Grading Trudeau on quantum computing
Update (4/19): Inspired by Trudeau’s performance (which they clocked at 35 seconds), Maclean’s magazine asked seven quantum computing researchers—me, Krysta Svore, Aephraim Steinberg, Barry Sanders, Davide Venturelli, Martin Laforest, and Murray Thom—to also explain quantum computing in 35 seconds or fewer. You can see all the results here (here’s the audio from my entry).
The emails starting hitting me like … a hail of maple syrup from the icy north. Had I seen the news? Justin Trudeau, the dreamy young Prime Minister of Canada, visited the Perimeter Institute for Theoretical Physics in Waterloo, one of my favorite old haunts. At a news conference at PI, as Trudeau stood in front of a math-filled blackboard, a reporter said to him: “I was going to ask you to explain quantum computing, but — when do you expect Canada’s ISIL mission to begin again, and are we not doing anything in the interim?”
Rather than answering immediately about ISIL, Trudeau took the opportunity to explain quantum computing:
“Okay, very simply, normal computers work, uh, by [laughter, applause] … no no no, don’t interrupt me. When you walk out of here, you will know more … no, some of you will know far less about quantum computing, but most of you … normal computers work, either there’s power going through a wire, or not. It’s 1, or a 0, they’re binary systems. Uh, what quantum states allow for is much more complex information to be encoded into a single bit. Regular computer bit is either a 1 or a 0, on or off. A quantum state can be much more complex than that, because as we know [speeding up dramatically] things can be both particle and wave at the same times and the uncertainty around quantum states [laughter] allows us to encode more information into a much smaller computer. So, that’s what exciting about quantum computing and that’s… [huge applause] don’t get me going on this or we’ll be here all day, trust me.”
What marks does Trudeau get for this? On the one hand, the widespread praise for this reply surely says more about how low the usual standards for politicians are, and about Trudeau’s fine comic delivery, than about anything intrinsic to what he said. Trudeau doesn’t really assert much here: basically, he just says that normal computers work using 1’s and 0’s, and that quantum computers are more complicated than that in some hard-to-explain way. He gestures toward the uncertainty principle and wave/particle duality, but he doesn’t say anything about the aspects of QM most directly relevant to quantum computing—superposition or interference or the exponential size of Hilbert space—nor does he mention what quantum computers would or wouldn’t be used for.
On the other hand, I’d grade Trudeau’s explanation as substantially more accurate than what you’d get from a typical popular article. For pay close attention to what the Prime Minister never says: he never says that a qubit would be “both 0 and 1 at the same time,” or any equivalent formulation. (He does say that quantum states would let us “encode more information into a much smaller computer,” but while Holevo’s Theorem says that’s false for a common interpretation of “information,” it’s true for other reasonable interpretations.) The humorous speeding up as he mentions particle/wave duality and the uncertainty principle clearly suggests that he knows it’s more subtle than just “0 and 1 at the same time,” and he also knows that he doesn’t really get it and that the journalists in the audience don’t either. When I’m grading exams, I always give generous partial credit for honest admissions of ignorance. B+.
Anyway, I’d be curious to know who at PI prepped Trudeau for this, and what they said. Those with inside info, feel free to share in the comments (anonymously if you want!).
(One could also compare against Obama’s 2008 answer about bubblesort, which was just a mention of a keyword by comparison.)
Update: See also a Motherboard article where Romain Alléaume, Amr Helmy, Michele Mosca, and Aephraim Steinberg rate Trudeau’s answer, giving it 7/10, no score, 9/10, and 7/10 respectively.
Follow
Comment #1 April 17th, 2016 at 10:27 am
I wondered when I heard him use the word “complex” whether he meant it only in the sense of “not simple”, or whether his prep materials had explained complex amplitudes and he had partially remembered that quantum states have something to do with being “complex”.
Comment #2 April 17th, 2016 at 10:57 am
dlyongemallo: Haven’t heard from you in a while; hope you’re well!
I interpreted “complex” to just mean “complicated” rather than “a+bi,” especially given Trudeau’s second use of the word: “A quantum state can be much more complex than that…” Also, most popular materials about quantum computing never mention that amplitudes are complex numbers—often, they don’t even mention amplitudes at all. I could be wrong though.
Comment #3 April 17th, 2016 at 11:07 am
“nor does he mention what quantum computers would or wouldn’t be used for.”
Doesn’t he say they will result in smaller computers… Which I don’t think is *quite* right.
Comment #4 April 17th, 2016 at 11:16 am
Before the speech, there was about an hour of Trudeau talking with faculty and students while having photo’s taken by the press, mostly in front of the ever-present equation-covered blackboard. I don’t think he had any private coaching time, but one of the faculty who spoke with him can correct me.
