B2B Tech Talk with Ingram Micro
B2B Tech Talk with Ingram Micro

Episode · 6 months ago

Power of 2: How to Harness AI with Lenovo and NVIDIA

ABOUT THIS EPISODE

With virtualization 20 years ago, you relied on virtualization companies to help you start the journey. When cloud computing came about 10 years ago, there were cloud providers that showed you the way.

Now, as the era of AI is upon us, you’ll need AI providers to help you down that path.

Shelby Skrhak speaks with Ed Soohoo, Innovation and Transformation Executive at Lenovo, and Jared Carl, Senior Enterprise Partner Solutions Lead at NVIDIA, about:

  • Why NVIDIA and Lenovo teamed up for the Power of 2 campaign
  • How to think about AI
  • What’s possible with the power of AI  

For more information, contact the Lenovo ISG Team (shelly.obrien@ingrammicro.com).

To join the discussion, follow us on Twitter @IngramTechSol #B2BTechTalk

Listen to this episode and more like it by subscribing to B2B Tech Talk on Spotify, Apple Podcasts, or Stitcher. Or, tune in on our website.

...you're listening to B two B Tech talk with ingram Micro the place to learn about new technology and technological advances before they become mainstream. This podcast is sponsored by ingram Micro's. Imagine next. It's not about the destination, it's about going someplace you never thought possible. Go to imagine next dot ingram micro dot com to find out more. Let's get into it. Welcome to B two B Tech talk with ingram Micro. I'm your host. Shelby scare hawk and our guests today are Ed So who of Lenovo and Jared carl of NVIDIA gentlemen welcome. Hey Shelby, nice to be here. Thanks Shelby, appreciate you having me on. Well. So we are talking today about the power of Two campaign with Lenovo and NVIDIA so Jared why have a video? And Lenovo really teamed up on this campaign to talk about the power of AI. That is a great question. Shelby people know in video from our background in computer graphics, we make pretty images. Right? Well, what we've done in the last 10 years as we found out that that power of making pretty images, the power of accelerating graphics and bringing that out to the mainstream is very applicable to HPC workloads. But HPC workloads, we need democratize those out become A I workloads as we continue down the path of AI which is training a computer to think like a human to recognize objects and to make guided decisions for us. That power is only possible within video technology as of today. Gpus accelerate that but we can't do that alone. We need partners such as Lenovo. We need an O E M. With the strength of Lenovo to be able to deliver that message out to us and actually help these customers on their Ai journey. Just like virtualization was 20 years ago you relied on the virtualization companies, the cloud story from 10 years ago, you ahead of the cloud providers there to help you figure out how to get on that journey with the AI providers today. As we get on the democratization of compute you need the brainpower of Lenovo and the ability to deliver an enterprise class system supportable anywhere in the globe with world class computing resources and the support necessary behind it to make it run Well, yeah, I mean the power of computing it, it's that's everything. I mean that's what, what is driving a lot of this. So, but I thought it was interesting Ed on our pre interview call, we talked about discussing technology, not in the very technical terms of of the capabilities and the tools but in real words and real instances and so that's I think what's fascinating about this campaign. So how do you think about Ai? You know, Shelby thanks a lot. I really appreciate that What's fascinating is I don't think about AI, I think AI should be almost behind the curtain just making things happen to anticipate what I'm looking for or thinking about or should act on and a I should be so transparent that I never really have to think about it. And I think that's the power of two where we make through the technologies of our infrastructure and our integration within video. It is seamless and transparent and more importantly power that power of imagination, that power of insight to me, AI is all of that. So when we look at these elements of AI, there's a lot that goes into it. So talk to me about the I guess analogies for how you see the elements of AI and what is needed to perform at its best. Yeah, so I...

