The Spirit of Possibility
A creator’s look at the breakthroughs that shaped how we log on, connect, and live. Gurdeep Pall, CVP at Microsoft, reflects on innovations at the forefront of networking, communications, and autonomous systems.
Episode Notes
Gurdeep Pall is a monumental builder whose constant curiosity, occasional audacity, and lifelong penchant for asking “what if” and “why?” positioned him as one of the key creators of and contributors to the most foundational Internet and communications technologies of the Third Industrial Revolution. Now in his 32nd year at Microsoft, Gurdeep has turned his attention to the scalable and ethical development of high-utility Autonomous Systems.
HIs conversation with Jesse touches on the sources of Gurdeep's fascination with technology, the inventive culture of Microsoft during the internet revolution, Gurdeep's role in inventing technologies central to today's digital economy like TCP/IP, VPN, and cloud-based communications, his personal reflections on working directly with Bill Gates, and a philosophy for moving from Comprehension to Creation that Gurdeep has applied in his stewardship of technologies and teams. The conversation closes on a discussion of the enormous potential of Gurdeep’s present-day work in AI.
Throughout, you’ll learn how Gurdeep brought together a lifelong love for learning, a disciplined technical expertise, and a deep interpersonal empathy to not only invent practical technologies - but to reimagine and help advance our world in the process.
Guest Bio
Gurdeep Singh Pall is the Corporate Vice President for Business AI at Microsoft and member of Technology & Research Leadership Team. He is an intrepreneur, a product thinker, and foremost a passionate technologist.
Gurdeep’s team of research scientists, engineers and business leaders is bringing digital transformation to business tasks through the power of AI, including a recent effort to train autonomous systems with reinforcement learning efficiently. Gurdeep is also responsible for Microsoft Garage and Hackathon, a hyper-scale, grassroots innovation program.
Teammates had the following to say about Gurdeep:
“A visionary builder of great products and businesses.”
“I was struck by his authenticity.”
“He never let hierarchy get in the way of engaging with us personally.”
“Gurdeep reaffirms for me that the role of a leader is to inspire.”
Helpful Links
Autonomous Systems: Gurdeep’s current world at Microsoft: https://innovation.microsoft.com/en-us/autonomous-systems
Gurdeep on the future of autonomy on the Microsoft Blog: https://blogs.microsoft.com/ai-for-business/autonomous-systems-vision/
A perspective on Gurdeep’s life published by The Global Sikh Trail: https://theglobalsikhtrail.com/story-english/gurdeep-singh-pall/
Gurdeep on LinkedIn: https://www.linkedin.com/in/gurdeep-pall-0aa639bb/
+ Episode Transcript
Gurdeep Pall [00:00:06] Very exciting times because you knew that if you could figure this out, you would make people's lives so much better, and that's probably been the most gratifying part of my networking journey that, you know, a lot of the code I wrote, my team wrote, it literally enabled billions and billions of people to get online and, you know, live their lives, work, study, learn, enjoy, whatever they do.
Jesse Purewal [00:00:49] From Qualtrics Industries, this is Breakthrough Builders, a series of conversations with people whose passions, perspectives, instincts and ideas fuel some of the world's most amazing products, brands and experiences.
Jesse Purewal [00:01:07] I'm Jesse Purewal, and on today's show, I'm proud to share my conversation with Gurdeep Paul Gurdeep is a monumental builder whose constant curiosity, occasional audacity and lifelong penchant for asking what if and why positioned him as one of the key creators of and contributors to the most foundational Internet and communications technologies of the third Industrial Revolution. Now, in his 32nd year at Microsoft, Gurdeep turned his attention to the effective and ethical development of high utility autonomous systems. Our conversation touches on the sources of deep fascination with technology, the inventive culture of Microsoft during the Internet revolution, Gurdeep's role in inventing technologies central to today's digital economy like TCP/IP, VPN and cloud based communications. His personal reflections on working directly with Bill Gates and a philosophy for moving from comprehension to creation that Gurdeep has applied in his stewardship of technologies and teams for three decades and counting. Enjoy Gurdeep Pall.
Jesse Purewal [00:02:17] I am here with Gurdeep Singh Paul. Gurdeep, it's an honor to have you. Thank you for being on the show,.
Gurdeep Pall [00:02:22] Jesse, thank you for having me. It's an honor for me as well.
