The Last Great Cloud Transformation Staying Ahead: Top Internet Trends Shaping Networking and Security – Webinar

Okay, here we go. Hello, everyone. Welcome, welcome.

com, and bunch of other things. But, um, more importantly, I am the co-host of the last great cloud transformation video series that we do here on Textron tv. And every month or so, we do a special version of this video series where we go live with a live audience, I wanna say a live studio audience.

But of course, those, that's like from the I Love Lucy days. We don't have, uh, live studio audiences anymore. But you are live if you're watching this, I hope.

And we want you to be involved in today's show. Uh, the last great cloud transformation is a, a video series we produce in partnership with our forensic CloudFlare. And we talk about the last great cloud transformation, hence the name.

But we really talk about something called the connectivity cloud. In a world where our data is stored in, you know, hyperscale data centers, on the edge, on endpoints, everywhere in between, sometimes it's, I think, living on the wires, right? How do you, how do you network all that together?

How do you secure this crazy mishmash of, you know, what's going on in the world today, the world we live in, and with AI and everything going on? And, and that right there is what we call the connectivity cloud, right? It's bringing all of this together.

CloudFlare is in an interesting place. They carry more than 20% of the world's internet traffic over its wires. So who better than to really understand what is involved in this connect connectivity cloud concept.

Um, anyway, that's what the show's about. Today's show's a little as you would expect. It is, you know, Christmas time, new Year's, everyone's kind of looking back, looking ahead.

We're no different. Uh, we're gonna be talking today about staying ahead, top internet trends, shaping, network and security. Before we jump into that though, let me quickly, uh, give you a little, kind of guided to here of what we expect from you.

Number one, we use this big marker, uh, webinar program. It's a, it's nice. You don't have to download any plugins.

It plays in your browser for most of you in the right hand corner of your browser, where no, you're gonna see the big market communications panel. And again, for most of you that means chat. You'll see the word chat in black bold, and it says, public private, public public has a red dot, which means that there's fresh, fresh meat in the public chat.

And it looks like Mitchell's already busy typing away out there. This is how you can communicate with us, with four of us here on the panel. But just a warning, when you put in a public chat, it is public.

So everyone on the webinar today would be able to see that. Um, but, you know, feel free, let us know where you're from today, where you're joining us. As Mitchell said, we'd love to hear from you.

Um, if you've got a comment, a question, a thought, anything you want to just make public, put it on in there. We got people already from Colorado and Canada and, and other places. So that's exciting.

Um, if you have something specifically that you want our panel to take up next to the word chat grayed out, you should all see q and a. And if you click on those, that q and a, uh, icon, it'll light up black. And it's a very similar to the chat interface, another interface where you could type your question in.

The thing is, if you put it in there, I'll see it immediately and do my best to try to kick it up here to the, to the, uh, panel and, and get their thoughts on it. When you put it in public chat, you'll, we'll get, we'll see it too. We'll get, you know, we'll do our best to get our thoughts on it, but everyone might jump in and add, give their 2 cents too, and that's a lot of fun as well.

So please do that. Um, so that's how you communicate with us. We'd love to hear from you.

We encourage you to hear from you. We'll, we'll, we've got plenty of stuff to talk about, but we want to hear what you want to talk about. Let me introduce you to our panel today.

First of all, let me introduce you to John N. Gates. John's been with us before.

Key person to the CloudFlare. He's gonna tell you about it, but he's a great person to have on these shows. He has some great opinions, great knowledge.

John, welcome back. Why don't you introduce yourself. Hello, Alan.

It's good to see you again, Mitch. You too. Um, I am John and Gates, as you said.

I am a field CTO at CloudFlare. And what that means is I spend a lot of time with our customers, big customers, uh, complicated customers, and, and really the outside world, uh, that, that intersects with CloudFlare. So I, I do a lot of, uh, panels and talking and traveling and, and, um, keynote speeches and that kind of thing.

