Episode Transcript
[00:00:00] Speaker A: SAM welcome to AI Today. I'm your host, Dr. Alan Bideau. And this week's show, we're going to go out on the edge a little bit. We've talked about it a lot over the last year.
We've, we've made some implications that we were, we were going to have a show about this, and then finally we're, we're going to do it. We're going to dive into quantum AI. Now, I know it's a little, you know, it's, it's, it could get a little crazy.
It's a fascinating world. Quantum mechanics is really awesome. It was actually one of my favorite courses that I took as an undergrad. And it really is starting, just starting to have an impact. We hear about it every day, folks are talking about it, and then it's going to revolutionize computing and revolutionize artificial intelligence. And so this week, we're going to just dive right into it. And so the easiest way to think about this is to ask ourselves, what is light? That's the easiest way. Is it a wave or is it a particle?
[00:01:34] Speaker B: No.
[00:01:35] Speaker A: Depends, right?
In quantum mechanics, it can be both.
And that concept is called wave particle duality.
Now, I'm not going to get into too many technical terms, but I'm going to explain it to you as best I can so that we can, you know, have a decent understanding of some of the things that we're going to need to really get just a basic, you know, grasp on how, you know, quantum computing can work and what the implications are going to be to artificial intelligence. And so in classical physics, we know that the everyday world, if you think about it, a baseball is a baseball, you know, good representation of a basic particle. If you think about it, ocean waves or waves on a pond are waves, right? And, you know, those are pretty easy concepts to think about.
But in quantum mechanics, it gets a little bit more fuzzy.
A tiny particle can, you know, it's got, you know, electrons, it's got photons, and it can behave both as a wave and a particle. Now, the easiest way to think about this is, you know, there was a famous experiment, you know, Young's experiment, called the double slit experiment. And let's go ahead, and we will show you what, what that looks like. And what we have here is. Let me start. There we go.
Imagine that you're shooting just photons, particles, you know, through two different slits, and they're going through simultaneously. And you see how in one hand it can behave. You know, because we're measuring it, it looks like it's just two slits. The particles are going right through the two slits. It's fine.
You know, the pattern that you see are just those slits in the background.
[00:03:27] Speaker B: Right.
[00:03:28] Speaker A: Now, the reason it's doing that is because in, you know, quantum mechanics, Heisenberg's uncertainty principle says that, you know, once you measure it, you are changing its state. You're forcing it to pick a side.
And in this case, you know, it can pick, you know, really, you know, up or down. And you know, it, you start to see what some of those behaviors look like. And you know, on the left hand side is when we're measuring it. But the thing is on the right hand side is when we're not measuring it.
So these activities are taking place and you see that there's a distribution.
And the easiest way to think about that at the same time is, you know, water flowing through two slits. You see the waves, you know, you have your different, you know, troughs that are in there, different, you know, activities that are taking place in there. You've got your constructive capabilities and those are really the lighter, you know, areas. And then you've got your destructive capabilities which are in the darker area. And, you know, you start to think, oh, boy, you know, so they can behave as both particles and waves, you know, simultaneously.
And that is very counterintuitive because, you know, as you're, as you're going through these and you see these, these patterns that are taking place and you see that, you know, the bright and the dark trick fringes that are, that are there.
You know, you see those, those different subtractions. Then you start to expand that. Then you start to look at other principles that become important, such as superposition, meaning I've got a coin.
Normally it can be heads or tails.
[00:05:16] Speaker B: In the real world.
[00:05:17] Speaker A: In the quantum world, it doesn't behave like that.
You know, a classical coin, it can be, you know, like I said, heads or tails.
Quantum, it can be heads, can be tails or it can be heads and tails.
And that's not an easy concept to think about because, you know, as it's, as it's moving through and you're, you're forcing it to pick something, then you know, what that means is, is that it's holding its position, which can be multiple possibilities all at once.
Now, when you measure it, it forces all those probabilities to collapse down into one.
And when that happens, that's why you can see a heads or a tails. This is a perfect example. You've got multiple spins that are taking place.
You've Got multiple.
Really an infinite amount of possibilities.
If you think about it, to go from 0 to 1.
You never can go to 0 to 1, right? Because there's an infinite number of possibilities in between 0 and 1.
Now, that doesn't do us really any good. And, you know, when you're measuring a distance or you're trying to measure something on your house or whatever that is. But in quantum mechanics, it becomes exceptionally important because the more states that you have, then the more ability that you can have, you know, multiple opportunities.
[00:06:45] Speaker B: And.
[00:06:46] Speaker A: And it also means that, you know, your. Your ability to compute becomes significantly more important because of this.