Jesse Purewal [00:02:26] Gurdeep, you are something of a human force multiplier in the world of technology and innovation. You're also a practical visionary who has helped to lay the groundwork for some incredibly important infrastructure and software. But you're also a really devoted and lovely teammate. We'll certainly unpack much of that today and give you a chance to answer those charges as we get into the conversation. But I'd really like to start with a simple question, which is just, how do you frame who you are and your place in the world?
Gurdeep Pall [00:02:55] You know, I believe that everybody has a purpose and it's not always clear what the purpose is at the outset. But, you know, when you get to be sort of old and doddering like me, you know, some things start to become clear. And I think for me, my purpose is to really look at what emergent technology, emerging science in the world and to see how it can make all our lives better.
Jesse Purewal [00:03:27] And Gurdeep, where did you discover that you had either the initial penchant to help make lives better or for technology? Did those unfold in parallel, in sequence where they back and forth throughout your experience? Unpack that a little bit for me?
Gurdeep Pall [00:03:46] Yeah, you know, I have to say in this journey, you know, it's been very serendipitous. It's been accidental. It's been luck, it's been, I think it was also the time. As I went through my undergraduate work as I went through my graduate work, the thing that I had was this curiosity. I was like, you know, a kid in a candy store and I don't care what it was. Right. I remember in grad school, I was working on this project on distributed Operating Systems, and and it was you know, I was like, so enthusiastic because I said, if you could do this, we could do that, we could do this. And some of those ideas were crazy. And, you know, one would never do that. But it was just this the possibility that you could do those things. I remember the spark sort of going off in grad school. I think it just it happened to me largely because of the time that we were in when this whole computing and digital third industrial revolution was sort of taking off and the environment that was in, whether it was grad school or it was this place called Microsoft, which was sort of this hub of so much of the stuff happening around me in.
Jesse Purewal [00:05:03] Gurdeep, where did you hone or where did you discover the curiosity as an essential attribute to who you are and who you would grow up to be?
Gurdeep Pall [00:05:12] Yeah, I think the the one moment I remember was the first computer science classes I took in my undergrad school. We were, you know, sort of taking on, like algorithm's introduction to algorithms sort of for the first time. And the professor was asking us if you had to do this, what would you do? And I it was amazing. I went into the zone where everything else disappeared around me and my brain was engaging on what this professor was asking. And it was such a simple problem. But I think that was the moment I got hooked on the computer science. As a problem solving medium, and I remember I came out of that and I was like literally like what just happened to me. I think that was probably the first moment I remember. But I have to say, you know you know, when I came to the US coming to grad school here, I was just blown away by the opportunity and the resources and everything that was happening around me. So I was like this, you know, fresh off the boat kid with bright eyes, just getting excited about everything. And in some ways, I still am that, that guy.
Jesse Purewal [00:06:25] I want to ask you about when you were getting ready to start your career at Microsoft, it was an an intense and far reaching kind of time of change in the world. What was your broader vantage point about what was happening and maybe where you fit in to some of the the change that was occurring?
Gurdeep Pall [00:06:47] The thing that was happening was most amazing was the digital industrial revolution. I mean, I was like in the middle of this river, fast flowing river, and I was in for a joyride. And I don't know why I was there, but I was there. And incredible stuff was happening around. I mean, this is the time when in the world around us, the Mark operating system was being built. This is what eventually all of iOS was built on Mark. The PC, the Windows 3.0 was imminent. Processors were getting much faster. You know, Sun Microsystems was putting out workstations, which are these incredible things, this company called Silicon Graphics. And I was right in the middle of that. And it was just pulling you in to engage and to do something. So it was at some level, it was this unbridled excitement and enthusiasm. But also I think there was a responsibility that we had to do something with this. And I found that at Microsoft.
Jesse Purewal [00:07:52] What were the variables in the ecosystem that had to come together for you to be able to come in and be a participant in and eventually a leader in this technology revolution?
Gurdeep Pall [00:08:04] Yeah, you know, the most I would say the fundamental one for me was, you know, having this agency to do something, to speak your mind, to share ideas. And to me, there was a moment I remember when I started in grad school, the first class was a class on computer theory, and it was taught by this professor is incredible mathematician. These students were engaging the professors, challenging them. The professors were challenging the students. And I can actually say something that is crazy. And then the professor will actually take time to either prove or disprove it or to get me to develop it into a better idea. Like that was a breakthrough moment for me, because I, I had not seen that in the of course, not only going to college, but I would say in the macro culture where I grew up and even looking at my parents and extended system around us, it was very, very there's a very clear demarcation between what you're supposed to do, what you question, what you don't question.