...think AI is kind of big Rubik's cube. You're twisting and trying to match things very, very quickly and there's a methodology to it. What is it, how do you break the code? The other analogy is almost like cracking a safe or a vault. What is the combination? What do you listen for? What do you feel for? As you turn the dials on the combination to unlock secrets, value insights and to be a i is all about unlocking and really finding the methodology to really help address the business issues. Right, right. We maybe got off on a tangent a little bit. But you mentioned that you actually see a lot of this in terms of movies. I do. I love movies and movies actually art imitates life and life imitates art in the context of a I the movie that stood out to me that kind of represented a I and it's kind of like essence is the movie Groundhog Day with Bill Murray and Andie Macdowell and what I loved about the premise of the movie was how do you get better every day? How do you learn the lessons as you iterate? You know, the funny scene that I thought was when he was talking to Andie Macdowell and he asked her about her background and she said I had a she has a degree in french literature and he scoffed at it. Oops, he learned really quickly the next day when the alarm clock went off at six a.m. And Sonny and Cher Song came on and said I got you babe, he actually started to take french lessons. And so that's a I iterating on the fly learning what would be the right way to answer in the next set of contextual, let's say, opportunities and the Groundhog Day, uh you know, story is about lessons learned. Iterating learning what the ultimate goal of having a building a relationship and in this context for a lot of our business listeners, it is your relationship with your customers and your employees, you know, as you're mentioning that it kind of occurs to me that isn't life a big AI set because aren't we trying to learn and adapt and take new information in and be able to improve the next day. We don't get that, that redo like we do in Groundhog Day, but do you see what I'm saying? I do and I love that part. Right to me ai is like a big wild hairy ride. It's like a roller coaster right? You just get your arms up above your head and just saying come on let's go to me ai should be something of an adventure. Something that you're going to learn every time and the technology is behind all of that to make it magical, powerful and relevant. Is this is why the power of two is so critical. What about data sets? The information that comes out is only as good as the information that goes in. So you know what are these data sets I guess? How would you think of those? And what can you explain to us in real words? So I'll use an analogy to me the goal of a lot of artists. The biggest conundrum they have is a blank canvas but what they have is a palette of colors. What they have is different mediums from oil to chalk to crayons and a like how do they paint the data? What what kinds of data do I get? Uh let's say access to the data that I own the data that I get from my partners and the data that I get from third parties. How do I pull all of those data streams together? Create something powerful relevant, serendipitous and discovery in new types of let's say reactions coming from my workforce or from my customers where I would never thought had I just used a single set of data Beauty is in the eye of the beholder, data value is...