It's called parallelism. Not an easy word to say, but it allows you to speed up a whole bunch of different activities because of the inference that takes place. It steers us toward the right answer.
Now, we're going to get even a little bit crazier here.
There's something called entanglement.
Now, it's exactly what it sounds like.
But imagine you've got two coins now, and we'll say that they're coins that are electrons, right? And they become entangled.
Now, what that means is it can be up, it can be down. It can have a whole bunch of different charges, right, in any certain direction.
But once those two become entangled, then no matter how far apart they are, if you take one coin and put it, you know, one light year, you know, we'll say out of the solar system into, you know, with the other one. If they are entangled. That means if one is heads and you're measuring it, then the other one across the solar system is also heads.
And it happens instantaneously. It just happens.
Now, that's not a very easy concept because you think, oh, it violates so many different things.
You know, general relativity, you know, it. You know, Einstein was not a huge fan of that concept because it is very much a challenge for folks. But really what it is is it's a basic principle.
It's just fundamental. And, you know, quantum mechanics has been pretty much one of the most successful and tested fields of physics for a very long time.
And, you know, just remembering some of these basic concepts that we have that allow us to understand the different fates of particles, the different characteristics of, of quantum computers, that activity is really what is going to drive the power that we're going to talk about in the next few minutes. And, you know, if you think about it, it's really just, just so important from an understanding perspective that, you know, there's no quantum computers. If there's no Qubits, right? A qubit is the measurement of pretty much the electron pairs that you have, just like a bit 0 and 1. Well, a qubit is up or down.
And it's that power that allows you to really scale, allows you to calculate things so much faster than you can theoretically in a classical computer. Now, I don't want folks to get too concerned. We don't have really a general quantum computer yet. We're going to talk about some of the theory behind it. We're going to talk about some examples, some things that we can do today.
And we're really going to deep dive into, you know, what that relationship is going to be between quantum computing and AI because some of us, including myself, are using certain types of quantum in some of the software and the tools that we're developing today. So I want you to stick around with us. We'll be right back after a few short messages from our sponsors.
Welcome back to AI Today. I'm your host, Dr. Alan Bideau. And you know, this week we're talking about quantum computing.
And I've got to give folks a little bit of an understanding of, of some of the basic principles. We talked about those in the first segment.
And, you know, it's that weirdness, though, that allows us to translate quantum computing into such, you know, a powerhouse of, you know, a system. And, you know, I've seen debates even this week, you know, should you call it a computer and should you do this and, you know, et cetera, et cetera. The fact of the matter is, is that, you know, it's a very strong, you know, and an exceptional capability that allows us to do an awful lot of things and the goal to solve a lot of problems that classical computers at least today just can't do. We've really maxed out GPUs, Moore's Law, really doesn't exist, you know, anymore. And, you know, that entire forcing function has really been the driver behind, you know, the need and the developments of, you know, of quantum. And we've heard about this for years. Many, many, many, many, many years. And it goes up, it goes down. We have progress, we don't have progress. Well, you know, over the last 10 years, it's been pretty exceptional. What some of the folks have, have done even as recently as, you know, a month ago, some, some releases and some new hardware have shown up. And so that's the power that we're hoping for. And, you know, we're gonna, we're gonna start. Let's take a look at the, Take a look at the. The movie here, real quick. But, you know, if you think about the differences Between a classical computer and a quantum computer. Like I said, it's around bits and qubits. You know, if you think about how these computers work. You know, a bit is like a light switch. You flick it on, you flick it off. But you can only have a zero or you can have a one. So that means you can only have one state at a time.
And that makes it very difficult. Whereas with a quantum computer, you can have multiple states. And I showed you that, right?
We talked about superposition and entanglement and, you know, interference.
Seeing how those three concepts all work together. To give you some sort of capability that you didn't have before. But, you know, as bits, you know, they're independent of each other. Light switch. Right. Well, qubits don't work that way, Right?
[00:13:38] Speaker B: They don't.
[00:13:40] Speaker A: Because of superposition. And because of your ability to look at the entire distribution of, you know, 0 to 1 or, you know, all in between. It gives you a huge, huge, huge advantage.
And that as you start to develop your systems. As you start to look at problems that you have.
That's why folks like myself are so hopeful that we get some more advanced systems in the near future. Because all those principles that come into play. They're all working across the same thing. They're all working across, you know, the wave function. Trying to figure out how we're going.
[00:14:25] Speaker B: To compute these things.
[00:14:26] Speaker A: How we're going to combine these qubits. And, you know, take advantage of them mathematically. Right? With constructive, you're adding things together. Destructive, you're subtracting. Right? And so, you know, when you have an infinite amount of solutions, that's a problem.