Gurdeep Pall [00:09:19] Software software is the most incredible medium. Humans have never experienced a medium where you could create so much without having to invent new materials, without having to build machines that are just the pace of innovation in software is like nothing else.
Gurdeep Pall [00:09:43] And then thirdly, I would say somehow the high tech industry got permission to operate differently. You know, take a look at Facebook, for example, like for the first ten years, like they never made any money and they just raised tons of money and they burned tons of money. But the industry gave them permission to do that before they started creating gobs of money. Microsoft was like that. Microsoft was one of the first companies to actually experience that. Microsoft had created this way of operating where innovation was premium, things could be tried very fast. You could throw many things at the wall at relatively low, you know, pretty efficiently, and you could see some of them succeed. And I think that was the another very big element of how all this sort of happened.
Jesse Purewal [00:10:34] And what kinds of things do you think set the conditions for Microsoft at that time to show up as a little bit of a rebel in the capitalist architecture?
Gurdeep Pall [00:10:43] You know, Jesse, great question, right. I've come to realize that there are two kind of things in the world. There are things that exist before. And if those things exist before, you have to leverage the learnings and the lessons and you have to build upon all that and evolve it further. And then there are things which just don't exist before. I think what Microsoft stumbled into, none of this thing existed before, and if if you ever find yourself in that place, which didn't exist before, the worst thing you can do is to try to lean into lessons or something else and constrain yourself in this new place. It's a new way to think about innovation, what you create, how you create, how you operate, what culture, all those things. They create a new and I think Bill is a culprit zero in everything that happened here. And we've seen 35, 40, 40 years later what it is.
Jesse Purewal [00:11:49] Culprit zero. I love that. Gurdeep, I want to ask you about some of the specific innovations that you've got the opportunity to drive or to help steward as you onboarded into Microsoft. I want to start with TCP/IP, one of the early opportunities you had at Microsoft. Now, TCP/IP had been specified over a few years by researchers and by academics. It eventually became a standard. To get to a standard. You have to have not only scale, which you did, given the number of PCs running the Windows OS at that time, but also the technology that you're building truly has to be breakthrough. Can you talk about the handshake between the scale that Windows as an OS provided you and just the specific wonder and innovation in this thing of Internet protocol that you drove?
Gurdeep Pall [00:12:43] Yeah, yeah. You know, it was a very interesting time. You know, just to take us back in the early 90s, there was a company called Novell, and Novell was the early leader in networking. And what they had, they had a protocol called IPX and they were pushing that they were getting a lot of early enterprise traction. In fact, Microsoft was trying to play catch up with them at the time. So that is one important piece of context. And the other piece of context is I had been working on a remote access work at Microsoft, first in a product called LAN Manager, which was our competitor for Novell's NetWare. And the first technology we worked on was remote access where you could dial up over a modem. So a lot was happening in this connectivity space. And, you know, people were trying to figure out how would this enterprise IPX NetWare Novell thing would resolve with this crazy thing that was happening outside of the Internet. There was this big gap between the enterprise world where all the money was and this whole Internet world where there was not that much money. It was sort of the coalition of the willing and and this gap existed.
Gurdeep Pall [00:14:07] Something interesting happened even within Microsoft. Just to give you a sense, there were two camps. There was a set of people who said we should keep pushing IPX, we should embrace it and we should really make it happen. And we should ride the wave and we'll work with Novell and so on. And there was another school who said, no, I think we should bet on on TCP/IP. And that was, by the way, I would say the weaker school. This group had less agency within Microsoft. And my manager of course comes to me and says, hey Gurdeep, I want you to work on this TCP/IP stuff for, know, remote access. And I you know, my my face fell. Obviously, the cool kids got to work on IPX. I got stuck with TCP/IP because, you know, hey, that was clearly the right direction for me career wise. And I remember Henry Sanders was one of the core architects for this work and I worked with him and a couple of people. It was just like four or five of us. We built the TCP/IP stack. Now this shipped actually shipped first NT 3.51, but NT at that time was a very small operating system. And then, of course, we built into Windows 95. You see prior to that, if you were on Windows you had to go buy a TCP/IP stack from somewhere and you would install it on your machine. And it was just painful and a lot of missing pieces and you couldn't debug it. And you know, if you talking to a Mac it would not work or a Unix. So we shipped this standard TCP/IP stack in Windows.
