The CIA and Jeff Bezos Bet on Quantum Computing

http://www.technologyreview.com/news/429429/the-cia-and-jeff-bezos-bet-on-quantum-computing/

by iProject • 14 years ago

103 49 14 years ago

dyeje3• 14 years ago

Stop reading after the second paragraph, quantum computing isn't some magic wand you wave at problems to make them disappear. It's simply a new way of approaching them, and it requires algorithms designed to take advantage of this new approach to computing to solve said problems.

Bakkot• 14 years ago

Those approaches are increasingly well-studied and a lot of them (Grover's algorithm, anyone?) look an awful lot like magic.

(For people who don't want to look that up, Grover's algorithm does searches on unsorted databases in root-n time, that is, in fewer operations than would be required to inspect every element of the database.)

jimwhitson2• 14 years ago

Grover's algorithm does indeed look rather like magic, but even better is the geometrical proof - WP's description looks good at a quick skim. In my (fairly limited) experience, this sort of geometrical thinking is the best way to understand quantum algorithms intuitively.

jimwhitson2• 14 years ago

Would it be overly conspiratorial to suggest that the CIA and NSA are in fact interested in a large-scale implementation of Shor's algorithm? Perhaps I underestimate how much use intelligence agencies have for optimization...

sp332• 14 years ago

I would be worried if they weren't interested in Shor's algorithm. That's something that fits their mandate at least.

jpdoctor• 14 years ago

> Would it be overly conspiratorial to suggest that the CIA and NSA are in fact interested in a large-scale implementation of Shor's algorithm?

The only thing conspiratorial is that it wasn't mentioned in the article. Hmmmm.

nikic• 14 years ago

Context: http://en.wikipedia.org/wiki/Shor%27s_algorithm

wallawe2• 14 years ago

Can anyone provide an explanation of this in layman's terms? I took a look at wikipedia but am having a hard time wrapping my head around it..

dyeje3• 14 years ago

A lot of public key cryptography methods used in the real world are based on the fact that factoring really big numbers is computationally unfeasible. Shor's algorithm makes it feasible, but it's a quantum algorithm that can't be run on a classical computer. Basically, alot of stuff won't be secure anymore when it's feasible to run Shor's algorithm.

mturmon• 14 years ago

They would be derelict in their duty to NOT be very closely tracking the (open literature) state of the art in factorization. I.e., it's not a conspiracy, it's their job, and it always has been.

I also presume they have people working to keep abreast of the state of the art in secret work of other countries.

cs702• 14 years ago

The societal impact of quantum computing would be immense -- the entire security infrastructure of the web would have to be rebuilt from the ground up.

However, many experts -- including Scott Aaronson (who is mentioned in the article) -- have cast doubt on D-Wave's claims, because the company has not yet proved that its machines are producing results by exploiting quantum phenomena like particle entanglement and superposition. In other words, D-Wave's devices may very well be different, but still classical, computers. D-Wave's engineers acknowledge as much: they are quoted in the article stating that "they don't yet know for sure what's happening inside the chip."

The fact that Bezos, a very smart guy, has invested in D-Wave, has changed my perception of the company from "these are likely crazy people making outrageous claims" to "maybe these guys are unto something." Exciting and a bit scary. Time will tell.

cynicalkane• 14 years ago

I don't think the quote

they don't yet know for sure what's happening inside the chip

means what you think it means. I think in context this is an attempt to describe how quantum algorithms work, where they need a nonobserved superposition to operate.

They may well not understand their chips, but that's not supported by that quote in that context. You don't need to know the exact operations involved to solve an optimization problem with quantum computers (or for that matter with classical ones).

dumb-dumb5• 14 years ago

"means what you think it means"

Can we find some other way to convey this message without using that same tired old phrase over and over and over again? As soon as I see this phrase by brain automatically shuts down because it tells me the author lacks creativity.

Are people who post to nerd forums really that deficient in English language skills? Or are they just hoplessly compelled to mimic the language used by others?

