Ive now been slightly obsessively reading and discussing quantum computers with friends (including a couple clever and informed ones) for two weeks and the theoretical possibility of one day feeding the traveling salesman to a QC is about the biggest real excitement that awaits us in the medium term (decades). Hence my Sim City comment. Seriously there is very little information and a lot of exaggeration in this video. I know great things are expected from QC I just dont believe 98% of whats in the vid has anything to do with anything.

...Maybe? It would absolutely annihilate at something like chess, or Go. I have a hard time imaging a good use case for having it actually run a video game, but I'm guessing few people working on early traditional computers could've envisioned any of the delightful diversions we now take as a given. Probably when I'm 80 kids will be playing quantum Minecraft in a layered omniverse of worlds, where removing a block in one world has consequences in nearby dimensions, with chaos theory realistically modeled and incorporated.

Some complex tasks a QC would absolutely rock at however. Feed it a long list of employees, hours of availability, and coverage requirements, and it should spit out a 100% optimum schedule immediately. Air traffic controllers (particularly at large hub airports) would likely find it helpful in coordinating flight plans. Logistics for manufacturing, shipping, etc. The downside is that encryption will likely be utterly fucked for a while, as a quantum computer with a sufficient number of qubits could try all possible options at once. So it'll be interesting, but we're still 10+ years from any sort of commercial products, and they'll be like the computers of the 60s: huge and expensive, big iron for custom purposes. Or at least that's my semi-informed guess, I ain't no technoprophet.

Someone who really wants to get involved in bleeding-edge tech would do well to dive into this field. Writing the algorithms needed to run a task on a QC requires a completely different mindset than programming a traditional computer. I don't think people with years of experience with current programming methodologies would adapt well. At best they'd be nearly starting from scratch, at worst they'd have to work to un-learn what they already know.

It'll be useful eventually, but I wouldn't bank on soon. My final project in college was related to quantum computing, which at the time (18 years ago) was effectively entirely theoretical. I've enjoyed seeing the steady, albeit slow, progress.

The areas where quantum computing will really shine are problems which involve a huge number of possible answers, but only one best or correct one. The traveling salesman problem is a classic of computer science, as you can scale it up in complexity to the point where any traditional computer will eventually choke on the sheer number of permutations to test. Great way to demonstrate the need for clever solutions and well-written algorithms vs brute force approaches. An adequately sophisticated quantum computer, however, will theoretically be able to solve the traveling salesman problem nearly instantly, regardless of the level of complexity / number of nodes to navigate. Because it just tests all possible answers simultaneously.

...then someone creates a stable, proven-theory quantum computer in their garage and bitcoins become worthless.

There is skepticism that the DWave even actually is a quantum computer:

https://www.wired.com/2014/05/quantum-computing/

I was hoping for more meat to his presentation, and was disappointed. I feel that he said absolutely nothing to help anyone in the audience understand what quantum computers actually DO or what sort of problems they'll help to solve. They'll absolutely not increase your FPS, as that's not what they're well-suited to do. What they are quite excellent at is taking a problem with many possible solutions and finding the correct (or best) one at an extremely high speed.

One example would be the Traveling Salesman problem. In brief, find the optimum route for traversing a number of points on a map. This is useful for things like scheduling package delivery routes, airline flights, etc. With a classic / current computer we write software that cleverly chugs through the possible solutions, throws out any that prove to be poor, and eventually gets to what appears to be the best or is at least a "good enough" solution. As the number of necessary points to be visited increases this problem scales in complexity quickly, so eventually a current computer would just choke on the problem and at best return an ok-ish solution in a reasonable period of time.

A quantum computer is a totally different beast. If it's "big" enough (IE, is comprised of a sufficient number of qubits), it takes the entire set of all possible solutions to the problem, and rather than iterate through them to find the best one, it checks them all simultaneously and immediately returns the optimum solution. It does this by using properties of quantum mechanics, and I think this is where the speaker was drawing his talk of parallel universes. If there are 3 qubits, they would exist as 000, 001, 010, 011, 100, 101, 110, and 111 simultaneously. The software would then define what the best answer would look like, and the computer returns the answer.

