Latest / Technology Now / The Future of Innovation
Transcript
- SAM JARRELL doing alright. I was actually curious, have y'all ever invented anything? AUBREY LOVELL I have not, I don't consider myself that cool. Although, if we're talking about music, I would say I've written songs, had a band, but nothing invented, I would say. What about you guys? MICHAEL BIRD Yeah, mine's not really a personal invention. Executive producer Izzy and I were chatting one day, we were chatting about breakfast foods. And in the UK, there is this breakfast spread it's known as yeast extract. And it's very love or hate it, people are the love or hate it. And I think that's Brits are known for liking said paste. And she's got me onto putting it on peanut butter. I think it's one of the greatest things ever invented. So I didn't invent it, but it's something I think about a lot. AUBREY LOVELL Nice. So we've covered music, food. Sam, do you have anything? SAM JARRELL I haven't invented anything, but I'm getting married soon. So I have been doing many a craft project and there's kind of an interesting invention for people who don't want to use real flowers for their wedding because they're very expensive. These are wood flowers. So they look like real flowers. They're soft too, but you dye them and there's no actual, you know, soil or anything in here, but it looks nice. MICHAEL BIRD Those are brilliant, Sam. Thank you so much for sharing. I love a good show and tell. We should have show and tell every single week on technology now, shouldn't we? AUBREY LOVELL I think so, I think we should add it into the block, into the script MICHAEL BIRD I’m Michael Bird AUBREY LOVELL I'm Aubrey Lovell SAM JARRELL I’m Sam Jarrell And welcome to Technology Now from HPE. MICHAEL BIRD Once again we have our reporter Sam Jarrell with us in the studio to record the second instalment of our two-part series on Innovation Sam. Thank you so much for joining us. Aubrey, I know about you, I thoroughly enjoyed the first episode. I thought it was absolutely fascinating. AUBREY LOVELL I did, I really loved it. It's a lot of fun between us three. MICHAEL BIRD Yeah, absolutely. And if you haven't listened to part one, go just go back and listen to it. You will not regret that decision. AUBREY LOVELL And like last week, we will be returning to the HP Garage in Palo Alto, the birthplace of Silicon Valley, where Sam met with the chief architect at Hewlett Packard Labs, Kirk Bresniker. SAM JARRELL Yes, he was so wonderful. And last week we talked all about innovation, the road to where we are today. But now we're looking forwards. What's next for the world of technology? SAM JARRELL Well, before we get too excited about what might be invented next, Aubrey, you've been looking back at some of the defining points of the history of innovation, haven't you? AUBREY LOVELL That's right, I have, and it's time for, as we know, technology then. AUBREY LOVELL so like everything, innovation happens in waves, right? We have a new discovery that kind of allows a leap forward in progress before innovation slows as it kind of awaits that next breakthrough. And we've seen this, you know, over and over again throughout history. But last week, I talked about events from the first three waves. Can either of you remember what they were? MICHAEL BIRD electricity, electricity maybe? Yeah, yeah. AUBREY LOVELL Getting warmer, no pun. SAM JARRELL Electricity is for sure. AUBREY LOVELL Yes, that's about right. Yes. So we're currently kind of sitting comfortably in the sixth wave of innovation since the Industrial Revolution, right? So we sort of covered the fourth wave during last week's episode with the advent of electronics and wave five included things like digital networking and software . And then we saw an increase in global internet usage from 2.3 million users in 1990 , which is up to 3.4 billion users by 2016 . So that's an increase from 0.04 % of the world's population, which is up to 45 % in just over 25 years . So that's pretty significant, right? And then the sixth wave of innovation only started around five years ago, so we were kind of still nailing down exactly what that looks like, but predictions generally think it will be focused on Internet of Things, devices, AI, clean tech, robots, drones, and responses to our changing planet. So it's kind of really exciting. Michael Bird (11:26.766) I don't know about you but I mean it does sort of feel like we're in this wave, like it does feel like a really exciting time, like I don't know, I feel like every time I open my news app there'll be some news of some big exciting innovation that's gonna change our lives. Sam Jarrell (11:45.73) I agree. think it's exciting. It also is sometimes nerve wracking. You never know when the AI robots are going to come take over. But that's why it's good to talk to people who know a lot about these types of things. And I was thrilled to be able to speak with one of those leading the charge into this wave of innovation when I spoke to Kirk Bresniker, the chief architect at Hewlett Packard Labs during our chat at the HP garage in Palo Alto. SAM JARRELL It is and I was thrilled to be able to speak to one of those leading the charge into this wave of innovation when I spoke to Kirk Bresniker, the Chief Architect at Hewlett Packard Enterprise Labs during our chat at the HP Garage in Palo Alto, but before we talked about innovation. I wanted to know a little bit more about what being a Chief Architect is all about… KIRK BRESIKER I'm the chief architect at labs, which means that I get to look over our whole research portfolio. Everything that goes into the next HP Aruba access point, in the next Cray supercomputer, and everything in between all of our research agenda. And specifically, I guess I'm kind of always in the uncomfortable middle space, you know, halfway between research and development and trying to understand how do we bridge when people are looking at five, 10 or 15 years into the future? And we have engineering development