Comment #5 April 17th, 2016 at 11:17 am
Isn’t wrong or very misleading to say that the non-binary aspect of qubits allows you to “encode more information” into a smaller space than regular bits, because of the result that a qubit could never be set in a way that allowed recovery of more than one bit?
Comment #6 April 17th, 2016 at 11:20 am
Yonemoto and Silas Barta: He says that QCs would let you “encode more information into a much smaller computer.” That’s true if you’re talking about communication complexity, or about how much classical information is needed to describe a quantum state, but it’s false if you’re talking about how much classical information can be retrieved from a quantum state. And yes, without further clarification, it could easily suggest a common wrong view, that a quantum computer is just like a classical computer but smaller or denser or something. That phrase is one reason why I can’t give Trudeau’s answer higher than a B+, even though his humor puts me in a generous mood.
Comment #7 April 17th, 2016 at 11:36 am
He seems to suggest at the end that he could explain them in far greater detail “Don’t get me going on this or we’ll be here all day, trust me.” I would be interested to know if he could have elaborated more on the description or if he just said it as a joke to wrap up the question.
Comment #8 April 17th, 2016 at 11:54 am
Jesse #7: I took that as clearly a joke. Now, if someone had taken him literally, would he, at that moment, have been able to add any additional true or semi-true statements about quantum computing? I have no idea! But maybe someone who met with him before the news conference could say.
Comment #9 April 17th, 2016 at 12:02 pm
Perhaps he reads your blog occasionally.
Comment #10 April 17th, 2016 at 12:13 pm
Look who was showing Trudeau around PI shortly before the press conference: https://twitter.com/Perimeter/status/721111011290431488
Comment #11 April 17th, 2016 at 12:18 pm
According to Jonathan Baugh on Facebook, the press conference occurred about 10min after Trudeau was talking with Ray: https://www.facebook.com/groups/qinfo.scientists.unite/permalink/10156843095955338/
Comment #12 April 17th, 2016 at 12:19 pm
who cares scott. post something more interesting.
Comment #13 April 17th, 2016 at 12:23 pm
#12: Just trying to get back into blog-practice after a long break—I’ve got some bigger stuff in the pipeline. 🙂
Comment #14 April 17th, 2016 at 12:31 pm
Scott #2,
When Trudeau says “A quantum state can be much more complex than that…”, I think he is hinting that he has discovered that quaternions and octonions are the key to progress in QC.
Comment #15 April 17th, 2016 at 12:49 pm
Seriously, what would you say if you had 10 secs to explain the principle to the press?
Comment #16 April 17th, 2016 at 12:53 pm
Hi Scott,
Why don’t you like the formulation “a qubit can be both 0 and 1 at the same time”? Whenever I’ve tried to explain superposition I end up resorting to similar phrases, and this one seems pretty harmless to me.
Comment #17 April 17th, 2016 at 1:15 pm
sustrik #15: I do that pretty much every week!
I say something about how a QC is a proposed device that would solve certain specific problems much faster than we know how to solve them today, by taking advantage of quantum mechanics, which generalizes the laws of classical probability. Then I talk about how you’d never talk about a -20% chance of rain tomorrow, but quantum mechanics is based on numbers called amplitudes, which can be positive or negative or even complex numbers. And how, if an event can happen one way with a positive amplitude and another way with a negative amplitude, the two possibilities can “interfere destructively” and cancel each other out, so that the event never happens at all. And how the state of a QC with (say) 1000 bits would have one amplitude for each of 21000 possible settings of the bits—an astronomical amount of information, if one wanted to write it down classically, for example in order to simulate what the QC was doing classically. But about how, when you measure the QC’s state, you just see a single random output (with its probability determined by its amplitude), not the gargantuan list of possibilities. And about how the goal, in QC, is always to choreograph things so that the possible paths leading to each wrong answer interfere destructively and cancel each other out, (say) some having positive amplitudes and others negative, whereas the paths leading to right answer reinforce. And how this is a very weird and specialized capability—it’s not nearly as simple as “trying all the answers in parallel” (if you did that, you’d simply observe a random answer), nor is it just a smaller or faster version of ordinary computing (a QC might even be “bigger” or “slower” than an ordinary one; all the hoped-for advantage comes from the QC’s ability to create interference patterns). Finally I talk about how a QC is known to give huge advantages over any known classical algorithm for a few tasks of practical importance (quantum simulation, breaking almost all the crypto used today…), and it might also give some advantages for broader goals like optimization and machine learning, but that’s an active research topic, and if the advantages exist they’ll probably be more modest and/or specialized.
