In an article on stopping a large spectrum of light with metamaterials in The Telegraph (research which is very cool, but isn't available online, yet, as far as I can tell), I find some lines that would make the Optimizer go bonkers:
By contrast, the switches in a quantum computer can be both "on" and "off" at the same time. A "qubit" could do two calculations at once, two qubits would do four and so on. Thus, it was theoretically possible to use quantum computers to explore vast numbers of potential solutions to a problem simultaneously.
Ouch, my brain hurts.
Okay, so I'm fine with, if a bit peeved, at describing a quantum computer having switches that are both "on" or "off" at the same time. I mean a superposition of two states is an odd object and saying good coherent English sentences about such objects is a slippery slope. But to say that a qubit does two calculations at once, well that's just plain silly. And the idea that quantum computers get their power in a naive way of exploring vast numbers of potential solutions, well that's not at all how we understand the speedups of quantum computers. In fact we know that this naive parallelism doesn't lead to quantum speedups.
Really I hate to be hard on these science journalists, but running this by a single researcher in quantum computing would have fixed this up. Or would it? Maybe there are quantum computing researchers out there who would let this pass. If so I'd love to hear their justifications!
To end on a happier note: at least the science writer spelled "Caltech" correctly :)
- Log in to post comments
Well, to be perfectly fair, quantum speedups require quite a bit of technical foreknowledge to grasp. If I had to write an article for the layperson on "a Fourier transform of quantum mechanical amplitudes", I don't know where I would start (I don't even trust that I fully understand it, yet). I don't think that excuses explanations that are fundamentally misleading, but I do acknowledge the inherent difficulty of the task.
The limit of my knowledge of quantum anything is also pretty low... It was the last thing we covered in my pretty basic physics class and really only covered stuff like how electrons and photons interact and then barely talked about uncertainty. To be honest, the bit you quoted about quantum computing is pretty close to my own limited understanding of the topic.
Is there any chance that you'll cover this misunderstood topic here? That's something I'd really like to read.
Scott Aaronson has a pretty good explanation of Shor's Algorithm here:
http://scottaaronson.com/blog/?p=208
This kind of thing irritates me not so much because it's incorrect, but because it basically adds nothing to the story. The last two paragraphs of the Telegraph article should have just been deleted. Even if that last paragraph had been reworded so that it was at least somewhat correct, a three or four sentence summation of quantum computing is simply not going to give the reader any useful insight into why quantum computing is different from classical computing.
If I was the journalist in question and my editor insisted that I include something about classical vs. quantum in the article, I might say something like, "By taking advantage of certain properties of quantum mechanics, quantum computers hypothetically will be able to perform some types of calculations much faster than classical computers." That may sound like a cop-out, but I think it's much better to let the reader know that they don't know something, rather than give them the impression that they know something when they really don't.
By the way, isn't Scott the optimizee referred to by his blog title, and not the optimizer?
Kurt said:
Even if that last paragraph had been reworded so that it was at least somewhat correct, a three or four sentence summation of quantum computing is simply not going to give the reader any useful insight into why quantum computing is different from classical computing.
For science journalists writing for general circulation, it is important to explain what things like "quantum computing" are, because their readers are likely to have never come across the concept before. Even a short summary is useful, and gives the reader an opportunity to either say "ok, I understand what they are talking about" or "I am interested and might go look this up on Wikipedia".
The problem with just saying "quantum computing is faster than classical computing" is that it doesn't differentiate quantum computing from parallel computers, multicore processors or faster clock speeds in the minds of the reader. Sure, a quantum computing researcher defines classical as "not quantum", but people with interest in other fields use the same word to describe other things.
The task of taking an entire subtle and varied field, like quantum computing, and reducing it to a small paragraph is an extremely difficult one. Journalists shouldn't be saying things that aren't true, but you can't expect them to distill the entire field down into one paragraph without losing some of the subtleties.
It's important that popular science articles are approachable and promote the public understanding and appreciation of science. It's not important that they support scientists egos by showing people how hard the field is and how smart the people working in it are.
The task of taking an entire subtle and varied field, like quantum computing, and reducing it to a small paragraph is an extremely difficult one. Journalists shouldn't be saying things that aren't true, but you can't expect them to distill the entire field down into one paragraph without losing some of the subtleties.
I agree that the task of the science journalist is hard, but really in this case the journalist is saying something that is pretty darn close to not being true. Does making people feel warm and fuzzy really justify lying to them? I don't think so, and I think this view of what the goal of science journalism should be is totally backwards.
And I don't see why Kurt's suggestion isn't a valid one: what does explaining what a qubit is (incorrectly) and explaining what quantum speedups are (incorrectly) add to the story? Kurt's suggested ending explains that quantum computers allow speedups (I'd personally drop the "hypothetically.") One could even add a sentence explaining that the speedup is different from parallel computers, multicore, etc, and that it relies on peculiar (for God's sake don't use spooky) properties of quantum mechanics that scientists are only just now begining to understand.
If you don't have room or time to do a good explanation (which is certainly the case in most "press release"/"Nature andScience embargo" driven articles) should you err on the side of broad statements that convey what scientists know and not warm fuzzy nearly wrong statements?