And being able to construct them the right way with superposition. And the probability that these state, you know, these qubits are in a state or in a different state. That is what really allows us to take advantage. Of the number of calculations that these things can really do. It's really a phenomenal thing. Because, you know, as they are working together. And you're extrapolating those probabilities across the entire system.
Then, you know, the speed that you.
[00:15:25] Speaker B: Can compute certain things is phenomenal.
[00:15:29] Speaker A: I mean, we're talking, you know, minutes in some cases.
[00:15:33] Speaker B: Where supercomputers today. That can take years and years and years to do the exact same problem.
[00:15:40] Speaker A: We'll have an example in the. In the next segment. But it's really about trying to, you know, drive the behavior.
It's trying to look at, okay? You know, I know that a bit can only be a 0 or a 1.
Now, that means that if you've got two bits, then the number of combinations that you can have is 0, 0, 0, 1, 10 and 1 1.
[00:16:09] Speaker B: There's four.
[00:16:11] Speaker A: Now, when you have to qubits, then we start thinking, oh, man, I have two of them. And. But there's an infinite distribution in between those.
So how am I going to, you know, attack that problem? How am I going to code that? What does that look like? What's that distribution look like?
And, you know, that means that, you know, you can represent though, if you think about it, what we talked about with superposition, you can represent all four states simultaneously at the exact same time. You can have all four states that you can have in a binary system.
Now, as you move these, you know, in impact or measure these, these qubits, that's when the distribution can change.
That's scary, right? You're taking advantage of the entanglement. You're taking advantage of the superposition that is there that allows you to really excel at massive combinations of data and problems, especially optimization.
You can see in the video, it's going back and forth and it's impacting what that distribution looks like. And that's exactly what you want because you're finding the optimal delivery route, meaning the optimal solution for whatever that, you know, situation that you're trying to solve is.
So then if you start to think about, okay, I want to, I want to solve, you know, some, some complex, you know, some maybe some logistics problems, supply chain, you know, those, those sort.
[00:17:57] Speaker B: Of activities, then those really difficult problems.
[00:18:01] Speaker A: Molecular biology and creating new molecules and.
[00:18:05] Speaker B: Drugs starts to really, you really want.
[00:18:09] Speaker A: To take advantage of these quantum systems. You know, as an example, you know, in 2019, you know, Google had a 53 qubit, I think. Yeah, it's 53 qubit system.
And I remember thinking, oh, that's pretty spectacular.
Well, what does that really mean? It sounded really cool.
But what it meant was that there were certain calculations that they were doing. It would take their system 200 seconds to do those. Whereas with classical computers it was 10,000 years they were estimating before it could complete.
That's unbelievable because that allows us to attack problems that, you know, we thought we would never be able to really.
[00:18:57] Speaker B: Solve it in our lifetime. And if you're a small business and.
[00:19:00] Speaker A: You'Re thinking about these things and you're thinking, oh, man, you know, this is.
[00:19:04] Speaker B: Going to be a fun show to watch in Star Trek, but it's not really Relevant.
[00:19:07] Speaker A: Well, let me, let me put those.
[00:19:10] Speaker B: Thoughts to bed because there are small.
[00:19:14] Speaker A: Businesses that are using some of these systems today.
And whether it's a quantum simulator, can be a quantum annealing machine, it can be, you know, any of those things in between, you know, from Microsoft and their Q Sharp programming to things that.
[00:19:31] Speaker B: Are in Python that you can use.
[00:19:33] Speaker A: And you can leverage, you know, D.
[00:19:35] Speaker B: Waves, quantum annealing machine, you know, anything.
[00:19:39] Speaker A: That you can use to optimize a lot of data is going to be huge.
These systems, they're coming, you know, they're getting more advanced. You know, D Wave has been around for a long time.
I believe it was 1999, I think so. I, I know I helped with the arl system in 2000. Oh my goodness, I'm getting old.
[00:20:02] Speaker B: 2004.
[00:20:04] Speaker A: And you know, it's, it's using those.
[00:20:07] Speaker B: Types of things for some, some very difficult things.
[00:20:10] Speaker A: The DoD, very interested. Intel agencies, FedEx, for example, very interested.
[00:20:17] Speaker B: Why? Because you can optimize delivery rates.
[00:20:20] Speaker A: You think, oh, it's not that hard. Well, I want you to stick around, come back with us.
I'm going to show you what a problem that you don't think is very hard is that is exceptionally hard and impossible for our classical systems to solve them in our lifetimes. So stick with us. We'll be right back after a few short messages and we're going to see just how easy some of these things are that are exceptionally hard to solve. So stay with us. We'll be right back.
