Quantum 101 • Murray Thom (D-Wave)

VP of Quantum Technology Evangelism at D-Wave, Murray Thom, joins Niki remotely to explore the basics of quantum computing and share some real-world, right-now applications of this leap in compute power. They talk AI, the impact that efficiency improvements have on daily life, and the exciting potential of quantum to facilitate breakthroughs in health care and drug discovery. 

“There's a really deep technical connection between quantum computing and artificial intelligence.” -Murray Thom

• Follow Murray on Twitter
• Learn more about D-Wave
• Build things with D-Wave Leap 

• Learn More at www.techedup.com
• Check out video on YouTube
• Follow Niki on LinkedIn

Transcript

Copy

Niki: I'm Niki Christoff and welcome to Tech’ed Up. Today's guest joining me remotely is Murray Thom. He's the Vice President of Quantum Technology Evangelism at D-Wave. That's a Canadian company that's deploying quantum tech to help businesses right now. 

In this episode, we're going to break down what exactly quantum computing is and the basics, the real basics, of how it works. More importantly, Murray and I are going to use this time to make the argument for why those of us who don't do linear algebra on a daily basis should care about this important technology. Thank you so much for joining me today. 

Murray: Thanks, Niki. It's great to be here with you. 

Niki: We will get started talking about what you do for a living because it kind of dovetails with what I do for a living. So, why don't you start with an intro of just who you are and what you're doing? 

Murray: Vice President of Quantum Technology Evangelism means I'm helping organizations and global businesses understand quantum computing and the transformational impact it can have for them today.

Y’know, I just finished reading this burn book from Kara Swisher, where she was sort of saying, “What's happening in technology is not about the technology itself. It's about how it impacts our daily lives.” And I'm really trying to help draw that out of, of these deep technical subjects in quantum computing so that we can all appreciate how the technology can help us. It can bring us value. And also, it's tremendously cool. 

Niki: Yes, so speaking of tremendously cool, on your LinkedIn, it says that at the University of British Columbia, you studied [reading] “photonic nanocrystal fabrication and quantum dots and waveguides.”

Murray: Yeah. Yeah. Yeah, 

Niki: That's [pauses, jokingly] you bring that up at cocktail parties? 

Murray: No, [Niki: laughs] Well, it's you got to read the room, right? Sometimes if you start bringing up photonic nanocrystals people, like, “Thank you and good night, you know,” [Niki: laughs] 

Niki: Exactly! Okay. So, this is how this dovetails with what I do for a living. I'm not a technical person, but I've worked in tech for almost 20 years. My job is to take complicated issues and technology and explain them to regulators, policymakers, staffers, policy people, reporters, PR people. 

So, having your brain and your ability to communicate this is going to be really helpful because nobody knows what quantum is. Well, some people know, but most of the people I know don't. 

What is it? But I want to start with kind of what you said about Kara Swisher. Why does it matter? Why is this type of technology important? Maybe if we get to the why, and then we explain the what, it will grip people more. 

Murray: Yeah, absolutely. So, let's start with some practical examples. At the port of Los Angeles, a company named SAVANTx built a solution to help them optimize how they were moving cargo onto trucks.

And that was a challenging problem for them. It was technologically challenging to solve that problem and it was also organizationally challenging. It turns out that the key performance indicator for those really large cranes they keep at the ports, they needed to keep them busy all the time.

And so, in changing the way that they were operating, they were able to use our systems to optimize the movement of each of these containers and increase the amount of cargo they were handling by 60% each day. And while they were doing that, they also managed to reduce the truck turnaround time by 12%. The port has this tremendous challenge, which is that like the amount of real estate that it has there is not getting any larger in Los Angeles, right?

It's really a question of with this increasing flow of goods and the kinds of challenges we see with our supply chain. Y’know, how can we maximize the flow through a fixed area? And I think there are a lot of businesses and, and public sector applications with very similar type of challenge.

And also do that in a way where they can respond in an agile fashion and with resilience. So that was just a, y’know, a fantastic example that I think that is very real world in terms of like the type of challenges that people find very difficult where we're able to bring them some new solutions.

Niki: So, I don't think people think about ports nearly enough. I think about ports all the time. I'm sort of obsessed with supply chain. I think we only think about it when, like, the Endeavor got stuck in the Suez Canal. But you're right. 

There's limited space. You see these containers kind of piling on each other. You see ships potentially arriving at different times. You have customs and border protection here in the U.S. checking every single shipment that's coming into the US for things like bags of rice could have a moth in it, right? So, they're, they're slowed down by that. 

