Cullen Improves the TalkingHeadz Standard

In this TalkingHeadz podcast Dave and Evan Go deep, well beyond the app layer, with Cullen Jennings of Cisco. This is the final episode of 2021. 

Cullen Jennings is the CTO for Security and Collaboration Business at Cisco. How do you become the technology leader at a technology leader? Practice! Cullen is Cisco Fellow (1 out of 19) with over 30 years of experience across startups, large organizations, and Internet standards. He has a thing for disruptive, real-time technologies. 

Cullen holds over 100 patents. Cullen is a member of the IEEE and ACM and has published numerous technical articles. He was heavily involved in making WebRTC the standard that has enabled browser-based communications. Cullen tends to have his hand in a lot of standards and solutions impacting real-time communications. Just weeks ago, the new Cisco Webex Hologram solution escaped from Cullen's lab.  

This podcast has a few more acronyms than usual, but will also leave you ROTFL. 

Follow Cullen here: 

• Twitter
• Github
• IEEE

Transcript

Dave Michels: 0:12

welcome to Talking Heads today we have with us Colin Jennings, from Cisco. But first, Evan, this is our only episode for December. Happy holidays to you.

Evan Kirstel: 0:24

Happy Holidays and I'm not sure why it's our only episode but we'll double down in January. How's that?

Dave Michels: 0:30

You know, this is part of our attack on Christmas. We're gonna say happy holidays, Season's Greetings. Keep it all generic

Evan Kirstel: 0:36

Happy Hanukkah, Kwanzaa, whatever, you know, add your holiday in here.

Dave Michels: 0:42

Because you and I never offend anybody. That's part of our creed. Well, I

Evan Kirstel: 0:44

don't but you do more than

Dave Michels: 0:48

offending people. What are some of the more ridiculous gifts that you have given or received in the holidays?

Evan Kirstel: 0:56

You know, these people who give gift cards should be banished from the kingdom. I mean, God's a great gift cards are the most useless innovation. The only thing worse than a gift card is cash. One because no one uses cash anymore. And have you heard of a thing called breakage? Do you know what breakage is? Is an industry term on gift cards. It used to be like the phone cards, a certain percentage of the cash on gift cards is never spent. Yep. Because you lose the card. you misplace it. You forget about it, it expires, the entire thing is a sham and I'm not going to have it anymore. What's your issue with cash? Because no one uses cash. Cash is crap. It just sits around as well. Everything is debit or mobile payments, you know, Venmo and Cash App. And that's another thing that just accumulates useless.

Dave Michels: 1:48

I'm going to have to introduce you to some more illicit activities, my friend. The various gifts that I have received and given, I want to share with you one of the best gifts I ever gave somebody an avocado

Evan Kirstel: 2:02

gift gifts this year.

Dave Michels: 2:05

It was last minute thing, all the stores were closed, went down to the grocery store, couldn't figure out what to and who doesn't like an avocado. But I gotta tell you, it's hard to wrap.

Evan Kirstel: 2:15

Is it the one that's ripe or the non ripe? It's

Dave Michels: 2:20

that's irrelevant. Right, but But yeah, they're a little hard to wrap. I think it's a great gift idea. I want to suggest that the people give the gift of it's green, you know, everybody wants to be green this year. It's a green gift. It's got, you know, travels. Well,

Evan Kirstel: 2:33

this is just terrible. So I let's just crypto that that's all I want. I'll give you my coin base wallet and pick your crypto and that's it.

Dave Michels: 2:41

There's only one crypto I like it. You know which one it is, are you to be

Evan Kirstel: 2:45

alright. Alright, let it rain.

Dave Michels: 2:48

Let it rain. Well, let's get to our interview.

god: 2:51

Let's do it. Talking. It's a semi monthly podcast with interviews of the top movers and shakers and enterprise communications and collaboration. Your hosts are Dave Michaels and Evan Kersal, both of which offer extraordinary services including research, analysis and social media marketing. You can find them on Twitter, LinkedIn, or at talking points.com. That's points with the Z and Devin curse go calm. That's K IR STL.

Dave Michels: 3:21

Today we have with us Colin Jennings, the CTO for security and collaboration at Cisco. Colin is the chief geek I guess you could say a Cisco are heavily involved in industry standards such as web RTC and shaken and stir. He heads up the secret lab at Cisco. And in this episode of talking heads, he's going to share all of Cisco's secret projects. Welcome Colin watts,

Cullen Jennings: 3:47

thank you very much good to be talking to you, David. Evan and I are passing on absolutely every last secret that we're doing in the secret lab right now. But one of them for sure. The hologram stuff we're finally really happy to get to talk about. Yeah,

Dave Michels: 4:00

we're going to talk about that one for sure. I just want to be sure you're comfortable. We tend to get kind of technical here. So if you haven't if we confuse you if you don't understand something you'd be be sure to let me know. And he'll clarify that for

Cullen Jennings: 4:11

you know, these acronyms just go on and on and on. So to be cleared up the acronyms.

