– [Patrick] So, we talked a little bit about Metcalfe’s Law. What are some of the other laws that govern connectivity?
– [Robert] Well, I think of two right away there’s Moore’s Law, which is very famous and Moore’s Law both makes it easier to implement networks, but also generates demand for networks, it makes processors much faster and able to process much more data more quickly. Then there’s Shannon’s Law, which came out of the professor Claude Shannon, MIT professor at Bell Labs researcher, which measured the bit rate through a noisy channel basically. And while we were hoping to use AT&T’s cable network, their copper that they’ve been installing for 100 years, to use it to carry our data between buildings, AT&T informed us that that copper was limited in bandwidth, according to Shannon’s Law, it could only carry 14.4 kilobits per second. So, any hope of going faster than that would be a violation of Shannon’s Law. Not too long after that, a company started selling 50–kilobit–per–second modems, which you’ll notice 50 kilobits per second is quite a bit higher than 14.4 kilobits per second.
– [Robert] So, what happened to Shannon’s Law and what happened is the telephone company over its 100 years learned how to install copper better and better. So, the capability of the copper network was beyond what they thought it was. And the internet people immediately exploited that with 50–kilobit–per–second modems, which teaches you that you have to be careful with these laws that they probably are very useful, but they also have a limited range, and you have to be careful not to overuse them. And I’m sure that goes for Metcalfe’s Law.
– [Patrick] Yeah, it’s interesting when reception sensitivity and error correction really allowed the reuse of copper in a new way. And I can remember, I think we all remember listening to modems connect and going through their tests to try to uplevel the comms and multichannel. And now with DaaS’s, what 3.1 here at the house, a gig down and half a gig up. And that just seems really hard to believe over copper that’s buried in the backyard for 40 years. But I want to go back to something that you said earlier talking about famous characters or famous contributors in innovation. You talked about Bill Gates and especially Steve Jobs. And not everyone can create, but we certainly all consume and utilize, we tend to notice those standards that have lasting effects on tech. So why do you think we’re so fascinated by the creator, what is created and really invention as a standalone moment.
– [Robert] So, you’re asking a why are we so fascinated by it?
– [Patrick] Why do we remain fascinated by some of the creators, the individual people, or maybe famous people who are proponents for technology or drive it forward?
– [Robert] Okay. So, the big goals are freedom and prosperity, and they form, those two form a virtuous circle. And the thing that drives that circle is innovation so prosperity through innovation drives freedom and freedom through innovation drives prosperity. So, this is important stuff we’re talking about, we’re talking about freedom and prosperity and innovation is the principal driver of freedom and prosperity. So, we want to know how to do more of it and do better of it because it will improve our lives greatly. So, then there are people like Gates and Jobs who played a pivotal role in major innovations. We want to study them to figure out what they did, how they did it to advance innovation because we want more of it, we need much more of it. So, I think that’s the answer it has to do with the importance of innovation is so high that the people who do it well, and currently it’s Elon Musk is our champion of innovation these days. We want to know everything about Elon. So maybe we can all be little Elons and thereby advance freedom and prosperity. I think it’s as simple as that.
– [Patrick] So, we’re all seeking education about innovation. It’s really macro.
– [Robert] Yeah. Like how do they do that? I want to do that too, or no, some of it is just curiosity about their quirks, and fame gets involved and fame has a sort of a distorting factor, but no, I think deep down, we want to know how they did it because we want to do it too because it’s so important.
– [Patrick] So, we don’t have $7,000 laser printers anymore. Maybe the ink, but not the printer. The internet has extended connectivity and access in a way that really would have been impossible, maybe hard to imagine. Is it in its final form? Is the internet done or as you mentioned earlier Elon, like we’ve watched on stage with the prototype of some of the brain implant, is that where this ultimately goes or does the internet end up being reinvented at some point?
– [Robert] Well, you’ve asked a very important question to which there’s a very clear answer. The internet is not done yet. You could argue it will never be done, but it’s certainly not done yet. And so, for example, in 50 years, the internet has managed to reach 55% of the human race, 4.7 billion people in only 50 years, really sudden adoption, but I’ve just told you about reach how many people were reached. So that’s a dimension of the connectivity provided by the internet, reach. Then we’ve talked about speed, well, speed is another dimension. We used to reach them at 300 baud, whoever we reached, we almost reached nobody, but the ones we reached, we reached at 300 baud. Now we’re beginning to reach them at a gigabit per second, which is quite a bit faster.