Welcome back to AI Today. I'm your host, Dr. Alan Badot, and we have dove in to the quantum realm. This is, this is what weird physics is all about.
And it's taking advantage of those, that, that, you know, this weird physics that allows us to really accelerate what we're going to be able to do in AI. Now, we've talked about it a little bit. You know, some of the, you know, I gave some hints, I gave some nuggets, you know, talked about, you know, the, the different states where a classical computer is really only 0 and 1, and in a quantum computer it can be 0, 1, 01 at the same.
[00:22:02] Speaker B: Time and a distribution across all of those states.
[00:22:05] Speaker A: And if you think about what a state is, it's really, you know, a consistent energy.
[00:22:11] Speaker B: Right?
[00:22:11] Speaker A: You know, I'm gonna throw some terms out there and you're gonna hear some things adiabatic, you're gonna hear some other, you know, different types of machines because.
[00:22:19] Speaker B: They'Re aligned to the methodologies that they use. But think about you know, estate just.
[00:22:24] Speaker A: Being something that, you know, where it.
[00:22:27] Speaker B: Is, it's really constant in that moment. That's the. I'm, I'm taking some liberties and I know I'm going to get some emails around that. But just think about it from that perspective. You've measured it. Boom, it's frozen. That's the state that it's in. Okay, and that's the same thing with.
[00:22:44] Speaker A: Heads or tails, right? If it's tails, boom, It's a state.
[00:22:48] Speaker B: It's measured. Okay.
[00:22:50] Speaker A: So as we're talking about some of these systems that you see over here.
[00:22:55] Speaker B: This is where AI really can gain an advantage.
Now, you've heard me say over the last year that, oh, we don't have.
[00:23:06] Speaker A: To worry about that until we get a quantum system.
[00:23:09] Speaker B: Oh, we don't have to worry about that until we get a quantum system that you can run on your mobile device.
[00:23:16] Speaker A: Oh, you know, we don't have to.
[00:23:18] Speaker B: Worry about, you know, Terminator or anything like that for 20, 30 years.
[00:23:24] Speaker A: Most of that is fairly consistent with.
[00:23:28] Speaker B: What the, you know, the general scientific community believes.
[00:23:32] Speaker A: Now we are using AI to accelerate.
[00:23:36] Speaker B: Some of the development that we're doing on the quantum systems. And in return, we can use the.
[00:23:41] Speaker A: Quantum systems to expedite the utilization and the things that we need to do AI better, to do them faster, to do them almost in real time.
And, you know, as we. I'm going to start my video. I forgot to. But as we are going through these.
[00:23:58] Speaker B: These different types of activities, that's where we're really going to see this, this disadvantage. And I have said over and over and over again, whenever I talk about quantum or even hint about it, this is a race we can't lose because a winner of the quantum race is.
[00:24:19] Speaker A: More than likely going to win the AI race.
And if you win both of those.
[00:24:24] Speaker B: That means your phone systems can be jeopardized, your encryption activities can be jeopardized. Your ability to communicate securely is going to be jeopardized. There's a lot of implications around that, you know, and we are behind.
[00:24:44] Speaker A: We're doing a pretty good job, but we're still behind for a lot of different reasons. And I don't want to talk about.
[00:24:49] Speaker B: Those right now, but maybe that's another show later on. But, you know, the fact of the matter is, is that with quantum machine learning and the algorithms that we're building, then the amount of data that we can use and the time that it takes to train these models will be shortened by exponential, you know, values, you Know, you see, like today it takes millions of dollars, hundreds of millions of.
[00:25:17] Speaker A: Dollars, and some, in some cases to.
[00:25:19] Speaker B: Train these models, especially, you know, these.
[00:25:22] Speaker A: These really large, large language models.
[00:25:26] Speaker B: And that makes it very difficult, but with a quantum system because we can pass so much information through, then, you know, the amount of time that it takes you to get an answer will be seconds in some cases.
[00:25:43] Speaker A: And so then you start to compete.
[00:25:46] Speaker B: With what the human brain can do and the number of transactions. And think about it, the number of times the neurons are firing in your brain is huge. I don't know the number off the top of my head, but it's probably in the, you know, billions of times per second, right? With a quantum system, you can start to simulate the human brain and do those exact same things.
Now how do, how do these quantum systems work? We talked about zeros and ones. It's gotta be more than zeros and ones in distributions, right? Well, it is, you know, but it's, it's looking at solving these matrices for math.
Now if we apply it, let's just apply it really quick to a maze problem. Oh, that's simple, that's not hard.