So, OK, I can’t do it in 10 seconds, but I can do it in 10 minutes. 🙂
Comment #18 April 17th, 2016 at 1:17 pm
Here’s the lone critical take I’ve run across:
http://www.jjmccullough.com/index.php/2016/04/16/the-north-koreanificiation-of-canadian-political-reporting/
(I have no idea / opinion how trustworthy the source is.)
Comment #19 April 17th, 2016 at 1:46 pm
Ashley #16:
Why don’t you like the formulation “a qubit can be both 0 and 1 at the same time”?
Because my experience tells me that people will round that off to “a QC would get its advantage by exploring every possible answer at the same time—for example, every possible divisor of the number that you’re trying to factor.”
It also gives them a false assurance that “superposition” fits neatly into one of our pre-existing conceptual categories—like the concept of “this and also that” (happy and also sad / working on task 1 and also working on task 2)—and that all the other stuff is just technical details. Which sounds totally plausible if you’ve never seen QM before, but is wrong—and will leave even an otherwise-smart person an easy mark for hypesters later on.
So, in whatever time I’m allotted, I try to get across the truth: that in the mid-1920s, physicists discovered an entirely new conceptual category used by Nature—one that’s neither “this and that,” nor “maybe this and maybe that,” but “a complex linear combination of this and that.”
If you had to round this new concept down to an everyday one, “maybe this and maybe that” would probably come the closest. But superposition generalizes the everyday concept of “maybe this and maybe that,” because the “this” and the “that” can actually interfere and cancel each other out or reinforce each other—which then forces us to accept that, in “maybe this and maybe that,” the “maybe” wasn’t merely in our heads; it was out there in the universe.
You can then define a QC as simply a computer that would exploit this new kind of “maybe”: the one that was discovered in the 1920s and involves complex numbers and is out there in the universe.
Some of this just comes down to individual taste. But personally, I’d much rather have a layperson half-understand or quarter-understand that the world itself coughed up a new kind of “maybe” (or a new kind of “and”), which all of physics is now based on and which a QC would take advantage of, than completely understand what ain’t so, that a qubit is “both 0 and 1” or “maybe 0 or maybe 1,” with “and,” “or,” and “maybe” retaining their intuitive meanings.
Comment #20 April 17th, 2016 at 1:55 pm
Also could someone explain, does the fact that “things can be both particle and wave at the same times” has anything to do with quantum computers?
Comment #21 April 17th, 2016 at 2:04 pm
Thanks for the detailed reply! I look forward to the successor to “Quantum Computing Since Democritus”: “A New Kind of Maybe” 🙂
Comment #22 April 17th, 2016 at 2:13 pm
leo #20: “Wave/particle duality” is just an old-fashioned way of talking about the central assertion of QM: namely, that physical systems evolve as “waves”—that is, as complex linear combinations of classical configurations—as long as they’re isolated, but then probabilistically “collapse” to a definite outcome (e.g., a “particle” with some definite location) when they’re measured. The language pays homage to Bohr’s ideology of everything having multiple complementary aspects.
In quantum computing, we instead use a language derived from Dirac and von Neumann: superpositions, state vectors, amplitudes, unitary transformations, etc. (With various modern additions: qubit, qutrit, gate, Hadamard, CNOT, ancilla, POVM, superoperator, …) This language is more abstract and general, and in my opinion clearer, but is ultimately talking about the same thing.
Comment #23 April 17th, 2016 at 2:21 pm
How he dodged the tough question about Canada’s response to ISIL in favor of the “easier” one about quantum computing — now that was funny.
Comment #24 April 17th, 2016 at 3:10 pm
“He does say that quantum states would let us “encode more information into a much smaller computer,” but while Holevo’s Theorem says that’s false for a common interpretation of “information,” it’s true for other reasonable interpretations.”
Could you elaborate a bit (or at least name) some of the other interpretations?
Comment #25 April 17th, 2016 at 3:14 pm
Shai #24: I did so in comment #6: “That’s true if you’re talking about communication complexity, or about how much classical information is needed to describe a quantum state [as opposed to how much can be retrieved by measuring the state].”
Comment #26 April 17th, 2016 at 3:56 pm
Steve #23: “he dodged the tough question about…ISIL”
One news report I read said that he answered *both* questions in the order presented. The viral clip only shows the first answer, and not the second. I have not seen any clip that shows the second answer, however, only mention of it.