Gurdeep Pall [00:15:36] It was late 1996. We were visiting. I think this is the Lawrence Livermore Labs right above California, Berkeley, and there was a gentleman there called Van Jacobson. Van Jacobson was one of the coauthors of the Congestion Control Protocol. And it is a piece of work. I mean, the whole world is running on that stuff now, right? I mean, he was a big mind. He was a scientist. And so we went and had a chat with him. You know, he was very complimentary about our TCP/IP work. At that time, I was the development manager for all the networking stuff, and so it was great to hear that from him. But he said, you know, people don't realize what role you played in making Internet happen. So he he shared two charts with us. The first chart was the number of machines directly connected to the Internet. And the second chart was the packet size of of packets on the Internet. These are the communication packets that are going back and forth. What turns out, we shipped a specific protocol inside the TCP/IP Stack, called Path MTU Discovery. MTU is the Maximum Transmit Unit. That means the what is the biggest packet you can send between these two endpoints? And if you send out a big packet and if a network segment could carry it, it would fragment it. It would break it up into smaller and then would reassemble on the endpoint. We added this protocol, which for the first time and we were the first stack to have it, we would negotiate end to end and we started sending 1500 byte packets prior to that to avoid fragmentation. Every other TCP/IP stack used to have 576 byte packets. He showed me that the Internet traffic had shifted significantly towards 1500 byte packets and that is conclusive proof that on one hand you see the explosion of the Internet in terms of number of endpoints connected. And the second was that it was because of the TCP/IP stack we shipped in Windows 95. And that was a pretty humbling moment. And at that point TCP/IP was mainstream, Windows 95 was mainstream and we were just seeing an explosion happen right in front of us.
Jesse Purewal [00:17:59] And Gurdeep, to to what did you then ascribe the success of TCP/IP, which by rights at the time sounded like it was characterized as an underdog. Somehow you you turn the narrative around. How how how did you make that happen?
Gurdeep Pall [00:18:15] Yeah, I mean, I want to be very clear. I mean, everything about the Internet, the credit belongs to IETF. It belongs to all the people who worked on the IETF protocols, way more than anything any one of us did at Microsoft, we delivered a great stack, which was an alternative to the IPX stack. We didn't go and say, we're going to do our own crazy thing with TCP/IP. We were true to the standards. In fact, we added something helped push some parts of the standards, which I think completed the story, but that's about the extent of it. I would take credit we should take for what the Internet is and the places where we extended it is. I'll give you you've probably heard of something called DHCP Dynamic Host Configuration Protocol. Prior to that, you had to enter in an IP address. Only then your machine could be on the Internet. We said this is crazy. Why can't the machine just request an address when it boots up and get it itself and then get going? So this was like the you know, it was a usability thing. And I remember I was sitting in one of the IETF meetings and there was a gentleman, brilliant, brilliant man. He worked at Berkeley, I think, for the Unix BSD 4.4. And he told me, why do you need DHCP or this ability to get an address and use it? He says, You know what I do? It's very simple. I just open up this file. I type in the IP address, I rebuild the kernel and my machine is on the Internet. And I'm like, no, you know what? People sitting at home who are not, technologists are never going to do that. There were moments like that where we you know, we filled in some gaps. But I would say we really lived off the bounty of this incredible work done by academics and researchers and some of the early technologists.
Jesse Purewal [00:20:04] So let me take you to 2001, Gurdeep, because this hyper growth of the Internet would continue to be something that you and the team at Microsoft would play a role in with the release of XP when Windows XP, Talk about what that experience was like and trailblazing not just the OS technology, but being that the first OS that would really start to connect people in the citizenry beyond academia and inside the upper echelons of the private sector.
Gurdeep Pall [00:20:37] Yeah, you know, one thing which has been really interesting to me is. It wasn't again clear when we were going through it, but looking back at it is that I felt that, you know, we had to cross this massive ocean, but whenever we wanted to take the next step, a rock would come under our foot and we would step on it. And then we would take the left foot up and another rock would emerge. And we crossed this ocean. Like it started with the remote access. The fact that you could dial up with a modem, you know, hey, I can dial over the modem now. The next thing we did, great, now it's broadband. It's a modem which always stays connected. But you're on the Internet. Well, how do I connect to my corporate network, which was separate from the Internet? Well, that's how VPNs emerged. And and then when VPN emerged, one of the things which happened is that we had to wire the entire operating system of the networking stack to deal with networks which came and went away. Because when you were connected, you had the network, when you were not connected VPN, you didn't have the network. When Windows 2000 shipped, it had this plug and play capability. So devices, network cards could come and go and you could have USB attach network cards, you could remove them and the network went away. So these all of these building blocks were already in place. And then we come into this technology called Wi-Fi. Now, this is again, Bill was very involved. He again, driven by this idea that how can I make computing more accessible to people like he was all over it and he was just beating us up left, right and center, like, when are you shipping this thing? And we finally said, OK, Windows XP is is the next big milestone. And this was a tough project because the Wi-Fi standards themselves were not actually finalized. So you would get one chipset, which would improve the standard in one way. Another would interpret another way you would have chipsets would fail the network drivers. What are the right network driver models to have for this?