...in the eye of beholder as well. So when when you look at Jared, when you hear, you know ed describe data data sets and uh ai in these terms, I guess what do you think of those? And and are there any that you would add to those? It's it's a very interesting context when we look at data and especially as we apply it back to a I you know, the Arthur C Clarke said any sufficiently advanced technology is indistinguishable from magic. Ai is not magic. It's just advanced technology. It has advanced technology that is the leverage years and years of data That's flown into a system. And if we look at customers nowadays, we look at the businesses adopting these technologies. Every single technological input into the computing world has come from streaming from data. How do we deal with the data? How do we solve the data problems faster. Now, the problem is we have way too much data that's coming in and out. We generate on average like 10 petabytes of data a day, Probably 15. If my teenage daughter went to Taylor Swift concert, we'll see where that flows in, right. But it's what you what data you have, what you've learned from your data, how much in front, how many models you can throw out the data to learn what it means. But then what do you do with the data? And that's really the click over point there of making ai real. Um when we look at simple Ai program I put simple in air quotes here. Right? Because ai is complex but it doesn't have to be chatbots. What is somebody meaning to ask a computer? These are everywhere. NLP natural language processing where you can scan a document page and it gives you a summation of what those are. Those are fairly straightforward because they're constrained environments. But when we look at the more complex Ai is a self driving car. You're dealing with really fast streams at 60 frames a second of video image of radar data. Of car data, right? Your car is already collect so many 100 points of data every given second to make sure that it's running well. And then you have to control the car and something that's very natural to humans, right? And I was driving this morning and a squirrel ran out in front of me and I immediately knew it hit the brake, right? The amount of programming that goes into just do that is immense. But it's dealing with the data and doing that together with somebody to understand at what points during the logical flow of data. You know when we talk about the data story here, we have two sides, the conversation, one is the data scientists that are actually manipulating the data. One is the infrastructure team that's actually building the infrastructure and supporting it to go with it. And I said there's two, there's actually a third one and the third one really is the business units, right? The business users, the people that understand what what do I want out of this experience to understand what's going on. So the data scientists look at the data flow, the infrastructure teams really look at the physical connectivity of it. Are you in a well connected network? Are in a disconnected network, self driving car? Example for example, you could be in middle of a desert. There's absolutely no connectivity out there. It has to be fully self sufficient. Right? You don't want to counter anything that's gonna mess it up. But then of course there's the business units of people that really understand that user interface that goes back in. It's like, well how can we build the story around this Ai model ai solution? But they all have to integrate with the data. Some of it is hard data, some of it is, yes, that's a squirrel that ran out in front of me. Some of it is empirical data. Something like personal data is like, well I don't want them to slam on the brakes because that's uncomfortable for the person. Right? How we interact with the ai is absolutely critical. One of the funny things about chatbots is they're very they're very functional now. They're very interactive. We still cannot program a computer to catch sarcasm yet. And I find that absolutely hilarious. So there are still some gaps in there of getting a computer understand. Oh, they're just joking. Coming again from somebody who has a teenage daughter, um...

...two teenage daughters to actually getting a functional chat bot that can figure out that when we're joking about O. G. Thanks for your help versus G. Thanks for your help. She sounds like a scene out of the Big Bang has the inability to really deduce sarcasm, which is which begs the question, who should be in this boat as you build out A I. And you threw out some very interesting personas of the data scientist, the infrastructure and the business unit. Actually, those should be three in a box, those three areas of connected, let's say best practices and points of view. The data infrastructure or the data infrastructure, technology has the power to making sure that it can provide at speed at the right time, at the right person etcetera. The scientists to really break things down and the business unit to interpret it into something that is relevant to drive, let's say the goals of the business, which is to delight the customer. And when I thought it was rather fascinating, Jared, you mentioned something about NLP and chat box uh you know, reading text for example, a I at some point is gonna understand tone intent and yes, eventually sarcasm at some point it'll get there. It also occurs to me that of these three personas if the business unit the end user, if that's the person that has that final purpose in mind. You know, I know ed you had talked about with data being it whether it's um it's like a canvas and so, you know, if they are the ones that are create well, who is the artist? Who is the, I guess, which is the canvas and where is the beginning of all of that? You know, It's fascinating. It I think it was one of the impressionists that he dreams his painting and he brings the painting to life. Each artist has a different style and interpretation from Claude Monet and Impressionism which uses usage of light, His amazing interpretation of that Jason Pollock abstract, you know, sort of expressionistic action art. Uh and then you have Pablo Picasso, which is very surrealistic. So every interpretation is very unique. And those personas of the end users are really kind of driving some of these sort of ai goals and uh objectives. And you don't know until you put it out there because the end user will find a different way of interpreting that piece of art or that data to do what they want to do. And it's a series of objectives now of the power to to capture that synthesize that and then deliver something iterative to take that customer down the path of the journey. so Jared drive the home drive home. The point for us when, when we look at the power of two with Lenovo and NVIDIA, what's possible then for for our listeners? Oh, that's a that's a big question. What is possible with the power of AI? Right. If we look at the implications of AI inside of the business, right? They have been doing pretty much if then statements, very basic programming language, right? If then if this happens then that very, very simple competitions at that point, when we look at the power of AI, you can put 100 different ifs and then have 100 different things and let the AI really figure out what the right then is right, Which is one is the right one that we're looking at getting that defined. Like we said, there's three people in the boxes. Three units in the boxes. Three decision makers in the box, understanding what that box looks like on the outside, understanding the shipping labels on that box, helping understand where that box is going, how...