Comment #27 April 17th, 2016 at 5:55 pm
So, there’s a probability that he answered about QCs first, and a probability that he answered about ISIS first. As we increase entropy of the situation by bombarding it with posts, we force the arrangement of questions to collapse into a single sequence, which we can then observe here. Ok, I think I’m getting this 🙂
Comment #28 April 17th, 2016 at 6:03 pm
One thing’s for sure: If it had been Stephen Harper instead of Justin Trudeau who had clumsily dodged the Islamic State question and changed the subject like that, the media would have torn him apart.
Comment #29 April 17th, 2016 at 6:40 pm
Since the destructive interference is the computationally interesting part, why not focus exclusively on that? IMO the quantum terminology just gets in the way.
In a sense, a quantum computer is like a classical computer that can do less. There are fewer execution paths it can take, and we map the wrong answers to the impossible paths.
It’s like a computer that only takes execution paths of even length, or that don’t terminate at 5PM on a Saturday.
A QC is really just a computer that takes some time off, that’s what it is. It’s a lazy computer.
Comment #30 April 17th, 2016 at 6:44 pm
Job #29: No, because if you really wanted to define the QC model, you’d need to say something about unitarity, and tensor products, and the Born measurement rule. And if you wanted to understand the model, you’d need to talk about mixed states and complementary measurement bases and all the rest. So you’d end up reinventing a huge amount of “quantum terminology,” just calling it something different.
Comment #31 April 17th, 2016 at 7:09 pm
That’s pretty ambitious, depending on the audience.
I think what ends up happening is that people stop halfway, leaving stuff like quantum superposition and 2^1000 in the air, before getting to the key parts.
Comment #32 April 17th, 2016 at 7:48 pm
🙂 I wondered what you’d say, and I was also watching his face and waiting to see if he’d attempt the “both 0 and 1 at the same time” part. And I was slightly disappointed that he didn’t, but I agree – he had that “in over my head” look, and it looked to me like he was trying to figure out what sound bite he could put across to the press without getting himself in trouble. I also agree with what Jennifer Ouellette said, about how this shouldn’t be a spectacular moment, that this shouldn’t be too much to expect of a national leader.
(I was trying to figure out why your pocket explanation seems so familiar and then realized I’ve been reading something similar to do with explanations of 2D IR spectroscopy and 2D NMR…and now my eyes have crossed again. I think I should stick to abstract expressionism.)
Comment #33 April 17th, 2016 at 8:07 pm
Turns out that, despite what has been suggested in some reports, it wasn’t an impromptu question; Trudeau prepped an answer and invited reporters to ask him about it. From earlier in that press conference:
“when we get to the media questions later, I’m really hoping people ask me how quantum computing works, because I was excited to deepen my knowledge of that this morning.”
https://www.youtube.com/watch?v=Y34jkcX-Q3A&feature=youtu.be&t=13m50s
Comment #34 April 17th, 2016 at 11:58 pm
That’s fine by me. PM wanting to show off sci…er, well maybe not chops exactly, but any kind of reasonable interest in sci: I’d say get it while it’s hot. Those are some good media instincts, too. Don’t know who they belong to, but they’re good.
He’s pumping the nerdcore pedal a little heavy, though. Feminism, quantum computing, next he’ll be bicycling to work on a Dutch folding cycle that generates electricity as it goes. I’m thinking of his dad, and how he stayed in power partly because of his tremendously broad suave appeal.
Comment #35 April 18th, 2016 at 6:48 am
Nobody asked me, but I’ll try nevertheless to give a “10 second explanation of QC”:
Quantum computers are cool because they can solve problems using tools that classical computers simply cannot use; it is like a bird that can simply fly from point A to point B, while a car has to miserable navigate the streets from A to B. And the “wing” of a quantum computer is quantum superposition: it allows wrong answers to cancel out and right answers to be reinforced, if you manage to manipulate it well.
It’s much less precise and informative than Scott’s answer, but at least it’s closer to the 10 seconds mark =)
Comment #36 April 18th, 2016 at 7:57 am
Can you identify the remarkable formulas on the blackboard behind Trudeau?
Comment #37 April 18th, 2016 at 10:28 am
Difficult to give a mark without first establishing the desired outcome of an assignment. The assumption here seems to be that 10/10 would require true, informative, non trivial statements about QC in ten seconds. Seems like a pretty tall order. I think a better assignment for “Quantum Computing for Politicians” might be: care enough to ask about the subject, listen to the answer, regurgitate without making any serious mistakes. On that basis I think he gets closer to full marks.