And when you put together all of those things, it's really, really complicated to get the container in the right order to make it efficient. And obviously, it matters to businesses because it lowers their costs and makes them more efficient if they can turn things around faster. But also it matters to the average person when you've ordered all this crap and you're wondering where it is and [chuckling] you get like the delay thing.

If you can actually keep those cranes moving you're making everything flow better. It's more efficient. It's going to be cheaper. It's going to be to your point, more agile if there's weather events or geopolitical events. 

So, I am obsessed with this as an example. What does quantum do that makes solving that logistics problem easier? 

Murray: To help us answer that question. Let's begin with what kinds of problems that classical computers find easy to solve. [Niki: ok] So, for instance, if you, if you had a problem where you were trying to, like, adjust the flow through a pipe and you had like a valve that you could turn you can turn that valve anywhere from all the way off to continuously all the way on.

If you had several pipes, and they were all sort of pouring into different separate, not connected reservoirs, that's the kind of thing that classical computers find really easy. You're always kind of varying things like in continuous degrees and, and they are very effective at optimizing problems like that.

Where you run into difficulty is if the valve can only be on or off and they're all basically hitting a collected network of reservoirs and if they have an impact on one another. Quantum computers, on the other hand, are natively built to do that. So, they have these, you know, fundamental building blocks that are kind of like in one state or another, and they interact with other elements.

So they're solving problems like the types of things we see in supply chains where it's like, “Okay, I have a stack of these cargo containers. I have to choose which one I'm going to move. And if I'm going to choose container one, then I'm not moving container two.” Those things don't have any continuous flow elements to them.

They're stacked, just like you said, like in the game of solitaire. If you want to get to the cards underneath, you have to get to the one on top. So, the, the sequence of cargo that you can move is affected by every choice that you make. So, that is the type of thing that at small scales, y’know, when ports were just handling individuals, ships with lots of capacity in their rail network. That was relatively easy, but when you started to get up with many more sort of players in the rail system and infrastructure that sort of coming in and coming out and many different shipping companies bringing material in you get this complexity, which emerges even at small scales. And that's why I think what kind of locks us into rigidity and inefficiency.

Niki: And so quantum computing, tell me if I'm getting this right, helps us take these really complex multi-factor optimization issues and solve them. It doesn't replace classical computing, but it's an accelerant or an additive way to solve basically math problems. 

Murray: Yeah! Yeah, exactly!

If you think about it, when you're, when you're doing these kinds of assignments, it's, it's easy to imagine a simple strategy, which is, like, just greedily assign things until you can't anymore. And then start to make some substitutions. So, a quantum computer is going to try to, like, make big changes It's going to try to, like, sacrifice a number of decisions in order to collect a larger set of benefits. And that's very difficult for a classical computer to do so. That's that kind of it's the role that it's playing in that optimization problem solving and it's working alongside classical computers when it does that.

Niki: Okay, so let's use another real-world example. I've heard you talk about this, and I don't know if D-Wave works on this specifically, but, like, grocery delivery. 

Murray: Yeah, yeah. So, there's a company, Pattison Food Group, here in Western Canada. They're operating their business across 3 million square kilometers.

They have been deploying now, two applications on our systems. One is to help them with the scheduling of their grocery delivery drivers, and through doing that, they've actually managed to reduce the time it takes to produce those schedules by 80%. 

Now, that was a huge value to them, particularly with the way that their operations needed to adjust with the pandemic, because at-home grocery delivery was, like, very important during that time. And so, now they've taken it to, in stores where they're producing schedules for their team members and they're rolling that out right now. It's currently running in dozens of stores, and they're expecting to see their workforce save 50,000 hours of times annually. 

What's great about that is that wasn't their favorite thing to do. [Niki: right] Y’know, so it brings less stress to the workforce. The other thing was that it had to take into account knowledge about a lot of different areas of the business. It was the most senior people who are tied up with that kind of work. So, now those folks are actually able to get out of the back office alongside their colleagues and also work alongside their customers, which is probably the most valuable part of the, of the work that they do, and also the most rewarding. 

Niki: And you guys are working on enterprise solutions, right?

So, businesses can probably more easily understand, like, I have this headache with figuring out delivery and supply and getting just-in-time inventory and so they can understand the benefit. Obviously, with grocery delivery; same thing that we just sort of said about getting your stuff on time or faster. I worked at Uber for several years, and people think, like, they get so frustrated. “Where's my driver and how are they assigning which driver is where?” And it's incredibly complicated. 

