I somehow feel like quantum computing will forever remain a theory and never make it into practice. How much reason is there to believe that I might be wrong?
Well quantum computing is obviously physically possible. Somehow systems evolve from a prepared input to an output predicted by quantum mechanics.
However the difficulty of scaling quantum computers goes up exponentially with the number of qubits, thus we may not achieve enough success to make them practical.
We have every reason to believe that quantum computing is possible, but it could happen that it's impossible (from an engineering point of view) to have stable coherence in sufficiently large numbers of qbits to do anything useful.
> Well quantum computing is obviously physically possible.
Well, since you mentioned this: would you mind linking me to a single demonstration of it that has been conclusively established as evidence of quantum computing? I've seen lots of D-Wave stuff but to my knowledge the experts don't actually really buy those.
> However the difficulty of scaling quantum computers goes up exponentially with the number of qubits, so we might not achieve enough success to make them practical.
But yes, this is more in line with what I was thinking. Even if it's possible I don't expect it will be practical (hence why I used this word in the earlier comment).
> would you mind linking me to a single demonstration of it
I mean physics is quantum. At least, the interface between our experimental apparatuses and the micro world demonstrates all the expected speedups. Every known experiment backs this up.
> Supposedly ~17 qubit devices are on the way. None has demonstrated quantum supremacy, practical or not.
That's exactly what I mean. I haven't seen any evidence of quantum computing being demonstrated. All I know is it's hypothetically possible. I'm not sure what people were downvoting when I said this, I'm just stating a fact. It's not like I wished it wouldn't happen or something.
Quantum mechanics is a very well established theory, and quantum computing is squarely within its orthodox predictions.
If quantum computing turns out to be impossible, and not just because engineering is hard, that would need a great and exciting new theory. Perhaps more exciting than quantum computing would have been.
In the next decade: Both won't happen. Building a fusion power plant would take at least a decade if we already knew how to do it. Quantum computers are still in the "We can find primefactors of numbers < 100" stage of research.
I'm not sure which of those (if any) demonstrate conclusively that the computation had quantum speedup. I know experts have disputed at least some of the claimed demonstrations.
1998: First computers with 2 and then 3 bits of memory.
2000: Computers with 5 and 7 bits of memory.
2006: 12 bits of memory
2011: 14 bits
2017: 17 bits
There is too little data to suggest anything else than a linear fit. But then again, the early stages of exponential, or logarithmic, growth can look linear. We should wait 10-20 years to get more data.
Flying cars not existing has everything to do with the resources they use (public federally controlled airspace which requires every driver to be a pilot.)
Quantum computers have the same barriers as other private space technology. The question to me is how long they will take to go from research toys to hobby toys which has a lot to do with cost and something to do with consumer safety.
Actually I mentioned flying cars due to fuel inefficiency (I automatically assumed it was too low to make them practical) but I neglected to actually check their energy usage, and now that I just did it seems like the maximum efficiency could be on par with a normal car and it would depend on the actual trip, so my mistake on that.
A better example might be artificial intelligence and the AI winters. Like the failure of machine translation despite high hopes in the 1960s. It took us some 50 years to get to where we are now with Google Translate, and it still only works well for some languages. I have no idea how close to human accuracy it will get, so I would neither be surprised if it never does, nor if it does someday.
Anyway, hopefully instead of picking on the particular examples you can see that my point is obviously that people imagined we'd have a lot of things now that we still don't have the technology for today. So the fact that we've done amazing things in certain areas doesn't really say anything regarding this particular issue.
> my point is obviously that people imagined we'd have a lot of things now that we still don't have the technology for today. So the fact that we've done amazing things in certain areas doesn't really say anything regarding this particular issue.
And there's plenty of things that people said were impossible that were achieved.
That doesn't tell us anything regarding quantum computing, either.
Looking at what people said about past things is completely irrelevant to the prospects for some particular potential technology. The only way you can try to judge the prospects for that technology is by actually getting into the details of that technology and how it relates to our current understanding of the world.
I'm not refuting anything. All your comments have focused on things that people have promised that haven't turned out, which gives one the impression that that somehow applies to quantum computing as well. I'm trying to paint a more balanced picture of the situation.
I see your point. I'm just trying to point out that failures of general tech to emerge are quite different than failures of use cases. You need more of a limit preventing the emergence of a Moore's law on entanglements to really make quantum computing go away and not just solve other problems than those originally assumed to be "easy" and months away.
> I replied that, even though we’ve by now been around this carousel several times, I felt like the ball was now firmly in the D-Wave skeptics’ court, to reproduce the observed performance classically. And if, after a year or so, no one could, that would be a good time to start taking seriously that a D-Wave speedup might finally be here to stay—and to move on to the next question, of whether this speedup had anything to do with quantum computation, or only with the building of a piece of special-purpose optimization hardware.
Anybody who claims to know when we'll be able to build practical quantum computers doesn't know what they are talking about. Scaling is a very difficult problem.
Quantum Supremacy doesn't require practical quantum computers. Special purpose quantum computers that can only do one task better than classical computers, even if that task is useless, meet the definition of quantum supremacy.
There will be 2 answers to that question and they will both be right and wrong at the same time :)
Maybe not at the end of this year, but very soon, quantum supremacy will be claimed by one of companies that are developing a quantum computer similar to the d-wave computer used by Google and others. That is sort a single purpose computer that can only be used to perform a very limited set of work loads, but it still qualifies as a quantum computer, but will most like be a very limited in what it can do and how useful it will be.
A few years later you will see quantum supremacy be claimed by one the companies trying to build a general purpose quantum computer like Microsoft. It is first then the quantum revolution will take off.
Both Google and IBM are building general purpose computers not D-Wave-like quantum computers. They will likely both reach quantum supremacy before Microsoft, which itself has said could take another decade to make something useful out of its topological quantum computing architecture.