So that’s a dimension of the network and then there’s latency, how long does it take to do a round trip? And that’s key in this video conferencing we’re doing is very latency dependent. If the latency were much higher, it would be unusable, but thank goodness, it’s not higher and here we are talking long distances with no discernible latency. So, latency is also a dimension of this connectivity, but the list of dimensions goes on and on and on like the security, some networks are very secure and others are not. Cost, the cost of internet connection, cost alone is very important dimension. So, my point is, we’re proceeding down all of these dimensions in parallel at different rates. And there’s a lot to go and you can see it now, when the internet started, it was a kilobit internet. The transcontinental trunks of the original internet ran at 50 kilobits per second.
And then through the 80s and 90s, we enjoyed the megabit internet. And now we are going over to the gigabit internet and it’s going to create a lot of opportunities for new unanticipated, surprising applications as it has in previous advances. Also, there’s been progress in the arrival of new models of networking. Right now, we’re networking people, 4.7 billion of us, but there are 8 billion so-called things on the internet already. So now the internet is going from… Today’s internet is video, by the way, most of the traffic on the internet is video. So, we have a video megabit internet, we’re now going to a gigabit internet, that’ll carry video, of course, but now it’s going to start carrying the traffic among so-called things. We call them things because we don’t know what they are yet, but as we discover what they are, we’ll use their real names. So, the gigabit Internet of things is the next phase. So, is the internet done? No, it’s not nearly done because think of all the headroom we have and all those dimensions costs for example, lot of headroom there.
– [Patrick] Okay. So, here’s a question about the scope of the internet. I think a lot of folks think about, let’s say you’re at home, you got a decent cable connection that’s way faster than it was even five years ago. You’re working online with this fantastic application at your bank that’s way more responsive and it’s mobile compliant and it just seems great. But behind that, there’s a lot of technology that just doesn’t change. We saw that situation during COVID where the state of New Jersey was desperately seeking COBOL programmers because they needed to make changes to their benefits systems in order to get support out there. So, are some of those technologies like COBOL just forever and that’s okay?
– [Robert] I wouldn’t say forever, but for a very long time, the RS-232 standard is still with us, and it’s been, I’ve forgotten now, but decades of RS-232 hanging on there. Well, just look at Ethernet, Ethernet was invented in 1973. Now it’s evolved considerably, but there’s still identifiable, like the Ethernet packet format is pretty constant over that entire period. I wouldn’t say forever, but I would say for startlingly long periods of time, things persist like COBOL, I remember when COBOL was new. So how about all the Ethernet cables that are holding together all those buildings out there, how long will those cables, I guess they’ll be there as long as the buildings will be there, which is measured in decades. Do buildings last forever? I’m not sure.
– [Patrick] Well, yeah, most don’t.
– [Robert] Well, the ones held together by Ethernet cable, those are going to last forever.
– [Patrick] So, Ethernet will outlast COBOL then.
– [Robert] Yeah, but it also started later too. So, COBOL had a head start.
– [Patrick] So, I was looking at an article the other day, this is a dev ops team and they have come up with a mechanism to wrap COBOL code into Kubernetes containers. So, there’s also a lot of reinvention around the platform and delivery of technology.
– [Robert] Well, in the network business, we call that encapsulation where we take an old protocol and we carry it in a new protocol or the reverse, we take a new protocol and we carry it in an old protocol. So, the old architecture of the AT&T systems called SONET, Synchronous Optical Networking. And so, for a long time, you be carrying internet packets on SONET links and vice versa. You be carrying what looked like SONET links on top of packet systems that’s called. And all of that you should think of it as aiding the transition by preserving the install base. You don’t have to throw it away right away; you can use it. It also accelerates the new stuff because it can just lay right on top of what’s already there. So, the internet currently runs on TCP IP, a protocol that was standardized in 19, I’ll be generous to say 1980. So, it’s been, so what’s that, 40 years? 40 years of TCP IP. And that’s not going away soon either, that’s going to be hanging around for another 40 years probably.
