William Hurley – Quantum Computing for Dummies

So, hey everybody, it's Whurley. I'm super happy to be here at Techstrong Con today to talk to you about quantum computing. So, I want to jump into it with a little bit of an overview, my personal story coming into quantum computing and, specifically, I want to say, you know, the session is quantum computing for dummies, and quantum computing is absolutely not for dummies.

When I first got into it, I understood that there would be a lot of physics, but it can be infinitely more complex than you think at first. And at the same time, there are so many ways to start getting into this new programming field, to start understanding how the physics work, how these machines work. And what we really want to do today is kind of break it down for you, and hopefully leave you with a sense of empowerment that you can go forth today.

Use some of the free tools you mentioned, use some of the resources we mentioned and actually get started playing with quantum computing. So with that, I want to jump right into it. I believe that this not artificial intelligence is the space race of our generation.

That is to say that I don't believe that you can have artificial intelligence without some form of non-von Neumann architecture. And I believe quantum computing will be a big part of that. S.

Everywhere around the world, Japan, France, anywhere you can name, then you kind of start to understand why this kind of computing is so important and why we need to be spending more time educating people earlier on this new technology. But again, I think this is the space race of our generation. This will shift global power.

It will change the way everything from stocks are treated to drugs are discovered. So it's incredibly important that we focus on this technology. And I appreciate you taking the time to listen to this talk for the next 20 or so minutes.

So, what is quantum computing again? But Whurley, you just rambled. You said all that stuff, you didn't tell us what it is.

Well, one of the easiest ways to explain it is with a coin. So, I have this coin here and I have a tails and I have a heads. And so, imagine that classical computing, your iPhone, your iPad, your computer you use everyday, even high-performance computers in very large data centers.

These things are all driven by a binary von Neumann architecture. Meaning, if I take that coin, I place it on a flat surface. If it's heads up, it's a one.

It's tails up, it's a zero. And on that flat surface, it can only be in one of those two states. So, think of that kind of like the binary state function of classical computing.

And now, imagine that with quantum computing, I take that same coin and I flip it in the air. And when I flip that coin in the air, the at the apex, is it heads or is it a tails? And the fact is that it's in a superposition as it spinning of both heads and tails.

And so, if you think about that, then you can start to understand how qubits work. And the science of qubits is pretty simple. You basically have a two-level mechanical system.

It allows for the superposition we just discard with that coin. And that means that we can measure probabilistic outcomes. And so, therefore, a qubit can be — think of it like a black sphere spinning and it can be in a straight up one or down zero or anywhere in that 360 degrees, as you're manipulating it.

And so basically, this allows for an incredible new area of computational capabilities. Problems it would take tens of thousands of years, millions of years. They could now potentially be solved in a day or a week or a month.

And this can revolutionize society. This can revolutionize technology as we know it. This is the first time in the history of computing we deviated from some pretty simple binary target processing functions.

Right. This is going to be the biggest change in the next 10 years, 20 years in computing. Since its existence, since it began.

So, understanding how qubit operations work is pretty simple. We have different applications that we can apply to these qubits, and we can spin these qubits in various ways. And because this is quantum computing for dummies, I'm not going to delve too far into the math.

But think of it as in these brackets. You see this possible states and outcomes with each of these, with X, Y, Z and then H. And you can see what this little yellow line kind of deviation and where we're spinning.

So, it's pretty complicated the way these machines are built. There's iron traps. There's adiabatic systems.

There's a classical circuit gate type systems and topology systems as well. You can build a quantum computer in any one of a dozen ways. But the basic idea is that you're taking some particle.

You're spinning it somehow. Maybe you're trapping it with lasers and you're cooling it down to absolute zero, and then you're manipulating that again. And then you get this black sphere by spinning it in different operations and therefore getting different probabilistic outcomes.

So why do we need these computers? That's a great question. A lot of people in the press either overemphasize the implications it has on security and other topics, or they underplay this dramatic shift that we're about to go through.