...you're going to draw the box, Is your medium charcoal in canvas or are you? You know, like I think was Michelangelo? Who said, I could be wrong here. I just carve somebody. I just knock away the bits that weren't that weren't the object that was in the objects already hidden inside of there, I just chipped away the bits that weren't o or simplify it. Absolutely. Because I can't trust stick figures but that is where the power of two is. We help you with that block of concrete and figure out that person, you know that person that's going to be in there, right. The business units know what's there, the data scientists are going to know, I've talked to all three of the people inside of those persona world and they each have the same objective as the power of two comes together. Lenovo and NVIDIA help customers bridge that. We both talk the logical flow of data. We follow the data From data to decisions on the logical side, we follow it from the, from the technology side, from the, from the physical side, from the edge to the data sent from edge to core, you can hear that term a lot but from the business side we figure out the value prop right? We figure out what it is. I talk to customers like, oh well we only save 2% a year, it's not worth doing it. 2% of their whole overall bottom line could be a couple of billion dollars of benefit, right? We in video Lenovo have the expertise to help them figure out to help them bring that triumphant of success together to make sure that there are going to be able to do that to reach the solutions that their companies are going to need to go forward and just like virtualization and cloud was, you know, 10 and 20 years ago every company is being asked what is your Ai solution? The ones without an Ai vision, they're the ones that are going to fail and I've seen so many companies get in their own way trying to figure out what it is right? And then by the time they adopted they're already behind on the curve. So Lenovo and video together we can get them on the curve, get them in front of their business. And really as our Ceo Jensen huang said if data is your currency, how are you going to invest it? And I love that the data aspect 75%. If you're going to remember one number out of this, 75% of data is going to be generated at the edge. What do you do with it? What's the context was the sin text? What's the culture of the person that is experiencing that data or generating that data you know Jared you talked about ai coming up with the right answer. There's always the challenge of ai in the hands of people who may not have the relevance or the insights to different nuances which begs the question of bias. What's your take on that Jared? So when we look at bias there are very interesting ways that people do that confirmation bias I think is one of the worst. When I was reading a study this morning about ai studying genomics. We have a talk, sorry we have our session coming up soon. Call Gtcr global Gpu technology conference and in there we have brain date sprain dates are where you can sit there and talk to somebody. I'm doing one on the ai historical contexts in history. Um and how we're using AI to uncover new historical bits of information so that probably won't make the podcast. But anyway I just want to pitch myself there. One of the ones I was reading this morning was around ai researchers using genomics data to uncover new insights back into history. One of the problems we have in confirmation bias is our limited datasets. Looking at the world from the western point of view right? You know less living in America. We collect DNA data. America is a pretty good melting pot but we are missing some of that DNA data that's out there. So what they had done is they've gone through and they looked at a lot of our ancient ancestors that came out of africa our ancient ancestors that came out of Asia and they went back through and looked at historical anthropological signs,...