Gurdeep Pall [00:22:49] All this was being figured out and we got it across the door and we shipped Windows XP with it. And the rest is history. Again, very exciting times, because you knew that if you could figure this out, you would make people's lives so much better, and that's probably been the most gratifying part of my networking journey that a lot of the code I wrote, my team wrote it literally enabled billions and billions of people to get online and, you know, live their lives, work, study, learn, enjoy, whatever they do, you know.
Jesse Purewal [00:23:30] What do you think it is within you that gives rise to being able to work your purpose of technology for bettering people's lives in that inventive sense? Why are you an inventor and a creator and a builder as opposed to someone who feels more hued to being a maintainer or a steward of something that maybe was built, that you just find excellence and continuing to push forward?
Gurdeep Pall [00:23:59] I would say it really starts with this a level of, you know, optimism, stupidity. I think those things are all kind of rolled into one because I remember one time my manager's manager, after a while, he realized that I was completely incorrigible and and I would say, yeah, we're going to go present this to Bill. And he would tell me, I hope you're bringing your own parachute. And this became his thing because he knew that I was kind of this crazy dude. And he he would not make those bets. He would not put himself out on a limb. You know, initially it was that, you know, I didn't care. It was exciting. I was going to go do it. But later on, I realized that, you know, frankly, Microsoft and the culture that Bill had created here, was actually going to reward that behavior. Now, I don't want to say that it was completely random, completely crazy. It was informed and it was an informed push into a direction. And and I realized that that was getting rewarded. Now, I would say I do think a lot before I pick, I'm still doing that, and that's the role I'm in right now. And even some of the you know, my interests outside of here are I'm pushing always, pushing the envelope, but I just tend to do a lot more research and think a lot harder now. In those days it was just jumping into it and then figuring it out.
Jesse Purewal [00:25:29] And Gurdee, how have you cultivated that kind of sense of whether it's optimism blended with with a certain brand of stupidity or or courage with the degree of foresight, how do you cultivate that in teams that work for you, that work with you as you've stepped into these successive generations of software innovations that you've been having a front row seat for?
Gurdeep Pall [00:25:57] I have this model back on it. The Four C's model, the first C is comprehend. You know, if you go into any area, any device, anything you're trying to do, you're to understand what is happening there, understand as well as you can. The next level is Critique. You should be to look at what works, what doesn't work of multiple things which are in that space which works better than the other, and why and so on and so forth. Once you understand that, then you Create. And Creation is like, OK, now how would I would do this? But there is one more level. It's called Captivate. And very few people, very few creators are able to Captivate. You know, and we know that. People like Steve Jobs, I mean, people like Stewart Butterfield, I think it is the artists who kind of their quest is to make it past Create and make it to Captivate.
Jesse Purewal [00:26:50] Gurdeep, I recognized we could probably do an entire episode of this podcast just on this question, but I'll ask it anyway. What was it like? What is it like, to work with Bill Gates?
Gurdeep Pall [00:27:02] He was a force to be feared. I mean, we all had stories of Bill just chewing you out. If he didn't agree with you or if you said something crazy. I was pretty fortunate that, you know, I got chewed out a few times, but it was a few and far between. What I got to experience with Bill was really learning about innovation, learning about really operating with confidence in uncharted waters. And the big thing I learned from him was big picture. And this is an area where I feel much smarter now than I was twenty five years ago. You have to keep peeling back the onion and say, what is the picture? OK, what if. Why? What if. Why? What if. And you keep pulling it back and eventually you get the purpose of what you're trying to do, you get the right scope in which to think about it, he forced you to think about things at that level. And I think a lot of people don't appreciate that about him. He totally got it. He connected the dots. He was able to see the big picture. And when you had the good fortune to work with him on the early days of communications work, I mean, once a week I would be sitting with Bill. He was pushing us. But he was an expert at the bigger picture.
