The Quantum Computing Speed Boost Is NOT What You Think

People say that Quantum Computers will take over the world, because they're faster at some types of problems than classical computers. But Why? What makes quantum computers Tick? In this video I discuss the concept of superposition, and how it relates to quantum computers quantum leap in speed.

Thanks so much for your support on the previous two videos, I really appreciate it!

Music by Vincent Rubinetti
Download the music on Bandcamp:
Stream the music on Spotify:

[ChronicSkooma]

I love this, all of what you're doing. I watched a TON of Stanford's videos from Leonard S. Its so dense and I just keep throwing myself at it. You are taking concepts I've been trying to get and making them so much more digestible. Please keep explaining the formulas and how they describe what's going on. You're a fantastic teacher. I want longer videos, but also that would come at a cost of production time and real value to each word, so its a wash.

[markr]

This is the clearest explanation I've seen yet. Please keep up with the video production. Subscribed.

[LordOfTamarac]

Please please make more videos, this is the perfect level of complexity. I’ve watched tons of quantum computing videos and these are the most elucidating by far.

[Futureflux]

Excellent work 🔥Your explanations of how the whole thing works — with a discussion about quantum gates — are so wonderfully clear!

I'm currently scripting a video about quantum computers and their possible industry applications, and I will definitely reference yours as recommended content. Thank you for putting this out there.

[AhmedMostafa-euup]

great video, waiting for the entanglement video😁

[Nagria]

instant new subscriber here.
love your work - keep it up.
you came out of nowhere and made the best Qantum computing youtube videos.

[vdotfour]

You should make a discord! I think your videos take grand concepts and put them in a manner that's very digestable and if I'm being honest, entertaining as well to learn about. I'd be glad to join and I'm sure many others would be too!

[Bt_R]

Finally someone explaining me this! Thank you. I've always thought it works something like your waterpath analogy, but all other explanations i found seemed to not confirm that. They've only always talked about the fact that a qbit can be at 0 and 1 at the same time, but with that alone, how would this be so special. Now i got it. can't wait for your explanations of some algorythms! subbed.

some questions i hope will be answered in this series:
- does the amount of available qbits determine the complexity of a problem a quantum computer can solve? or could any problem be solved already with a few qbits, by some sort of "chaining"?
- can we try that out ourselves, either on real systems or simulated ones?
- what are specific tasks that a quantum computer can solve currently, detailled examples with your toolbox we've seen in this video.
- what are tasks that a quantum computer might be able to solve in the future with better algorythms and/or more qbits?
- are there tasks that a quantum computer will never be able to solve, or much slower than a classic one?
- currently we seem to work with only o|1 states. could this be extended to more states in the future? would it be useful?

[GopalSharma-jcjp]

Thanks for a wonderful video and a flawless explanation. Have a good one!

[simplyme]

Very well explained. My sincerest congratulations!

[KipIngram]

6:42 - We "just" need an algorithm... That's an awfully big "just" - it's *the* problem of quantum computing. In a few nice cases we've found such an algorithm, but we really don't understand how to generate them "generically."

[KipIngram]

I don't understand the title... What you described is exactly what I was expecting. What do most people "think" is the reason for quantum computing power?

[CraigMiddleton-cg]

Clear explanation even for someone with a limited physics background

[szilardecsenyi]

it would definitely be worth mentioning QFFT and the Hidden subgroup problem in your next video.

[Adam_Wegert]

When I watched this video until 2:39 my immediate thought was: Michio Kaku should be put in dilution refrigerator and shielded from the environment for producing (dis?mis?)information theoretic noise about quantum computing xD

[veervishalmishra]

Nice editing and explanation, by the way
How many quantum computer giveaway at 100M subs😂

[snap-off]

Classical computers: Program for a little bit, wait for an eternity for an answer.
Quantum computers: Program for an eternity, wait for a little bit for the answer.

[SuperMaDBrothers]

Is the walsh hadamard transform not just n hadamards applied in parallel?

[edcorns]

Excellent explanation of how quantum computers are *supposed* to work... but actually don't. Let me explain.