Dave Michels: 4:17

You already brought up the hologram. Normally, they don't let you out of the lab. And then you can hear you had a big project actually see the light of day. It was just a few months ago at the Cisco WebEx one event. Cisco introduced holographic conferencing, other than the fact that it actually works today. What makes this more interesting than the metaverse

Cullen Jennings: 4:38

Sure. So I mean, there's a lot of excitement about different things going on with 3d meetings right now and different sort of approaches to them. So there's there's a whole bunch of different tacks that people have taken for this. They give you different properties and different things. So we can probably talk about some of the range of that. And I'll talk about the range from sort of going from some people are looking at very much displaying tarz and some virtual space where we all go and meet together. And we've seen many

Dave Michels: 5:03

solutions. Evan likes cartoons. Yeah, I

Cullen Jennings: 5:06

mean, that, hey, there's times where there are advantages to being completely anonymous and not yourself and whatever, and you know, the furry character in Second Life. But the sort of tasks that we were really focused on was like, Well, how can you show somebody something really in 3d? And how can you really understand what something looks like and bring photorealistic people into that. So we very much want to create an experience that was like being there. And then, you know, this comes back to all the way back to our telepresence days, we've had aspirations of trying to figure out how to make a very 3d experience that does that. I'll talk a little bit more about tech, we used to do that in a minute and how that's different than some other tech choices. But we want something where you can understand whether the person liked what you were saying, or what their expressions were, that we could get around some object and talk about it. People could point at it and understand I mean, you've had that experience of trying to use words to explain some 3d relationship, like the thing over there and other right now your other right, I mean, it's just a disaster, you need to be able to really see things in 3d in a 3d world. So that was the type of thing that we're going after with that.

Dave Michels: 6:11

And you're doing that with the Microsoft HoloLens right now. Is that something that I mean, Cisco has always been kind of a hardware company. I don't know if you knew that. So is that something that we could expect more of a more? Is AR going to be the approach? And is there a standard? Or what's going on there? Yeah. So

Cullen Jennings: 6:27

we've gone down an AR approach, because we think that most of the task we're interested in, why don't you want to bring somebody into the real world that you're in, you want to bring them into your world and have them in that environment. So this system would take a hologram of me. And if you were wearing an augmented reality headset with projected me as a hologram, sitting across the table from you talking to you right now, like we are. So right now we're working with a range of headsets, the two lead ones that we're going with right now are the Microsoft HoloLens and the magically headset, the the ML two. And the headsets, that's an area, we think there's going to be lots of variations and vendors and stuff. And we really like those, I think I totally missed the memo that Cisco is a hardware company, I just work on the software side of it. And that's a major, we've grown that from a small portion of the business to the bulk of Cisco. Nowadays, it's a pretty big chunk, it's fun to bring that part along. We've also always just tried to work with a range of hardware that people have available to them and be fairly agnostic about that and interoperate with various things. So this is very much interoperable across headsets type solution.

Dave Michels: 7:27

Just one last question on the metaverse, what was the reaction? It was just a few months ago, you kind of unveiled it,

Cullen Jennings: 7:32

it's been a phenomenal reaction. I mean, some of the reactions are like, ah, that's not real, yet that doesn't really exist. And I was like, Well, you know, sign up and come do a demo of and you can try it yourself and see how real it is. And it's true that we are moving it out slowly. We want to do it in a in a limited way. So we can really understand the use cases, because I think this is, you know, it's not often that we invent a new media on the Internet. I mean, we moved audio and then video and now we're adding this like, you know, hologram experience. And there's a bunch of different ways people are doing it and different approaches to it. We'll talk about those a sec. But you need to figure that out slowly and figure out what use cases are, what are the important things, you know, what's the size of things we need to capture? What's the range? And so we're using that to sort of iterate the designs as we move forward on this and rolling up with a handful of customers right now. And that's growing over time.

Evan Kirstel: 8:18

That is exciting. But I must say I'm not sure which is worse. Dave Michaels is a hologram or David Michaels in real life, I'll have to give that some thought. But before we look more at the future, let's look at the past column. I first met you something like 10 years ago, where you shared this vision of web RTC that was gonna change the world. And it has in many ways, but many companies, including yours, and zoom and Microsoft still use these big fat clients as their preferred approach for video meetings. Did something get off track? Or did I misunderstand the original vision of web RTC?

Cullen Jennings: 8:55

No, I think you got it. And I mean, I would argue web RTC is a huge success and change the world. Right now. We're recording this call. I'm on Web RTC, and we're recording the call over right. And it's during COVID. I don't know how many COVID tests you've taken all in the last while

Evan Kirstel: 9:10

but too many, too many.

Cullen Jennings: 9:13

Most of those were over web RTC. It's really made it easy for people to put together incredibly high quality applications that bring voice and video and applications on the internet and move those applications across. And that includes all of the major video conferencing vendors are using it heavily for certain things. But you point out something is really interesting about it is that is not their preferred client. I mean, lots of the major vendors, WebEx included in this guide you towards a thick app versus the client. There's probably a few reasons on that. I mean, why like, you know, some of the areas where I think we went on Web RTC where it just it didn't end up as well as I hoped would be around camera permissions and microphone permissions and the difficulty of people understanding how to grant access how to select the right camera how to select the right microphone, that's a big stumbling point that we see we look at people coming on and trying something for the first time and where it goes right and where it goes wrong. That's actually one of the areas that causes the most issues for WebRTC. And it's a tussle in some sense between making the privacy be very much in the end users control and making it really easy to create an application that has a great user experience with a high success rate with the user of a first time user. So I think that that's a, that's an evolving area. And then over time, we'll see rousers and W three C and others work together to try and solve that problem to make it easier. And then you'll see that remove some of the barriers to entry on the Browser Stack.