The algorithm for that is incredibly complicated behind the scenes. And I, who also use Uber, get so mad [chuckling] when I can't figure out why it's taking forever for my driver to get from A to B, even though I understand that you've got supply of driver issues, you've got traffic congestion, you've got things that are unanticipated.

And I think this could be like an overlay on that sort of problem too, where people look at it and they don't understand why it's so hard, but it could just make it much, much more efficient. 

Murray: Well, I mean, the interesting thing is that we're building quantum computers that you don't need to know quantum mechanics to understand how to program them.

I'm talking about Python programmers work on their laptops. We've got examples in GitHub that allow people to build software quickly. And effectively what they're doing is they're creating a map between what represents frustration in the problem and what represents frustration for this quantum system.

And so, I mean, in this process of turning quantum mechanics into useful work we basically take our frustrations we pass them to the quantum computer, the quantum computer is trying to make itself less frustrated and in so doing it's making us less frustrated. So, it's, kind of, a very nice way to harness the technology.

Niki: I think what you said is so important. We have a big Silicon Valley listenership. Programmers may not be looking at this because I think programmers are just as intimidated by the concept of quantum computing - well, maybe not just as, as intimidated, but are intimidated by quantum computing also. It seems abstract; it seems way down the road I remember reading a Time magazine article, and they had this quantum computer and it was, the article said things like, “Even smells and sounds can screw up the quantum computer and they can only use,” tell me if all of this is wrong, “they can only use plain wax dental floss to, [laughing] to fix this machine.”

And I was like, “Wow, this seems a little dicey to be doing anything with. Okay!” 

Murray's laughing. 

What does this mean? Am I getting all this wrong? 

Murray: [chuckling] No, no, no, absolutely not!  I mean, you're, and the other thing that's amazing and maybe people will be surprised is the dental floss thing is totally accurate.

Niki: [laughs] It stuck in my head! It's such a visual, like, white, unwaxed dental floss is the only way this gazillion-dollar computer can be fixed. [laughing] 

Murray: Yeah, that has to do with the fact that dental floss works really well at ultra-low temperatures. But, but we'll set that one aside. Quantum effects are famous for being delicate and, like, easily disturbed.

And so, the notion here is that you want to try to harness the quantum effects and you want them interrupted as little as possible. And it turns out that how that impacts the application that you're going to build is affected by the class of quantum computer that you're trying to try to construct.

So, maybe just to demystify this, I mean, the definition of a quantum computer is very simple. It's just a machine that's going to use quantum effects to help accelerate calculations for us. At the most straightforward level, it's really about how is the computer going to harness those quantum effects as a resource.

Now there are, not surprisingly, although not maybe commonly spoke about, there are different classes of quantum computer and the class that you choose sort of impacts whether you need to know quantum mechanics to program it or what kind of applications that's going to impact. 

But let's say two that we're building at D-Wave, y’know, in one of them, it's harnessing the quantum effects to allow the computer to store more information. And in the other one, it's using the quantum effects to move between solutions more quickly. 

So, if you think about those two modes, it's going to respond to that collapse in a very different way. If you're putting information into quantum states and they collapse. Pfft, the information is gone. And a lot of folks will be talking about and working on error correction in that model to preserve the information that's being stored in that state. And that's going to be very important for quantum chemistry and, y’know, fluid dynamics and things, the calculations of that, sort of, those types of applications. That's what it's well suited for. 

The other technology, it's called annealing is the name of the technology, which has to do with the process that it uses to approach a solution. Annealing is using the quantum effects to move between that solution more quickly. And what happens is that when it collapses, it collapses on an answer. So, it could be like a grocery delivery driver schedule that you can use. If you hold it there, the quantum state will actually reemerge, and it'll continue moving through.

So it, it responds to collapse differently. And that's why those systems are really critical for near-term applications. And those are the ones that you don't need to know quantum mechanics to program and they're also easier to build. So, they're the largest, most highly connected quantum computers we have today.

Niki: To sum it up for me and what's important to folks in this town, we want to continue with the more complicated, slightly more fragile or harder to use technology. We want to keep putting money into that and developing it, but it is longer term and then annealing works now.

You guys are using it now for enterprise solutions. We don't have to wait for that to be applied because as you just said, I mean, I don't exactly know what this means, but it can collapse onto a solution, which sounds great. 