Not a big deal. A classical computer looks at the maze and you're gonna think, I'm gonna brute force it, I'm gonna send it down, we'll drop a marble down, we'll simulate going through it. When it runs into, you know, a wall or a dead end, we just start over and make sure that we don't go in the same, the same way. That's a brute force method.
Scientists love the brute force method. That's why these high performance computers are really, you know, really being taxed, especially weather models, weather simulations, those kind of things.
Now imagine that, you know, you're doing this over and over and over again. And this is just for a maze.
The number of times, you know, if you get lucky, that's great. But the number of times that you have to do this for a very simple case, you know, can be, you know, 24, 25, you know, we're just guessing out there, you know, even more than that. Potentially not efficient.
[00:27:41] Speaker A: Now with a quantum system, well, let's.
[00:27:44] Speaker B: Throw efficiency really out the window. And we're going to just say that we're going to represent all those states.
[00:27:48] Speaker A: At the same time and we're launching.
[00:27:50] Speaker B: Them and we're going to get there and we're going to, we're going to go multiple paths at the same time and find those dead ends.
[00:27:58] Speaker A: That means that we can get to.
[00:27:59] Speaker B: The exit Heck of a lot faster.
[00:28:02] Speaker A: Than what you can do with a classical system.
And as you increase that complexity, you start to say, whoa, that's huge.
Now we just say, okay, you know, let's, that's, that's amazing.
[00:28:19] Speaker B: That's not that big of a deal. Well, let's look at a real world problem. So imagine I'm a traveling salesman, very famous physics problem, and I've got 16 cities that I've got to go to now with.
[00:28:34] Speaker A: My goal is to say, okay, you.
[00:28:36] Speaker B: Know what, I want to go to.
[00:28:37] Speaker A: Each city exactly one time before I go home.
But, you know, I've got to go.
[00:28:44] Speaker B: To the shortest route possible.
[00:28:48] Speaker A: You start thinking, oh, man, okay, now.
[00:28:50] Speaker B: It'S, now it's gotten a little bit more real.
[00:28:53] Speaker A: Applies to logistics problems.
[00:28:55] Speaker B: It applies to, you know, a whole bunch of other optimization problems that are out there where you're trying to find the best one in a sea of similarity.
[00:29:05] Speaker A: And, you know, again, classical computing, brute force, it's going to go through, it's.
[00:29:11] Speaker B: Going to do all these calculations, and.
[00:29:13] Speaker A: With 16 cities, it'll do about, you know, I think I calculated earlier a.
[00:29:18] Speaker B: Little over 87 billion different possible routes.
Oh boy, that stings.
[00:29:29] Speaker A: So, 87 billion different possible routes.
That means that the bigger it got, the larger the problem.
[00:29:38] Speaker B: You can quickly see how classical high.
[00:29:42] Speaker A: Performance computing and supercomputers will not be able to do this in a reasonable amount of time.
And you can guess there's a lot.
[00:29:53] Speaker B: Of heuristic algorithms out there and stuff. And you can guess, but it doesn't.
[00:29:58] Speaker A: Necessarily give you the exact answer.
[00:30:01] Speaker B: Now, with a quantum computer, though, there's so many different ways that you can do this, like quantum annealing and D wave. You can use, you know, that kind of capability to solve this problem in a reasonable amount of time. And we're talking, you know, maybe minutes.
[00:30:15] Speaker A: That you can do this 87 billion calculations in, you know, comparatively speaking, in minutes.
That's, that's crazy.
See, and that's a simulation that we.
[00:30:29] Speaker B: Just saw that it's able to get that.
And I've heard some people say, oh, yeah, but just draw lines like in a coloring book and you'll be able.
[00:30:38] Speaker A: To see the exact thing that you want.
[00:30:40] Speaker B: No, it doesn't quite work that way. Right.
What we're showing is that using this approach, it really brings all of those uniqueness, you know, things that we talked about earlier to bear.
It's, you know, solving a technical challenge that traditionally we wouldn't even be able to, you know, to start in some cases.
[00:31:10] Speaker A: You know, thinking about drugs, thinking about.
[00:31:13] Speaker B: You know, solving diseases, thinking about inventing.
[00:31:17] Speaker A: New types of materials and the complexities.
[00:31:21] Speaker B: Around that, the human genome, all those things very, very, very good problems for quantum systems.
Now, I don't think, and this is.
[00:31:35] Speaker A: My perspective, I don't think we'll be.
[00:31:37] Speaker B: Running Microsoft Word on a, on a quantum machine unless it's maybe, I don't know, maybe Microsoft will come up with something that's three dimensional and you can reconstruct a whole bunch of different things. I don't, I don't know. But those applications are going to be different.
And the impact though, that we'll have is going to be huge in my lifetime. We're not going to get rid of classical computers and switch them up for quantum computers and, you know, vice versa. But we continue to be able to.