"means what you think it means" is like some kind of lingusitic meme.

Future posters: Let's be creative and devise another way to state this.

This comment is not personal. It is just a randomexample of a phenomena. Many, many, many nerds use this phrase. I'm just tired of reading it. I can't be the only one. Next time you are tempted to type "doesn't mean what you think it means", please stop yourself. Write it some other way.

greenmountin• 14 years ago

I think "don't yet know for sure what's happening" is a bit misleading to the casual reader. If they're actually being honest it's more like: don't know that anything "special" is happening.

Frankly, their devices just aren't quantum computers until they can show they perform quantum mechanically. Until then, they are just hype. Bezos' investment is perfectly rational as an affiliation move -- if he was serious, he'd donate to, say, the Perimeter Institute in Waterloo where real quantum computer development is taking place. But the press releases for their accomplishments get several orders of magnitude less attention, so why bother? (or, is it that you just can't buy in to IBM -- also real work -- or HRL)

Their job ads [1-3] focus is very telling to me; which are mostly for new algorithms people, with long unfilled placements for physicists and materials engineers. The big development in QC this decade has been that "materials matter", and that you can't just throw qubits together willy-nilly. D-Wave scaled their qubit count (see the ridiculous "Rose's Law" slide [4]) basically by ignoring all these problems.

[1-3] http://www.dwavesys.com/en/files/20120627_d-wave_algorithms_... http://www.dwavesys.com/en/files/senior_engineer_2011.pdf http://www.dwavesys.com/uploads/20091026_Design_Engineer.pdf (http://www.dwavesys.com/en/careers.html)

[4] http://www.dwavesys.com/en/dev-tutorial-hardware.html

neworder• 14 years ago

Just a side note - you're probably conflating Perimeter Institute (www.perimeterinstitute.ca) and Institute for Quantum Computing (http://iqc.uwaterloo.ca/). Both are in Waterloo, but it's the latter that is involved in developing quantum computing implementations.

tisme• 14 years ago

Bezos may be a very smart guy, but I don't see - short of him becoming an expert in quantum physics - how his investing or not investing in this company means anything with respect to them being onto something.

quonn3• 14 years ago

> the entire security infrastructure of the web would have to be rebuilt from the ground up.

I don't think so. It would probably suffice to add NTRU support to SSL.

fjorder2• 14 years ago

This is the first I've heard of NTRU. Has it been proven immune to quantum computational attacks, or has an efficient attack algorithm just not been found yet?

The former would be world-shaking, so that's probably not the case unless I've been living under a rock. The latter just means that it might be safe, but it might be broken at any time if somebody comes up with the right algorithm. If you're making long-term information infrastructure plans that's something you have to account for.

zmanian• 14 years ago

The largest societal impact of quantum computing would be to solve the 3d structure of biological molecules from sequence data.

This would be produce an almost immediate revolution in all areas of biotechnology whose implications would be difficult to predict.

jrabone• 14 years ago

But the problem (much like a qubit) can go one of two ways. Either D-Wave have built a 512-qubit quantum computer - in which case the CIA / NSA will be strongly suggesting that they don't tell anyone, or they haven't - in which case they won't be telling anyone. The fact that they're not supplying any proof, therefore, contains no useful information.

ecolak• 14 years ago

I don't think their big contract with Lockheed Martin should be a good indicator of actual quantum computing. What they built is probably optimized for some of the problems that Lockheed Martin is trying to solve...

RK• 14 years ago

I have heard from people familiar with the Lockheed project that the purchase was largely motivated by tax incentives. Something along the lines of a Canadian-American agreement, in which companies could claim 10x as much business expenses if they were making purchases of research equipment from Canadian suppliers. The implication was that Lockheed never had strong confidence in this technology, but had little to lose.

SeanDav• 14 years ago

Quantum computing and nuclear fusion power generation are probably 2 of the biggest problems to solve in order to make the next technological leap.