– [Patrick] Yeah. I remember a lot of when SOC troubleshooting back in the day when that was a second driver that you would then install on your PC and TCP IP wasn’t included. So how do you recommend technologists seize opportunities to not just focus on the tech? Because certainly for me and others and pretty much everyone, it’s fun to play with toys, it’s fun to work with leading edge technology. It’s fun to just play with technology, but the business actually is trying to achieve goals, is trying to make money. And there’s an opportunity for technologists to engage the business more. So, is that something that we should be encouraging folks to do? Is that something that’s a missed opportunity that there’s trapped value in organizations with a lot of technologists and IT professionals, or does that really start more with an innovative management team that goes and pulls that creativity and innovation from their team?
– [Robert] Well, only because you’ve used this phrase three times, am I going to mention it. You mentioned that it’s the purpose of the company to make money, and you’ve contrasted that with art. And I don’t think that’s the right way to look at it, to start answering your question. It’s the purpose of a company to serve its customers and making money is part of sustainability, that is profit and sustainability are synonyms although they’re often used as opposites, profit versus people, sustainability versus whatever. So, you need to make money, our goal at 3Com was to make 10% net profit on sales. It was like a law, a constant of the universe, we needed to make 10% net. That’s what we owed our investors; we owed our investors 10%. But the engineers at the company were more, including me, we were more focused on serving the customer.
That is, what do the customers need. And in some cases, we thought we knew better than they did what they need. And I think that sometimes happens, but mostly they know better what they need. So, for example, when we first came out with our Ethernet cards, people would say, what do you use them for? And we would say, it’s a platform, you can use it for millions of things. And then the customer would say, “Really well, name one.” And we had trouble naming one, that is spreadsheets were the killer app for personal computers, what was the killer app for networking? And the first answer to that, which we got from customers is printer sharing and file sharing and email, and that became our killer app. And then later that was replaced by multi-user accounting systems. And later that was replaced by internet access, people buy PCs now to have… Actually, they’re stopping buying PCs, they’re buying cell phones now, but the purpose is to be on the network, not to balance your checkbook.
– [Patrick] So, businesses then have an opportunity to encourage their IT professionals to just really focus on the customer and supporting them, and then use that as the guide to help them develop their careers?
– [Robert] Yes. And one thing we did to encourage exactly what you’re driving at is when we launched a new product, we sent the engineers who developed that product out into the field, so that was when we rolled it out, the development engineers went to the field to sit with the customers and see what happened. And one of our mottos is, no news is not good news, so you had to go out and see how customers were. The mistake we used to make was shipping stuff to customers and then if we didn’t hear any bad news, we assumed we had been successful. That is a dead end. You got to go out and find out what the hell is going on out there. And doing it with the development engineers is a way of closing the loop on what customers really want and being sure engineers are developing what the customers really want.
– [Patrick] So, making customers part of that feedback loop.
– [Robert] Making engineers part of that feedback. Sending them out, now, you don’t want to send them too long because they don’t like to get up early.
– [Patrick] No, because if you stay up on a hacking run until two o’clock in the morning, that doesn’t work.
– [Robert] Oh, is that your excuse?
– [Patrick] Yeah. Well, and there’s this-
– [Robert] Why is it that customers get up so early relative to engineers?
– [Patrick] Maybe it’s going back to your time at PARC, are technology professionals viewed as slightly different and their natural work tendencies are maybe a little more accommodated than sort of a more uniform enterprise experience?
– [Robert] Everybody’s a little bit different and engineers have their differences too. And of course, then now we run the risk of stereotyping and generalizing, but I think customers get up earlier than engineers as a general rule. And that creates stress because they want to have their breakfast meetings prior to the presentation. And we don’t want to get up that early. Oh, by the way, if I had to do over again, I would start my company in New York instead of Palo Alto, because it’s a lot easier to attend a seven o’clock breakfast meeting in Palo Alto if you’re in New York than it is to do the reverse because of that time zone thing. So, imagine a 7 a.m breakfast meeting in New York for the California engineer, that’s four o’clock in the morning for her.
– [Patrick] So, you talked about the internet and how it is continuously reinvented, are we reinventing education? Because these days you’re really focused on helping startups turn what looks to be really promising technology into useful tools, and that’s for education and energy and healthcare. So, talk to us a little about what you’re doing now.