Evan Kirstel: 10:41

Fantastic. Well, it's great to see continuing innovation as always, and is it fair? Can we call you the godfather of web RTC yet? Or do we have to wait another

Cullen Jennings: 10:50

year? I don't think we can go there. But definitely been one of the people I was lucky enough to just be from the very beginning involved with sort of creating that concept and the idea and pushing it along. And it has moved to the number of applications and the number of things that impacts is amazing. And there's parts of that, that is voice and video. But the data channel is an amazing real time communication challenges. It's changed gaming in many ways for web games, it had impacts far beyond what we ever imagined when we started it.

Evan Kirstel: 11:19

Absolutely. And I think it's changed the way many of us work. And what about you? What is WebRTC? Great, and when is it not the best in your personal experience?

Cullen Jennings: 11:29

I think that some of the areas where you see people doing interesting things right now you can't quite do in Web RTC is a lot of the machine learning techniques that are being applied, they're a little bit harder to do in a browser context right now. But that's changing, they will be added into there, and they will get there. But it's easier to evolve them outside of that. So you're seeing a range of very low bitrate codecs that have very decent human natural sort of sound starting to come out that are based on machine learning, you're seeing a lot of video processing, where people are trying to subtract out backgrounds or apply effects, those types of things that are happening in that type of environment as well. So it's still a space where there's a lot of room for innovation, as we see the particularly the mobile phone vendor starting to add support and silicone for machine learning models that opens up new things that weren't possible before. And that's part of what my team does is we look at stuff that like, you know, we try it 10 years ago, and yeah, sort of works. If you're okay with it takes, you know, three hours to compute each frame. And then we just watch how the silicon GPUs and everything move forward to the point where suddenly something becomes practical in a commercial sense, that just wasn't at the time it was invented.

Dave Michels: 12:40

So we're talking about WebRTC. In the standards process if you were involved in the number of different parts or committees of the WebRTC standards process. Is there some sort of standards process emerging for either the metaverse or even the holographic conferencing type of thing? Are we just too early on that?

Cullen Jennings: 12:57

Metaverse were were way too early. I'm not even sure anyone knows what they mean when they say Metaverse, and it is going to be one of the most fascinating from a sort of standardization points of view because it brings more diverse communities than it was sort of ever imagined. I mean, my end to end encryption, people are going to hate it when I say that crypto refers to something involving blockchain not not encryption, but the crypto community is coming together, we have the gaming communities coming into that we have these enterprise real time communications, we have these people that built large digital twins and AutoCAD models and designs of whole plants and engineers. And all these people have some interest and has something to bring to this whole Metaverse type experience, you're going to see a huge argument about whether it should be a completely distributed model or a controlled by one company, one universe sort of model, I think that it will end up in the same way the internet ended up very distributed will end up very distributed. There's a huge bits of piece of that. So that's way out on the shorter term, the sort of hologram models and bringing those into things. I think that we're seeing a variety approaches sort of start to reach the standardization stuff. So a bunch of the companies, including Cisco have gone down this light field approach, because we think that provides a more flexible quality than a sort of texture map polygon approach. But there's also the texture map polygon approach, which some of the other companies have gone more down that type of approach. So Microsoft's a little bit more in that camp, we see some of this stuff, probably guessing roughly where medicine thinking is worth more around that polygonal model. And so we're going to see multiple standards that sort of fit together or maybe work together in some cases or don't fit together, coming down for that the entertainment people are sort of coming at it from a different point of view than the real time communications point of view. And that's something else that's converging right now. I mean, we've, you know, forever we've had large streaming media like Netflix and YouTube and those types of things. And the way it was done was done much differently than how a real time communication call like a voice or video call of the I've worked most of my career on and you see those are coming together right now as well. The next Generation of sort of technologies and standards, we'll probably score for those.

Dave Michels: 15:03

AV one, AV ones the next version of WebRTC, isn't it? Well,

Cullen Jennings: 15:07

AV ones, like a great video codec. When we were doing web RTC. One of the hardest decisions to make on it was, we all wanted interoperability. So what do you pick as a video codec? And some people were like, well, H 264 is everywhere. And some people were like, Yeah, but VP nine is royalty free. And I can't put 264 in my royalty free business model. And VP nine works for me. Eventually, that was really solved by a combination of Cisco Mozilla and Google working together to Cisco basically sets up H 264. In a way where we will end up paying the royalties for it if you use it in a certain way instead of you paying for it that allowed Mozilla and so we allow it to be in Firefox, that way we cover the royalties for 264 Being in Firefox. And that allowed the browser vendors to support two VP nine and 264. And everybody could use whatever they want. But in that process, we all realize that neither of those things were really what we want. And we want something that was really the best of both of those worlds. And started working on AV one and AV ones coming out now. It's starting to be supported in some hardware, but it's not supported and enough widely deployed hardware yet that you can full on move to it yet. But the writing's on the wall. I mean, that's just in this now just a matter of time. Like it's not what is going to be the next Kodak that is relatively obvious to me. The question is, when will it will it happen is where we end up?