If D-Wave comprehensively cracks Quantum computing I reckon Bezos may become the world's first Trillionaire.

swalsh• 14 years ago

In the past, technology would provide efficiency to a problem, and a lot of times up root a job in the process. Which is okay, because a new job was created in the wake. Astute observers though might notice that the skill level has been increasing each time a job is transformed. This is okay when an evolution in technology is equivalent to a single lifetime, but as the evolution of technology speeds up the adaption becomes more difficult. I feel like we are seeing the effects of the last technological revolution today. As a software engineer, I get 4 emails a day from recruiters... but those with no technical skills are being unemployed in troves, and finding a job is near impossible.

So my question. What happens when an assumed intelligence explosion disrupts, an already disrupted world? I absolutely don't believe there's anything we can (or should) do to slow technological progress... but its an interesting topic to see how it affects society. There seems to be a disconnect, technology evolves exponentially, and society evolves linearly.

wallawe2• 14 years ago

Did anyone else catch the temperature miscalculation?

> It is actually a cold gun: the structure is a chilly -259 °F (4 °Kelvin) at the wide end and a few thousandths of a degree above absolute zero at its tip...

4 degrees Kelvin is approx -452 Fahrenheit is approx -269 Celsius. Absolute zero is −459.67° F

ars2• 14 years ago

4 Kelvin is probably the correct number since that's the temperature of liquid helium.

fjorder2• 14 years ago

You can go a bit lower than the temperature of liquid helium just as you can cool something to lower than the temperature of liquid water just by using evaporating water. However, 4 Kelvin is probably correct. The image they show appears to be of a cryostat that would almost certainly use helium as a cryogen. Such cryostats are realistically limited to around 2-3 Kelvin. They can go a bit lower when empty, but generally not if you stick a complex experiment with a lot of heat-conducting fibers and wires into them.

4K is probably warmer than they'd like for a quantum circuit, but getting their apparatus colder would require a much beefier cryostat or more exotic technologies like magnetic cooling built inside a helium cryostat. Such setups are capable of getting complex setups down to the hundreds of millikelvin, which is much nicer for this sort of work!

As for writing a scientific article using Fahrenheit... Please don't do that. Just don't.

batgaijin• 14 years ago

What happened with all of the posts from this guy mocking D-Wave?: http://www.scottaaronson.com/blog/?p=431

cing2• 14 years ago

"I hereby announce my retirement as Chief D-Wave Skeptic, a job that I never wanted in the first place." -http://www.scottaaronson.com/blog/?p=639

cing2• 14 years ago

On a related note, D-wave is sure posting some interesting job opportunities: https://www.mitacs.ca/o/2012/07/cognition-and-creativity-fra...

perlpimp• 14 years ago

No one has explained exactly what kind of problems would Quantum computer could solve well. From few articles I've read it is not for all kinds of tasks a magic ax.

if someone can expand on that, that would be great.

jcr• 14 years ago

> He (Arronson) and other critics say the company must still prove two things: that its qubits really can enter superpositions and become entangled, and that the chip delivers a significant "quantum speed-up" compared to a classical computer working on the same problem.

I'm curious how one goes about "proving superposition"? If one does the most obvious thing, opening the box containing Schrodinger's cat to observer and hence prove state, then you've just lost the superposition.

Anyhow, for those wondering about addressing the impacts of successful quantum computing on cryptography (i.e. security of most things, including your banking transactions on the web), you should check out the efforts of Dan J. Bernstein (DBJ) and friends on the "Post-Quantum Crypto" web site:

http://pqcrypto.org/

Though research is still on-going, at present it seems some algorithms are resistant to quantum cryptanalysis. It might take a decade or two to retool all the systems using crypto vulnerable to quantum cryptanalysis (assuming someone succeeds in building a quantum computer), but at least it seems like there's hope.

The CIA and Jeff Bezos Bet on Quantum Computing

Related Stories