qubit chip performance, Scott Aaronson stepped down for the second time as chief D-Wave critic. In the blog post where he announced this, he also observed that on “the ~10% of instances on which the D-Wave machine does best, (…) the machine does do slightly better (…) than simulated annealing”.

This puts in words what the following picture shows in much more detail.

Screenshot 2014-01-18 17.47.52
Instance-by-instance comparison of annealing times and wall-clock times. Shown is a scatter plot of the pure annealing time for the DW2 compared to a simulated classical annealer (SA) using an average over 16 gauges on the DW2. This is figure 6 of the recent benchmark paper. Wall clock times include the time for programming, cooling, annealing, readout and communication. Gauges refer to different encodings of a problem instance. (Only plot A and B are relevant to the settling of my bet).

Now, if you don’t click through to Scott’s actual blog post. you may take away that he actually changed his stance. But of course he hasn’t. You can look at the above picture and think the glass is ninety percent empty or you could proclaim it is ten percent full.

The latter may sound hopelessly optimistic, but let’s contemplate what we are actually comparing.  Current computer chips are the end product of half a century highly focused R&D, with billions of dollars poured into developing them. Yet, we know we are getting to the end of the line of Moore’s law. Leak currents already are a real problem, and the writing is on the wall that we are getting ever closer to the point where the current technology will no longer allow for tighter chip structures.

On the other hand, the D-Wave chip doesn’t use transistors. It is an entirety different approach to computing, as profoundly different as the analog computers of yore.

The integration density of a chip is usually classified by the length of the silicon channel between the source and drain terminals in its field effect transistors (e.g. 25nm). This measure obviously doesn’t apply to D-Wave, but the quantum chip integration density isn’t even close to that. With the ridiculously low number of about 500 qubits on D-Wave’s chip, which was developed on a shoestring budget when compared to the likes of Intel or IBM, the machine still manages to hold its own against a modern CPU.

Yes, this is not a universal gate-based quantum computer, and the NSA won’t warm up to it because it cannot implement Shore’s algorithm, nor is there a compelling theoretical reason that you can achieve a quantum speed-up with this architecture. What it is, though, is a completely new way to do practical computing using circuit structures that leave plenty of room at the bottom.  In a sense, it is resetting the clock to when Feynman delivered his famous and prophetic talk on the potentials of miniaturization. Which is why from a practical standpoint I fully expect to see a D-Wave chip eventually unambiguously outperform a classical CPU.

On the other hand, if you look at this through the prism of complexity theory none of this matters, only proof of actual quantum speed-up does.

Scott compares the quantum computing skirmishes he entertains with D-Wave to the American Civil war.

If the D-Wave debate were the American Civil War, then my role would be that of the frothy-mouthed abolitionist pamphleteer

Although clearly tongue in cheek, this comparison still doesn’t sit well with me.  Fortunately, in this war, nobody will lose life or limb. The worst that could happen is a bruised ego, yet if we have to stick to this metaphor, I don’t see this as Gettysburg 1863 but the town of Sandwitch 1812.

Much more will be written on this paper. Once it has fully passed peer review and been published, I will also be finally able to reveal my betting partner. But in the meantime there a Google+ meeting scheduled that will allow for more discussion (h/t Mike).

Update

Without careful reading of the paper a casual observer may come away with the impression that this test essentially just pitted hardware against hardware. Nothing could be further from the truth, some considerable effort had to go into constructing impressive classical algorithms to beat the D-Wave machine on its own turf.  This Google Quantum AI lab post elaborates on this (h/t Robert R. Tucci).

Update 2

D-Wave’s Geordie Rose weighs in.

 

 

 

 

 

 

Out of the AI Winter and into the Cold

D-Wave, Popular Science, , , , , ,
dwave_log_temp_scale
A logarithmic scale doesn’t have the appropriate visual impact to convey how extraordinarily cold 20mK is.

Any quantum computer using superconducting Josephson junctions will have to be operated at extremely low temperatures. The D-Wave machine, for instance, runs at about 20 mK, which is much colder than anything in nature (including deep space). A logarithmic scale like the chart to the right, while technically correct, doesn’t really do this justice.  This animated one from D-Wave’s blog shows this much more drastically when scaled linearly (the first link goes to an SVG file that should play in all modern browsers, but it takes ten seconds to get started).

Given that D-Wave’s most prominent use case is the field of machine learning, a casual observer may be misled to think that the term “AI winter” refers to the propensity of artificial neural networks to blossom in this frigid environment. But what the term actually stands for is the brutal hype cycle that ravaged this field of computer science.