teams that are thinking about next quarter, the next quarter or the next quarter. How do you get those two groups with very different time horizons to cooperate together? SAM JARRELL Wow. That is an interesting space to be in. I feel like you must have to be very forward looking all the time. KIRK BRESIKER It is. It is absolutely. Being forward looking is also understanding how you would tell a story. Because you know, innovation isn't just something that we do here at our company Well, here we are in the 21st century and innovation is definitely a team sport. So someone has a great idea. Fantastic. Who in the world of of of semiconductor fabrication are we going to partner with to make that device? Who in one of our semiconductor, processing partners are we going to get to incorporate that device? Who in the open source, kernel community, are we going to get to support that device? Who in the AI framework will be used to building the next generation great AI outcomes on top of all of that coming together? SAM JARRELL Wow. So you mentioned that innovation is kind of it's a team sport, but I am kind of curious because patents aren't necessarily. How many patents do you have? KIRK BRESIKER So I right now have 33 patents with a couple more in the pipeline. So people don't, you know, you think, what? How long does it take to get a patent the fastest time I've ever had was three years. it's intentionally, you know, a process where you want some scrutiny. You want someone to make sure that, yeah, this is really important because we're saying is we're going to give you the patent, the license, the ability to to capitalize on the technology exclusively for the patent period. SAM JARRELL And so, you know, obviously, you have to educate people quite a bit on some very cutting edge technology and what sort of things that you are you currently working on that you are allowed to tell me about? KIRK BRESIKER Well, you know, much of our work at Hewlett-Packard labs, is open research. So it is intentionally done with national labs. You know, something with DARPA or something with AI, RPA, something with researchers, both here in the United States, all through Europe, all around the world. And and then we're they're writing the paper and we're presenting it at conferences. But I guess for me, what is really interesting is when I think of the breadth of our company, we think about that little tiny wireless access point or that that, that 5G base station that's connecting sensors and people, and then all the way back through the networking, through the compute, through the storage, through all of those clouds. Because that's really what it takes to really have a great innovation these days. It's not just about one technology, it's about how you pull all of them together or to do something really interesting. And so when I see our labs teams, I see them working on the problems that almost are the the function of success. Here's a great AI outcome. It only takes a a, it only takes a giant room of computers. Well, what if I want to have that happen underneath the hood of my autonomous vehicle? Or in the head of a drone, or in that next generation access point? How do I reduce its energy footprint? How do I make it even more effective so that it's easier for us and it's possible for us for delivery, not just to do a few people, but what if I want great AI outcomes for 8 billion people? SAM JARRELL You brought up something that I really wanted to talk to you about. And that is AI. It's, you know, the thing that's on everybody's lips, it's the hottest topic in the world right now. But I feel like when we talk about AI, typically people think of software. But really, I haven't heard very much about physical AI. Can you explain to me what exactly physical AI is and its applications? KIRK BRESIKER Yeah. You know, right now we’re we’re in an age of marvels. physical AI is more, more like the intelligence that we embody. If you think of the of the entire range of intelligence here that we share our planet with, you know, it's not just us, it's right all of life that has that ability, that has that neuron that's making an interaction with its physical environment. And that's really what we're talking about here, is not just training an AI on text, on tokens, on that next bit of knowledge, but having it operate and learn the way that that we operate in learned as children, you infants, you know, they're sitting there and are immersed in their environment. They're immersed in the world. And they might not realize it, but every time that they pick something up and put it in their mouth, they're doing an experiment. They're gathering information. They're training their own neural network. When you pick something up and you drop it, you're learning about the physics of the world that we live in, and you're training your own neural network. That's why we have all those great toys for babies is to to begin to teach them. And that's really what physical AI is. What does that next robot that will we'll be training up, that understands the physics of the world in which it is moving. What is that next thing that's pulling it? All of that will pull information, not just the text. We're kind of like just what presents the visual information plus all of the physics of objects moving, occupying space, how things move, how they start, how they stop pulling all of those things together and giving that that autonomous robot, that physical manipulation device, that ability to learn from its environment, very much like the way that we learn. SAM JARRELL So it feels very sci fi to me. how close do you think that we actually are? KIRK BRESIKER I think we're being motivated by the fact that we've we've used up all the information. So all of the language has been digested. It's really that next opportunity for us. And in many ways, we might find that it is a completely new and novel and complementary way for us to train our AIs. And then we'll also have these AIs now that are being trained