Jesse Purewal [00:28:35] So then extend for me up the stack, maybe as you went from the networking software side into communication and collaboration. Talk a little bit about how you got attracted to to that part of the stack and and what some of the early visions you had were for the interconnectedness not just of data and information, but of the way that people would fundamentally be able to augment and amplify the ways that they work together.
Gurdeep Pall [00:29:07] I was tapped to take on this area because Bill thought, rightly, that one of the hardest problems you would need to solve to make communications work on the Internet are actually networking problems. And he was absolutely right, like some of the hardest things we had to cross over was how do you make jitter-free audio work on a very, very stochastic Internet? And and so his instinct was right, but he didn't have to have me to do this. And maybe I displayed some other crazy attributes, which I didn't know at the time, but they tapped me. But it became clear that this was a completely different communication system to be built and the system did not work around phone numbers. This system worked around identities of people. So I remember creating the slide and presenting it to Bill in the middle was a person, the next layer around that was presence, which is is this person available or not available? Are they in a meeting are they busy, which nowadays you see in communication tools, the presence, away, online, busy, etc.. And then we said once you have this whole system in place, whether you're making a voice call or a video call or you're sending a chat message to somebody, all those are just modalities. Once you found the person, you establish a session and you say, I want to do the voice call or I want to do a video call or I want to do a chat call, I want to do a meeting. We created the slide. This slide existed before any product existed. And I remember doing this presentation to Bill and all his directs, on our communication strategy ought to be about this. And this was the it was a very technically articulated vision of what we wanted to build. And I tell you, I got so much support. You know, we got a big team. I got to work with some other smart people at the right time. We did the right acquisitions and that created Lync, Skype for business. And what is Teams today, while we were doing this, unknown to us in a different part of the world, there were a bunch of crazies like me. Who were dreaming of building Skype and there's a big lesson here, by the way. The lesson is that and this also I learned from Bill that if it is possible, if something starts to become possible and you don't move on it right away, you will miss it. Now, Microsoft was always sort of business and enterprise focused, more so than consumer focused, we focused on the enterprise side, Skype focused on the consumer side. We were able to leverage some constructs like the active directory, et cetera, inside the enterprise. They focused on other constructs. And in the same way we said we're going to build it as part of our Office suite and all that. That was our super power, their super power was we're going to go after telcos and these telcos are very, very regionally bound. And we're going to create a solution which works across all these telcos that you can call anybody and you don't you're not beholden to their, you know, business models and so on. And they created that amazing thing. And, you know, Skype before, of course, later point in my career, we ended up acquiring Skype. I led the acquisition and ran it for a while. And that journey was a really long, complex journey because when I was running Skype, it was at one point 40% of the world's international calling, traffic was going on Skype. And if you say I'm going to go and change the fundamental architecture of Skype and while all this is happening and you're also building your Skype for Business product or Lync and Skype product, it is this complexity was just incredible. And we started on this multi-year journey. And that journey is why we have Teams today. The entire backend is the convergence of Skype scale, and all the enterprise capabilities. And now we have this cloud running service, which works on mobile. It works on your laptop. It works for consumers. It works for Enterprise all on one service. And Zoom is incredible. And, you know, they that's a great success story. But there's only like, you know, I would say a couple, three of these plants that exist in the world which have the capability of, you know, that level of communications and reliability and scale and so on.
Jesse Purewal [00:33:47] It's truly remarkable to see the fruits of those kinds of labors. Certainly in the pandemic world, we would have been in a vastly different place, had some of those doors not been opened.
Gurdeep Pall [00:34:00] You know, I was talking to Satya a little less than a year ago. We would talk about the pandemic and we were the thing we were reflecting on is it's kind of another way of making the point that if something is possible, if you don't act on it, you know, you will miss it. I think that is also true here. Right. Like, if Microsoft had not scratched that itch that we need to bring communication, where would we be 20 years later when this world needed this thing? I mean, all the blood, sweat and tears and innovation and everything, it all got realized like last year. And, you know, we only saw glimpses of that when I was running Skype. I mean, there was some incredible stories. You know, in fact, we had a whole storytelling of Skype group where we had stories like this person, you know, meeting their child for the first time over Skype and, you know, the last words you share with a loved one on Skype. And there was this incredible stories. But the pandemic has just made it like a mission critical, the world cannot go on without it, kind of a thing.