I wish I collapsed onto a solution more often! [chuckles] 

Murray:  Yeah, it's a it's a graceful exit, let's say! [chuckles] 

Niki: It’s a graceful exit! [laughs] But it's usable now. You don't have to necessarily know that for a programmer to use it and have the frustrated computer solve the problem for them. 

Murray: Yeah, and I think what's most telling is that these applications are being used regularly by businesses already.

So, y’know, that port of Los Angeles example, SAVANTx, as they were talking to me about the runs that they were doing, they were calling our system every 15 seconds, two shifts a day, six days a week. Pattison Food Group has had their production application running since October in 2022.

So, the technology is fun, and interesting, and cool. Through the benefits of all the software development, there's a cloud platform. It's got really high availability, greater than 99. 9 percent availability, you can call it 24 hours a day around the world. It's supported in 42 different countries.

And there's a really powerful open-source software development kit that people use to get started with programming examples. It's starting to become much clearer about where the near-term applications are. And as we build those up, then I think that's how people are going to be more interested to learn about the technology, right?

Start with the applications, get this to do useful work for you. And then, let's go deeper and deeper. And because it's open source, you can go [chuckling] all the way down the tunnel, all the way down the rabbit hole. 

Niki: Yeah, so I think this is really important, again, for our audience, the other big chunk of audiences in Washington, D.C. 

Do voters care? That's an important element to what D.C. cares about. But secondly, do businesses care? And if this is a competitive advantage for businesses or for the public sector, right? People managing traffic and storm response, and so on. If, if the public sector and businesses care, then Washington is also going to care.

So, I think understanding these real world applications is incredibly important to getting funding for also the kind of moonshot and longer-term vision for this tech. 

So, I'm excited about it. This is part of why I wanted to have you on.

Okay!  It's 2024. It's AI-palooza. [chuckling] Partly because people got, y’know, chat GPT, even though we've had artificial intelligence for a long time, they're super excited or panicked about it.

I'd like you to talk a little bit about how do AI and quantum work together, or how can quantum unleash some of the promise of artificial intelligence? 

Murray: Yeah. Okay. So, you're absolutely right. I mean, we've, we've had artificial intelligence for a while. There's, there's a really deep technical connection between quantum computing and artificial intelligence.

And, it has to do with the way that they're constructed. If you look at the architecture of the quantum processor and check it out, we've got stuff online to show this to people. It looks exactly like a quantum neural network. So, if you had access to a quantum neural network, how would that affect the machine learning algorithms that you would produce?

So, let's say machine learning algorithms, I mean, LLMs and Chat GPT. That's what people know the best: you submit a text query and you get text back, you get images back, y’know, maybe you get some video back or something like that. Of varying quality but when, 

Niki: [interrupts chuckling] I know that's one of the, that's one of the things people are panicked about right now is the varying quality of that, but yes!

Murray: Yeah, a lot of times it looks like it's dreaming, y’know, it's got that sort of nonsensical, like it's, “something's not quite right about this.” It's generating a huge amount of information. It's not helping us determine which information is valid and which information is not valid. You need to know logic to be able to do that.

Like if I said, you know, “Murray is in London today and there was another article that said Murray's in Paris today.” Y’know, are both of those lies?  Y’know, well, it turns out if you, if you find the date within the article and the two dates are different, then they can both potentially be true. So you need to know these logical relationships of, you know, this information relative to the absence or presence of other information.

And what's happening in artificial intelligence is that it's evolving based on the capabilities that classical computers have, so that shapes the way they're building technique, and that's moving it away from logic. And if we want to be able to -

Niki: [interrupts excitedly] I just want to say this. I just want to repeat what you just said. It's moving away from logic. We had a guest, Dorothy Chou from DeepMind, who said, tell me if you agree with this, “We have a lot of data in artificial intelligence. We don't have as much intelligence as we need.”

Murray: Yeah, exactly! Sure, we can pick out patterns, but the relationships between the patterns, the conclusions we need to draw are those we're still depending on ourselves. Right? 

So, a lot of folks are artificial intelligence and businesses are building, y’know, business analytics tools. So, they're just producing patterns and then the idea is show that to your executive team, and then they can make from it what they will. But there is a lot of removing noise from things that we could do.

If we could bring logic into things, y’know, if we can start to encode it and learn it in those those algorithms and fundamentally at a mathematical level that is not going to happen with just classical computers on their own. The types of models that quantum computers represent are the ones that can represent logical relationships: things that either happen or don't happen and when they do they have an impact on other logical conclusions.