You can hopefully see how this totally breaks encryption. With a current computer and a long enough encryption key, an encoded message would take the fastest machines a huge number of years to decipher. With a quantum computer you hand it a gibberish encrypted message, it loads all possible transformations of that message simultaneously, and it then returns the transformation which looks most like a coherent message.

I'm excited to see what these machines can do for us, but they're going to necessitate some significant structural changes in how we handle sensitive data.

One of those "problems conventional computers can't solve" is factoring primes (OK, they can, but it's very slow).

Quantum computers could make this very quick.... Which actually kinda sucks. Because, in a very simplistic sense, all encryption and by extension all electronic transactions depend on factoring primes being hard to do.

Basically quantum computers are going to break the Internet the second they become widely available.

With the amount of money being spent by really smart people I'm sure something about it is valid. It's just, right now, they're all yammering about facts not in evidence. Also, as their quantum computers are actually slower and less powerful than contemporary computers, occam's razor would suggest it's a elaborate black box scam with a couple Raspberry Pies burbling away inside. Until they start using them to increase my FPS, I'm not buying into the technobabble.

WHAAA?

<siftbot says...

Your video, Are You Ready To Be Outpaced By Machines? Quantum Computing, has made it into the Top 15 New Videos listing. Congratulations on your achievement. For your contribution you have been awarded 1 Power Point.>


Thank you very much...a wonderful surprise.

Your video, Are You Ready To Be Outpaced By Machines? Quantum Computing, has made it into the Top 15 New Videos listing. Congratulations on your achievement. For your contribution you have been awarded 1 Power Point.

Your video, Prime Minister Justin Trudeau Explains Quantum Computing, has made it into the Top 15 New Videos listing. Congratulations on your achievement. For your contribution you have been awarded 1 Power Point.

So if someone is gonna make a simulation of the universe..

It would likely be some Fermi Lab scientist who wanted to study the Big Bang.

They would reverse engineer the expansion of the universe as much as possible..

[ a thing that's already been done and being tweaked to get even better Planck length "resolution", as it were ]

And once they got the best estimations..

Would dump all those rules and variables into a quantum computer and run sim after sim, checking to see WHAT HAPPENS AT THE END!

Much like The Game of Life sim developed by mathematician John Conway.

Realistically recreating human consciousness - along with every event in the universe - is no small task.

It would require:

- quantum computing
- a data storage room the size of texas (if not all north america)
- easily more energy than is consumed by the entire planet in a year

So stating - "only one person would need to experience that simulation"..

..is like saying - "you would only need one person",

to recreate the Great Wall & Pyramids & Grand Canyon & Himalayas, etcetera..

Sure, I guess. But the amount of time & energy make it extremely improbable.

..Much more likely the civilization would collapse first.

The guy even concedes that point multiple times.

I'd love to get in on this conversation because this subject really interests me. This video touched on a lot of interesting theories.

@robdot - I don't understand people who think they "know" the answers to the universe. There are unanswered questions in every model. Do you know the answer to what dark matter and energy is? Nobody has yet detected it. Yet, our "universe" is supposedly filled with the stuff.

Let's also define what a universe is. My definition is; it's a place governed by the same set of physical laws.

So we have "our" universe, that we hypothesize about through our observations and measurements. We have theories that say "other" universes exist in some form or another. If their physical laws are different then ours, there would probably be no way to observe them, and therefore no way to prove their existence. Lack of proof is not proof that it doesn't exist.

I could write a book on what i "think" about what our universe is. For simplicity, let me just say that I moved from telecom engineering to software architect. In software, we create programs to run simulations. We create vast game worlds with whatever "physical" attributes we want to program into them. Lets assume we created artificial intelligence. In what context would "it" live? Most everyone assumes it would just be one conscience interacting with us in the form of a robot (Que cheesy Hollywood films).

Let's give it the power of quantum computing. It then decides to understand us (it's creator), it needs to program a simulation that mimics all it knows about our physical world. It wouldn't make one simulation, run it and be done. It would make many simulations, probably simultaneously, tweaking each new one based on the results of the previous ones.

Just imagine where this could lead. This intelligence could figure out how to create a multitude of different, very elegant universes. Its time scale would be different then our time. It's simulation could take seconds on its viewing scale, but appear to be billions of years when observing from within it. We have the power to pause, rewind, replay, tweak our simple creations. Imagine what this super intelligence could do with theirs?