Jesse Purewal [00:35:17] So, Gurdeep, we've we've started our journey at the networking level of the tech stack. We've gone into communications. I'd now like to go closer to cognition and within the artificial intelligence domain, ask you what you see in the software world that's changing and is dynamic and wonderful and interesting ways that you think will generate some really wide open spaces of new opportunity for the economy, for companies, for individuals going forward.
Gurdeep Pall [00:35:50] You know, when I was in grad school, we were still in the old world of AI which was very rules-based and and then, you know, we sort of all experience this thing called "AI-winter" where everybody lost faith in AI and and and a lot of the funding dried up, and so even professors were so are working on it. I moved on to other things and so on. And then something pretty major happened. It stopped happening, I would say, in the 2000s where this machine learning started to take off. And, you know, I do make a bit of a distinction in AI machine learning, but machine learning basically was that we suddenly had a lot of access to digitized data. Now, rules worked when he had no data, so you had to envision, codify, how you believe the system should work through rules and rules based systems. When you had data, people said, you know what, we have these algorithms which mathematicians have been working for a long time to find patterns and things here. If you have a lot of digitized data, you don't let me just pull out whatever algorithm and I'm going to throw it at this data and I'm going to show some results.
Gurdeep Pall [00:37:04] You know, we won't have Google search and Bing search and all that if ML had not been viable and had real results. But then something amazing happened in the late 2000s, this professor who had been working on deep learning, he showed some results and in using deep learning or deep neural networks, and he shook the whole community, which had been sort of plodding along with old school AI or ML and so on and that totally changed the direction of AI, and it made many things possible that were just not possible before, and many of these benchmarks like, you know, speech recognition better than a human, translation, better than a human, object recognition, better than a human. All these benchmarks started to sort of fall. And we are living in the big moment of deep learning and what is possible. And we are finally now aspiring to take on tasks that previously were considered not possible. And I think we are really in the midst of what I call the the Fourth Industrial Revolution, which is going to be driven by AI and it is going to fundamentally change the world around us.
Jesse Purewal [00:38:24] And what excites you the most about the applications of these general-purpose or specified AI's? I know there is a lot that you're personally working on. Maybe some of those are nearest and dearest to you, but perhaps you also have a broader commentary on what excites you the most, say over the next decade or so?
Gurdeep Pall [00:38:44] Yeah, I'm a sort of self-selected my main to an area right now, which I'm very excited about, which we broadly call Autonomous Systems. And I think there is a very practical value here which applies to everything around us. We have partitioned the world into dumb things and smart things, and we have put ourselves into smart things and everything around us into dumb things. Then you say, what is it about the smart thing which makes it smart? And why can't the dumb things be smarter? And the number one thing which pops is controlled, I now look at everything and say, why can't how can this thing be smarter? Like I came and sat on this table. Why did I have to pull the chair when I stood here, when the chair straightened itself and coming park itself behind me? OK, so what prevented the chair from doing that? And it is actually control. We have a partition of the world so that we make everything on as so we can control it. I believe we have gotten with AI to a place where the smarts can actually be put into the object itself, so it knows why it is there, how it is supposed to work and it gets the job done. Imagine everything, every switch, every air conditioned controller, every tap, every light bulb. Every machine, every computer, every camera, every cable, every desk, every thing actually has its smarts. It doesn't have to be infinitely smart. It would to solve math problems just has to straighten the chair and pull up. And so I think that is what is really exciting to me. I feel like the human race, it is a superpower like this would be the single biggest growth in our abilities if we were able to make the world around us smart so that we could self actualize a way into other things like writing, poetry or painting or other creative things or creating other new science and so on.
Jesse Purewal [00:40:52] It strikes me that there is a mildly dystopian version of your Four C's that has the same first three Comprehend, Critique and Create, but that instead of going into the Captivate lane, there are those who would try to wrest Control as their fourth using either behavioral data or sentiment data or some amalgamated view of a person, a group of people or even a society to say now this is going to be the thing that must occur next. And sometimes we know what that thing is and sometimes we don't. Sometimes it's a conscious thing that's playing out, sometimes it isn't. As we're learning with the way that social media technologies and other even hardware based technologies can can have an influence on the way that that behavior shakes out. So I think you're right to think about this in a macro sense. What do you advise people who are thinking about this problem outside the technology realm to consider, as some of the must do's given to your point from earlier, if something's going to happen, it inevitably will. So we're not really guessing about what we might be guessing about when, but we've got to think of how we're going to wrestle with this.