– [Robert] What I’m doing now, well, I thought you were going to ask me about education and what’s happening there. I guess what I’m doing is participating in the innovation of education. So, K-12 education has been in a steady decline for some time. And now you could even argue that university education is going down the same rat hole. But here comes the internet, it is almost as if the internet was developed in anticipation of COVID because when COVID hit, suddenly people turned to the internet to continue education. And it worked, we’re finding at the University of Texas, that many of the students, many of the faculty, think it’s working just fine over the internet.
Whereas just a few short months ago, I’ll never teach over video because I need to be in the same room as my students. Well, most of that has gone by the boards and we’re… So that gradual eroding of education, bricks–and–mortar education has been giving way to internet learning. And that trend is going to continue and accelerate now, we’re seeing it right now. And the course I’m teaching at UT Austin is now entirely online and I meet with 18 students twice a week online. I think degrees more and more students are picking and choosing the courses that they want to take or that they need for employment rather than a list of courses toward a degree.
So, I think degrees will diminish in importance. And then there’s the brands, the UT Austin is a brand, Harvard’s a brand, MIT is a brand. But now we’re beginning to see professors be brands like Christiansen, the disruption professor, I believe his brand was right up there with Harvard’s brand and he didn’t need Harvard, he was a Harvard professor, may he rest in peace. He might’ve needed Harvard initially, but now the Christiansen brand could stand on its own. So, there’s a whole different organization of credibility in education. So, the internet is having major impact on education. I guess I’m not an education innovator, but I’m watching, I’m a user and I’m also delighting that the internet is delivering the goods or delivering the video in this case. But then you asked what I’m doing, which is a slightly different question.
I’m doing many things, but the two most important is this new course I’m teaching to 18 students, twice a week for 90 minutes. And the course’s title is Startup Innovation and I’m using the invention of Ethernet and the founding of 3Com, two case studies in the use of startups as tools of innovation. And it’s a lot of work to teach a course, I’ll tell you. So, I’m doing that. The other thing I’m doing is I’m principal investigator on a Department of Energy contract to enhance the startup ecosystem for geothermal energy that is getting the oil and gas companies to drill not for hydrocarbons, but to drill for heat and then bring that heat to the surface and convert it to electricity. And so, our little project is the Department of Energy agreed with us that a good way to accelerate the advance of geothermal was to encouraged startups to take it up. And so that’s what I’m working on.
– [Patrick] It’s interesting, you tend to mention communications and energy in the same sentence when you’re talking about energy, especially where it’s produced in the field with renewables like with geothermal, is connectivity then sort of a first step in pretty much all innovation now?
– [Robert] Yes. Yeah, I have a talk I give called Connectivity, which argues that connectivity is a thing, and it’s a thing that’s everywhere, even in geothermal energy, you know how they communicate in many cases between the bottom of the well and the surface, they do so using sound waves through mud at single digit bits per second. So, the digitization of energy is progressing and we’re going to increase, we need to increase the bit rate from the bottom of the hole to the top of the hole so we can drill better wells. So even in energy connectivity has very positive effects on the advance of energy. I tend to view energy as a network problem and that we don’t distribute energy, we exchange energy different times of day different levels of demand. So, we’re building what I called an internet of energy sort of a play on internet of an energy network.
– [Patrick] So, for future energy technologies like 5G are within critical?
– [Robert] Definitely. Yeah. In fact, this conversation that we’re having, I think is going over a 5G network. I’m digging into it, I’m communicating from my Mac through Wi-Fi to my iPhone, which is connected to the telephone network over cellular network. I know there’s a 5G in there somewhere. I know there’s an Ethernet in there somewhere. You’ve heard of LTE, do you use LTE?
– [Patrick] Oh, yeah. The LTE started appearing what, five years after 4G?
– [Robert] Yeah. And so, what do you think LTE means? L-T-E?
– [Patrick] That is an acronym that I have forgotten.
– [Robert] Well, in the telephone world, they think that it means long-term evolution, but I know better, it means Leads to Ethernet. When you talk on the phone, the telephone company tries to get you on an Ethernet as soon as possible. So that’s how I interpret LTE. Anyway, I’m using LTE now to talk to you through Wi-Fi, through Macintosh.