Evan Kirstel: 16:28

Fantastic. Great to hear that history here on the podcast. What aspects of the metaverse do you think will apply to our space enterprise? communications and collaboration? I mean, my two cents I saw books, presentation on the future of work at meta and that didn't look very appealing, in my humble opinion. But what do you think in terms of working in the metaverse, I mostly

Cullen Jennings: 16:53

want to work in the real world, actually, when it comes right down. That doesn't mean I don't want to use a bunch of the types of tools that they're talking about there. I mean, obviously, I want to create 3d models, having digital twins of things is an amazing thing. Being able to create in a really creative way between engineers and artists to design a new product or build something is amazing. Being able to bring in people that don't have any expertise in these types of environments and have them be able to naturally manipulate things. So I'm very, you know, one of the customers we're working with that's using our hologram stuff right now is a medical company is doing training, where they're trying to teach surgeons who don't necessarily have a computer or technical or VR background, how to use a particular surgical instrument to do some operation. And that type of environment is just amazing to be able to do in a 3d space. So maybe that will fit into very much the type of idea of where a meta and some of the companies in the space are going with avatars and that type of space. But I think it's much more likely that it will be solved by people are doing sort of photorealistic humans brought into photorealistic models that fit into whatever the space you're in and using and bring that out. That's sort of the direction that I think it's going but I think the thing that's most interesting on this is there's gonna be a bunch of companies going different directions where it's all new, and they're all gonna play off each other and influence each other. What happens in one space is going to bleed into another one in ways we just didn't expect. And I think we're going to sort of up for a time of rapid innovation here, I think we're going to see some of the top hardware vendors bringing out newer, better, smaller, lighter, you know, sort of headsets and display devices that make this all easier to use. I think we're going to see a ton of different software companies invest in incredible amounts that I mean, if there's anything to take home from the Facebook discussion areas that they're investing a lot in this or medic question, I was mistake calling them Facebook here I can get with the program. That's a pretty big model. And they'll experiment within play with it and drive where the market goes.

Evan Kirstel: 18:51

Yeah, and clearly, we need new devices, innovative devices, I can't wear a VR headset more than an hour without breaking out of the sweat. Are there particular devices, AR VR haptic suits or other kinds of entities that you're particularly excited about? Maybe the apple rumored Apple glasses that we might see next year.

Cullen Jennings: 19:11

There's always rumors about the next Apple hardware. And everyone's always excited about that. But let me jump into the sort of VR and then other things. You want to know the deep, dark secrets of our lab, I will tell you that when we started our hologram project, we should have graft is one of our key metrics, how long you can use it before you want to throw up. I mean, I think it started at about 15 seconds. And in the last year I've not had a single person say Oh, this makes me nauseous or anything. And that was partially the hardware changing, but it was largely the software changing. The secret about these systems is you have to get all the cues to your brain that are telling you what's happening with this virtual world to match as soon as any of them don't match and are a little bit wrong. Part of your brain just is just upset with what's going on. It's just It's just wrong. You get headaches, you get fatigue, you get nauseous, you Get all of these problems of VR. So as we resolve bugs out of our system and figured out which parts of the visual illusion we're creating are important to the brain and which parts we could fake, and you wouldn't notice, that was a huge part in moving those things forward. So I think the world of you know, dial forward five years, I think we're all going to be using 3d headsets more often, I'm not one of the people believes you're going to have one on your head all the time and use it everywhere, I think you're going to occasionally have tasks where it's good, and you use it for that task, and you don't use it for tasks where it doesn't make any sense. And when we do that, we're going to have some stuff that just works really well. And we feel fine in and we're going to have other stuff that's just poorly designed. And he just like, you're like, I really don't want to use that software, it just makes me have a headache after half an hour. So it'd be a wide range that happens there. And a lot of our learning on this, this journey was around like how fast we have to update things, how to make that. And you hear that quite often comments about oh, VR won't work without 5g that we need three millisecond latency for this to work. And there are some parts in the system where you need really low latencies. But it's it's the trick of figuring out how to split this up. So like the holograms is the WebEx hologram system. And we're just projecting a hologram. With that today, they would work over a normal sort of good broadband home system to upload the hologram at about 30 megabits per second. But then for all the people who are viewing you because not everyone necessarily being captured as a hologram. For the people who are viewing it will download it more around six megabits per second. And it'll work over cellular and it works over latencies that are comparable to normal vise conferencing, like we're, you know, running in normal cloud data centers. And we had to figure out how to split the system up to split. So we could get the stuff that had to be fast, really close to the headset and fast and the stuff that can be farther away farther away. So I think would be a lot of learnings in that space. And there's a lot of complexity and making that type of stuff really working in it deployed engineering system.

Dave Michels: 21:59

I like that nauseous factor, a big fan of the movie sleeper that Woody Allen movie sleeper. In the future, if you're not feeling well, you eat candy and smoke tobacco and determine those are healthy in the future. You're like talking about in the future, I'm feeling nauseous, you put on these VR glasses, you'll feel better.

Cullen Jennings: 22:17

Yeah. Let's talk about some other suits and things other things that are coming in VR VR. So we look at every single technology we can get our hands on on any of these things of like what it might be to experience different senses of meetings going forward. So one of the more interesting ones are these ultrasound based touch ones. So these are haptic interfaces, where you put your hand out into thin air and you feel like you're touching something. And how it works is there's an array of ultrasound sensors on the table, which make sound and they modulate the sound waves such that the peaks of the sound waves all collide at the exact point where you want to feel something and it makes a loud enough sound that actually activates your nerve as you touch it. I know this all sounds really scary. They assure me it's safe. I'm not sure 100% believe that. But these systems you can, I mean, it's not like you can make it feel like all kinds of different surfaces or anything. But if you have a visual illusion of something, and then you move your finger into it, and right when you hit the surface of it, you also get this sensation, your finger, it very much has a feeling of touch. So I think that like that's a fascinating technology. There's tons around the suits that side sort of hand gloves and things that give feedback and force feedback to you. Those might be interesting for really specialized use cases, but they're harder to get everybody excited about putting on because they're more of a commitment to get into them and out of them and use them for specific things.

Evan Kirstel: 23:40

It reminds me of the old at&t ad reach out and touch someone that's we're gonna have to bring back the advertisement back from the 70s.