Gurdeep Pall [00:42:05] Yeah, I mean, really well-framed. I mean, now when you do something, you have to look at the bigger picture. You know, the fact that we we have this whole effort on ethics of A.I. and it is not fluffware, it is so critical. I in my role, for the first time in my technology career, I had an ethicist and an economist reporting to me. Why? Because what we were doing had implications in those dimensions, and we ran to them like literally like we're not even out of the door and we ran into them. I'll give you one example. We were building this ambient computing solution where we could observe it, sensors, you could observe a space and then you could do all kinds of things with it. It is amazing project. We said, hey, here's a simple idea. Why do people have to swipe cards, let people walk in. With sensors that can observe a space and I know it's Joe and it's Mike constituents, Jill and OK, and I know that all this person is actually not a Microsoft employee. So we beep red flash and they can be whatever mechanism to get them to register as a guest or whatever. And that we ran into like, right. This brilliant idea ran into number one issue. H.R. told us the employee agreement that is signed does not allow us to video employees. So we said, well, then what do we do? So we worked with the lawyers, worked with H.R., and we said, well, you could have an opt out. No, you could have an opt in model for this. But then on top of that, we had to make sure this was fair. Can you imagine, like, if we were not able to recognize somebody in a wheelchair, walked in, drove in? I mean, that would be brutal, right, I mean, or somebody with some kind of skin or gender or age or, you know, some malformation, you know, whatever may have happened and it would be brutal, so suddenly, we have to think about it very differently than this. Oh, let me just go quickly build this out. So I think thinking through and approaching it wholistically is the best way that we can all make progress here.
Jesse Purewal [00:44:29] Gurdeep. I want to go from the future back to the past, specifically back to the day in 1987 when you flew from India to L.A., connected up to Portland on your way to Eugene and the University of Oregon. If today you could write the Gurdeep Pall of 1987 a letter and stick it in his luggage and have him discover it and read it on that flight, tell yourself anything at all, with the benefit of nearly 35 years of perspective as a leader and as a builder and as a human being, what would you tell him?
Gurdeep Pall [00:45:03] Don't hesitate in taking risks. Do the most that you can in the space. Do not have your own things, hold you back. Which luckily I didn't have too many, but I, you know, I still could have done better and, seek environments, seek people where you can create the most, where you are encouraged, where you are supported in in this process of creation. This is got to be one of the most amazing gifts we have as humans. The second thing I would say is that I learned this from my parents and my grandparents. That people matter. People you work with matter. How people feel, matter, and you will never go wrong in treating people well. I've had relationships at work with people who came back to work with me again and again and again, and these are super smart people, smarter than me. And because of that, I've been able to do so much more myself. That is a lesson I would tell myself any day, even do more if you can. But those those are two things I would say,.
Jesse Purewal [00:46:26] Well, Gurdeep, it's been a privilege. Thank you for all that you've done to make our technologies in this world more powerful and our communities more connected and our experiences with one another ever more delightful and meaningful. So appreciated and can't thank you enough for being with me today and sharing your story with our our listeners. I think we'll all be better off for it.
Gurdeep Pall [00:46:48] Jesse, thank you. You know, I think this was a wonderful conversation. Your questions were great, and I hope that if somebody takes the time to hear this podcast, they don't see this as a thing where, you know, I'm really talking about myself. I've been very fortunate and lucky. And I did happen to get on this path, which has given me these experiences. And I'm sharing them with the spirit that others, whatever they can get out of it and do great things, bigger things with amazing things. So thank you, Jesse.
Jesse Purewal [00:47:36] Thanks so much for listening to Breakthrough Builders. You can subscribe to the show wherever you get your podcasts. If you enjoyed the show, I'd be grateful if you could spread the word by leaving a rating and a review. It really does help other listeners find us. And please tell your friends, Breakthrough Builders is a production of the industry's team at Qualtrics. The show is written and hosted by me Jesse Purewal. Mastering by Nate Crenshaw. Post-production and music by Clean Cuts Audio, part of the Three Seas Collective. Design by Baron Santiago and Vansuka Chindavijak. Website by Gregory Hedon and Photography by Christy Hemm Klok. Special thanks to the entire Breakthrough Builders crew at Qualtrics, including Ali Rohani, Jeremy Smith, John Johnson. And Kylan Lundeen.