– [Patrick] That is an interesting experience. We went from switched POTS networks with really great quality MOS, mean opinion score, where you had operators listening to random calls and then grading them on a quality from one to five. And to this day we tolerate maybe less quality and reliability in mobile devices in exchange for that mobility and that freedom to move.
– [Robert] That reminds me of the old days when the Bell Labs people insisted that in order to communicate, you needed five nines of reliability and that made everything clunky and expensive. So, the internet people, we delivered nine fives of reliability.
– [Patrick] Yeah, resiliency. And to this day, that’s one of the big advantages, well, one of the big focuses for cloud maybe is that you were talking about things like eventual consistency and being able to quickly recover and being able to make sure that you can restore from a backup. And then that reduces risk so resiliency to this day remains one of the things that seems to make that user experience really good.
– [Robert] And that’s a dimension of connectivity that is this resiliency. That is, there’s going for efficiency, which sometimes is the opposite of going for resilience. So, for example, having two copies of something makes the system generally makes it more resilient, but less efficient because you’re using twice as much storage. So that’s a dimension, one that should be urgent right now, our network is pretty fragile.
– [Patrick] Yeah. I think I might steal that from you. Instead of just saying that availability is expensive or five nines is expensive it may be five nines is expensive, but users may actually like nine fives better. So, I got a question for you as we kind of come to the end of this conversation around predictions. And obviously as a tech staple, you’ve seen plenty of those, especially as a CEO and editor of InfoWorld. So, you were talking about cloud now, and a lot of IT teams, especially in operation side, they’re really rethinking multi-cloud for example. And we’re predicting that they’re going to consolidate operations around fewer clouds or even single ones. So, what do you think is on the horizon for tech pros? If you really stepped back from specific technologies like hybrid or connectivity technologies, what really is the big picture for a technologist? What should we be preparing for?
– [Robert] Well, we’ve already touched on several answers to this question. One of which is resilience, one of which is security, one of which is aggressively using connectivity. If you have a little doodad in your house, that’s standalone that you bought at Home Depot, as soon as you put it on the network, it becomes much more valuable than before you can do many more things with it. So, in terms of mindsets, I would bring up a very connectivity, intensive mindset. How can more connectivity help here? Because it can generally, adding connectivity. Connectivity can also add to resilience. Here’s one of my pet peeves, when my internet goes down, why do I have to call the cable company and tell them that my network is down? Why don’t they know ahead of me? Why don’t they know within milliseconds, whether the network’s down or not?
So, there’s a lot of fragility out there that we can… So, I would take an anti-fragility or pro–resilience approach, connectivity intensive. It’s funny, cloud is sort of after my time, I never sold cloud. I solda client server, which is a little bit like cloud, although the servers tended to be in the same building. So, cloud is like a pendulum that swings back and forth, the intensity of computation switches from the big servers to… these little telephones have a lot of computation. And now I think the pendulum is now swinging back toward bringing stuff back down from the cloud and putting it on the edges, that seems like a natural swing of the pendulum.
– [Patrick] So more local data and knowledge creation and edge services instead of evermore centralized systems?
– [Robert] Because of the emergence of latency. So low latency is required of the Internet of Things, the things like to have very low latency and the extent to which you centralize things in the cloud, you add latency. So, you want to bring things back from the cloud in order to reduce the latency and make things work.
– [Patrick] Yeah. The speed of light does tend to be an ultimate kibosh on everything, right?
– [Robert] You don’t feel limited by the speed of light, do you?
– [Patrick] I remember once upon time, we had a modem that we called Luna because the total roundtrip time averaged around 500 milliseconds, and I have no idea how long the routing was all that.
– [Robert] That was the original spec for the internet.
– [Patrick] 500 milliseconds?
– [Robert] You needed to be able to send a character, an eight-bit character needed to go from the terminal across the United States, be read by a timesharing system, and another character sent all the way back and printed. And that round trip had to be under half a second. That was a spec.