because they live and occupy and exist in the same three dimensional world as we did. Or maybe they'll be the first ones that are really trained and natively occupying in environments that we cannot exist in. So, you know, having an AI that learns what is left live at the bottom of an ocean, to have an AI that learns live and is trained in the vacuum of space. SAM JARRELL Okay, okay. And are there any places that you can think of that physical AI is already potentially being worked into practice, or is it is it so early stages that it's not even it hasn't had applications quite yet. KIRK BRESIKER Oh no. There are definitely real applications. And if you think of some of the things that we have demonstrated from our own labs, work and research on sustainable energy, and we're working, you know, on the wave motion, you know, we've we showed those at our discoveries in past years the wave motion. Well that is actually that is a reinforcement learning mechanism that’s learning from the environment. So that way motion device isn't just programed once in the factory. And with the if the ends the pieces of of some control loop that we might have crafted, it is learning what it is to be that way. Motion engine in that spot in that ocean, SAM JARRELL So at HP labs, that's really at the center in the heart of our innovation, where do you see the tech world and even our innovations going in the future? KIRK BRESIKER We’re definitely the ones looking out three, five, 10 or 15 years. But, you know, we have innovation all across the company. And whether it's not just not just the the engineers either, it's also understanding how do we run the most agile, efficient company. SAM JARRELL Are there are there any sort of more sci fi or wacky ideas right now that you think have, are interested in, or think that maybe ten, 15, 50 years in the future will have legs? KIRK BRESIKER Well, I think I think right now, I mean, we all are experiencing what it's like to have these AI agents that are doing remarkable things. I mean, it seems like almost overnight we went from, well, when it can paint a picture, when it can write a poem, it's like, dun dun, check and check. And now understanding how those will continue, I think we'll still see as a society how that's going to improve. Now, the one of the piece of that, and some of the work that we're doing at labs is understanding, okay, we've done inductive reasoning. We've taken literally all the world's information. And now we can predict the next word, the next word, the next word, the next sequence, or the next brushstroke. Fantastic inductive reasoning. Take a lot of a large amount of information and generate the patterns behind it. But we don't necessarily have right now is deductive reasoning. The logical reasoning that we have in mathematics and in theorem proving, and so understanding how we begin to bring in those technology and what compute devices are necessary for those types of calculations that might be quantum, that might be quantum inspired . But some of that, some of those breakthrough computational approaches being brought in to complement what we see with the large language models, especially given how voracious they are for resources. SAM JARRELL So I might be transparent. Whenever I hear about quantum computing, it goes way over my head. Can you just give me a like, what's the couple of sentence explainer on what quantum computing is? And then I want to know how it's supposed to maybe work with AI. KIRK BRESIKER Yeah, absolutely. So, we'll go back to 1982. Richard Feynman, Nobel laureate, probably the most famous physics teacher in the history of the world. And back in 1982, he said, you know, the world isn't classical. Damn it, it's quantum. So if you want to simulate it, it better be quantum mechanical. And by golly, it looks like a good problem because it's not going to be easy. And he was right on both accounts And so really what we're talking about quantum computing is how can I use quantum mechanical behavior in order to simulate quantum mechanical systems? And why I care about this is because that's how we're going to get the next, material science breakthrough, the next great pharmacological breakthrough next battery chemistry, SAM JARRELL Thank you. That was that was very helpful I so then I'm curious, you know, because it seems like we're not there yet. Then with it how how far off is proper quantum computing. KIRK BRESIKER So I think that, you know, this part of the work that we're doing, labs as a matter of fact, we were doing the basis in quantum computing is something we call a qubit, which is the smallest system that exhibits all these quantum mechanical behaviors and that we can control and then harness to do a quantum mechanical calculation. We were doing, qubit design, over here at labs about 15 years ago. And that team actually stopped and pivoted away from the work Now four years ago after we had acquired SGI, and Cray and became the world's predominant, preeminent supercomputer, provider, a lot of our customers are saying, I love what you're doing with the exascale system, but for the one after that one, where do I plug the qubits in? Where do I integrate quantum computation into this? Because I really want to do this. I want to have all the physics and science I can do today plus plus quantum. And so that's really where we started. We want to reengage with the quantum technology community, globally. And that's where I got to hire a small team, and made some critical, critical hires. And we began to explore what is it like, what is it like for us to have the world's fastest supercomputer combined in with these quantum devices for computation? And so that's where we are right now. We wrote, a position paper how to how to create a quantum supercomputer. How do you scale from the tens of quantum devices we have today to the 10 million we'll need to, in order to affect one of these real problems SAM JARRELL . Okay. So, Kirk, we've talked a lot about different innovations. We've talked