[00:32:10] Speaker A: Solve hard problems and look at new.
[00:32:13] Speaker B: Areas of research that, you know, these types of systems are perfect for. Now, I don't want to, I don't want to sugarcoat anything. So I want you to stick around though, and we're going to talk about, you know, some of the limitations, some of the challenges that it has and.
[00:32:31] Speaker A: We'Re going to talk about some realistic.
[00:32:32] Speaker B: Timeline that I'm going to try to get you prepared for. So stay with us. We'll be right back.
[00:32:58] Speaker A: FOREIGN.
[00:33:06] Speaker B: We are back.
[00:33:08] Speaker A: Welcome back to AI Today. I'm your host, Dr. Alan Bedot.
[00:33:12] Speaker B: And we have been hitting quantum computing.
[00:33:15] Speaker A: And really quantum mechanics really hard. And I'm proud of you if you.
[00:33:20] Speaker B: Stuck around because this is not an easy subject, but we're hearing about it more and more and more all over the place.
[00:33:27] Speaker A: Some good, some bad, some a little.
[00:33:32] Speaker B: Unrealistic in their projections and expectations. But that's my job, is to really set it straight so you guys can understand what some of the implications are, the real story behind it. And you know, just to, just to give you all a sense of, okay, this is another technology that's coming.
In some cases it's here already, but it's coming and it could impact how you're planning, what your strategic growth looks like, all those things.
Now we know with what we can do from an AI perspective today, it's pretty good.
But we have maxed out what GPUs can really do. Now. Nvidia and folks, they still are able to, there's, they keep squeezing more and more out of these GPUs, but just like large language models, we are going to see a time when you really can't do much more.
They keep coming out with some unique architectures. And there's a lot of different things out there that could really impact what their performance is going to be as we continue to move forward.
But at the end of the day, you know, we're coming to a point where we, we don't have any choice. We have to have some more breakthroughs and we have to do a lot of things to, to really accelerate that.
From a real world example, you know, if you look at, you know, the auto industry, you know, folks like Volkswagen is a perfect example. They are looking to really optimize traffic flow for some of their buses that they're providing to certain cities.
That's one of those examples where there are so many different variables that go into calculating the traffic flow of some of these cities that you can't do it on a classical machine. You just can't.
[00:35:37] Speaker A: You could try.
I'm pulling for you.
[00:35:40] Speaker B: That's all I'll say.
[00:35:41] Speaker A: And you know, other areas though, like.
[00:35:45] Speaker B: Financial institutions, some of those financial institutions are really leading the way in some of these quantum algorithms that are coming out. And there's a whole bunch of reasons for that, but it's really looking at risk.
How can they minimize the risk?
How can they make sure that whatever portfolio that they're managing is really optimized and, you know, it's, it's going to give them the least amount of risk.
Those types of problems, again, huge manufacturing, same thing, you know, quality control systems. Big, big, big, big. And that's looking at trying to defect, you know, detect defects in some of these, these sort of systems. Things that I'm, you know, using, therefore are along those same lines. Damage assessment. Being able to get better, you know, understanding of damage, looking at depth, being able to get the depth of some of the damage more accurate is fantastic. And it really can be a game changer in certain industries.
These optimization capabilities really can flow across every single vertical market, especially when it's coupled with AI. And healthcare is a great, a great example of that. Because if you think about, again, in drug discovery, think about being able to simulate how molecules interact and, you know, how they, you know, how you build them. In some cases, you know, you're starting to think about personalized medicine.
Oh, you know, here's this person's, you know, genome, and we're gonna, we're gonna tailor something specific to them, a medicine or something that will enable them to get better. And for somebody else, we're going to tweak it just a little bit.
I mean, it feels like science fiction.
[00:37:43] Speaker A: But you know, Quite honestly, we're getting.
[00:37:45] Speaker B: Pretty close to being science fact.
And that's what's exciting because, you know, our, you know, push our desire to, you know, do things better, faster and cheaper is really what we're trying to, to do. And quantum computers allow us that, that capability. Now, let's, let's set the record straight, though, a little bit, right?
There's no such thing as a general quantum computer right now. There's no quantum computer that fits in your. On your mobile.
[00:38:19] Speaker A: Right?
[00:38:20] Speaker B: So we don't have to worry about that right now.
But what I want you to do is think about these, these quantum systems, where they are today.
Think about them as kind of like your Model T or your Model A from a car perspective, very early in their infancy or, you know, they're, you.
[00:38:40] Speaker A: Know, it's, it's thinking about, okay, you.