...the skeletons and stuff like that they discovered and they discovered a brand new human dead end of evolution that they said humans had broken off from our typical Neanderthal chain and had went off on this other path and they just didn't evolve. And what it does is that it explains that contextual data and confirmation bias. That if we just run these sets on just american data we would not have captured it because they pulled in european data and they pulled an african data which unfortunately there's not a lot out there for a genomics data yet and made a brand new discovery. The other side of this is a very interesting one when you start looking at broad historical context. Again, I love history, right? And they're sometimes you can reach a right conclusion by looking at the data but it's actually an incomplete conclusion. And one of my favorite ones is you don't invade Russia in winter right? Because napoleon failed the german army in World War two failed right. It makes sense. Russia is a really nasty place and win or whatever blah blah blah. But they miss the point. And that point is is that genghis khan or Prester john or whichever one, whichever way you want to call him, invaded from the east and you can conquer Russia from the east because the way the geography is they missed the data point that it's not necessarily invading Russia in the winter is that you should invade Russia from europe because it funnels you in and Russia can trade land for time. But doing it from the east and that's an interesting point of contextual data that they have gone back through and analyze and said yeah it is purely possible. I'm not sure that's a good example. But you know I like that. I think there are a lot of instances in military engagements that teaches us many different things to do the unexpected to do the left hook. In many cases there's a, there's an analogy that is used, which is to catch your opponent off guard with the left hook in boxing. And well as in the military coming from a advantage point of speed, uh, intent to drive chaos and disarray. There's some fascinating things where the left hook really comes out of nowhere and there are some great examples and I'll defer to Jared and from his uh, degree and thesis around military history. I think it's fascinating to use Ai to kind of run some different models on what would have happened if something else were sort of uh, integrated in as a different dataset. Yeah. And uh, going, going back to the left hook and there's always the mike, Tyson quote of uh, everyone has a plan until they get punched in the face. Right. That's exactly the same thing that we see with Ai a lot is that, oh, we're doing down this great path. Then all of a sudden they get punched in the face with a new data set and they start over from scratch. So, and that's why self driving cars have taken so long is because they realized originally they started, it's going to be a million hours. Then it was 10 million hours. Tesla is now saying a billion hours of self driving car data to get there. But yeah, when we look at historical context, right? Egyptian high rock cliffs, fortunately are fairly easy image recognitions. They're all relatively the same shape, but they evolved over, you know, Egypt uh empire or Egypt civilization range forms, you know, 3000 years using ai to figure out how the context of those higher glitz have changed over the years, based off back to what they know from the original translation thereof. They're using a eyes and this is one of my favorite ones. They're able to fly drones and fly planes over old battlefields. We're talking old battlefields like roman battlefields. Now one of the greatest things that humans ever figure out how to do during warfare is to dig a hole and hide yourself in it. So what they found out is that as these romans dug these trenches, it loosened up the soil well after the romans were done with it, all that soil and back into the farmer went back to plowing that soil is better air rated. So therefore the crops grow just slightly better. But because of this year's Yeah, so it starts to plowshares...

...exactly. So what they're able to do now using ai and imaging and these these these super high accuracy models is figure out just a minuscule difference of where grain is growing slightly better as they pull that out, they're able to see the zig zags of the roman trenches, They're finding new mayan temples deep in the jungle, where the jungle doesn't match up where it's supposed to be right. The signals in the noise, right? Or or you know, when you find this needle in a stack, a needle in a haystack, Typically you're trying to find a needle in a stack of needles. I mean, that is really how hard a lot of this data is to go off and try and find. I have a whole bunch of other, you know, weird historical ones that we can throw out. There was a It's difficult, it's not true. But back in the 60s, there was a one of the first giant supercomputers in the world was built in the basement of the pentagon. Uh, Robert McNamara was a huge data guy. He wanted to run the wars based off the data. And you guys experienced during World War II in Bomber Command. And she figured out that well, we're gonna run the Vietnam War based off of status based off the statistics, bullets, bombs, casualties, horrible, horrible things like that. And so the joke went, is that robert McNamara took all this data fed into computer, uh, 1967 and asked when we're going to win the war and it said you won in 1965 because the one thing, it could not context was human sentiment and that is one of the key things of capturing human data. Funny story here as well. Uh, I have a friend, great guy, roman historian as well. He loves his history. He does not like cheese. It's not lactose intolerant. He just doesn't like cheese, right? I'm like, I'm fine. I don't like asparagus, I'm good with that. But everybody likes cheese. That's my that's my opinion. But if somebody and here's the thing is when you break that down, you're like, you don't like cheese, you're weird. Well, I don't like all cheeses. I don't like stinky feet blue cheese, right? You know, some people do, Some people don't. So when you break down cheese, there's a lot of data points inside of cheese that can make it interesting. But now is there an association of his dislike of cheese to something else that he does like? And that to me would be an interesting thing to figure out is because you don't like cheese. Let's go figure out. You know, because he also likes dr pepper. Okay, that's a good one. He also likes tequila, but he doesn't like tequila and dr pepper. I don't think that's a real drink. He likes steak, right? So there's all sorts of things. They're of human context that there's no scientific reason. Well, at least that we know he doesn't like cheese, but there's so many other data points in there that really make you go figure out that what is the association between these and how can we better target them. Right? And since then I've started asking people use that as a reference. There are actually decent amount of people out there that just don't like cheese. It's just that. And I thought it was, I thought it was just one of these really weird things are the only person I ever met. And so then I started using as an example and I've met more people who just don't like cheese. And that to me is a fascinating data point because it's like, okay, since you don't like cheese, are there are things you don't like? Are there things you do? Like how can we associate that is one of those unique data points? Because it's not all that uncommon. Whereas me not liking asparagus, I don't see why anybody would like Asparagus. I see nobody likes asparagus. Um, but yeah, that's the point, right? Is our likes and our dislikes. That is really hard for technology to figure out how those data points play back in together. You know, McNamara asking the computer when we're going to win, he could not capture the fact that the North Vietnamese would never give up. Right. And that's why when we look at modeling of human data and the military aspects in history, that is something especially ancient history. We have absolutely no context on these stories. There's so many dead ends of history that we just give up on and we don't know yet because we don't have any primary sources sitting out there right now, I find it fascinating you brought up the NBA, the North Vietnamese army,...