Comment #38 April 18th, 2016 at 10:46 am
Chris #37: It was more like a minute—but in any case, treating the assignment as “explain quantum computing to laypeople in one minute,” I’m comfortable with the grade I gave. And I like to think Trudeau might appreciate that I respected him enough to not give full marks just because I like his politics (to whatever extent I do—I don’t follow Canadian politics as closely as I should), or because I think being obsequious to politicians is the way to ensure continued funding for science. It may “only” have been a B+ answer, one any bright teenager could also have given after hearing people explain QC for an hour, but it was an honest B+. 🙂
Comment #39 April 18th, 2016 at 11:51 am
Jeb Bush did worse when asked a trig question- though to be fair he really could not have expected the question. See
http://blog.computationalcomplexity.org/2013/08/how-much-trig-does-your-governor-know.html
Comment #40 April 18th, 2016 at 5:36 pm
re gasarch #39:
Excellent post. Does anyone reading this blog know the angles of a 3-4-5 triangle without help from a calculator? I doubt it. I hope the majority know 1/3 of the answer!
Comment #41 April 18th, 2016 at 10:09 pm
CNN asked on Twitter “Can you explain quantum computing like Justin Trudeau?” This got me to wondering whether, in a single tweet, one could address both of the two questions Trudeau seemed to be answering: how does QC improve on classical, and what would that be good for? My best shot at this can be seen at
https://twitter.com/PrattWhittemore/status/722225433584869376
Challenge: improve on this. (I paid 5 characters including the space to reply to @CNN, replying to me would cost you 17 characters so either reply to @CNN or to no one—you’re welcome to the extra 5 characters. 🙂 No fair replying with a link to an explanation with more than 140 characters.)
Comment #42 April 19th, 2016 at 2:48 pm
How about “QC lets us use parallel universes as scratchpaper”
Comment #43 April 19th, 2016 at 6:25 pm
Dave #42: Doesn’t get across the restrictions on how you can access the scratch paper.
Comment #44 April 19th, 2016 at 8:41 pm
Grade D- , if he showed the same vague misunderstanding of economic theory he would be ridiculed, and rightly so.
Comment #45 April 19th, 2016 at 10:28 pm
James #44: Firstly, much as it pains me to say so, it’s probably more important that politicians understand economics than that they understand quantum computing theory! But secondly, if you asked economists, I’m sure they’d tell you that politicians have a long history of misunderstanding basic economic theory much more egregiously than any misunderstanding Trudeau showed about QC, and being not only not ridiculed but lauded for that.
Comment #46 April 19th, 2016 at 11:57 pm
Regarding the 8 quantum researchers attempts: Scott is the only one who actually attempted to answer the question. And he did OK in the sense that if you know the subject already, you can see that some key aspects were touched on. Not likely that anyone “outside the loop” would have a clue what he is talking about.
My first impression is to give Scott a B+, a couple of the others a D, and flunking everyone else.
About half of the answers are totally loony, about what you would expect from Time magazine. Maybe Krysta Svore heard the question wrong, perhaps thinking they asked “what is magic fairy dust?”.
I suppose it is not fair to grade the responses as if they were answers on a test that people had time to prepare for. Probably people got asked with no time to think about it, and since it seems like an impossible task, just blabbed some generalities about how wonderful QC will be. Hardly the time to point out the inconvenient fact that there might never be a working quantum computer.
Comment #47 April 20th, 2016 at 12:21 am
Listening to the answers is more enlightening than reading them.
Many of us will recognize and perhaps identify with Scott’s delivery: the excitement of an expert with a chance to talk about something that she or he really likes. I get into that zone when teaching discrete math or algorithms, but not so much with programming.
Next up comes Krysta: A great imitation of a big city TV station weather person reading a press release. It is a lot more fluid than than, say, a baseball announcer reading a mini advertisement between pitches, but the cadence of reading an announcement is clear.
Comment #48 April 20th, 2016 at 5:26 am
Raoul: I don’t agree that I’m the only one who tried to answer the question. For one thing, Martin Laforest gave much the same answer I did. For another, “explain QC in 35 seconds” is such an absurd request that it’s no surprise people would interpret it very differently! Aephraim answered “is a QC just like a classical computer but faster, and if not, why care about it?” while Krysta answered “what are the conceivable applications of QC that would most benefit humanity?”
But more to the point: if no one got above a B+, then OK, great, how would YOU do this in 35 seconds? Put up or shut up! 🙂 A test that it’s impossible to get an A on is just a badly-normed test.
Comment #49 April 20th, 2016 at 8:25 am
I was at PI that day, so I wanted to just point out two things about that speech:
(i) He *did* answer the ISIL question immediately afterwards. People who are saying he dodged it just watched an edited version of the video where the second part was cut.