So that's, I think, what's really exciting is that through incorporating quantum computing and artificial intelligence, we have an opportunity to begin to pick out relationships and logic and, hopefully, verification so that, y’know, we're not just bombarding ourselves with random noise all the time.

Niki: So, just to recap this. So, I'm making sure I get it at a really basic level. You've got this large language model or other enormous data set, and it's basically doing predictive work, right? Quantum can help you actually make fast logical connections for a better outcome. An outcome that is more usable because it doesn't have these random hallucinations or noise in the results. 

Murray: Well, [sigh] 

Niki: No?! Tell me how I'm wrong cause I’m, I’m sure I got that wrong. 

Murray: Well, when you're building an artificial intelligence technique, that's meant to be like artificial intelligence. The hallucinations part is, is part of that process. You're teaching it a set of relationships. And then if you give it some input that's going to cause it to remember some patterns. That's a lot like what dreams are like for us. And that turns out to be an important component of learning.

Often people will ask me, you know, when would I use an artificial intelligence approach and when would I use optimization? Because we can do both of those with our quantum computers. And I would say, “Okay, well, let's say you have a task in your business and you're going to hire somebody new and you're going to teach them how to do that task.” Are you going to teach them by example with some data? Are you going to say, “Look at these two scans, this one's healthy and this one's not healthy. And then I'm going to show you a lot more scans that we're going to, we're going to classify them.” 

That is what machine learning is trying to do. You show it data and then you help it discern what's happening in the data, as opposed to, let's say, like, a hospital administrator who's scheduling nurses on the ward, that person has got like a series of rules they need to follow, which is like, “I need to schedule nurses based on seniority. This is based on their training. I have three shifts. If somebody works a night shift, they can't work a morning shift the next day.”

So, there's a series of rules there. And as I was mentioning, artificial intelligence, machine learning today is abysmal at discovering those kinds of logical rules. The way we use our quantum computers in concert with classical computers is in those optimization tasks is to basically describe those, pass them to the quantum computer, and say, “Explore a huge world of possible decisions that could be made and quickly zero in on a set that make me less frustrated, make my workforce less frustrated.” 

And in the way that it's contributing those kinds of trade-offs about things happening, not happening, that's the role it can play in artificial intelligence where it's trying to create and, and learn relationships about things that y’know, “if one thing happens, the other one does not.”

That piece is missing right now, which is why we get just a lot of data that works really well when you're analyzing it qualitatively, but not like from a logical mathy kind of point of view. 

Niki: Okay, this is really helpful, so I understand this.

It's the difference between making radiology scans more accurate, what you're looking at, and the nursing schedule, which is incredibly complicated and involves a lot of OSHA rules, and overtime rules, [chuckling] and hospital systems, and you are trying to fix a frustration in the business itself because of the if this, then that, and you do it over and over and you get a much better schedule that comes out of it.

Murray: Yeah, yeah, exactly. 

And we've, we've been relying on humans to do that. And we are fantastic at it. It's just that when you get beyond, y’know, five nurses and a single shift per week and you start getting into, like, dozens or hundreds of nurses with three shifts per week, we're really pushing what humans can do. We start creating a bunch of policies and rules that limit the effectiveness that we can have, our resilience, and our agility. 

And so, what we're exploring with our customers, y’know, those two examples I gave. NTT DOCOMO in Japan, y’know, working in telecommunications, y’know, MasterCard that's looking at offer allocation and fraud detection. What we're exploring with them is “How can we actually go beyond what you're capable of doing right now, incorporate quantum computing with your IT infrastructure in a way that's bringing value immediately?”

Niki: Right. That it's not as abstract. 

So, maybe this is abstract and maybe it's shorter term. I'm not really sure, but when I think of AI and some of the real promise of it, I often think of drug discovery and the amount of data that's involved in coming up with new solutions and new treatments.

Is there a way that quantum can in the future or could now be used with that? I definitely get the idea of optimizing these really difficult real world problems of essentially like logistics, humans meeting, scheduling, meetings, computing, meeting data, but what about things like drug discovery?

Murray: Absolutely. Well, I mean, at D-Wave, we've got a partnership with Zapata AI to investigate exactly that. To create new quantum generative artificial intelligence techniques to be applied in that space. Personally, I find it unbelievably motivating to work in health sector and the life sciences sector because I would love it if this technology that I've been working on for 20 years can positively impact people's lives.