about quantum computing. You've told me that innovation is a team sport. I want to know what innovation outside of quantum computing are you most excited for in the next 20 years? KIRK BRESIKER I think I'm really excited when we can take our human intelligence, our human intuition, embody that in a physical AI device and have it go where we can't, And still I look at the challenges of of ourselves leaving the earth, and I look at the energy and time it takes for people to go there. And I wonder, what if we could go not as ourselves, but as our as our intelligence, as our intuition, as all of those capabilities physically embodied in an AI that we can send out in place, you know, not as, as a partner to us. And I think for me, that's the real the real interesting thing when humans can leave the Earth, at least in our intuition, in our understanding, in our belief, in our aspirations, embodied in something other than our own selves, and still feel that direct participation and kinship with expanding out human knowledge, into, into, you know, whether, again, whether it's the bottom of the ocean or deep into space, understanding how we will move forward, hopefully also as we spent time to, to heal the Earth, because we still want to understand how to live here and thrive here, but understand, you know, again, how we expand that. So for me, that embodied intelligence heading out into into the far reaches of space crossings, crossing the emptiness of the void, but carrying humanity's spark forward. MICHAEL BIRD Man, how good is Kirk? He's got so many stories. I don't know about you, but one of the things that really struck me was the line, innovation is a team sport. Like did that strike you Sam? SAM JARRELL It did. mean, we're standing in the HP garage and even the initial oscillator invention that is in the background there happened only because of the support that Bill and Dave had and that continues on today. You can't just do everything by yourself. It truly is a team sport to bring these things forward AUBREY LOVELL Definitely, and I think there's something to that as well, right? Because if it is just one person, it's only from one person's experience and viewpoint, but when you do have a team and an innovative team at that working around the clock to do these types of things, it just makes your outcome stronger. And I think you see that with AI, right? Like we talk about that a lot around your inputs are only as good as your outputs. And if the inputs are being programmed with all these different perspectives and information versus it coming from one source, it's going to be a stronger AI MICHAEL BIRD Yeah, and when you said basically we've used up all the information, I think sort of what he was alluding to is that these large language models, we fed in as much information that's freely available. But Sam, I think that the sort of conversation around physical AI was quite interesting. The fact that actually, okay, we sort of looked at all the text based on maybe all the information that's available on the internet digitally. But actually, if we can send an AI out into the physical world and they can... interpret the physical world and learn about the physical world like a baby will learn about the world. Actually like what innovations come out of that? SAM JARRELL Probably a robot that can help you do the dishes. But it sounds like you'll have to teach them first. You'll have a little robot intern. MICHAEL BIRD Did you say robot intern, Sam? SAM JARRELL Yeah, because you'll have to teach it. Same way they don't have to learn basic things like a baby, but you might have your robot interned for a while that can only wash the dishes before it can maybe cook or something. AUBREY LOVELL See, Michael, we put it into the universe. We put it into the universe. It's happening soon. SAM JARRELL I agree, I agree. Now, there was one thing which Kirk said that I thought was particularly interesting about how long it can take to hear about all of these breakthrough discoveries all the time. But for most inventions, it's actually a bit of a slower process. KIRK BRESNIKER I was once I was in the back of a conference room sketching out some ideas because it was this new kind of customer, a service provider, an internet service provider. And so we were listening to them. I was sketching out my green engineering ideas about 1998. By the time that everything on that pad had shipped as, as, as one of our products, it was 2016. So I think that's for all the innovators out there. Just be prepared for the long haul. Prepare to nourish an idea. Prepare to look for other ways to get your ideas more fleshed out. AUBREY LOVELL Okay, well that brings us to the end of Technology Now for this week. Thank you to our reporter, Sam. Love having you with us. And our guest, Kirk Bresniker, thank you so much. And of course, to our listeners, thank you so much for joining us. MICHAEL BIRD Yes, and if you've enjoyed this episode, please do let us know, rate and review wherever you listen to podcasts. And if you want to get in contact with us, please send us an email to technologynow at hp.com. Subject line, can we have Sam on the show again? Maybe that's a good subject line. And don't forget to subscribe so you can listen first every single week. Technology Now is hosted by Aubrey Lovell and myself, Michael Bird. And our reporter this week was the fantastic Sam Jarrell. Sam, thank you so much for joining us. This episode was produced by Harry Lampert and Izzy Clark with production support from Alicia Kempson-Taylor, Becky Byrd, Spencer Trinwith, Mikey Nissenbaum, Alyssa Mitry and Renee Edwards. And our music was composed by Greg Hooper. AUBREY LOVELL Our social editorial team is Rebecca Wissinger, Judy-Anne Goldman and Jacqueline Green and our social media designers are Alejandra Garcia, and Ambar Maldonado. MICHAEL BIRD Technology Now is a Fresh Air Production for Hewlett Packard Enterprise. (and) we’ll see you next week. Cheers! AUBREY LOVELL Cheers. SAM JARRELL Bye y'all.