(ii) People are complaining that he didn’t *really* know what he was talking about, but I don’t think it was supposed to sound like he did. It was more like he rehearsed this so he could joke around with that particular audience. Everyone laughed, then everyone cheered, but no one thought “Oh wow, our PM is so smart, he knows what quantum computers are”. How this has propagated into the world’s media afterwards is a different story.
Comment #50 April 20th, 2016 at 9:26 am
The fist two “explanations” are even more embarassing than Trudeaus’s since they come from “experts”. Hype and half-truths are nowadays far too common in order to promote certain parts of sciences. Scott’s answer is probably the best and more honest.
Comment #51 April 20th, 2016 at 10:09 am
Scott, bet you just loved the answer from Barry Sanders, no? 🙁
Comment #52 April 20th, 2016 at 10:12 am
It’s notable that Justin Trudeau’s 96-word explanation of quantum computing has a Flesch-Kincaid Grade Level of 7.58, which is a markedly lower grade level than any of the other explanations offered (the metrics are summarized below).
Here is a quantum computing explanation that is crafted to be even simpler than Justin Trudeau’s, in the sense that its product of “words times Flesch-Kincaid Grade Level” (henceforth the “wtfk unit”) is even less:
Here the notion that “the world is made of short lists of equations” is reflected in two of the Perimeter Institute’s research initiatives “Tensor networks: from entangled quantum matter to emergent space time” and “From discretuum to continuum: computational tools that bridge the gap”.
Metrics Some comparative reading-difficulty metrics are (in order of increasing wtfk-score):
• 0625 wtfks: the above quantum computing explanation
(96 words times Flesch-Kincaid Grade Level 6.51)
• 0735 wtfks: Justin Trudeau’s explanation
(97 words times Flesch-Kincaid Grade Level 7.58)
• 1097 wtfks: Barry Sanders’ explanation
(67 words times Flesch-Kincaid Grade Level 16.37)
• 1133 wtfks: Murray Thom’s explanation
(98 words times Flesch-Kincaid Grade Level 11.56)
• 1369 wtfks: Aephraim Steinberg’s explanation
(148 words times Flesch-Kincaid Grade Level 9.25)
• 1583 wtfks: Krysta Svore’s explanation
(103 words times Flesch-Kincaid Grade Level 15.65)
• 1675 wtfks: Davide Venturell’s explanation
(101 words times Flesch-Kincaid Grade Level 16.58)
• 1700 wtfks: Martin Laforest’s explanation
(118 words times Flesch-Kincaid Grade Level 14.41)
• 1917 wtfks: Scott Aaronson’s explanation
(131 words times Flesch-Kincaid Grade Level 14.62)
The above Flesch-Kincaid Grade Levels are computed with the interactive “Readability Calculator” that is hosted at the web-site “Online-Utility.org”.
A quantum challenge Can quantum computing be reasonably explained in 500 wtfk units or fewer? (my own attempts at this level of quantum simplicity have failed)
Comment #53 April 20th, 2016 at 10:29 am
mike #51: Barry, who I know, could explain as well as I could everything that’s wrong about the “run all possible answers at the same time” explanation. I think his mistake was not that he actually believes such a thing (or was trying to mislead anyone), but rather, that he simply thought that if you’re summarizing in a few sentences for a layperson then there’s no possible way to do better. Personally, I think it IS possible to do better, but the way to do better is not obvious, and only emerges from a lot of trial and error.
Comment #54 April 20th, 2016 at 12:13 pm
John Sidles #52: I can do even better than that, and achieve a Flesch-Kincaid grade level of 4.1 in a mere 22 words:
Duh. Duh. Quantum computers very small. Also very fast. Exciting! Solve all problems. Try all answers in parallel. Me Tarzan. You Jane.
Comment #55 April 20th, 2016 at 12:53 pm
Here is a humorous take on Justine Trudeau’s QC lesson:
http://syruptrap.ca/2016/04/delirious-exhausted-justin-trudeau-still-explaining-quantum-computing-to-journalists/
It’s great when political leaders are not afraid of science.
Comment #56 April 20th, 2016 at 2:37 pm
Did John Sidles just define a “wtfk unit?” 🙂
Comment #57 April 20th, 2016 at 2:42 pm
Extending Lisa Adams’ comment (#55), how many Shtetl Optimized readers recall that (former) US President George W. Bush is also a pretty fair amateur physicist?
Lisa’s reminder that “It’s great when political leaders are not afraid of science” is grounds for sober reflection too. 🙁
Comment #58 April 20th, 2016 at 5:01 pm
Scott #48:
I agree that I should not be grading anyone’s response, because I cannot come up with a passing answer.