...right? And what's fascinating about that is that they were fighting a war from an infinite point of view, not a finite point of view, right? McNamara had an endpoint, the N. V. A. They fought their battles. They've been fighting it for generations. So there's a it's a whole different mindset and what is the strategy and the syntax and then, you know, you know, McNamara was one of the original whiz kids from ford during World War Two. So they use all the numbers to kind of recalibrate and re tool, all of the capabilities to deliver in the supply chain. Using data to drive faster production at higher quality and at higher speeds. It's just an amazing set of let's say, dynamics that McNamara and his counterparts actually delivered. They took data right? And they really mastered it, but they didn't have the nuances, right? The sin text. The emotional the turning point of the Vietnam war was walter, concrete walter conch. I walter Cronkite, thank you. So walter basically said, the world of the war has changed and that changed the sentiment and that was enough to flip the bit and exactly. And it's really interesting because when you look at the context of war, human conflict in general right? There's many things that go on. But if you just focus on the beans and the bullets, getting the beans and bullets, the frontline soldiers. The other thing that they missed was the turning of the civil rights movement inside of America at the same time, the distrust of the war through walter not walter. Cronkite was responsible for it. But he really opened people's eyes. He told the truth about the war. That guy's just basically isn't winnable. The myth of the good war is gone. And the fact that the americans were starting really protest against it. That was not in any of their calculations because they couldn't get it. But it goes back to, what problem are you trying to solve? You know, and and the the good old Hitchhiker's guide to the galaxy answer life universe and everything is 42 because that's six times nine, which of course it isn't unless you use base 13 math. So it's very interesting. There is like when you try to find that problem you're solving, make sure you're solving the right problem, make sure you're actually truly do understand your problem. A lot of these early ai adoptions. Yeah, they're really straightforward. Um, you know, industrial inspection, you're trying to make sure your product comes down the product line that doesn't have a dent in it. Right, Those sort of things. We understand those problems. We try to solve when we get more into the human sentiment, the recommendation engines. What captures your interest that that is really where those data scientists really come into play, which is why they run off of mountain dew and coffee. So and I love that mountain dew coffee. It's sort of like the spectrum that you need to kind of you know the canvas, what is the canvas really sort of contain If you look back in the sixties right? It was Camelot and the counter was civil rights, how do you square that? Right? How do you go from the Beatles to Motown, all these different types of things when you pull them together, it's still music is still society. But when you have Camelot as one point of view and civil rights as the other, wow. Yeah. So the you know the the the challenge and the opportunity of a I coming to help us sort of navigate through this, understand it right and revel in the discoveries that come from it. I think they're gonna be amazing as we go forward. It's it's interesting because data is neither right nor wrong. It still just data. Right at the end of the day it's still zeros and ones you can skew data all you want that goes back to the confirmation bias right? If you only present a certain kind of data of course you're only going to get a certain kind of answer and that is where the data scientists fortunately can step in and say look we need more data sets. We need additional data sets. We need to look at it this through a...