Cullen Jennings: 23:47

I think in today's world, that ad is never coming back. But yes.

Dave Michels: 23:54

I want to get back to the present here. The Metaverse is really interesting. But it's about as interesting as warp speed. You know, it's a little too futuristic for me right now. So let's get back to Current Conversations and enterprise communications. The UCAS industry is really excited about hybrid work. And when they tend to talk about hybrid work, they tend to talk about places work at home, work at the office, work at both. And that's good. I'm all about that. I love that portability. I love that that approach. But I think it misses the more important topic about when you work or the async work. And I've already spoken about this a little bit. I want to consume meetings and other content on my time. So what are your thoughts about async work? And is that going to get easier in the future?

Cullen Jennings: 24:43

It can't get any worse. It's got to get better in the future. Let me just back up on that a little bit is for sure. That's super important. And I think that where we're coming into this from is hybrid work has certainly been pushed to a new level in the next couple years and we might be going back to a time where we go partially back into the Office and lots of people will be in the office out of the office. And you'll always have distributed teams. But many people, many companies, my team for the last over 10 years has been completely distributed around the world, and everybody's in different locations. We've been doing hybrid work for a long time. So I think there's a lot of experience in this more than a lot of the vendors, semi lead on. I mean, lots of companies have been doing this. And sure some things are changing, it might be coming a little bit more important. But we can go back and look at the things that we know work and don't work. I think most people have had this experience, if they're honest about it, they will tell you a meeting where everyone's in person in front of a whiteboard is great. And a meeting that everyone is remote is maybe not as good. But that also works. And it's actually this hybrid use case that is the hardest, where some of the people are there and some other people aren't there. And you know, that's true. You've been in that experience. And we keep pretending like there's a whole bunch of new things to say here. But there, there isn't that much. We know that. So we need to figure out how to fix of some of those problems. And we need to figure out what was the hardest thing about the hybrid work? It was always time zones. Okay, I mean, figuring out how you work with people across a diverse set of time zones is so hard. So I think that async has been proposed as an answer to that, in many ways. You could argue email is an example of the original async communication that sort of drove business in strong ways. And it did drive businesses greatly improve people's productivities to be able to communicate that way, and how things run around the clock. And people deal with things on their own time. It's just an awful communications forum for any sort of human way of dealing with stuff. So I think we need tools that allow us to do it in voice and video better. And you can see, you know, Cisco has some fast moving experimental labs that are bringing out things with asynchronous video that we're playing with and experimenting with. And I think you'll see a lot more of that across the whole range. The other thing that I think is important about async is one of it is recording it, so you can consume it on your time. But I think it's also consumed in your way as well. I watch lots of people who I imagine this podcast, many people watch it at various speeds, right? I mean, I knows I know, when I've gone to listen to it, I can set the player at a speed that works for me. And I think that it's not just speed, there's a lot of things that we can do around making the output media and the result of it work for the person consuming it, not the person producing it.

Evan Kirstel: 27:18

Yeah, great point. And I also think real time comms is way over hype. There's lots of content in the enterprise presentations and video and training that I prefer on demand. Yet, you know, most enterprises aren't really thinking this way. It's what do you see you at Cisco, you're connected with every enterprise on the planet? Do you sense any trends here, I mean,

Cullen Jennings: 27:41

we since a very strong interest in trying to figure it out more than they have before and a focus on understanding it's a problem at the very top and, and trying to go do it. Now, some of the issues are around finding and categorizing content and being able to have it in the right spots and be able to get it into a form that's right for the person who's consuming at that time. So if you have a training video that's recorded, some people just want to review some training, get a certain part out of it, some people want a very condensed overview of the whole situation. Some people need the training and need to know the details. And that's probably the simplest example. If we're talking about, I'm trying to review the architectural design of jet engine, and I need to go back and find the various places where various decisions were made and understand why these tools aren't greatly suited for that right now. And I think that trying to figure out how we create that is really important. And it'll be a range of changing the tools, many of the vendors will come up with better tools, but people will come up with different processes and different ways of working and understanding what's norm. Just a willingness to accept that that's important and work with it that way will be really key to

Dave Michels: 28:47

I'm gonna put you on the spot Colin, can that new hologram stuff be done on an async? Way? So

Cullen Jennings: 28:53

right now, of course, we're developing and testing it, we do do recordings and play them back internally. But right now, we have not added recording as an async thing to it to put out onto into the marketplace. And partially that's because we can't figure out exactly how people would want to consume that media in a way that really covered the original experience. So we start with the easiest thing that we know well, which is, you know, these interactive conversations that we've been working, you know, it's we know a lot about what a telepresence, you know, trying to be as good as being there type conversation works like and we'll move that over. But we definitely need the holographic version of that too. And it is amazing, then playing with something to be able to walk back through a factory that was recorded holographic or something like that. And after the fact look for something you didn't see in the same way and look for something that wasn't even the point when the data was recorded, but turned out to be important later. And now you can go back and look at it from a different angle. Look at it from the viewpoint that makes sense from you to see it. And so that's part of why we're really focused on this light field is because we end up recording the whole life Feel them all the light in the room going every different direction, and we can extract any view from any point of view later. That's.

Dave Michels: 30:07

So the third time you watch the recording, I didn't see that candlestick in the library before. That's interesting.