But when you get right down to it, the state of high-performance computing is changing. We're starting to run into some barriers. We have high-performance compute centers.

We have TPUs (tensor processing units), GPUs (graphics processing units). We have FPGAs and we have combinations of them. IBM's new supercomputer is a combination of classical processors and some graphic processing units.

Right. There's a bunch of different things you can do. But if you start looking at the problems, we're reaching the computational bottleneck.

That is to say that data is growing exponentially, but our processing power isn't keeping up. And then, on top of that, you have a series of problems that are very computationally complex, meaning that it's not that a classical computer couldn't solve them. It's just that it may take them 10, 20, 30 years, 300 years, three million years to solve.

So when you look at that, you look at the past, at the present, at the future. We kind of have this interjection of quantum as a technology years ago. This is something that was worked on in the labs.

IBM has, I believe, a 70 year history in the space. It is not new, but we're starting to get to the material science breakthroughs and the physics discoveries, et cetera, to make it real. So today we're just seeing a potential, a fraction of a fraction of what quantum computing can do.

You have very small machines. Fifty three qubits, 17 qubits. You know, we're hoping to get to a couple hundred qubits in the next 24 months.

But, for each qubit, it is exponentially more powerful than a classical bit. So by some people's calculations, one classical bit would equal about one qubit, rather would equal one billion classical bits. So if you think about that, these things are very, very powerful and it's super important that with everything we're doing, creating more data to creating more problems from healthcare to the environment, to arterials sciences, to going to Mars.

All of these things are going to require the kind of computational capabilities that quantum computing offers. So, the way I like to think about this is this: We used to cross from one side of the United States the other, and that journey took weeks and months sometimes, and someone died along the way. It was horrible.

And then the train came along and it crossed the country in about a week. All right. And trains are great.

Many places all over the world still use them. But no matter how far you go, eventually you run into an ocean. And so, we have airplane travel that augments train travel.

C. We didn't get rid of trains, but airplanes provide this completely new capability. You can fly from New York to Japan, for example.

And so think of this the same way. So classical computers haven't — they're not going to go away. Quantum computers aren't going to replace them, they're not with their function.

But eventually they reach a limitation. And that's where quantum computing comes into play. And we want to see these things work together.

So this is not a situation where you have a computer that, you know, comes in and it replaces all classical computers. That's going to happen. In fact, I believe these things would have been more like TP use, more like GP use.

So to be co processors, so to do what we call a full program execution, we'll write a program that goes through and does all of the work we want and then it'll stop, send off the computational complexity of quantum computer, get an answer back and continue along the path. That's probably the most likely scenario for the way these things are going to end up working, but continuing on. You know, let's talk about the power of quantum algorithm.

Let's put it into a perspective where maybe you'll understand. So we understand that we're using quantum mechanics and quantum physics to create these machines. We understand that they are non-binary system that we can use that blocks here rotated.

Any one, any position we want that could possibly disability probabilistic outcomes. But where does the rubber meets the road? How is that actually practical?

How does that provide value to us as users? Well, think about the classical drunk example. So that to say that we're going to walk into a bar, we're going to look for our friend and everybody's wearing the same team shirt.

There's 50 of a line across the bar. And we're going to walk up to somebody with a tap on my shoulder and see if it's our friend. And if it's not, we're going to flip that coin.

We have we flip that coin. If it's heads, we're going to go one step to the right. If it's tails.

Go one step to the left, and so when you look at that. Basically that means that we're going to have this very random, jagged way of going through that line to try to find the thing we're looking for. So what the first step?

We go 50%, have a 50% chance of going right. A 50% chance of going left. And with each concluded step, we go the same way.

So we're kind of like, randomly almost, going through this crowd, looking for this person that we're looking for, trying to find them with the flip of a coin. So remember, we talked about that classical system earlier, being its heads up, then it's a one, if it's tails up, it's a zero — kind of right and left. Well, turns out basically that ends up being like a drunken walk.