...different lens of trying to solve the same problem and that's really where, you know, a lot of these rules and structures around AI needs to come into play to make sure we don't end up with, you know, with with biased a eyes or you know, they get back to the beginning of the podcast you said is, you know, the media is the media into art imitates life, you know, they do love a bad guy, right. AI becomes a great bad guy, But we don't really understand yet, you know, the goodness the day I really can bring in because Ai is powerful, we use ai in our daily lives now um which is amazing to think about it, but where it's going to be in 5 to 10 years, that's gonna really change our are the way we look at things. Well, a fascinating conversation today and I'm glad we're able to get a little bit off off the script and really kind of, you know, brainstorm this is something that to a I bots would not be able to deal in a podcast. That's funny. If you want to go watch the Youtube videos of two Ai bots talking to one another, they go off, they go off script a lot more than that and I can at least more contextual. Right? Right. Uh well, as we start to wrap up this episode, uh we always ask our guests, where do you see technology going in the next year, so, and I'm going to start with you. I think it's going to be um it's going to go in what I would call the E. R. S faster, deeper, wider and thunder. I think we're gonna see a lot of really interesting things in technology that we're starting just to discover now because of the power of two were able to do it you know in a in a manner where we can stand up systems faster. We can go deeper with analytics and we can go wider with all sorts of context. So as we do that it's going to get funner. I think that's where the market is going to be headed embrace ai because that's where the game's gonna be played. Isn't faster, deeper, wider fund or daft punk song. Think so that's a good one. Yeah harder. Better faster, stronger. Yes. Yes. Where is technology going in the next year versus the next five years is pretty interesting. As we come out of our covid bubble. What are we looking for? Out of technology? I think people are going to be looking at technology on how to re engage. I hope they're looking at to break their own tribalism. And I think we're going to see a lot more, oh how do I put this oversight into how that data has changed? There's a lot of bad data being shared and I think that we have to protect ourselves from it. Um I don't think most, I don't think most engagements are true to the fact that they present the right data. There's way way, way too much biased data that's going out there. And I hope to see more technology, more oversight for that in the next five years. You know I brought up self driving cars. I don't think we'll have fully autonomous personal vehicles. We'll definitely have point to point vehicles. Uh you know distribution system, there's a huge truck or problem and uh I'm sorry lorry driver problem in the U. K. Right now for delivering fuel. Well you typically fuels delivered from the pumping station, sorry from the distribution center out to the pumping station. They drive that same route once a week. You can program a truck to do that so easily and it gets out to the trucking station and does that. We're going to see more ai in our customer interactions with our facilities. You know there are power company or phone company, our bank, whatever they're still going to be that we're going to really see that take over. And it's only going to become better. The last one I really see is when we look at I. V. A. Which stands for intelligent video analytics. It's just a data stream when you look at a camera. Now it really is just a data stream. We're going to see that where it's going to really start optimizing our life. We're having really big supply problems right now. And a lot of that is is tracking where the product is tracking where it needs to go, how we get it off of the ship's most efficiently, right? That is exactly what Ai is optimized to do using those cameras to track the products to...