Evan Kirstel: 30:13

This is fascinating holographic communication. This is kind of an interesting tech story I've heard in a while, frankly, I've been underwhelmed with the amount of innovation over the last couple years, you know, really a lot of meat to products, nothing particularly new from Apple. And as a gadget geek, gadget fanatic, nothing really kind of super interesting to buy. So yeah, I can't wait to get my hands on this, this demo this product. He may

Cullen Jennings: 30:41

interesting point there about it seemed like he hadn't had as many new things or something. I can't speak for any of the other companies. But I can certainly say that, you know, this whole COVID Time has been there certain things that have worked really well and have been great. You can bring teams do and get them done. But there are other things where boy, you wish you could get everybody in front of a whiteboard in the same room. And so we've been thinking really hard about what was it that made some of those things work fine when everyone was remote? And what was some of the things that weren't working fine? And how do we go and fix those and provide the environments to make those interactions work better, because we're going to have to have more of those.

Dave Michels: 31:16

It's not just creativity, I think I think you raised an important point, both of you. But it's also relationships, it's been interesting to watch the online dating change during the pandemic. Not that I know anything about that, of course, but the issue has been now that we've started to go back to events again, that, you know, I sit there at an event listening to a PowerPoint presentation. And I think this was better remote, I preferred being remote and skipping the travel. But then you go afterwards, you have a meal with some friends, and you build new relationships. And and that's been really hard during the pandemic. And so is that going to get better? Are you working on top secret projects that will allow relationships and maybe more creativity type of these things that beyond the PowerPoint presentation? In an online event? Or online meeting?

Cullen Jennings: 32:07

For sure. And I mean, the real issue that you've hit us like those things that happen outside of the room where somebody was standing up giving a presentation, like how do you encompass any of those, and I mean, Jeetu has talked about the experience he would like to have with the hologram system of he and his friend both order the same pizza, and it arrives at the table, and they are in different cities. And they're both in different cities eating a fairly similar looking pizza, having a conversation with each other across from the table. And with this, you wouldn't I mean, sure, lots of people had family meetings and meals over zoom in the last couple years. And it was pretty clear, that was not the experience anyone wants. So the question is, what is the experience people want? How do we get closer to that? How do we build those relationships? We've looked internally in our teams, and you think about how much is the innovation this team is producing based on relationships that were made long before COVID started? And it's just sort of running on that momentum? Right? And how much are we missing by not being able to create the new relationships and new people. And there's people on my team who I had never met in person until quite recently. But we knew each other quite well, because we've been using these tools. And we had a sort of relationship that was, you know, far beyond like somebody that I just met, I met on a few video calls. So I think that this is an area where the we're like, I think that in 20 years, we will look back at what we have today and go oh, that looks so awful. It's amazing. Why didn't anybody do the obvious things to try and make some environments better?

Evan Kirstel: 33:38

Grant coin, shifting gears to security. And related topics. Kellen, you're a thought leader on encryption, if I do say so. And we've seen a supercharged interest in encryption working at home and from the consumer standpoint, and our whole variety of new apps for secure communications. But while this focus on the enterprise side of encryption, when there hasn't really been that much more emphasis on the importance of the enterprise, when that seems to be the focus we should have in terms of keeping secrets and data protection, data privacy. So why is the corruptions pretty rare? Still, you know, whether it's email or messaging or various over the top apps in the enterprise, we seem not to have made that much progress.

Cullen Jennings: 34:35

In some ways. I think we've made tons of progress in some ways. I think we've made you know, there's so much more. Okay, so let me talk about some things I think are important that were driven by ideas that largely first appeared in consumer apps, but are now standardizing and moving to much more industrial strength solutions that are moving into enterprise applications. So Let me just call out something, I think that we've done a fairly good job on, actually encrypting the bits that go over the wire by putting them over a TLS channel has widely happened. Okay, that's, that's going pretty well. So that stops, some just attacker who can listen on the wire from seeing what's going on. And that's a good thing I've added improves things, it means that, you know, these man in the middle don't just automatically get all your data. And we've tried, you know, at one point in time, it used to be hard to roll out certificates and stuff. Cisco is one of the founding sponsors of Let's Encrypt. And the whole game of that was to try and make it free and easy to deploy certificates in an automated way. And that, that is now the number like there are more websites with Let's Encrypt certs than anyone else has had it, you can actually see when they started and really got going, not just the rate of encrypted traffic on the internet change, there was a rate of change change, you can see an inflection point where that service changed how fast we were encrypting the internet. Okay,

Dave Michels: 35:59

I got a call on that. And I've got, you know, probably one of the most popular websites on the Internet called Talking Points calm, and it's running with certificates, and it's an HTTPS, and you get the little padlock and everything on that. So people come to my website, and they find out all these incredible secrets and value that I offer. And then they send me an email saying, Hey, here's my credit card. And the email, of course, is unencrypted. What good is it to encrypt a webpage, with SMS and email and all these primary communications are not encrypted.