– [Patrick] And now the FCC is going to require Starlink to deliver sub 20 milliseconds in order to get some additional funding for rural internet connectivities. It does change a lot. So, let’s wrap up with a couple of final questions. So, the first one is I have seen you interviewed several times, and you’ve also certainly been interviewed in a lot of articles, you were a regular guest on panels. You’ve been asked probably just about every question, maybe not. So, what’s something that you wish that people would ask you, but they never do.
– [Robert] Well, here’s one new technologies, like we’ve been talking about can have one of two effects. One is to reduce cost and the other is to increase revenue, which is more important? In some degree you need both, but I’ve always said increasing revenue is the priority, not reducing costs. So maybe that’s because I was selling stuff that I wanted you to buy. So, I wanted you to, but I’ve always felt that there’s more upside in increasing revenue than there is in decreasing costs, that is you can innovate maximally by just focused on cost reduction, because there’s limited… You can’t innovate by just cutting costs, you need to grow revenue and that’s a different mindset, the revenue growth versus the cost cutting. That’s a question I’ve always had an opinion about that, but no one’s ever, very few people, no one has ever asked me that question.
– [Patrick] So, for our audience, who are IT professionals and IT is obsessed in a lot of cases with cost consciousness. You’re saying that one of the first paths to innovation is to really start thinking about how those technologists are supporting revenue, not just reducing cost of operations.
– [Robert] Yeah. The growth of revenue, which often means the introduction of new products that are different from the ones you’ve been offering, but a focus on that versus cost cutting because cost cutting is limited. You can bring your costs to zero eventually, then you can’t do anything, but revenue you can grow that infinitely into the billions. So, I think IT people should emphasize, you can’t give up on costs, but you should emphasize revenue growth rather than cost reduction.
– [Patrick] Thank you. All right. So, here’s our last question. You’ve had so many different roles over your career and the thread between them is really entrepreneurial. So, looking back, what do you see as your legacy? We know you for the Ethernet and for Metcalfe’s Law and 3Com, but what do you want to be remembered for?
– [Robert] I don’t think about that subject but let me try to answer. I invented Ethernet, of course, that statement depends on what you mean by I and what you mean by invented and what you mean by Ethernet. But I think I’m stuck with that as my obituary will begin, he invented Ethernet and there is not a lot I can do about that. I would prefer that my legacy be bigger than that and include having been an internet pioneer. So, the internet is bigger than Ethernet. And I was lucky enough to arrive in January of 1970 and the internet had only been operating since October 29th, 1969, so I was two months late for being there at the beginning. The internet has just been such an incredibly positive experience. You talk about, or I talk about freedom and prosperity.
The data is overwhelming that the internet has had dramatic effect on freedom and prosperity and the reduction of poverty and freedom of speech. And so, the internet is… So, if I were to be remembered, it would be for the internet, but unfortunately, I’m stuck with, he invented Ethernet, and that’s that. Oh, and what I hate is he invented Ethernet, but Wi-Fi is now wiping it out. Wi-Fi is wireless Ethernet, that was its original name, they changed the name in 1999. So, I tend to count, when I say ethernet, I count Wi-Fi. Of course, there’s 100 people who invented Wi-Fi and they don’t count Ethernet as the same thing. The whole notion of bringing the internet to the very end to desktops to cell phones, that’s the idea of Ethernet.
– [Patrick] Well, conversations like this really make me think of you more as the father of available connectivity and a proponent and an advocate for it because you believe deep down in your soul that increased connectivity of people and things, and energy is how we actually move forward. And so, I think of you as much more of an advocate for connectivity with Ethernet just being a mechanism to actually deliver that.
– [Robert] I appreciate that. Yes, there are people who frequently questioned connectivity, especially this day, a day and age with fake news and all that stuff going on, people are questioning the value of connectivity and they want to shut it down. I’m not over there, I’m over here, which my job is to connect things. And I think connecting things leads to freedom and prosperity and we should do more of it. And there are side effects that need to be taken care of, but they shouldn’t set the tone.
– [Patrick] Well, I think that’s a great point to end on, so Bob, thank you so much for all of your time and being a part of our event. And thank you so much for sharing all of your knowledge with our audience.
– [Robert] Thanks for having me. It’s been fun.
– [Patrick] For more information about TechPod, visit orangematter.solarwinds.com/techpod. For Solarwinds TechPod, this is Patrick Hubbard. Thanks for listening, and until next time.