I would have trouble giving a 35 second comparison of John Sidles’s new “wtfk unit” with the traditional computer world “WTF” unit. Maybe I should go with an example:
“The Canadian Prime Minister can explain QC with 0735 wtfk units. The American presidential primaries achieve a rare 9 on the 10 WTF scale.”
Comment #59 April 20th, 2016 at 7:06 pm
Ok here, goes, genuine 35 sec typing attempt:
QC is a
Ah fuck, start again (this is genuine)
QC is is misguided attempt to aplly technonlogy to a not very well understood scientoicf equation – the scroedinger eqaution, assuming it is continuos in time and allows almost infinite information to be harnessed.
Oh dear, well there you go, probably correct apart from the grammar and spelling. 🙂
Comment #60 April 21st, 2016 at 1:37 am
You know, at least in authoritarian regimes the journalists have to be ordered to give the Leader embarrassing, fulsome praise for trivial accomplishments. Western democracies are maybe worse: here, the journalists do it voluntarily.
Comment #61 April 21st, 2016 at 2:48 am
Scott #19:
You mentioned about the “maybe” as it wasn’t merely i our hands. It was/is out there in the Universe.
Is this “maybe” thing lacking due to the decoherence effect.
Comment #62 April 21st, 2016 at 8:28 am
OK, I can’t help giving this a shot:
“Normal computers are built of bits, switches that can be either on (0) or off (1). When the computer computes, all those bits flip back and forth in a complicated dance (an algorithm) that eventually produces the answer. If you make the bits small enough, they obey different rules — quantum mechanics. Quantum bits can access some new, totally different states (not 0 or 1). Imagine being an engineer who only knows about solid matter, and then you discover liquids. “Wow, this stuff FLOWS! I can use liquids to solve some problems much more efficiently!” Quantum computers could solve some problems much, much faster by taking shortcuts through the unique states that only quantum bits can be in.”
About 110 words, FK grade level of 8.5. Less than 1000 wtfks. 🙂
Comment #63 April 21st, 2016 at 4:42 pm
Nice blog and I am enjoying the comments, especially those that criticize what I said. Scott is right about my comment: I don’t think of quantum computing as parallelism but I tried to say, in 35 seconds without rehearsal, the essence and point of quantum computing so that a layperson could follow. My reference to parallelism is somewhat allegorical in referring to quantum evolution with interfering pathways, but such a description is naturally open to misinterpretation. I agree with Scott that there are better ways to explain in 35 seconds than what I did but, as Scott says “the way to do better is not obvious, and only emerges from a lot of trial and error”. Anyway trying to explain quantum computing in 35 seconds was a fun challenge and really hard. Anybody who doesn’t like my short version should let me give my 20-hour version and then judge 🙂
Comment #64 April 22nd, 2016 at 1:39 am
Meanwhile, here’s the Australian PM’s attempt …
http://www.theage.com.au/federal-politics/federal-election-2016/malcolm-turnbull-opens-new-quantum-computing-labs-at-unsw-20160422-gocnno.html
Comment #65 April 22nd, 2016 at 12:44 pm
As for John Sidles’ challenge — it took a little massaging of my usual spiel, but I got it in the end:
Normal computers are built to be understandable with classical physics, for the most part. A computer which took more advantage of quantum physics could perform single computational steps, that normal computers couldn’t. It could solve problems much more quickly by taking shortcuts, solving problems in many fewer steps.
[48 words] × [10.07 Flesch Kincaid level] = 483.86 WTFKs (ftw)
Comment #66 April 22nd, 2016 at 4:34 pm
Nielde Beaudrap’s explanation of “Quantum Computing” (#65) is really excellent (as it seems to me). Anyone who has tried to write with similarly siccant simplicity will agree!
It’s natural to wonder, can John Preskill’s idea of ‘Quantum Supremacy’ and Scott’s ingenious paths toward it (like BosonSampling) be summarized similarly?
After all, the key ideas(s) of Quantum Supremacy are plausibly (as they seem to me) even more subtle than the ideas of Quantum Computer.
E.g., in which respects are the grounds for skepticism of Quantum Supremacy more general than the grounds for skepticism of Quantum Computing? In which respects are skeptical grounds more specific?
Conclusion Short yet well-crafted explanations of Quantum Supremacy — say of length ~1 kiloWTFK or so — would be a valuable contribution to STEM discourse.
Helpful to students and reporters, for sure!