But behind the scenes, um, I'm really connected also within the company to lots of different folks like our product team, our engineering team, and, uh, all the folks that are building CloudFlare. So I have this sort of unique vantage point of being a connector between what's going on behind the scenes and, and the outside world. And, and it's been great for the last three years I've been at CloudFlare.

And, uh, it's, it's really, uh, amazing to see what, what we're up to and where we're where we're headed. Absolutely, John, that's fantastic, and thanks for joining us again. Sure.

Also, joining us today from CloudFlare is David Belsen. David, he's a first timer here, so we'll go easy on him. David, thank you.

Thank you. Thanks for being on. Why don't you introduce yourself?

Sure. So, thank you, Alan. Thank you, Mitch.

Um, so, so picking up on John's vantage point comment, um, I have a slightly different vantage point at CloudFlare. Uh, I am the head of data insight and the, uh, product manager for CloudFlare Radar, um, which is where we get the data from that we'll be talking about today. Um, I've been in the industry for about 30 years.

Been at CloudFlare for the last three. Uh, and as my, uh, my boss once described it in, in an introduction, um, I stare at data about the internet. Uh, so, uh, that is my primary focus of my role at CloudFlare.

And, um, I've been doing that for about the last 16 years. Fantastic. 16 years.

Good for you, man. That's great. Um, last but not least, my cohost here on the glass.

Great Cloud transformation, it's future Vice President analyst, uh, for security and DevOps. It's also been someone I've worked with for the last 20 years. Mitch Ashley.

Hey, Mitchell. How are you? Good to be On here, Mitch.

What's that? Since I'm closer to 25 years now that I think about it. Yeah.

You know, but who's counting Yeah. Amongst friends. You don't look that Pleasure to talk, talk about network and traffic and trends and security and cloud.

So, good stuff. Quick topic. Yep.

Absolutely. And thanks for joining us Mitchell. Mm-hmm.

So, as I was saying in the green room when we were waiting, you know, there's no shortage of naval gazing and pitif pontificating this time of year when people do their year in reviews and their look aheads and, and you know, what's coming and what to expect and what you need to do. Fortunately, or unfortunately, you know, my grandmother used to have a saying about opinions. They were like, covers on pots.

Every pot has one, and everybody has an, she used to say something worse than that, but I'm not gonna say it. Um, but everybody has one, right? And, um, A lot of people have opinions, but when you ask 'em where, where's the beef?

Where's the facts, where's the metrics to back up these opinions? And pontifications, you find out there's a lot of subjective pulling stuff from their gut, from other body parts and, and there's really nothing behind it. We're gonna do it a little differently here today.

We've got David Belson, who, as he said, sits and stares at the internet for the last 16 years or so. Right. And David, you guys, you know, tell us a little bit about what you guys actually do as part of this program, and then let's talk about maybe some of the key sort of trends and, and things going on that John and Mitch and I can kind of pull some, you know, predictions or some, you know, thoughts around, if that's okay.

Yeah, absolutely. I mean, so I mean, just to start and give you a little bit of background, uh, you know, CloudFlare has millions of customers. We, we service millions of websites, about 20% of the, the, the websites that are out there.

And, you know, based on working with these customers, um, that, that throws off what we, what I call at least, uh, a lot of data exhaust. Uh, so a lot of information about, uh, you know, when you go to a cloth, uh, delivered website, we have information about, uh, you know, what country you're coming from, what network, what kind of browser you're using, you know, all sorts of interesting bits and pieces. Uh, and what we do is we aggregate a lot of that data together, whether it's usage data or adoption data, security data.

Uh, and we present that in aggregate on CloudFlare radar, um, in, in, you know, in some cases, uh, a near real time view. Uh, so you know, what's going on from a traffic trend perspective in a given country or a given network or things like that. Uh, and then at the end of the year, like you mentioned, we kind of do a little bit of navel gazing, takes us a couple months to pull together.

Uh, but we pulled together the year in review, uh, microsite and a couple of companion blog posts that, uh, look at a lot of these, uh, trends and, and, uh, you know, observations, uh, over the course of the year, uh, we break it out into, to traffic to adoption, uh, connectivity, security, and email security. Um, so we can, you know, I can jump in with a couple of, I can start off with a couple of stats, I guess. Sure.