We can check after N steps the probability density and somewhere in the standard deviation, we might find our friend. So if our friend is out on the right or all without the left, we might not find it as easy with this method. And so we can use quantum superposition — again, the ability for that bit to be in kind of one, kind of zero state, until we measure it and it collapses.

" So, unlike our classical drunk, our quantum drunk can actually go right and left at the same time. And because of that, then that allows us to search this massive area a lot faster. So think about that again.

We're sitting at the bar. We taps on the shoulder. We flip a coin.

We go, right. We have someone to show. And we flip a coin.

We go left. We're back where we started to flip it again. We go left again.

That's going to end up in that kind of bill kind of search. With the quantum drunk, we're actually able to be exponentially better. So with the side I'm showing you now, take a look at the blue guy.

That's going to be our classical drunk. So we're flipping that coin. We're kind of randomly going around looking to where we're going for to get to this exit.

And eventually we get there. And you made a spill out here along the way. But with our quantum drunk, we're able to actually go through all of those possibilities in a much shorter time frame, meaning that it's exponentially better and faster to get results.

So there's a lot of examples of this: searching for information, using things like Grover's search algorithm. You know, it would be a good one. There's a lot of ways that this can improve your life.

You may not use a quantum computer, but you're going to be affected by them. Google will be using them to search and categorize information. So you'll be able to search it from your phone.

You won't know you're using quantum computer but it'll affect you. Drugs you may take to cure some condition you have might have been discovered by this technology, the way financial trading works, the way pricing discovery works, portfolio optimization. This technology is poised to affect everything in your life and everything that affects your life.

And again, in this example, it's super important to understand we have the bell curve (our classical random walk). But we have this vampire things (our quantum walk), meaning that we search mostly on the side and we search a wider area, faster. So we get to the result we're looking for.

Now, what that means, in short, is the power of a quantum algorithm is that we can explore more steps and exponentially faster in the same amount of time. Right. So the power behind Grover's search algorithm that I mentioned earlier, we're looking for an oracle.

That's where you really start seeing this increase in efficiency. And, as we all know, search is something that affects all of our lives right now. So where is quantum computing today?

So we talked about kind of what it is, we can do these supervision, these probabilistic outcomes. You know, we've given an example with this power of a quantum algorithm with the kind of drunken walk versus the quantum walk. But where's the reality?

You hear about 53 qubits, 70 qubits. Well, you know, the processor in my phone has three billion transistors at almost five nanometers. So, that sounds like it's way more powerful.

And that's not necessarily true. Again, these are completely different ways. The number of bits don't necessarily equal the usefulness or the power.

But the problem is that with the qubits, you can add qubits easy, but you get more noise and the noise becomes a problem. The fidelity of the qubits becomes an issue. But, at Strangeworks, we see a different problem.

And that's why we're so interested in giving this session at Techstrong Con; because we see the problem as not being the computers, but being the work force. If I had a million qubit machine today and I dropped it in the middle of Tokyo or San Francisco or London or New York, there would be no one who knew how to program it. There would be no one who can take advantage of it.

And anybody who tells you that they can is completely out of their mind. Right. There is not enough education on the computer science side and not enough education on the quantum computing side, on the quantum physics side.

We need multi-diverse teams based in all of these sciences and data science, and understanding what an enterprise needs working together. And so the problem with quantum computing today is you. Is the fact that you not involved.

The people you know aren't involved. That people aren't taking enough interest in the biggest technological shift in the history of computing. And so there's limited access to the computers.

But we're going to talk about how we're addressing that. But there's sparse talent, maybe 100 people worldwide that matter in quantum. Maybe another 100 or 200 outside of that.

And that's it. So there's 23 million developers worldwide and we're on a mission to get every single one of you involved in quantum computing. And that's a long-term play.

So the other reason you don't see some interest is it's not like you start learning quantum computing today and next year you can have a new high paying job. It's going to happen sometime between today and some point in the future. And we don't know when, but we do believe that preparing now is the wisest choice.

So, what do we think should happen? Well, we need to make it available. We need democratize access to quantum computing and the technologies.