Comment #67 April 22nd, 2016 at 9:34 pm
[…] note that the much–acclaimed and soberly–evaluated answer on quantum computing by Canada’s Prime Minister, Justin Trudeau, had […]
Comment #68 April 23rd, 2016 at 6:30 pm
I liked Martin Laforest’s most.
I also liked yours. I think the best way to explain it is to explain briefly probabilistic algorithms and say now probabilities can be complex numbers and then follow something similar to Martin’s that this allows wrong answers to cancel each others probability and therefore leave only the correct solution while a traditional computer would normally check each of these solutions separately.
Comment #69 April 23rd, 2016 at 6:49 pm
For people who are saying that quantum computing is not going to be very useful, people said the same thing about deep neural nets and machine learning. It is important for the government to support basic science. Here is what Geoff Hinton, one of the central figures says about funding for basic research:
Comment #70 April 23rd, 2016 at 7:15 pm
so uncertainty is not information?
Comment #71 April 23rd, 2016 at 7:44 pm
eli #70: Yes, Shannon information is a measure of uncertainty. But Trudeau was trying to explain why a quantum state can encode MORE information (in some sense) than a classical string of the same size. There, it would’ve been clearer to talk about the exponential number of amplitudes, or something, than about “uncertainty” (presumably alluding to Heisenberg’s uncertainty principle—another central feature of QM, yes, but not one that’s directly relevant here).
Comment #72 April 24th, 2016 at 3:03 am
That’s really the unintuitive part, and it’s not just because the concept of a negative probability is strange.
Suppose you flip two coins. There are four possible outcomes. Classically, we can perform operations to transfer probability from one outcome to the other.
For example, if we want to transfer p(HH) to TT we add a conditional operation that transforms HH into TT. We can even transfer only a portion of p(HH) by flipping a coin first. And we can carry out this operation, and any other similar ones, without having to evaluate all possible outcomes.
Transferring probability from one outcome to another involves a subtraction and an addition. The subtraction is just adding a negative probability. So even classically we can talk about negative probabilities.
IMO the part that’s difficult classically is performing a ternary operation. For example, suppose that, whenever the outcome is HH we want to transfer p(TT) to HT. We can add a conditional operation on HH but we don’t have p(TT) at hand. And even if we did, once we’re in HH we can’t eliminate TT as a possible outcome.
For a ternary operation we’d need to evaluate two paths at a time. And if we add a second ternary operation then we need to evaluate four paths at a time. For a chain of n ternary operations we’d need to evaluate 2^n paths classically.
IMO It’s one reason why a quantum circuit consisting only of Hadamard and CNOT gates is easy to simulate classically, but not so when Toffoli or Fredkin gates are added in a non-trivial way (e.g. Toffoli chains).
My guess is that it’s still possible to simulate a quantum circuit classically if the quantum circuit uses Toffoli/Fredkin gates in a trivial way (e.g. not chained), or only uses log(n) of them, where n is some tractable number. Am i wrong?
I’m curious whether there is a quantum circuit that can be efficiently simulated by a classical machine that still improves on the best known classical algorithm.
Comment #73 April 30th, 2016 at 11:28 pm
How does Scott Aeronson answer Shor’s Factoring Trilemma:
1. the Extended Church-Turing Thesis is false,
2. textbook quantum mechanics is false, or
3. there’s an efficient classical factoring algorithm.
Comment #74 May 1st, 2016 at 8:38 am
#73: Don’t know how “Aeronson” answers it, but I’m betting firmly on 1.
Comment #75 May 2nd, 2016 at 9:35 am
Scott #74: I agree that option 1 is the most likely one, but let’s assume for a moment that you just happen to discover that there is a classical factoring algorithm that is efficient enough to make RSA as used today completely insecure. How would you go about publishing it?
Comment #76 May 2nd, 2016 at 11:41 am
Aula #75: Applied cryptographers actually have a well-developed protocol for what to do when you discover a security flaw in a deployed system. The basic idea is that you first contact the affected companies, etc. and give them some lead time to patch the flaw, and only afterward do you publish your result.
If you discovered a fast factoring algorithm, contacting companies in secret wouldn’t really apply anymore, because now the whole world is affected!
But the underlying principle could be adapted to that case as follows:
1. You announce to the world that you have a fast factoring algorithm, and you prove it by factoring challenge numbers given to you.
2. You hire bodyguards, and you keep all notes and files related to your algorithm in a very secure location. 🙂
3. You give everyone 6 months or something to switch over to lattice-based cryptography or something else.
4. You publish your algorithm.
There could also be a step 0, where you first notify your nation’s government, and ask to be placed under its protection in the months before your algorithm gets published. I’d consider that an optional step, depending on how much you believe in your nation’s government and military.