Alright. Uh, so I think one of the first things we, we always highlight is, uh, traffic growth. Uh, and in 2024, uh, we saw about 17%, uh, global internet traffic growth.

Um, so, you know, not, not too shabby. Uh, it's not the, uh, you know, purported, you know, doubling every month that we saw, you know, 25 years ago. Uh, but I think things are a little more realistic now.

Uh, and and that's compound growth really on top of the, the growth we've seen over the past several years as well. So, you know, definitely we have, we have not, I I I, we have not hit the end of the internet yet, I guess you'd say. Uh, you know, there's, there's always more going on.

There's always more to, to watch, to view, to, to interact with, you know, especially with AI now. Uh, and, and, you know, bringing not only, you know, text-based ai, AI interaction, but we're also now starting to see more, I guess what they call multimodal. And I think John has probably played with this more, but, you know, you know, voice interfaces to, to AI using AI to generate video and audio and, and, and pictures.

Uh, so, you know, there's, every year there's another driver for, for internet traffic. So we're not at peak internet yet. No, definitely.

No, But you know what, 17% growth is nothing to poo poop. Right? That's a healthy clip.

If my business, well, we, but most businesses, right? If they grew 17% a year, if my portfolio grew 17% Right? Yeah.

We'd all be, we'd all be happy with that growth. Exactly. Right.

Um, though this year is a good year, but anyway, um, but 17% year over year is great, but what can we expect, John, David, I mean, you know, because the, the, the issue is, look, when you look at Western countries, when you look at Japan, when you look at China, they're not growing people, right. Our, our populations are relatively stable. And when you look at, okay, what percentage of the population has, has gotten online, has access via your cell phone from a lot of these people, I don't know if we can expect 17% growth.

So I one other. So, so I mean, jumping from that, one of the things that we also saw was significant increase in, uh, starlink traffic that we, that we traffic from starlink that we saw on our platform, on our network. Uh, so I think what we, one of the things we definitely see there, to your point, is a lot of pent up demand.

So in some cases, in some countries, I think there's a situation where people can't get online, uh, for a variety of reasons. You know, whether it's it's they can't afford it or whether they, you know, a lot of times it's, it's more physical. Like, I don't have a good cell signal or, uh, my ISP won't run fiber across the mountains to where I happen to live.

You know, things like that. And, you know, starlink now is providing effectively a lifeline for, for those folks. And even we saw that in the US earlier this year, uh, with the, the hurricanes in the southeast, and, you know, so I think that what we're, what we're seeing certainly with, um, startling traffic in a lot of these countries is, you know, you can almost pinpoint the date when they got approval to deliver service, because from there we see a hockey stick, and it goes almost straight up, uh, in terms of the traffic we're seeing from those countries, uh, over that, over that network.

Absolutely. Um, John, so I look at it and say, okay, so if we're going to see growth on the internet, that growth is gonna come from company companies that really haven't been on before. So let's say non and areas of countries that haven't really been on before.

These are people who are not digital natives. They may not be savvy to cyber criminals or cyber scams or phishing or, you know, have good hygiene, have good digital hygiene. I'm sure they take baths.

But, um, you know, know what, what does that mean? What does that mean for CloudFlare? What does it mean for the industry as a whole?

How do we, Yeah, there's so much in that question because there's a lot of dynamics that you can start to think about when you, uh, think about new people coming onto the internet. New people that are contributing in some way to the build out of, of the internet in their part of the world, you know, new applications, new languages, um, they're, they're now consumers. They're gonna buy potentially from services that exist or interact with services that are already out there.

Uh, they're also potentially new sources for, uh, cyber threats because people's networks become part of botnets, either intentionally or unintentionally. So sometimes they, um, create new, new places where attacks can come from, right? And so CloudFlare sees that kind of a thing across its network too, where the botnets are, where the, uh, compromised machines that make up those botnets are, um, in some countries where there's not a ton of law enforcement, they become, you know, the, the home for, um, you know, potential new attack, uh, sources.