[00:38:44] Speaker B: Know, I know it doesn't have air conditioning. I know it can't do this. It can't do that. But darn it, it can drive me to downtown. It can, you know, do a lot of things that I had to do by myself.
And then you start to have AI in it. Then you're starting to really be able to achieve some things that we thought were really in the future, but maybe not.
[00:39:12] Speaker A: And, you know, you'll hear a lot.
[00:39:14] Speaker B: Of terms, you know, and we've talked about it. We hit on it a little bit, right? Oh, it's 50, 52. You know, it's 53 qubits, or it's 100 qubits or it's 140 cubits or whatever that is. Oh, you know, this, this machine finally reached supremacy where it can go faster than a classical computer. You know, you'll hear Shor's algorithm and Grover's and whoever else's methodology that's out there.
Just take a step back.
[00:39:44] Speaker A: Don't be afraid of the qubit.
[00:39:46] Speaker B: Now you understand what it is, then you can start to say, okay, the more qubits that you have, the faster.
[00:39:55] Speaker A: You can do things.
[00:39:58] Speaker B: What's preventing us from having more qubits? Well, it can be a whole bunch of different things. It can be the noise that's in these systems because then, you know, you see those giant devices that are on TV and they're. Oh, they're pretty cool. The lattice work is crazy. And you've got all these tubes and pipes coming out from everywhere.
That part is not the quantum system. That's the thing that keeps it cold.
That's pretty much their refrigerator.
The actual quantum piece is probably about the Size of or, you know, less than the size of your phone.
[00:40:36] Speaker A: Now, if you're, you're D Wave and.
[00:40:37] Speaker B: Their methodology is, is, you know, quantum annealing is different. They've got some big machines, but still same sort of principle, though. It's not.
[00:40:46] Speaker A: The whole system is not the quantum system.
[00:40:49] Speaker B: Most of that is the cooling piece.
And so we've got limitations from noise, we've got limitations from error, we've got hardware limitations, we've got methodology limitations.
And so, you know, it's not something where a general computer is going to be ready to go tomorrow or even ten years from now.
But if you use them the right way, then you can take advantage of some, you know, problems that are very applicable to small businesses especially, you know, again, logistics, financial, healthcare, that mathematical capability.
[00:41:36] Speaker A: Where, you know, if you think about.
[00:41:38] Speaker B: If I can solve the traveling salesman problem, and that traveling salesman problem applies to, you know, how I ship things out or the routes that I deliver things, or, you know, how I am choosing different materials to work with, or, you know, how I do, you know, X, Y and Z from an image perspective, then you can use these systems to again, gain a little bit of an advantage.
Couple that with your AI models now.
Now you're starting to, starting to get pretty aggressive with things and people are going to look at you and they're going to say, oh, you know, what are you using for X, Y and Z? And you're going to say, well, I'm using some convolutional neural nets with, you know, some quantum annealing and allows me to really get accuracy on, you know, damage, you know, and, you know, predict what. Some of those other, you know, activities are going to be around damage prediction and those kind of things significantly faster than anybody else.
[00:42:50] Speaker A: We're gonna think you, you've, you've lost.
[00:42:52] Speaker B: Your mind because they hear the word quantum and they think it's not real. It is real. It's very real.
[00:43:01] Speaker A: And I would encourage everybody to go.
[00:43:03] Speaker B: To D wave, go to IBM, go to, you know, IonQ, go to Microsoft, you know, whoever, look up that information that they have.
You can use some of these, you know, modeling systems today. In some cases, like Deep Wave, you can pay for it through Amazon and get some cycles around that. So if you have those kind of.
[00:43:24] Speaker A: Problems and it takes you a long time to calculate them, there's other ways.
[00:43:29] Speaker B: To do it and you don't have to be afraid.
Again, though, people might look at you funny and think you're crazy. It's okay. Come over to the fun side It's a lot more fun over on this side because we're pushing the bounds of what we can do with the tools that we have. And all you can do is show them the results.
That's it.
And if you can do that, there's nothing wrong with that at all.
Now timelines, because that was really the biggest question that I got this week. I got a whole bunch of questions.
[00:44:00] Speaker A: Around timelines and I think it's because there were a lot of articles and.
[00:44:04] Speaker B: A lot of TV coverage last two weeks on Quantum. Oh, we've got this or we've got that or Microsoft has, you know, can do quantum computing now at room temperature, stuff like that.
Folks are kind of excited about it. Usually I only get about, you know, 10 to 15. This week I got about almost 100 questions around quantum, so and they were.
[00:44:25] Speaker A: All pretty much around timelines. So I'm going to talk a few.
[00:44:30] Speaker B: Like I said, D wave, quantum annealing, something that can be done today.