Cullen Jennings: 36:28

Totally agree, let's tear that apart and dig into the next let's peel the next level the onion on this because you're you're writing what you're saying there. So just putting the bits over the wire encrypted, great first step, however, all the data in the databases and everything that got to that website still there. And when a cloud site is a large cloud vendor, it gets compromised, or has a mistake, or whatever it all gets released. And look at the continual series of security disclosures we have, okay, that, you know, this person lost all this private information, all this information was lost. And those are the ones you hear about the ones you less hear about is like how much like financial loss particular company had more and more, you're seeing these malware encryption attacks, but these attacks on the data on the site. So my belief is that we have to, and I'm going to get to messaging here in a minute, because it's sort of the next level of this, my belief is that we probably need to have end to end encryption. And so when I talk about end to end encryption, I don't mean that the service you're using, whether it's your website, or whether it's Google's or WebEx, or whatever it's encrypted up to that, I mean that they don't have access to the data, that the end user that your data goes from one user to another user, and the services and websites that you're using don't have access to data they shouldn't have, and therefore there's nothing to lose, they don't even have it, right. If they're compromised, they're not losing your data, they might take their service down, but they're not losing your data. So that's the level where end to end encryption sort of is working a little bit right now. But the problem with that is we do not have a really strong, really usable identity and authorization system. So that's why you constantly see, oh, give us your phone number, we'll send you an SMS, which is pretty insecure. And if you can tell me the code and the the five digit code, which is not very many digits, we'll pretend it to you. Or I'll send you an email with this link unencrypted over your unencrypted email system. And if you can hit the password reset link, you can change your password, right? So we have a real problem with authorization identity. And this exists even in the consumer market as well. There's not too many good solutions to this. And the ones that are sort of trying to do it have huge privacy problems. So I think that one of the big challenges for the industry here and a bunch of people want to be working on this, there's and you know, this is one of the areas where I think that people like the browser vendors, Microsoft's identity, Cisco and our dual identity system, those type things can work together to try and solve the problem of coming up with end user identities, either email address likes, or phone numbers like are probably one of those two, and using those as as a global day. And you see lots of the companies in this space, try doing acquisitions, trying to make things zoom sounds acquisitions of space, Microsoft and acquisitions in this space Cisco's done acquisitions in this space, that are trying to figure out how to bring a strong identity solution and authorization solution, so that you know who you're talking to it when we're talking about candy earlier, I always refer to it like if you're encrypting something, but you don't know who you're encrypting it to that's candy from strangers, right? Like you don't know, if you're encrypting it to the attacker or whatever, we have to have a strong identity system, we have to have an end to end identity. So the Indian identity right now, the solution that I think is probably the best technical solution to it right now is this message level security. And it's a standard at ITF from a bunch of vendors. Cisco uses it for WebEx and data security, but other vendors use it as well. And it was driven out of the type of stuff signal was doing and it provides some very strong ways for a group of people to all come up with common keys to encrypt each conversation. And as people get added or leave the group the keys can change and rotate and you can make sure that people who came, who got kicked out of the group don't have access to stuff that's discussed after the point that kicked out, it solves all of those types of problems that you need to solve. Now it doesn't solve identity authorize like identity, do we need an identity system outside of it, which I think is still using, but the solution, that MLS to encrypt messaging, I think that that's the way that we're going to see all of our secure messaging go. And though you could do it on top of SMS or email, I don't think that'll ever really happen or be widely deployed, I think that those systems will just be phased out as a way we communicate, people will use different applications for secure communications.

Dave Michels: 40:35

Are you a fan of RCS replacing SMS?

Cullen Jennings: 40:38

I mean, it'd be better than SMS. That'd be great if they could pull it off. But really, I think iMessage and those types of things where it's just like, we just took your phone number and made an app that ran over the internet sounds like a better plan to me. So I'm actually where I am. And this I certainly don't speak for Cisco on this. As you know, my point of view is, the internet is going to beat the telcos when it comes to messaging apps. And you can look at the consumer argument and tell me I'm wrong. I mean, like, I don't know. But it looks to me like, people will be using apps as the primary bulk of their messages, particularly obscure things. However, let me add a sub caveat to that. I think phone numbers will survive long past the telephone network, phone numbers are an amazing identifier to use. And if you think about how many of the consumer apps, you identify the people you want to connect with, you do it with phone numbers pulled out of your address book, you have a personal address, book, the maps that phone number to a name, phone numbers are inherently unique, there's a good way to get them and allocate them. And they're internationalized. I don't have a problem of I can't represent my name, and like, you know, whatever. So they have a lot of really positive properties. You can tell somebody won, people know what they are, people know how to deal with the privacy of them. Like, you know, if you have somebody private, if somebody's phone number, you sort of understand who else you might be able to give it to and who else you might not be able to go to. I think phone numbers are awesome. But I don't know, I think the SMS and stuff, it has a purpose, it's a tool that will be used RCS is an extension of it that will help and fix some of the problems in it. But long term, I think that just this is a Innovators Dilemma, sort of incremental improvement versus just something completely new to replace it. Well,

Dave Michels: 42:15

we're running out of time, I want to be sure to leave plenty of time for you to share all the Cisco secrets here. That would be a crime not to ask you about shaken and stir since that's a bit of a big topic. And you've been heavily involved in that. Evidently, it has nothing to do with double Oh, seven. And also Evidently, it has nothing to do with reducing Robo callers. So can you kind of give us a little overview of chicken and stir and where we're at with that.