...understand where they're going is a relatively simple process for AI. Once you get the models trained and being able to implement that out into production, it will solve a lot of our supply chain problems. There's some that it's not going to solve. I'm not saying it's going to solve every problem, but there are definitely some out there that that it's going to going to help with. Yeah, I think the big search for toilet paper over the next six months is going to be quite the ordeal and you know, it's supply chain notwithstanding, it's it's taking the mindset and best practices of just in time, which which has been time and memorial for supply chain for decades. It's now just in case what is just in case really look like, right, how do you anticipate volume velocity and variety of different types of things in the shelf in the store, in the home, in the distribution centers. These are all really ai opportunities to really dig dig a lot deeper in regards to what are the art of, let's say the possible in regards to getting the right product to the right person at the right time at the same time a big challenge. I think Jared is p squared right privacy and personalization. I want it personalized but I want my privacy as well. Can that be accomplished with ai? That's a very important question. I always take the stance of, I am in control of my own data. I only share my data with personal people with only certain people I know where my data is that I only signed up for certain things. We represent a higher level knowledge of person and technology. I would not expect somebody who does not live inside the technology world to understand all the complex privacy concerns that are out there. They don't understand how their data can flow back in the organizations and between organizations and be shared etcetera etcetera. But then again, I talked to some millennial friends, I'm not picking on millennials, I'm just saying their generation younger than I am. They don't have concerns about privacy. You said I'm absolutely fine putting videos of myself on Tiktok doing silly things for free goods and services. So it is an interesting expectation out there. But companies such as in video Lenovo, we will always follow the privacy laws set forward by the governments to make sure that we handle that data correctly. And that is where I see the important happening as you see companies such as Lenovo and video following the right privacy rules to make sure let the people do the people stuff and they can protect themselves and we try to educate them as best as possible. We follow the right rules and we build the right ai models and we work with the right customers to make sure we solve the right problems. No, I like that. I think three in the box, the data scientists, the data infrastructure team and the b you really have to sit long and hard amongst themselves to understand what is not just the art the art of the possible, but the art of privacy, governance, you know, uh and resilience and risk as we look at the compliance requirements from countries to states, governments, uh and to even ourselves. So it's I think it's gonna be a fascinating journey. You know, the fastest growing job. One of the fastest growing jobs. The data scientist. Right? I mean they make six figures coming out of college now. Seven figures if you live in California slide me up one. I'm not that smart. One of the more interesting jobs we see out there is an ai ethicist or data ethicist, right, who is presenting the data ethics and they're going back to those three in a box. Right? Infrastructure, data scientists, business unit more and more business businesses are starting to hire data ethicists, right? How do we handle this? It's a cross between data science? It's a cross between ethics. It's a cross between privacy and legal. It's a very interesting position out there that more and more companies are starting to realize that they need this...

...and they need to take it, take ownership of these of these problems and I hope to see that extracts quickly expanded to talk about the business units, right, these business units trying to get the right results. How do we get there? Because if you take all things considered, if you throw an ai a historical problem, we need this product built. Why don't we go employ a bunch of 12 year olds to go build it. Why not? Their free labor? Right, But that is morally wrong. Right. And that is really where I start seeing a lot more data ethicists come into play as part of that important business motion as they monetize their data. Wow, that sounds like another episode to me Jared. Sign me up. Only if we can have it with dr Pepper and Tequila. There you go. There you go. Well, George Clooney to kind of come up with that company. Yeah, there you go. All right then. Well I surely appreciate all of your time and insights Jared and Ed, thank you so much for joining me. My pleasure shall be was blast as always. Shall we appreciate the time you gave me today and thank you listeners for tuning in and subscribing to B two B Tech talk with ingram Micro if you like this episode or have a question please join the discussion on twitter with the hashtag B two B tech talk. Until next time I'm Shelby scare talk. You've been listening to B two B tech Talk with ingram Micro. This episode was sponsored by ingram Micro's. Imagine Next B two B Tech Talk is a joint production with Sweet Fish Media and Anger Micro. To not miss an episode. Subscribe today to your favorite podcast platform. Mhm.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (395)