One of the original first casualties of the collapse of artificial intelligence research in 1969 was the ancestor of the kind of learning algorithms that are now often implemented on D-Wave’s machines. This incident is referred to as the XOR affair, and the story that circulates goes like this:  “Marvin Minsky, being a proponent of structured AI, killed off the connectionism approach when he co-authored the now classic tome, Perceptrons. This was accomplished by mathematically proving that a single layer perceptron is so limited it cannot even be used (or trained for that matter) to emulate an XOR gate. Although this does not hold for multi-layer perceptrons, his word was taken as gospel, and smothered this promising field in its infancy.”

Marvin Minsky begs to differ, and argues that he of course knew about the capabilities of artificial neural networks with more than one layer, and that if anything, only the proof that working with local neurons comes at the cost of some universality should have had any bearing.  It seems impossible to untangle the exact dynamics that led to this most unfortunate AI winter, yet in hindsight it seems completely misguided and avoidable, given that a learning algorithm (Backpropagation) that allowed for the efficient training of multi-layer perceptrons had already been published a year prior, but at the time it received very little attention.

The fact is, after Perceptrons was published, symbolic AI flourished and connectionism was almost dead for a decade. Given what the authors wrote in the forward to the revised 1989 edition, there is little doubt how Minsky felt about this:

“Some readers may be shocked to hear it said that little of significance has happened in this field [since the first edition twenty year earlier]. Have not perceptron-like networks under the new name connectionism – become a major subject of discussion at gatherings of psychologists and computer scientists? Has not there been a “connectionist revolution?” Certainly yes, in that there is a great deal of interest and discussion. Possibly yes, in the sense that discoveries have been made that may, in time, turn out to be of fundamental importance. But certainly no, in that there has been little clear-cut change in the conceptual basis of the field. The issues that give rise to excitement today seem much the same as those that were responsible for previous rounds of excitement. The issues that were then obscure remain obscure today because no one yet knows how to tell which of the present discoveries are fundamental and which are superficial. Our position remains what it was when we wrote the book: We believe this realm of work to be immensely important and rich, but we expect its growth to require a degree of critical analysis that its more romantic advocates have always been reluctant to pursue – perhaps because the spirit of connectionism seems itself to go somewhat against the grain of analytic rigor.” [Emphasis added by the blog author]

When fast-forwarding to 2013 and the reception that D-Wave receives from some academic quarters, this feels like deja-vu all over again. Geordie Rose, founder and current CTO of D-Wave, unabashedly muses about spiritual machines, although he doesn’t strike me as a particularly romantic fellow. But he is very interested in using his amazing hardware to make for better machine learning, very much in “the spirit of connectionism”.  He published an excellent mini-series on this at D-Wave’s blog (part 1, 2, 3, 4, 5, 6, 7).  It would be interesting to learn if Minsky was to find fault with the analytic rigor on display here (He is now 86 but I hope he is still going as strong as ten years ago when this TED talk was recorded).

So, if we cast Geordie in the role of the 21st century version of Frank Rosenblatt (the inventor of the original perceptron) then we surely must pick Scott Aaronson as the modern day version of Marvin Minsky.  Only that the argument this time is not about AI, but how ‘quantum’ D-Wave’s device truly is.  The argument feels very similar: On one side, the theoretical computer scientist, equipped with boat-loads of mathematical rigor, strongly prefers the gate model of quantum computing. On the other one, the pragmatist, whose focus is to build something usable within the constraints of what chip foundries can produce at this time.

But the ultimate irony, it seems, at least in Scott Aaronson’s mind, is that the AI winter is the ultimate parable of warning to make his case (as was pointed out by an anonymous poster to his blog).  I.e. he thinks the D-Wave marketing hype can be equated to the over-promises of AI research in the past. Scott fears that if the company cannot deliver, the babe (I.e. Quantum Computing) will be thrown out with the bathwater, and so he blogged:

“I predict that the very same people now hyping D-Wave will turn around and—without the slightest acknowledgment of error on their part—declare that the entire field of quantum computing has now been unmasked as a mirage, a scam, and a chimera.”

A statement that of course didn’t go unchallenged in the comment section (Scott’s exemplary in allowing this kind of frankness on his blog).

I don’t pretend to have any deeper conclusion to draw from these observations, and will leave it at this sobering thought: While we expect science to be conducted in an eminently rational fashion, history gives ample examples of how the progress of science happens in fits and starts and is anything but rational.