A little expensive, but still can be done. I love it.
[00:44:37] Speaker A: And you know, you just got to.
[00:44:39] Speaker B: Make sure you got the problem right. The, and you know, the specifications and you'll be able to really accelerate some of the things that you're doing looking at other machines like the videos that we, or that image that I showed earlier.
You know, it's really going to depend on the breakthroughs that we have in material science and even some of the breakthroughs from an AI perspective that are coming.
I think in by, you know, 2028, we will have much better error correction methodologies, a lot better ways that we can control that. And if we can control the system like that more, then the number of qubits that we'll be able to have I think will probably approach in the upper hundreds.
And remember again, the, you know, the problem that we showed earlier when you got two qubits, it scales exponentially, right?
Start having hundreds and hundreds of qubits.
Now you're able to really attack some problems and use some AI that will, will really change how you are fundamentally operating and calculating things.
In the next probably after that, I'd say around 20, you know, 2030 to 2035 ish. We're going to start having some different, you know, more robust quantum machines. And I think it's really going to be across all of those, those systems that I showed you, not all of them are proven, some of them are still pretty hypothetical.
But you know, like the topological, you know, system Microsoft, you know, just in, geez, I think it was January, came up with that Article saying that they could do a whole bunch of things now with that, that chip that they had developed.
That's phenomenal because, you know, topological, you know, topological machines have been theoretical since, well, since I even started hearing about them. And that was years ago.
And it's those kind of breakthroughs that we're starting to have. But you know, even beyond that, when you start looking at what it can apply to, you're talking about fault tolerance systems, you know, where you've got these, these quantum computers are becoming more mainstream, right? And you know, that's going to transform a lot of different industries and it's going to do it even more so than what the Internet was, was able to do.
Cyber security is a huge.
[00:47:23] Speaker A: Opportunity.
[00:47:25] Speaker B: The reason is exactly because of what we were talking about. Entanglement, superposition, inference.
Remember, if you measure it, then it's in a state, it sets that state.
So imagine a hacker comes in and is trying to get into your machine and you're in a quantum computing machine and they're doing something and boom, that, that something that they're doing is messing with your qubits that you have, setting them in state. Well, guess what? That means you've got somebody in your machine.
Just those kind of things become, become really, really important. And how we, how we start to plan out, how we start to look at problems, how we, how we are, you know, driving to a solution for some of these complex problems that we didn't even think we'd have an opportunity to solve. And so watch how this market evolves.
Don't spend a ton of time, I'll.
[00:48:27] Speaker A: Do that for you.
[00:48:27] Speaker B: But don't spend a ton of time trying to redo your, your strategic roadmap or those kind of things.
If you have a, a certain problem and you're curious about it, send me an email, I'll help you. We'll look at it together, try to figure out maybe quantum will work, maybe it won't work. Maybe there's some things in classical computing that you should try first.
[00:48:49] Speaker A: And you know, if you're still interested.
[00:48:52] Speaker B: Or something that could apply. We don't look at something, I don't mind doing that, that's fine. I love doing that kind of stuff.
But don't forget about it, don't ignore it, because you're gonna see a lot of articles.
You know, quantum communication, quantum satellites, you know, all these things are going to be, start to take over some of the news and applying it to AI is going to be another game changer.
So then you start to look at AI and you combine that with quantum and infinite compute capability with some of these algorithms, then you start to say, okay, we may not be too far 10 years from actually being able to get to a system that can think for itself, that can, you know, become self aware, that can recognize its own emotions, those kind of things. Because now we're computing at the brain speed, which is really fantastic.
[00:50:02] Speaker A: But it does bring up a whole.
[00:50:05] Speaker B: Bunch of questions that we don't have time to talk about today. That's going to be another show. But I hope I was able to convey to you some of the basics that you don't have to be afraid of the qubit that you know, some of the spookiness and the weirdness that you hear.
They are not Star Trek, they are real and they're proven quantum systems. Huge impact already. We wouldn't have gps.
Solar panels would be a lot more difficult to use and develop your phone. You know, there's so many different technologies that it has impacted that you can't ignore it now. It's not easy again, but you can't ignore.
So I hope, you know, everybody enjoyed the show. I love talking about these things. Keep bringing the questions to me. If you need any clarifications, I'm happy to do that. So I appreciate you all as, as usual.
So it's fun topic for me to talk about and we'll, we'll talk about the bad side in the show in the future because it will have some, some empathy, patience that we all again have to keep thinking about. But thank you and we'll have another great show for you next week. I appreciate you and we'll see you all later.
[00:51:32] Speaker A: This has been a NOW Media Network's feature presentation. All rights reserved.