We need to encourage collaboration and we need to focus on that diversity. You need physicists, and researchers, and developers, and mathematicians, and finance people, and maybe chemical people and all of these problems are going to take teams to solve. The algorithms were developed most likely the open source, most likely the open, most likely by geographically dispersed teams collaborating across a quantum platform.

So, when will quantum arrive then? Well, as I said, could be tomorrow or some amount of tomorrows between now and some point in the future. But we believe that you kind of take that old Viking phrase and paraphrase it, say 20 years is when the pessimist think it will be here.

Optimists think it'll be here in three years. Some people say a year even. We believe that you should be preparing today.

We think you should be learning how to do this, learning the tools, learning the technologies, understanding the parts about the physics you don't understand and really getting yourself in a position to take advantage of this. So, why? Can we put it in the financial terms?

Yes. Why is it so important? Creating massive opportunities.

Boston Consulting Group did an amazing analysis. It is the best one so far. I believe this is the next three to five years are kind of the noisy, intermediate quantum area.

And that's going have about two to five billion dollars of estimated impact on operating income for enterprises. So there'll be material simulations that reduce expense and time consuming trial and error lab testing. That's kind of where that'll be used.

In 10 years, they'll have a broad quantum advantage. We have 25 to 50 billion dollar impact on operating income of enterprise. And that's going to be near normal time risk assessment for financial services, quant hedge funds, things like that.

And then 20 years plus, there'll be fault tolerant, full scale quantum computing — general purpose quantum computer, if you will. And that's 450 to 850 billion dollars in impact to the operating income of enterprises. And that'll be custom drug designs and material sciences and all manner of things.

So this is going to make a tremendous impact technologically. It's going to make a tremendous impact financially. And I believe it's going to make a tremendous impact environmentally and in society in general.

So, how can you get started? OK, Whurley, you've talked to us about quantum computing. It's really hard.

You thought you're going to learn it easy when you started. Turns out you didn't do so well. It's all this complexity after more than one person in these teams.

Well, that's what we wanted to do as kind of our gift to the community in general. And we started by forming a partnership with Stack Overflow, free website, free resource. Many of you probably already use it.

" But we now have over 8,900 developers and physicists working together, answering quantum computing questions. It's about 1,300 people in there every day and about 85% of the questions are answered. In the last two years that we built this out, there's over 2,500 questions, everything from how you use IBM's Qiskit or Google Search, to whether a D-Wave annealer is better than an ion-trap machine, to the physics, to the math, the linear algebra.

There are questions that cover everything. It's an amazing resource to start. But we wanted to take it a little further and certainly do that.

We said, look, there's this big mess of questions. Where do I start? Should I start a program on ion-trap machine?

Is it different than a circuit game machine? What about topological? Should I use Python?

Do as Qiskit? Frameworks like forest, frameworks like ocean, leap. There's so many questions.

com. You can sign up for the beta today, we're going to be launching it. As long as you share your work with the community at large, as long as you partner in a good community member, it's free.

It's a very easy way to get started. And you can access IBM's Qiskit, Google's Cirq, D-Wave, slip in Ocean, blueqat from Japan, Regetti's forecast, on and on. And we continue to add new capabilities, and with a library of experiments from you to draw from.

I mean, things like Shor's algorithm to teleportation to grow research that we mentioned earlier. You can go in, take the code, read about it, for QIT, start your project, play around with it and run it in all these simulators. And, if you have access, run it all the way up to an actual claw machine.

com are going to be the tip of the spear of that we have even more planned for the future. So with that, I'll thank you for your time. And if you have questions, please feel free to contact me.

com. You can even text me, that is my cell number, it's everywhere on the internet, so it's pointless to try to keep it private at this point. com, sign up and if you have questions put in a sport ticket or go to Stack Overflow and ask a community question, get an answer there.

So thank you very much for having me at Techstrong Con today, I look forward to seeing all of you being involved with quantum computing in the near future. Thanks and have a great day.