The interesting thing when you look at drug discovery is you have this aspect of, let's say you're building a protein, or something like that, where you're, you're constructing a sequence of amino acids and you've got a set of 20 amino acids you can pick. You can't continuously vary them. You either have one or you have the next one. And each decision you make affects the next series and also how the resulting protein is going to fold. 

So, it's also filled with these problems, which have these discrete issues with them. So, I think it's an, it's a natural fit there. The second thing, just zooming out here, is that I think too many people are focused on, like, just discovering the drugs themselves. 

And I think we need to consider how unbelievably impactful it would be if we could reduce the cost of testing and validating the drugs are working. That program, the thing that costs hundreds of millions of dollars over many years or decades, if that cost can be reduced, that's going to have a transformative impact on the types of medications or drugs that can be developed and how quickly they can be brought to market. And also, being able to develop things that don't affect everybody, but actually affect a small group of people with very severe consequences. 

I'm really motivated to work on it. It's just a matter of how can we get more people into the space, understanding it and applying those kinds of solutions, at this point. 

Niki: I think I'm so glad you talked about that example of how drug discovery can be impacted because you're right everybody's focused on the new next drug. And I often say, maybe this is right, maybe it's wrong, but I feel like when people start talking about efficiency, everyone kind of zones out.

It's like talking about the national debt. Like, nobody knows what it means, but if you say “Efficiency means that these drug companies are going to have lower costs as they're discovering drugs and shorter timelines, which means they're going to have more resources to put into additional drug discovery.” That is a really concrete example of how efficiency improves our quality of life. 

I often want to get people excited about efficiency because I'm excited about it [laughs]. But I think that's a really good way of thinking about rather than just the, y’know, the shot-in-the-dark cancer drug, the idea that you're creating more space within these companies to explore more cost-effectively.

Murray: Yeah. Yeah. Let's keep the, the goals and objectives of the same. Let's help people and keep them safe. Do no harm. Let's keep that the same. I think sometimes people confuse jumping through hoops with achieving those objectives, right? Let's measure ourselves by the objectives and then let's find a way to measure the process and improve the process.

And if we can improve the process, shorten it, make it less expensive, the impact would be enormous because it would apply across all pharmaceutical companies. It would probably apply across all sorts of human ailments that we'd be investigating. And that's the kind of, y’know, revolutionary, world-changing impact that we seek to have with quantum computing here at D-Wave.

Niki: Well, I will compliment you on a couple of things. First of all, for someone who's incredibly highly technical and has worked on building quantum computers and been in this space for 20 years, you just explained clearly. 

So, you have incredible communication skills on top of your technical skills and obviously business acumen. And I'm also excited for you because the fact that I'm talking about quantum means what you've been working on for 20 years is now [chuckling] hitting the headlines enough that I'm like, “What is this? We got to figure it out!” 

What's the final call to action to our listeners? What's something you'd want to point them to so they can learn more, try this out. If they're in programming, get sort of excited about it. 

Murray: Yeah. Well, first of all, thank you so much, Niki. That's absolutely the goal I'm looking to hit. And it's such a pleasure, y’know, coming out, and talking with you about this, and sharing this information with people. 

What I want folks to realize is how easy it is to get started. Everybody who's listening is going to have either themselves or a colleague, or a brother, or a sister who's interested in tech, who wants to try this kind of thing out.

You can basically just do a Google search for D-Wave Leap, and any of the top hits that are going to come, they're going to take you to our cloud platform. You can quickly sign up and get trial access for free and begin to explore the types of applications and see how those are being used.

There's demos that you can run with no programming, and then it takes you to the open-source code that those demos are run on for those of people who want to program, it can start modifying the program and exploring the boundaries of their learning. And I think we're only limited by the number of folks who are taking this new capability and applying it into these application spaces with a transformative impact.

I really look forward to that and, and to helping people and engaging folks with the technology. I'm expecting, y’know, this is a community of teachers and learners. I'm going to learn a lot from them through their experiences and what they're capable of doing. 

Niki: This is fantastic. I'm unafraid to say when I don't understand something, ‘cause I feel like in this town in D.C., you have to pretend you understand everything. And I don't find that particularly constructive [Murray: Right] to the public service. [chuckles] 

Understanding this is great! I am going to drop some links just for people who want to go into the show notes and click directly and experiment and try it out so they can go down the open-source series of tubes and try it out themselves.

Thank you so much for taking the time and coming on. 

Murray: Fantastic. Niki, thanks for having me here and I look forward to talk to you again soon time.