Not just the machines, but the people that are launching the attacks. And so we have to think about that. Um, that's, it's not all bad though.

Again, there's a lot of people that are doing good things, you know, driving new, um, applications forward, and with ai, uh, with people that, you know, we talk about that digital divide being people that are relatively new to these platforms. Um, they can accelerate that, uh, quite a bit with the help of ai. You know, when, when someone sits down with a chat bot or a, a chat GPT type capability, um, they can quickly become developers and start to build out, uh, you know, applications that might be very useful to the people that are, uh, you know, newly minted on the internet, so to speak.

So I think there's, you know, good and bad from, you know, uh, the, the addition of all this new traffic, some of it's going to be also just growth of, of existing businesses. You know, it's not just all new people that are creating this additional traffic. It's also existing, uh, platforms and businesses that are growing in terms of their, um, presence on the internet, whether it be, you know, taking advantage of AI or video or some other kinds of, um, you know, multimedia type act activity.

So it's, it's hard to tease out exactly where to start on that question, but it becomes this sort of virtuous cycle, this flywheel effect where one is additive to another, to another, to another. And I think that's where the growth is coming from, and that's what will propel us forward. John, just thinking about the areas that, uh, you mentioned so far, starlink, and of course we see a lot more, um, you know, AI related traffic, generative AI being probably the most visible to people externally, but we've had a lot of AI kind of applications running on networks.

Um, and, and, uh, certainly the applications moving into the networks, it, it also means that those might bring some security issues with them, maybe an increase of some of the things we already see today, for example, gaming. I know you see an increase of that in your, in your networks. Does that make you think about security any differently?

Or is it just handling the scale and it's popping up in a particular industry, but it's not adding anything unique to it, it's just more of a growth? Or there new challenges coming through those new applications and uses? Well, there are challenges for sure.

I mean, one of the things that I saw in, in, um, in one of the reports David did on radar, uh, across the platform, is really the rise in scrapers associated with ai. And you can sort of think about scrapers mm-hmm. As a good thing.

If you want them to come and discover your content, you can think of them as a bad thing. If you, if they, you, you perceive them as stealing your content or stealing your inte intellectual property, um, you can see them driving new, um, additional traffic in the form of ai, uh, being a proxy between end users and websites, instead of just going direct, you may be going through an AI search engine of some sort. So it creates more demand, it creates new security potentially, uh, you know, challenges, so to speak, from, from a perspective of a company trying to defend their, uh, site or their application.

Um, it creates new tools for the attackers themselves to leverage, to launch more automated attacks and more, uh, attacks at scale. So yes, it, it drives all kinds of things. Again, this, this sort of flurry of, uh, you know, your brain sort of starts to take off when you, when you think about all the potential ramifications of, uh, AI and, and these sort of, uh, modern tools that are starting to emerge.

So I think just even going, going beyond AI too, I think, you know, we, we as an industry almost, uh, refuse to, to learn from the past. And I think, you know, Mitch, you know, as we, as we deploy more and more applications, and as these companies and organizations and startups, whatever build and, and, you know, roll out these applications, inevitably, you know, there's software in there that really should be patched that isn't, uh, there's, there's inputs that really should be sanitized that aren't. Uh, so we, you know, there are best practices out there.

Sometimes people do a good job of following them, sometimes they don't. Um, but, you know, from our perspective, you know, that's also why we're here is to be able to, to watch for those issues, watch for those vulnerabilities, and, you know, help, help the customers, uh, avoid being impacted by them. You know, David, the old saying, right, if you don't learn from your mistakes of the past, you're doomed to repeat them.

Right? And that as an industry, that's what we seem to be really good at. Absolutely.

Hey, I, I've got a couple things I want to throw on here, including your question from the audience. First of all, I want to talk for a moment as a publisher, right? I'm the publisher for all of the text drunk sites.

AI has not been a friend, because what's happened is we used to get so much of our traffic, organic search traffic from Google. com. If you search something on DevOps or platform engineering, You should come up.

We always come up pretty high, just because we have the top level domain. com. It's a lot of security content just organically.

We got a lot of traffic. Well, what's happening with AI is Google and the other search vendors. But Google is, you've seen this when you do a Google search.

Now, the first thing that pops up is their own ai, I guess it's Gini, the, the summary. Yep. The summary of it then is the, now then you got the three sponsored things.

You don't see what I would call an organic search until three quarters of the way down the page. That's absolutely right. And so what's happening is Google's using AI to capture and keep more of that traffic on Google.

On Google, where they serve up their ads and their content and send less to publishers like, like Textron. And, and so as a result, we're saying, well, look, if Google's not gonna send the traffic, what do we have to do differently? What trends are shaping our world?

And, and we're still kind of grappling, but that's not the topic of today's show. But I, I wanted to throw that out there. We do have a question, though from Francis G and I want to, uh, I'm gonna try to paraphrase it 'cause it's a long one.

But basically, Francis uses proton mail. I think we're all familiar with it or heard of it, right? And it's his main email.

But, you know, he has Google in exchange kind of mail coming into it. But when he goes to search for content and then has to re, you know, they make you register to download. We get this from our vendor customers all the time.

We don't want non-work addresses. If it's not a work address, you can't download the paper. And because he uses proton mail, and it's, so, um, it's France, it's a post proton mail or whatever.

They reject him and don't let him download this. com or Gmail for that matter. I guess, um, what, you know, what, what do you say to, to Francis here?

What could, what what, I mean, other than, Hey man, you gotta bite the bullet. You know, what, what is the, the right answer here? David, do you have any thoughts on that, or you want me to?

I, I mean, I, my thoughts are not probably very useful, which is, I'm not sure there is a good answer. I mean, I've encountered it as well, and, you know, it, it, it's frustrating. It's 'cause 'cause you know that the next step is they get my, my CloudFlare trust, and the next step is that I start spam from, you know, their 847 affiliated vendors.

And it's like, that's, no, I just want the damn white paper. com. com and you know, that proprietor, that sole proprietor should also be allowed to register for your white paper and, and download it.

And it, yeah, I don't know that I have a good answer for Francis. But you know what, David, it goes to what we were talking about, where's the new growth on the internet coming from, right? com or what have you.

I think just like in real, you know, the internet in many ways mirrors real life. And just like in real life, we have seen the gulf between the haves and the have nots, Right? Really widen up.

I mean, we're sitting here talking about ai, like it's, you know, of course we're using ai. There's a lot of people out there who, you know, they haven't even, they're lucky they're doing web browsing and Google searching. com email address means that, okay, this is a company, it's got x billion dollars in revenue, and, you know, there's potential sales opportunities and so on.

So there's, you know, there, there's a, a chance for it to convert into some amount of money. Um, you know, and I can also then take this registration email and sell it onto my partners. And, you know, presumably they will pay me more for a legit enterprise email address than a, a, you know, generic Gmail address.

So, you know, me as a, as a white paper vendor, a white paper provider, I think that's, for me, I'm looking at it from a money perspective and, and sort of end user experience be. So I didn't to step on you, John. No, I mean, I, I had a thought when you said haves and have nots.

I mean, there's so many aspects of, of that. When you think about the internet itself, I mean, what, what you might be talking about there is really just the, the nature of that email address, but also behind the scenes, there's the reputation of the IP addresses that are associated with certain email addresses. People have a hard time sometimes sending emails if they're not on a platform that is well regarded or well known.

I was also thinking back to your, uh, David, your, your radar data. There was a lot of haves and have nots on the internet in, in regards to connectivity and the, you know, the, the speed and the reliability and the nature of, of how you interact with the internet. And so really, I think there's a lot of dimensions when we think about how important the internet is to all of us, how important it's become, uh, in terms of, uh, how we do business, how we, how we interact with banks and, and, uh, other, you know, companies that we work with.

Um, and then the expectation, we sit here, I mean, I assume we're all in America here, we sit expecting an open internet with certain freedoms associated with it, and, you know, relatively good, reliable connectivity. But that's not the case all over the world. No.

And I think everybody has a different experience with this big thing that we call the internet. I mean, I think depending on who you talk to, we may not have such an open internet in America going forward. Right?

I think that's one of the trends. I mean, what, and who better to talk to about it than CloudFlare? I'm glad you brought it up, John.

You got me going. Um, you know, I, I worry about what I call the balkanization, the balkanization of, of the internet where, whether it's because we're, we're claiming it's for free speech or it's for the betterment of society, or we're, we're, we're, we're not happy with some foreign power or whatever is a trade agreement. Are we, you know, China's done it live for many years, the Great Wall of China, right?

But are we heading for the balkanization of the internet? What does that mean? What does that mean for us?

It's a complicated question, you know, and I, and I think, um, it's gonna mean different things to different people around the world based on the laws and the regulations and the, the restrictions that they have. Um, we don't all have the same experience. And, and there is always risk that we could lose, um, you know, something.

So we all have to pay attention to this as well. It's not, uh, it's not a given that we'll always have an open internet or one that, uh, that works the way we want it to or expect it to. And so, um, you know, this is one of the other trends that I'm kind of paying attention to as we sort of tail off the year, is just the, um, the challenges with adoption of something like AI in different geographies.

Because every country has a different perspective on what kind of level of control they have. David's data shows regularly, internet shutdowns that happen around the world based on a government deciding that for one reason or another, they're gonna turn off traffic. I mean, it, the one that surprised me the most is when they turn it off during, um, testing, during the, uh, during the, the exams.

Exams, right? Certain parts of the, the country and the world just turn off the internet so they can stop kids from cheating on their exams. Now, that's just one crazy kind of out of the blue example, but it happens for lots of different reasons.

Um, different countries might be in conflict with others. They might be trying to protect their power, uh, by limiting the communication of their population. They might be, um, trying to gain some sort of advantage in terms of, um, you know, limiting the flow of traffic or the, the timing of around elections.

We, we also saw tons of disruptions, um, in the context of elections. Oh yeah. And we think, you know, CloudFlare itself, uh, was brought in to, to help in many cases, uh, protect the integrity of elections, at least here in the us.

Um, uh, we were involved in many of the, uh, regional, local, uh, candidates, all the different pieces and moving parts of an election. Um, you know, the internet now plays a bigger role. And so, uh, as a company, we have been, uh, very invested and involved in that.

And it just, again, leads you to, to think about where things are going, what, what's happening. And, and I think, uh, there will be more regulation. It will be more complicated to comply in different geographies.

You will need help on that kind of a thing. And that's a another area where cloud and, and connectivity cloud may, um, may end up helping and and aiding companies in, in making that transition. You know, I was, I was just at, uh, black Hat in London about two weeks ago in Black hat Europe.

And, uh, one of the things that I noticed, 'cause you know, maybe like you guys ele know, I've been going to Black Hat for quite a few years now, and you kinda see the trends of how that's grown changed a lot. But one of the things I noticed about Black Hat this year was there are a lot more, you know, they call 'em briefings. They're talks a lot more of the old briefings about security, vulnerability research, what people are finding about new, uh, new approaches and attacks.

And a lot of 'em were about AI hacking, AI hacking into generative AI models and things like that. And, and to be honest with you, it's a little, it's a smaller conference than in the us. Uh, but between that and, and many more of those kinds of topics, uh, it was a little bit nostalgic for the old, uh, black a days where that was the main thing that we talked about.

But what it made me think about is we have a new re resurgence or new surgeons of, of kinds of attacks that maybe that it's early, but if, if AI does and becomes as prevalent as we, what we think it will, that could easily become a big, if not majority, part of our attack vectors that are coming across the network. So you could see that rising pretty quickly. I don't, it's probably too early to see that yet.

