Hacker News new | past | comments | ask | show | jobs | submit login
“Qutrit” Experiments Are a First in Quantum Teleportation (scientificamerican.com)
49 points by gyre007 74 days ago | hide | past | web | favorite | 22 comments

So, I guess this means that "qutrit" computing is kind of like ternary (or "trinary") computing [1] with 3 well defined states. I got interested in trinary when I read somewhere that Russian scientists had at one point built a trinary computer (can't remember where I saw that) and I sat down and did the trinary math because it was interesting - and it worked.

Perhaps in qubit computing (speculating as I am not an expert), it is harder to detect the difference between the states which are somewhere in between 1 and 0, leading to the needs for error correction, but I also thought that, that particular property of qubit computing was interesting for its unpredictability, if it could be harnessed. Since a lot of machine learning models (example, Monte Carlo) use randomness, I always wondered how lack of precision might be an advantage. Out of my level of knowledge here a bit if anyone has any thoughts.

I had to look it up and, I'm not a quantum expert, but it seems my gut feeling was basically right: "Similar to the qubit, the qutrit is the unit of quantum information that can be realized in suitable 3-level quantum systems. This is analogous to the unit of classical information trit of ternary computers. Note, however, that not all 3-level quantum systems are qutrits. The term "qu-d-it" (quantum d-git) denotes the unit of quantum information that can be realized in suitable d-level quantum systems." [2]

[1] https://en.wikipedia.org/wiki/Ternary_computer

[2] https://en.wikipedia.org/wiki/Qubit

You are right about qutrits. Essentially, anything that exhibits quantum behavior is in infinitely many levels at once. It is our ability to measure discrete levels that leads to qubits, qutrits, etc. Generally speaking, the more levels the more unreliable the measurement.

As for Monte Carlo with respect to ML, are you referring to the “random walk” aspect when you say “use randomness.” This refers to the fact that the method samples one possible sequence of events and uses that sequence to update its model. As opposed to dynamic programming methods where the value of all possible sequences is estimated and used to update the model. Not sure that the randomness from quantum is useful here. Where it could be useful is to encourage exploration of the state space. So, normally Monte Carlo methods have various tricks to make sure every sequence is sampled infinitely many times. These tricks could be implemented by the error in quantum computing, but I don’t know enough about the field to really be sure of that. And of course it would help in that you could compute the results of many sequences at once which is critical bottleneck in dynamic programming.

You're thinking in the same direction I am.. but I am definitely too far out of my knowledge level to take it much further without sitting down and really working hard on the problem. I'm going off the deep end here into what is probably pop science or science fiction and might not make any sense at all. I wish I understood more about this in detail.

I just always observed that quantum computing was interesting because certain states seemed unpredictable, which is why scientists studying quantum computing struggle when they don't get the expected states. If that unpredictability is actually random due to the function of quantum states, then it would provide a unique tool since true randomness really is not easily found.

And the ability to potentially generate many random/expected states quickly - yes, I was thinking about being able to generate/explore the space faster. Or to use that in logical processing somehow as an advantage.

I have always been fascinated by a simple concept in logic - "maybe". I'm borrowing from "fuzzy logic" a bit here and twisting it the way I wish for my thought process. If one formalizes "yes","no" and "maybe" (i.e. 0, 1 and something else), can one then make complex logical statements involving probabilities, randomness, etc. by exploiting the "maybe" case.

For example, if I have one of those magic eight-balls, but it represents quantum states - if I can clearly detect 1 or 0 but sometimes I get a 0.5 or a 0.8 unpredictably, how could I use that to my advantage...

Another line of thinking for this is neural networks, another thing which I wonder could be accelerated by quantum computing. If I make the analogy that, as a human with a brain that is like a neural network, when I am faced with inputs that I have not processed before, my brain has to come up with a solution to process it from my various experiences and it might product an output that makes sense, or it might not.

Knowing my own fallibility here, if I know that sometimes I am faced with inputs that I don't know how to process or didn't expect, I can find a logic for how to handle those situations that might let me work around them.

[1] https://en.wikipedia.org/wiki/Fuzzy_logic

What I think, measurement of a qubit returns 0 or 1 always. The unpredictability is in the part where you were expecting 0 and got 1 because of noise I guess. But the measurement of a qubit in superposition does result in a purely random result.