Quantum Computing Hype Cycle and a Bet of my Own

D-Wave, Quantum Computing, Uncategorized

box in troubleThe year 2013 started turbulent for the quantum computing field with a valiant effort by long time skeptic and distinguished experimentalist Michel  I. Dyakonov  to relegate it to the status of a pathological science akin to cold fusion (he does not use the term in his paper but later stated: “The terms ‘failed, pathological’ are not mine, but the general sense is correct.”).

Scott Aaranson took on this paper in his unique style (it’s a long read but well worth it). There really isn’t much to add to his arguments, but there is another angle that intrigues my inner “armchair psychologist”:  What exactly is it about this field that so provokes some physicists?  Is it that …

  • … Computer Scientists of all people are committing Quantum Mechanics?
  • … these pesky IT nerds have the audacity to actually take the axioms of Quantum Mechanics so seriously as to regard them as a resource for computational engineering?
  • … this rabble band of academics are breeding papers at a rabbit’s pace, so that no one can possibly keep up and read them all?
  • … quantum information science turned the ERP paradox on its head and transformed it into something potentially very useful?
  • … this novel science sucks up all sorts of grant money?

The answer is probably all of the above, to some extent.  But this still doesn’t feel quite right.  It seems to me the animosity goes deeper.  Fortunately, Kingsley Jones (whom I greatly admire) blogged about similar sentiments, but he is much more clear eyed on what is causing them.

It seems to me that the crux of this discomfort stems from the fact that many physicists have a long harbored discomfort with Quantum Mechanic’s intractabilities, which were plastered over with the Copenhagen Interpretation (which caused all sorts of unintended side effects).  It’s really a misnomer, it should have been called the ostrich interpretation, as its mantra was to ignore the inconstancies and to just shut-up and calculate. It is the distinct merit of Quantum Information science to have dragged this skeleton out of the closet and made it dance.

The quantum information scientists are agnostic on the various interpretations, and even joke about it.  Obviously, if you believe there is a truth to be found, there can be only one, but you first need to acknowledge the cognitive dissonance if there’s to be any chance of making progress on this front. (My favorite QM interpretation has been suggested by Ulrich Mohrhoff, and I have yet to find the inspiration to blog about this in an manner that does it justice – ironically, where he thinks of it as an endpoint, I regard it as allowing for a fresh start).

Meanwhile, in the here and now, the first commercial quantum computing device the D‑Wave One has to overcome its own challenges (or being relegated to a computing curiosity akin to analog neural VLSI).  2013 will be the year to prove its merits in comparison to conventional hardware. I’ve been in touch with a distinguished academic in the field (not Scott A.) who is convinced that optimization on a single conventional CPU will always outperform the D-Wave machines – even on the next generation chip. So I proposed a bet, albeit not a monetary one: I will gladly ship a gallon of Maple sirup to him if he is proven right and our dark horse Canadian trail blazer don’t make the finish line. The results should be unambiguous and will be based on published research, but just in case, if there should be any disagreement we will settle on Scott Aaronson as a mutually acceptable arbiter.  Scott is blissfully unaware of this, but as he is also the betting kind (the really big ones), I hope he’d be so kind as to help us sort this out if need be. After all, I figure, he will be following the D-Wave performance tests and at that time will already have formed an informed opinion on the matter.

The year 2013 started off with plenty of QIS drama and may very well turn out to be the crucial one to determine whether the field has crossed the rubicon.   It’s going to be a fun one.

The Unbearable Lightness of Quantum Mechanics

Popular Science, , , , ,

Updated below.

Gravity and Quantum Mechanics don’t play nice together. Since Einstein’s time, we have two towering theories that have defied all attempts by some very smart people to be reconciled. The Standard Model, built on the foundations of quantum mechanics, has been spectacularly successful. It allows the treatment of masses acquired from the binding energies, and, if the Higgs boson confirmation pans out, accounts for the elemental rest masses – but it does not capture gravity. (The current mass generation models that involve gravity are all rather speculative at this point.)

Einstein’s General Relativity has been equally successful in explaining gravity as innate geometric attributes of space and time itself. It has survived every conceivable test and made spectacular predictions (such as gravity lenses).

On the surface this dysfunctional non-relationship between the two major generally accepted theoretical frameworks seems very puzzling. But it turns out that the nature of this conundrum can be described without recourse to higher math (or star-trek like animations with a mythical sound-track).

Much has been written about the origin of this schism: The historic struggle for the interpretation of Quantum Mechanics, with Einstein and Bohr being the figureheads of the divided physics community at the time. Mendel Sachs (who, sadly, passed away recently) drew the following distinction between the philosophies of the two fractions:

[The Copenhagen Interpretation views] Schroedinger’s matter waves as [complex] waves of probability. The probability was then tied to quantum mechanics as a theory of measurement – made by macro observers on micro-matter. This view was then in line with the positivistic philosophy, whereby the elements of matter are defined subjectively in terms of the measurements of their properties, expressed with a probability calculus. […] Einstein’s idea [was] that the formal expression of the probability calculus that is called quantum mechanics is an incomplete theory of matter that originates in a complete [unified] continuous field theory of matter wherin all of the variables of matter are ‘predetermined’.

(From Quantum Mechanics and Gravity)

These days, the Copenhagen Interpretation no longer reigns supreme, but has some serious competition: E.g. one crazy new kid on the block is the Many World Interpretation.  (For an insightful take on MWI I highly recommend this recent blog post from Scott Aaronson).

But the issue goes deeper than that. No matter what interpretation you favor, one fact remains immutable: Probabilities will always be additive, mathematically they behave in a linear fashion. This, despite its interpretative oddities, makes Quantum Mechanics fairly easy to work with.  On the other hand, general relativity is an intrinsically non-linear theory.  It describes a closed system in which the field, generated by gravitating masses, propagates with finite speed and, in a general, non-equilibrium picture, dynamically affects these masses, in turn rearranging the overall field expression.  (Little wonder Einstein’s field equations only yield to analytical solutions for drastically simplified scenarios).

There is no obvious way to fit Quantum Mechanics, this linear peg, into this decidedly non-linear hole.

Einstein considered Quantum Mechanics a theory that would prove to be an approximation of a fully unified field theory.  He spent his last years chasing after this goal, but never achieved it. Mendel Sachs claims to have succeeded where he failed, and indeed presents some impressive accomplishments, including a way to derived the quantum mechanics structure from extended General Relativity field equations.  What always struck me as odd is how little resonance this generated, although this clearly seems to be an experience shared by other theoretical physicists who work off the beaten path. For instance, Kingsley Jones approaches this same conundrum from a completely different angle in his original paper on Newtonian Quantum Gravity. Yet the citation statistic shows that there was little up-take.

One could probably dedicate an entire blog speculating on why this kind of research does not break into the mainstream, but I would rather end this with the optimistic notion that in the end, new experimental data will hopefully rectify this situation. Although the experiment on a neutral particle Bose-Einstein condensate proposed in Kingsley Jones’ paper has little chance of being performed unless there is some more attention garnered, other experiments to probe the domain where gravity and quantum mechanics intersect get a more lofty treatment: For instance this paper was featured in Nature although its premise is probably incorrect. (Sabine Hossenfelder took Nature and the authors to task on her blog – things get a bit acrimonious in the comment section).

Nevertheless, it is encouraging to see such a high profile interest in these kinds of experiments, chances are we will get it right eventually.

Update

Kingsley Jones (who’s 1995 paper paper I referenced above) has a new blog entry that reflects on the historic trajectory and current state of quantum mechanics.  I think it’s fair to say that he does not subscribe to the Many World Interpretation.

 

 

 

Scott Aaronson Visits D-Wave – No Casualties Reported

D-Wave, Quantum Computing

Those following the quantum computing story and D-Wave are well aware of the controversial scuffle between Scott and the company.  So it’s quite newsworthy that according to Scott’s own accord the hatchet has been buried. No more comparing the D‑Wave One to a roast beef sandwich (hopefully BLT is out of the picture too).

Scott is still taking on D-Wave’s pointy-haired bosses though.  He wants them to open the purse-strings to determine more clearly how “quantum” the Rainier chip really is.

Unfortunately, I think he overlooks that pointy-haired bosses are only interested in the bottom line.  At this point there is nothing stopping D-Wave from selling their system as a quantum computer. Not a universal one but who’s counting.  Any closer inquiry into the nature of their qbits only carries the danger to add qualifiers to this claim.  So why should they bother?

In terms of marketing, the label “Quantum Computer” is just a nice nifty term signifying to potential customers that this is a shiny, cool new device.  Something different.  It is supposed to serve as a door opener for sales.  Afterwards it just comes down to the price/performance ratio.

At this point D-Wave One sales won’t benefit from further clarifying how much entanglement is occurring in their system – I see this only change once there is actual competition in this space.

Update:  I finally don’t have to develop a cognitive dissonance about adding Scott’s and D‑Wave’s blog to my blogroll. So to celebrate this historic peace accord this is my next step in the development of this most important website.  BTW if you still need an argument why it is worthwhile to pay attention to Scott, treat yourself to his TedX talk (h/t to Perry Hooker).