I'm just curious your perspective on this. Well, we know that the data is sort of the, um, um, you know, the, the thing that people are trying to protect, it, it's sometimes the data models that they're training with ai, sometimes it's the data that's being exposed via ai, the data's, uh, ultimately sort of the, the, um, the, you know, the things that really companies have value associated with. And so there will be attacks, there will be attacks against AI trying to pry out data that shouldn't be, um, you know, access for one reason or another, that people will find data that's been trained into the models that they can now extract in some interesting way that, you know, maybe it was accidental.

You know, I, I talked to a company the other day, an enterprise who was, um, entertaining number of, of vendors who all had AI solutions that they were bringing to the table. And they were, you know, benchmarking them against one another, looking at their, their own AI that had been trained. And they, they had a battery of tests that they were running against these vendors to see if their own proprietary data was within these models, right?

To see if it had leaked or if it had made its way into, uh, you know, if you have a multi-tenant SaaS platform, is the data leaking between tenants of the platform. And so we're gonna have a whole new set of, of, uh, vectors of attack, and we're gonna have a whole new set of, of attack surface that we're going to have to pay attention to as cybersecurity professionals. Um, we're going to have new tools for the attackers to accelerate their capabilities in terms of attacks.

You know, what used to be a script kitty that was only good enough at pushing the button down on the, you know, on the, the toolkit can now maybe write their own based on the, the fact that the AI tools give them sort of that developer superpower, they can, you know, start to build their own tools and, and don't count on, you know, the, um, uh, the guardrails because they're the open source models, the models that people are training, the models that are freely available on, on various platforms. Um, you, you, you know, they don't always have guardrails. And so now they, you have the ability to have sort of an unconstrained, um, AI model in your hands and use that for whatever purpose you might, might, might imagine.

Guys, I, I got two things I want to talk about. One directly to what you said, John. Look, it probably wasn't a great year for open source.

Open source is ubiquitous. It's become ubiquitous. It's become the dominant form of software.

Most applications, F 75 or 80% software, open source components now, but it's not been a great year for open source security, right? Software, supply chain security has really come to the forefront. Things like SBOs, everyone knows what it, if I would've told you three years ago, had an sbo m you would've thought something very different than, than a software bill of materials, right?

Mm-hmm. But everyone knows what an SBO is now, um, how does that play into, you know, CloudFlare strategy, for instance, right? You use a lot of open source yourselves.

We all do, you know, how, how do you, how do you get your head wrapped around it and feel good at night? Or do you, maybe you don't. I I mean, David, I'm gonna let you, you answer first.

Let me hang myself first. Um, I think that, you know, from what I've seen being a cloud player for, for three years and, and, um, is that, you know, we, we've got a lot of really smart people in the engineering team, and, you know, I, I think the work that they do probably to dig into that code, and I've seen them do some really just amazingly detailed stuff. Like I, I also work very closely with the blog, and some of the posts are just insanely deep and insanely technical.

And so I, I, there is some level of trust that I have that says, okay, these guys are not just downloading the latest whatever package from GitHub and, you know, putting it on the network. Uh, they were actually, you know, auditing the code. And, and I know there's a lot of work we do around, you know, integration tests and, and audit tests and running, you know, uh, software, uh, you know, scanners against it to make sure that certain things are not, you know, are not, are not present in that code.

So I think there's, there's, that's part of it. But I think we also, over the years, have demonstrated, um, the ability to react quickly and, and to address issues quickly so that, you know, if there is something in the underlying code that we didn't catch, uh, you know, as soon as we discover it, find out about it, you know, whatever that, that we're gonna, we're gonna make sure that gets resolved in, in, you know, hours if not minutes. Uh, and then, of course, write a long detailed blog post about what happened and how we did it, and, you know, what we're doing in the future to make sure it doesn't happen again.

Uh, What's what's in my mind right now as we're talking about this problem, is just the visibility for a company into what they have, what's theoretically, um, you know, a target, potential target of an attack. What are they exposing via API, whether it be on an open source platform or on a closed source platform, or, you know, what's behind the scenes and what connects to what. That becomes a challenge because, you know, we're living in this highly distributed world, SaaS vendors, everybody's got an API, more people than ever are writing code, integrating with a, a APIs or leveraging open source packages that they, that might be in their Python code, just, you know, PIP install and, and you know, you've basically got, you know, a bunch of packages now.