Cullen Jennings: 42:41

Right? So first comment on the names. I mean, that came out of a time when I was Area Director at ITF. And in the real time application center structure area where many of the names have a drink like context, I can't imagine when shaken and stirred would refer to but it may have actually had to do with beverages, I don't know, or double O seven. So this is a technology to authenticate phone numbers and understand that caller idea that you understand whether the caller ID is authentic or not. And the simple idea is that when you make a phone call somebody along the chain here, maybe your telco, maybe you maybe somebody will sign an assertion that says the phone number you put in your caller ID really is your phone number. And that will come to the other end. And that's what we're deploying today and rolling out. And that's rolling out across a ton of countries. If you're seeing your caller IDs you get from spam calls just sort of changing over the last six months in the US, Canada, UK. That is the reason why is that technology. Now the confusing message that gets out there all the time is like and this will solve our spam problems. No, this just tells you who the numbers coming from. Once you have that, you can start to build reputation service and build like, these numbers are good, these numbers are bad. That number is bogus. Like, if there's a number coming out of some country, that's not even in that country, it's probably more like mostly bogus. Now, there's some complications that like there might be a call center that's allowed to use that number in that country. Now boys do it. So there's lots of sort of weird use cases that we have to support and make work and that what makes it so difficult to build one of these systems. But stir shaken is just a first step along the way of getting us an identity of who the caller is. Once you have that, you can start working on really reducing the spam. And I've spent a ton of time thinking about wild ideas do spam. I mean, actually, one of the early one was to create a crypto token where you have to mined some Krypton, put it with your call to prove that you did work. So it was more expensive for the spam callers to call you. Now, that idea is totally bad idea and completely failed, hasn't gone anywhere. Don't get excited about it. But it's just we've tried all kinds of range of things and the stir shaken films a real basic building block that will allow us to really attack this as a better model. One thing that might happen about it might be really simple. It might just be like once you trust the phone numbers, you'll only have your phone ring for the phone numbers that are in your address book. every one else will go to voicemail. I mean, you know, the solution might turn out to be that simple. And I think that that's one of the things that we as technologists need. I mean, we get into all these complicated crypto solutions and various things, when really what really happens in the real world turns out to just be a combination of technology coupled with changing human behavior. And you couple those two together, you get pretty amazing solutions at times.

Evan Kirstel: 45:24

Yeah, it's funny in a blog post on talking points, calm and one of the regular contributors suggested we make all calls cost one cent, that would change the economics of Robo calling. What do you think?

Cullen Jennings: 45:37

Yeah, I mean, Jonathan Rosenberg, and I wrote it an internet draft and a lot of stuff about exactly that about the charging for microtransactions. And all those things, and it would really change things. You could even have ways where you could refund the one cent if it wasn't spam call, a lot of those get called the sort of something at risk, you know, in modern crypto, since probably staking a claim and putting that at risk, or just putting some money at risk for making the call and those types of things. I think those types of solutions could work. The problem with deploying them is when you look at all the people that would need to change to employ them, there are some people on that chain that do not have the incentive to deploy a solution like that. So it will never work. And I think that's one of the really hard things in solving in changing email or changing calling behavior or some of those things is you have to have a solution that everybody has an incentive to play. And that's one of the things that changed a little bit in stir shaken, it's one of the first Internet standards ever, which didn't directly have an incentive for certain parts of it to deploy other than regulators to regulate it into existence, which has happened, okay. And this is a very weird combination of the regulatory powers working with the internet technologists to cause something to happen, which, you know, may or may not be a good thing, that dust hasn't settled on this yet. It hasn't succeeded yet, but I think it's an interesting approach that you're going to see more of in the future is how we combine technical solutions with regulatory solutions cause things to happen. And that's, that's a double edged sword. It's got great parts and terrifying parts all put together.

Evan Kirstel: 47:09

Fantastic. Well, this has been quite a tour de force of telephony and security and futurism. So thanks so much for joining us, particularly on short notice and tell us what are you doing outside of work to keep saying, Are you still a kite surfer as snowboarder? Like what else are you doing in your quote, unquote, spare time these days?

Cullen Jennings: 47:32

I'm very passionate about getting out and kiteboarding. I mean, when you're when you're on that your minds just in a completely different space than that stuff that my day job takes me through and you have to be 100% focused all the time, or you instantly crash which reminds you you weren't focused instantly. I love it. I've been doing some of that locally. I used to be able to travel a lot to do that. And it's been sort of sad and weird to not be able to travel and see friends. I mean, most, I think for before COVID for like five years before that I'd been to Perth every year because I love kite boarding. There have been a lot of other places. I love kite boarding. I've changed my life a little bit. I've been doing it much more locally on the lakes in Alberta and Canada where I live and we just out to snow season the lakes just got frozen and gotten snowing. I'm going to go out on a lake that I hope is frozen, but not sure about this Saturday and see what happens.

Evan Kirstel: 48:19

Alright, well stay safe. Send us some pics, and keep up the amazing work. Thanks for joining us, hey, it's

Cullen Jennings: 48:25

just such a pleasure to be with you guys where you can just talk across all different things and bring these all these threads of these conversations together as always so much fun. Thank you.

Evan Kirstel: 48:34

Wow, Cohen is such an extraordinary talent and thought leader. And there's a reason he's been at Cisco for over 20 years. I mean, he is the foundation for so much of what's done in collaboration, security and standards and innovation. Really. I mean, pretty honored to have him on the show.

Dave Michels: 48:54

Well, you know, he's very selective. I've heard that he's turned down all the other

Evan Kirstel: 48:58

podcasts. Yeah, well, I think we're the only one in this space. Other other podcast anyway. I don't think enterprise communication is exactly a top 10 New York Times podcast.

Dave Michels: 49:11

It was so fun to go to talk about so many different types of things. You really threw in a while one there without kiteboarding thing. I did not expect that. Good job.

Evan Kirstel: 49:19

I would try kiteboarding once and I think I would end up in a tree or something.

Dave Michels: 49:23

Yeah, maybe in a haptic suit, but I'm not even sure they're

Evan Kirstel: 49:27

in the metaverse until the next show. All right.

Dave Michels: 49:30

Happy New Year.

Evan Kirstel: 49:31

Happy New you awesome man you gotta get out of the phone if your phone no man knows me