While I can’t comment on qutrit, the whole idea behind ternary computing is that not knowing something communicates information unto itself. For example, if a waiter asks one of a pair of diners if they’re both ready to order. Their reply of I don’t know, would communicate to the other diner that they are in fact both ready. I imagine the same ideas could be embedded in quantum computing.

Not sure if I read it on HN or where, but:

'Three logicians walk into a bar.

The bartender asks: "Do you all want a drink?" The first logician says: "I don't know." The second logician says: "I don't know." The third logician says: "Yes."'

Haha, yeah, exactly what I was trying say above, although I feel like this does it a lot better.

For those thinking Star Trek, they literally address it a few paragraphs down :)

The name quantum teleportation brings to mind a technology out of Star Trek, where “transporters” can “beam” macroscale objects—even living humans—between far-distant points in space. Reality is less glamorous. In quantum teleportation, the states of two entangled particles are what is transported—for instance, the spin of an electron. Even when far apart, entangled particles share a mysterious connection; in the case of two entangled electrons, whatever happens to one’s spin influences that of the other, instantaneously.

> in the case of two entangled electrons, whatever happens to one’s spin influences that of the other, instantaneously.

This is a very lazy way of saying what happens in the experiment and what eventually leads to confusion. "influences" is the most problematic word here.

A non-incorrect statement would be, "the outcome of measurements are found to be correlated when the two distant observers meet at a future time and compare notes."

I think at our current understanding of the philosophy of physics, saying more is not something most people should be attempting.

I think most people can intuitively understand the observed behavior by thinking about the process as if the two particles had been "pre-programmed" to correlate with each other at the time of entanglement. With that explanation, things seem pretty mundane, then it can be explained that even though the behavior appears like the particles have been pre-programmed, the nitty gritty of Bell's Theorem shows us that this isn't the case, and that's why it is so fascinating to scientists.

This is correct, but this also doesn't really capture how significant and incredible of a feat quantum teleportation is. If we have a system in a quantum state A, it is 100% impossible to actually determine A with any precision. The best you can do is say that when you inspect A, you get will get this collection of classical outcomes with these probabilities. The state A itself, while arguably quite real, is fundamentally unknowable. However we can move this state A from particle to particle or system to system, with theoretically absolute fidelity. This is quantum teleportation. It really is quite incredible. This thing that you can't know can be moved around. This is kind of philosophically non-trivial. Theres a pretty solid philosophy of science argument which is basically if you can't actually know something via experiment then what claim does it have to being called real? At first glance this might seem to apply to quantum states, but being able to move them around exactly atomically even though you cant really glance under the hood kind, gives some serious weight to there being something "real" there.

Yes, this is interesting because it seems like information would travel faster than light in this case, but based on our current knowledge about universe this would break Einstein's Theory of Special Relativity, which could mean there is some other out of band channel of communication between entangled particles.

I've heard it analogized something like this: You and I have a deck of two playing cards. We know which cards they are. We shuffle them, and each take one without looking. Then we can travel as far from each other as we want, and whenever I take a peek at my card I'll know instantaneously what card you have (and vice versa). We remain connected at a distance, but no information really traveled faster than c and you and I couldn't use our knowledge of the other's card to communicate.

Yeah, I've heard that analogy, and I've also read somewhere it's wrong, despite sounding good. But unfortunately I didn't absorb the explanation of why.

Specifically, it violates the no-communication theorem. Not only can't you signal faster than light, you can't signal using measurement at all.


What is “it”? Where is the violation?

Just that classical information cannot be sent faster than light using measurements of entangled particles.

Of course, quantum teleportation works, it's just not FTL, and the spooky connection between entangled particles can't be used to send information either.

So there is no violation anywhere. We agree :-)

Your parent thought that there could be one (“it seems like information would travel faster than light in this case”) and your response (“specifically, it violates ...”) was somewhat confusing.

One thing I always wonder with these quantum teleportation/communication experiments is if they are actually dealing onesy twosy with photons or if it's like a particle accelerator where they are dealing with millions/billions and are just looking for the statistical outcomes and/or anomalies.

They deal with one photon at a time. The total number of events that are used to estimate probabilities usually number in the hundreds or thousands, usually collected over the course of hours or days.

Awesome thanks!

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact