Quantum computing will never happen. Discuss.

[philip_hughes]

Hello the ladies and gents I am back and I am ready to kick off a few topics which I have been thinking about over the last I don't know decade the first of them is the fact that since we last chatted I've gotten myself a degree and PhD and a whole heap of other things and what I point out I use voice to text a lot which means no punctuation sorry.

My degree was heavily based in statistics and what I've discovered is that physics and statistics are very different Beasts. Quantum computing is very interesting because it's one of the only situations I know of where a huge amount of money has been spent on something that has never actually been observed what I mean by this is that the concepts behind quantum computing a basically Tunneling and the ratio of two versions of polarized light. This is problematic not only because it's not really that well measured but because if you simply polarize a laser and point it at the wall you will discover that the point does not shimmer that means there is no ratio between the light. I'm actually building a rig to formally confirm this and in case you're wondering about the actual experiment and I can't remember the actual names and I do apologize for that but basically what happens is you fire the laser at a polarized sheet and you split the light and then you measure the two dots that appear from the light split and polarization from that you get a ratio.

Seeing this is the start of the discussion I'm not going to prove very much stuff I'm just going to put the points out but basically my belief is that this ratio is nothing but a statistical artifact known as Simpsons paradox I will discuss it later if you're interested and also I have some criticisms of quantum Tunneling as well but I will let other people talk for a while first have a wonderful time and this is supposed to be fun it's not supposed to be evil evil angry discussion it's a fun debate.

[Usenko]

Hey, where did THIS clown spring from?

(For anyone who wasn't here back in the cretaceous period: Philip is my little brother. He's the one that got me into X games in the first place. Thanks mate! )

[philip_hughes]

Hey, where did THIS clown spring from?

(For anyone who wasn't here back in the cretaceous period: Philip is my little brother. He's the one that got me into X games in the first place. Thanks mate! )

I'll be streaming soon.

[decifer]

Difficult discussion when it's already done almost 30 years ago.

https://www.spinquanta.com/news-detail/ ... 0214081413

[philip_hughes]

Difficult discussion when it's already done almost 30 years ago.

https://www.spinquanta.com/news-detail/ ... 0214081413

Invented and works are 2 different things. It still has a frightful error rate- 80%... worse than flipping a coin

[decifer]

Difficult discussion when it's already done almost 30 years ago.

https://www.spinquanta.com/news-detail/ ... 0214081413

Invented and works are 2 different things. It still has a frightful error rate- 80%... worse than flipping a coin

I disagree. The concept is proven, so it "works". "Just working" and "working well" are different things, indeed. So your question should more be like "Will we ever be able reduce the error rate to a reasonable level?" instead of "will quantum computing ever happen?"
It's similar to nuclear fusion. The concept is proven and it works, but whether it will become a feasible power source at any point in the near future or not is still up to debate.

[philip_hughes]

Difficult discussion when it's already done almost 30 years ago.

https://www.spinquanta.com/news-detail/ ... 0214081413

Invented and works are 2 different things. It still has a frightful error rate- 80%... worse than flipping a coin

I disagree. The concept is proven, so it "works". "Just working" and "working well" are different things, indeed. So your question should more be like "Will we ever be able reduce the error rate to a reasonable level?" instead of "will quantum computing ever happen?"
It's similar to nuclear fusion. The concept is proven and it works, but whether it will become a feasible power source at any point in the near future or not is still up to debate.

Let's explore this the way I see it I don't think that the principles by which they say the computer is working are the principles by which the computer is working that's my problem. I accept that you accept that the computer is working as they say on the tin so I suppose it's my job to convince you that it's not actually doing what they say it's doing. Is that a fair enough proposition?

[decifer]

Is that a fair enough proposition?

Sure. But I guess I have to disappoint you. That's already the limit of my knowledge about that topic. I know, it's working according to predictions and the current understanding of QM, that there are some existing QC and that the only real application right now is research. Other than that, if you have a theory about it and - if I understand correctly - are even working on an experiment to prove it, you're already far ahead of me.
I mean, I could throw some googled or AI generated stuff at you, but I guess that would be pretty boring if not annyoing for you.

So all I can say is: No idea. But my money would be on the quantum physicists.

[euclid]

Yes, noise it a problem but it was also a problem in the first attempts of building a gravitational wave or a neutrino detector. In time, both have been overcome, not in a perfect way (yet) but both delivered decent results. And yes, you are right: It is based on statistcs but hey, it's QM, what else to expect.

Cheers Euclid

[philip_hughes]

Is that a fair enough proposition?

Sure. But I guess I have to disappoint you. That's already the limit of my knowledge about that topic. I know, it's working according to predictions and the current understanding of QM, that there are some existing QC and that the only real application right now is research. Other than that, if you have a theory about it and - if I understand correctly - are even working on an experiment to prove it, you're already far ahead of me.
I mean, I could throw some googled or AI generated stuff at you, but I guess that would be pretty boring if not annyoing for you.

So all I can say is: No idea. But my money would be on the quantum physicists.

That is literally my problem. I can't just say I have a death Ray that works and I use it for research purposes and you know you can't see it if I did that would you be a bit suspicious?

The kicker here is i have observed and recorded the EXACT PRINCIPAL that is being used to leverage the qbit. It's on YouTube. I discovered this when i was building my own double slit experiment. The experiments that confirm this action in quantum physics report results but do not release the full dataset so there is no falsifiability.

In other words, the "working" computer may well be operational, but is just reworking random nonsense. This is not a working quantum computer... it's a glorified beer fridge.

[mr.WHO]

Yes, noise it a problem but it was also a problem in the first attempts of building a gravitational wave or a neutrino detector. In time, both have been overcome, not in a perfect way (yet) but both delivered decent results. And yes, you are right: It is based on statistcs but hey, it's QM, what else to expect.

Cheers Euclid

I read somewhere that noise is still a problem.
Aparently the higher/more computing speed you run with QC much more noise come into action (similar to where in normal CPUs get into errors due to quantum tunnelling the smaller the CPUs get).
Unless this is solved it seem QC will have a narrow band of computing speed where it's useful (e.g. much faster than normal computing, but still provide valid results without too much errors due to noise).
If noise issue won't be solved, QC might remain very niche and specific area of application.


Oh and most of people don't understand that QC is not a holy grail - quantum processor won't beat your old CPU in running Witcher 3 (QC is used for propability/dustributed/paralel computing, while normal processors are applied for deterministing 0/1 and sequenced computing).

Imagine this in Witcher 3:
Event A - player swing the sword.
Event B - player hit the monster.
Event C - hit takes 100% of monster life killing it.
Event D - player get 100 EXP and level up.

Classic CPU will go with A-->B-->C-->D with 99,999999% accuracy (the are minor calcualtion errors still accounted in classic CPUs, because 0 or 1 in wrong place can produce fatal error).


Now with QC imagine you're doing faster calculation because you calculate all 4 events in the same time...which of course make no sence due to messed cause and effect.
Now then why not put them in sequence? Well, you can, but it will looks like that:

Event A - 95% chance of success (5% error/different result due to noise)
Event B - 95% chance of success
Event C - 95% chance of success
Event D - 95% chance of success

...and you go from A to D SLOWER than with classic CPU,while using much more expensive and complex Hardware (e.g. imagine having latest most expensive alienware rig...and go play OG Tetris).
...and it will run worse as at each action all you need is bad RNG roll to have completely different outcome (or worse, a fatal computation error).

Now imagine you actually want above to be super fast, like milion times faster than classic CPU - you can, but then you don't have 95% chances but 50%.

Still there are different things that are perfect for QC to be applied.
I assume that if there will be a technological breakthrough, then in 10-15 years in future, we might have QCPU as a standard along CPU, GPU and RAM in our PCs.


...unless f*cking Bezos and Co steal all RAM and make us cloud slaves with Stadia 2.0

[chew-ie]

I'd say as soon as it happens we'll get all the answers and forgot about the question.

[decifer]

This is not a working quantum computer... it's a glorified beer fridge.

Yeah, I mean, that's what it is. Even Hannah Fry says this

I thhink, there are only a few options:
a) It's all a hoax.
b) They're correct, the reality holds true to the predictions and theories and you're overlooking/misinterpreting/missing something.
c) You're correct and probably provide a major milestone for QM.

I'd rule out a) for myself. I doubt that, because the people involved are smart enough to know, it would become a massive boomerang at some point. And I don't think the whole scientific community of maths and physics involved in this is suddenly throwing all ethics overboard for some tech-bro money.
B) would be good for you, as you're learning something no matter what and are applying science the way it should be done. And maybe there are still new insights for everyone in your results, who knows?
c) would obviously be good for the whole community and yourself.

So, go for it.

I myself have no stakes here, could not argue for one or the other side and see only good outcomes. But as I wrote, my money would be on the guys who try to figure this stuff out since decades. And I have some trust into things Feynman thought where true. I know, "don't appeal to authority", but I do in this case

[Chips]

Must admit, it's all way over my head. Seen/read the layman's explanations for what's going on, but I've zero *real* understanding of how they actually achieve it, and a keen awareness that I'm no-where near intelligent enough to *really* understand it either. That's also ignoring that if I were, then how much time would be required to build the knowledge to understand to the point of being able to offer any form of critical assessment of the field and contribute here. I'd certainly be in the wrong job and catastrophically underperforming life potential if I were, too

However, when you mentioned error rate it triggers memory of a Microsoft publication a short while back that claims to have improved error rate vastly. Zero idea if it's only theoretical, actually practical, or overstating (within a specific confines). Sorry I can't contribute anything meaningful

[philip_hughes]

This is not a working quantum computer... it's a glorified beer fridge.

Yeah, I mean, that's what it is. Even Hannah Fry says this

I thhink, there are only a few options:
a) It's all a hoax.
b) They're correct, the reality holds true to the predictions and theories and you're overlooking/misinterpreting/missing something.
c) You're correct and probably provide a major milestone for QM.

I'd rule out a) for myself. I doubt that, because the people involved are smart enough to know, it would become a massive boomerang at some point. And I don't think the whole scientific community of maths and physics involved in this is suddenly throwing all ethics overboard for some tech-bro money.
B) would be good for you, as you're learning something no matter what and are applying science the way it should be done. And maybe there are still new insights for everyone in your results, who knows?
c) would obviously be good for the whole community and yourself.

So, go for it.

I myself have no stakes here, could not argue for one or the other side and see only good outcomes. But as I wrote, my money would be on the guys who try to figure this stuff out since decades. And I have some trust into things Feynman thought where true. I know, "don't appeal to authority", but I do in this case

An absolutely balanced approach thank you for saying that. I am currently building an experiment I haven't performed The Experiment and if I go in saying x is true and Y is false then I'm a crap researcher no if ands or butts.

Let's get a few things out of the way. The reason why I believe quantum Tunneling is not a thing is because it's perfectly explainable with other mechanisms. Is probably better explanations to this but I'm going to start with Rutherford and say that his explanation of radiation required Tunneling. The issue as I see it is because radiation and photons and any other subatomic particle have always been viewed as a stochastic measure not a point form. These have and always will be measured in distributions because of our capacity to detect the things. If there is a distribution there is a probability that the Tunneling event you're seeing is just something that's happening at the edge of the statistical envelope in other words no tumblings occurring it's just people need to tighten the measurement.

Regarding entanglement that I think is a worse statistical mistake and it's the same mistake that got a woman convicted for Murder when her two children died of cot death. Here is not the calculation of probabilities but the fact that if you have a binary event or a Boolean you can't have part of the bullion you must have the whole boolean no matter what the probabilities are. To put a crudely you can't have a fraction of a dead baby as an aside if someone wants to be literal about that point you're a sick person.

Moving on the exact point that I'm trying to explain is something called collapse of the wave function the collapse is not a collapse in the traditional stent since it is the object being detected if you do not have enough statistical strength then that object will be detected as a whole object and therefore bias the statistics just from either this is sampling along part of a distribution or just not having enough samples.

Either way I have good reason to doubt the explanations of Tunneling and spooky action at a distance they are real concerns they are statistically based and because these concerns still exist to create multi-billion dollar computers of concepts which are not set in stone seems strange to me but it's okay what we now have is a test if quantum computing works and quantum physics is a thing then I'm wrong if it's not a thing they're never going to get them to work so we literally have a good test regime up and running now

[decifer]

Well, not the easiest read, especially as a non-native english speaker. So I apologize if I misunterstood something.

I mean, most of those principles are still debated, at least partially, so you are not alone. See Sir Roger Penrose, whom I highly respect and mostly not understand. He's not really a fan of the Copenhagen Interpretation or the Many Worlds Theory.

Quantum Tunnelling on the other hand has practical applications, like QTC touch screens for example, that would need an alternative explanation.

Even Einstein wasn't happy with entanglement, but apparently it is proven in an experiment - which I assume is the one you're talking about. I still don't understand, why it proves that the outcome of the measurement can't be a predetermined value of each particle, but that 's because I don't understand either the statistical part of it, nor the quantum physical one. I've never seen, what the same distribution of each spin measurement is, if you discard the other partner completely without measuring it, compared to no entanglement at all. Did anyone check, if the process of splitting/entangling alters the distribution? But I highly suspect they did and it's just not as exciting as the actual experiment and thus not reported to the same extend. I also have no idea whether that would even be relevant or prove anything.

But since I don't understand most of the stuff going on there, I also accept that's just a feeling about something I can't fully grasp and therefore trust the countless people recreating that experiment and studying the topic more than my gut. And even they admit it is a weird concept. So if you doubt it, I can absolutely understand that and I'm actually excited for whatever you find.

[philip_hughes]

Well, not the easiest read, especially as a non-native english speaker. So I apologize if I misunterstood something.

I mean, most of those principles are still debated, at least partially, so you are not alone. See Sir Roger Penrose, whom I highly respect and mostly not understand. He's not really a fan of the Copenhagen Interpretation or the Many Worlds Theory.

Quantum Tunnelling on the other hand has practical applications, like QTC touch screens for example, that would need an alternative explanation.

Even Einstein wasn't happy with entanglement, but apparently it is proven in an experiment - which I assume is the one you're talking about. I still don't understand, why it proves that the outcome of the measurement can't be a predetermined value of each particle, but that 's because I don't understand either the statistical part of it, nor the quantum physical one. I've never seen, what the same distribution of each spin measurement is, if you discard the other partner completely without measuring it, compared to no entanglement at all. Did anyone check, if the process of splitting/entangling alters the distribution? But I highly suspect they did and it's just not as exciting as the actual experiment and thus not reported to the same extend. I also have no idea whether that would even be relevant or prove anything.

But since I don't understand most of the stuff going on there, I also accept that's just a feeling about something I can't fully grasp and therefore trust the countless people recreating that experiment and studying the topic more than my gut. And even they admit it is a weird concept. So if you doubt it, I can absolutely understand that and I'm actually excited for whatever you find.

QUANTUM TUNNELLING

@ -> .-iilii-. -> O
Atom photon distribution detector

(Best diagram i could manage)
Photon may be discrete but we can't see it. We only see the distribution....

The photon is somewhere in the distribution, in reality is

. , / | \ , .
. | |
. ^ *
Anywhere | in here!

But let's assume it's where the asterisk is....

If that photon hits the detector, it does so ahead of time. Not because it tunnelled, but because it left earlier than expected.

Now let's think of atomic models...
Rutherfords experiment only mentioned something small and dense within the atom. It was assumed to be the centre. There's nothing stopping it from being to the side or even embedded in the walls! Given this level of uncertainty you need to eliminate this stuff before you assume something breaks causality.

*edit... my diagram was awesome in composition. The printed version compressed the spaces

[decifer]

Sorry, but I can't really follow you anymore. But there are multiple experiments providing evidence and even multiple practical applications for quantum tunnelling. Another one would be flash memory present in a lot of modern devices.

But I'll leave this topic now I guess. I've basically said everything I know from following the occassional physics vlog or talk every now and then. Good luck with your experiment.

[philip_hughes]

Sorry, but I can't really follow you anymore. But there are multiple experiments providing evidence and even multiple practical applications for quantum tunnelling. Another one would be flash memory present in a lot of modern devices.

But I'll leave this topic now I guess. I've basically said everything I know from following the occassional physics vlog or talk every now and then. Good luck with your experiment.

If you must but flash memory etc is not exactly the breakthrough that you imagine for example a seven segment display holds 256 states if you get two seven section displays all of a sudden thats 60 odd thousand bits of data, 3 moves into the billions. so flash memory just doesn't impress me. It's inefficient at best. I have recently spent a lot of time studying computing for various reasons and it's occurred to me that what we have as computing is literally abstraction on abstraction and it's mostly control based.

Consider this:

A comparitor compares a conditioned pulse. It's over 2v.

It shunts a charge to bits of a segment display that light up a 2. That's computing.

Changing from voltage to "does this capacitor hold charge?" Creates a system which is neither binary or analogue but is ridiculously powerful.

If capacitor full, dischsrge, move to next cap and fill..

If that logic is applied pulse wise:

Pulse 1
1
P2
01
P3
11
P4
001
P5
101....

Ripple adding in hardware! And yes I'm building it as we speak!

Binary is abstraction.

Computation itself is understandable if you just look.

[esd]

Some electrides use quantum tunnelling to move electrons. Mayenite is a calcium aluminate oxide that used to be an insulator but in its new electride state is a conductor. And a catalyst.

So quantum tunnelling is a thing.

As for quantum computing, it's real. The biggest problem seems to be error correction, essentially a Signal to Noise Ratio problem, which generally is a matter of practise to filter the errors. Linkypoo to Ars' latest QC news.

It's not like our non quantum computers don't have error correction too! Single Event Upsets are caused by various radiation hitting a chip and flipping a bit, and need error correcting where possible.