Want some pessimism about quantum computers? Try here this article where a survey of 700 IEEE fellows is made:
Seventy-eight percent of respondents doubt that a commercial quantum computer will reach the market in the next 50 years.
700 IEEE Fellows can’t be wrong can they?
It’s not so much that the responses are wrong as the question is wrong-minded. In my view, there shouldn’t yet be promises, roadmaps, or Manhattan projects for quantum computing. QC is basic research and there is no telling where it will be in 50 years.
Let’s prove them wrong.
That is excellent news! It means you can milk this thing well into retirement!
Imagine if quantum computers started showing up in Radio Shack ten years from now, being a quantum computing researcher wouldn’t be so sexy anymore.
How about we first have them all sit the final exam for John Preskill’s QI course, and then ask the opinion of those who pass…
I wonder if any of these folks are aware of the existence of liquid state NMR quantum computers. Companies have been selling hardware into the QC research community for quite a while now. Bruker sells NMR machines into this market for about $1M a pop. I think this qualifies as a commercial QC doesn’t it? A company makes money selling QC components? That’s what commercial means I think.
Also Greg: Most QC research is basic research. Not all of it. There are three serious efforts that I’m aware of whose objectives are to build special-purpose QCs that can beat conventional systems in a relatively short (2-3 years) period of time on some well-defined problem set (interestingly enough, none of these are in the US). Whether or not these efforts will succeed is still unknown, but this work is definitely not basic research.
The only reason that experts are no better at predicting the future than non-experts is by definition, namely that “the future” is things that no one can predict. The date of the next solar eclipse is technically the future, but no one calls it predicting the future precisely because experts can give a reliable answer.
Experts are much, much better at predicting the “present” than non-experts are. For example, an expert can tell you the population of Armenia to within half; most non-experts don’t know. Or the dates of solar eclipses. (But of course, they can easily find out from experts.) Again, that is by definition. Expertise is by definition exactly this predictive ability.
Any QC which is not described as basic research is a mixture of good vitamins and snake oil. Not necessarily in equal proportions; some of it is all snake oil, while some of it is very good basic work only slightly tainted by hype.
rrtucci: The extent of the capabilities of NMR QCs isn’t relevant to the question of whether or not they are commercial QCs. (Which they are). My point is that these questions about when commercial QCs will arrive are too vague to have meaning. We already have commercial QCs. A meaningful question of this kind could be: What will be the date when a QC pure play’s market cap first exceeds $1 billion? I am willing to bet that the date is prior to December 31, 2010.
It seems likely to me that a lot of the respondents are ignorant to the progress that has been made over the last few years.
Suppose it was a survey of computer scientists rather than engineers, would you really expect to get an informed answer? The fact is that most don’t work in the field, and so while the well informed majority makes an informed vote, the uninformed this drown out by shear force of numbers.
I’m not sure what way it would go with physicists, as I suspect a larger proportion deal with quantum control in some form.
I meant to say informed _minority_ not majority.
Ooops.
Richard…is that you? I loved your path integral book. How’s the Hibbster doing?
I always find it interesting to note that it took over a decade from Turing’s universal computer to the invention of the transistor, and another decade to invent the integrated circuit. I suspect we are about at the point where the invention of the quantum equivalent of the transistor is about to happen. But that is just as wild a guess as the pessimism of the IEEE fellows.
Don’t tell the folks at D-Wave. They’ve got real money on a viable QC coming along a lot sooner than 50+ years from now.
On the other hand, it took just over a hundred years from the invention of the analytic engine, to the integrated circuit.
Actually there’s a great deal of evidence that ‘experts’ don’t do any better than non-experts when it comes to predicting the future. I’m sure if you polled people working in the field you would get a very different answer, but not necessarily a more accurate one.
I’ll offer a different metric from Geordie’s billion-dollar market cap:
When will a paper appear in Science or Nature using results calculated on a quantum computer, in which the results are the focus rather than the machine or process that creates them?
Put another way, when will a quantum computer solve a problem for us that can’t be (practically) solved classically?
Perhaps it hinges on the word “commercial”. Will QC ever compete with Dell for desktop/laptop machines for the masses? Thats a much different question than will there be a few hundred/thousand specialized machines. If “commercial” means “sell one”, probably yes relatively soon. If it means compete with silicon/GaAs, well……thats a longer time scale for sure!
Gordie said:
“I wonder if any of these folks are aware of the existence of liquid state NMR quantum computers”
I’m sure they are aware of it, and of the fact that these QCs are not scalable beyond about 10 qubits. Factoring 15 into 3×5 is the full extent of their abilities
“There are three serious efforts”
What would these be?
The LSNMR quantum computer is a quantum computer in the same sense as the microchip in your thermostat is a personal computer. Scalability is everything, and this is where Rod’s comment comes in.
@Geordie
“There are three serious efforts that I’m aware of whose objectives are to build special-purpose QCs that can beat conventional systems in a relatively short (2-3 years) period of time on some well-defined problem set (interestingly enough, none of these are in the US).”
Which would be the other two?
Also I think the fact that the NMR hardware is dual use rules it out.
“Will We Let Those 80% Use Our Quantum Computer?”
I do not think, Dave, it is right or fair that people who expressed either pessimistic or skeptical views about quantum computers will not be aloud to use them when they are built.
Certainly it would not be fair and _obviously_ it will not happen. However, do you really think these 700 have any real idea about the challenges of building a quantum computer? Criticism grounded with good justification is fine. Criticism without such justification is the same category as those who push creationism.
“Certainly it would not be fair and _obviously_ it will not happen.
I am relieved to hear that, Dave, thanks.
“Criticism without such justification is the same category as those who push creationism.”
It was a survey, Dave, they had to answer based on what information they had at the time. If an inversor is relauctant, say, to invest in a company that promise a comercial QC in less than 50 years, I wouldn’t necessarily compare his action to pushing creationism.
Okay the creationism thing was a bit heavy handed 🙂 But I get very sick of this large abundance of these negative views of quantum computation by people who I’m very certain have spent no time themselves invested in understanding why a large number of people are optimistic about building a quantum computer.
Begon natter nabobs of negativity!
The good thing is that a small piece of evidence can completely change the views of most of these people since their opinions are highly correlated.