The math behind quantum computers goes all the way back to that initial *conjecture* of quantum mechanics that a particle takes all possible paths *simultaneously* ... which, in fact, is *not* how physics works in this universe. The actually correct model of quantum physics is the *pilot wave* one, where a particle is an actual "particle" (a region of space with particle properties, like specific position, specific energy content, etc.), and that "particle" is being guided by a pilot (guiding) "wave" that it produces by its mere existence (via the perturbation of space/changing of potentials in the surrounding spatial "field"/spatial nodes).

What actually propagates in all directions ("takes all possible paths through space", though that's not really a correct way of looking at it) is the pilot "wave", whereas the "particle" itself takes *one and only one path* through space, and that path is determined by the interactions of the particle's own pilot wave with pilot waves of all surrounding particles (plus any "reflections" of the particle's own pilot wave caused by the effects of that pilot wave on those other surrounding particles).

What all of this means is that, yes, one can never know the *exact* path a particle will take, because the very act of "observation" (measuring the particle via interaction with another particle) will perturb the measured particle, and thus affect its behavior (via its pilot wave), so quantum mechanics is correct in that one regard that one *must* model the behavior of quantum systems as *probabilistic*, since one does not have *enough information* about the system to use a deterministic model (plus, some of the probabilistic behavior of quantum systems have their origin *outside of this universe*, and generally, one can never model events *external to the (measuring) system* as deterministic, since one has no information, whatsoever, about anything external to one's own system [of existence]).

So, here's what I really have to say about "quantum computers": *Quantum computers are finite-state machines with probabilistic execution* (meaning, a machine with *deterministic states* and *probabilistic transitions* between those states).

To break it all down, what the above statement means is that:
a) quantum computer can only take *finite number of states* (which what is usually called "the collapse of the wave function"),
b) quantum computer does *not* search the whole problem-space *simultaneously* (since a particle will *always* take *one and only one* path through space),
c) particle's pilot *wave* will take all paths through space simultaneously, yes, but that will only give one *the most likely* (with the highest probability) path that a particle *would* (but actually won't) take if it was *completely stationary* (didn't move at all), and thus *didn't interact with anything* ; i.e. if that particle wasn't used to perform any computation at all (by interacting with other particles), and was, therefore, *utterly pointless* to be used in a systems called 'a (quantum) computer' (which is supposed to actually compute something)
d) since a particle takes one and only one path *with each execution* of the "quantum computer" algorithm, a "quantum computer" actually searches the problem-space *sequentially*, just like a classical computer does(!), but that searching is done *not deterministically* (or iteratively, one step at a time without repeating any step more than once, as classical computers do), but *probabilistically*, with *no (user) control* over which branch is going to get searched (at any execution step), and the non-zero probability of the same step being searched more than once (which is a lesser problem), and *some branches not being searched at all(!)* (which is a much bigger problem).

The latter (bigger) problem means that, not only those 1, 000 executions of the IBM's "quantum computer" algorithm may actually not be enough to get a (correct, let alone optimal) solution, but that *solution is not guaranteed for any number of executions that is less than infinite* (number of executions)!

To summarize: A "quantum computer" (machine) is to the 21st century what "perpetuum mobile machine" was to the 19th century (before the laws of thermodynamics were formalized), when everybody and their grandmother was trying to get *free energy* out of literally nothing. With "quantum computers", (almost) everybody and their grandmother is now trying to get *free computation* (or, in practice, *free time* not having to be spent on computation) out of literally nothing.

More specifically, U.S. "supremacy" in "quantum computing" would be an equivalent of the (early-to-middle) 19th century America claiming to be on the verge of building a free-energy perpetuum-mobile machine, at the moment when all American engineers had already been fully aware of the laws of thermodynamics (and the literal impossibility of such a machine) for some time, but their Chinese and Russian counterparts weren't yet aware of any such laws, so American intelligence services were using that *gap in knowledge* to make Russians and Chinese waste their money, time, and energy (any and all possible efforts, in general) trying to build a literally impossible machine (called "quantum computer").

[ShirleyMcCoy-ttbn]

So these quantum computers are like the old time analogue computers?