So it's a, to me, it it's a more complicated world. It's more distributed, and it lends itself even more to, um, the kinds of things that Cloudflare's trying to solve, which is connecting, protecting, providing visibility, logging, you know, uh, understanding what's connected to what, uh, putting some controls around that kind of a thing, uh, becomes really important. You know, API security is a big area of investment for CloudFlare, and, and I think this ai, uh, trend is going to only make that more, uh, of a challenge because everything is going to be a, a producer and a consumer, and it's gonna all try to connect, you know, together.

And, and, uh, the only way to get some visibility back is to, uh, you know, get a little bit of, uh, logging in in places that, that you might, you know, currently be ha you know, having a blind spot, so to speak. Absolutely. Guys, I want to go in a different direction for a moment if it's okay.

Sure. You know, Mitch and I, almost every day we do this Monday to Friday, we do Textron Gang in the morning, it's about 45 minutes. We cover three, uh, three topics every morning.

It's kinda like the view meets Fox and Friends meets Morning Joe, no po Well, sometimes it's politics, but we try to stay away from politics. We stick to technology. Thank you.

Um, we had a discussion yesterday about network net, not necessarily network security, but network. We're of an age, all, all four of us here, some, a little more than others, but we're, we grew up in the same era, right? When we first got into technology, if you had a Novell networking certification, you were golden.

You made a good living, you were a Novell certified, right? Younger people on here probably don't even know what the heck we're talking about, but it was a thing. Mm-hmm.

And, you know, boom, that, that disappeared with the, in onset of TCP and, and, and internet protocol, right? Um, then the Cisco, you were Cisco certified network engineer, again, you were guaranteed pretty much a good living, right? Cisco certified network engineers and Man Juniper came out high speed networking, right?

And, and some of the other extreme networks. And you know, we, we went through that period nineties, two thousands where networking people were hitting the love button, people were loving networking, right? Network was a huge part of the, of the, of the game.

You know, we all lived behind our Moten castle and we were tied to the network. The cloud changed all that. CloudFlare is a, a result of it in many ways, right?

It is. But now with the Edge, with endpoints doing anything from everywhere, all at once, all of a sudden networking's important. It's the connectivity network.

How, what are the, what are the, and David, I'll ask you based on radar, but John, your opinion, what are the trends and what, what is, what is this new age networking, if we can call it? Yeah, well, I, I think, you know, you can sort of look back at, at the early days just a little after your, your, uh, Novell network example. You know, we really started to see the uptick in the use of the internet, but the early internet was mostly, um, connecting your private network to the internet and then putting, it Was a network of networks.

Yeah. And putting, putting a firewall between your private network and that outside untrusted internet, so to speak. Mm-hmm.

And then trying to keep the bad guys out, right? And, but the, the fi the world we find ourselves today is the internet is much more, um, it's become the, the backbone for a lot of enterprises, because a lot, lot of us work remotely. A lot of us are no longer on a company owned and controlled, uh, network.

Maybe the device is company owned, but the network, you know, it's my home network. It's, you know, at and t's network or spectrum's network or somebody else's coffee shop network, if you will. And so we have a very different model for building and protecting those kinds of networks because it is kind of all internet at this point.

Um, I mean, yes, there's pockets where inside of a data center, you're gonna have a very much traditional switched ethernet network. Um, AI obviously uses, uh, high speed networking when you're training, but once you get beyond a certain, you know, sort of subset of, of, of use cases, now you have to think about how you position AI near users like you did with CDN. How do you get, uh, AI at the edge so that the workload, whatever it is you're doing, whether it be a chat bot or something else, how, how can you make it perform well, how you can make it adhere to local compliance requirements?

And that's a very different model than trying to route everything back to some central data center and then route everything back out, um, like, like we used to do. And so that, that, that model has just fundamentally died, you know, and networking is being re uh, uh, reformed, remodeled for the modern era. The pandemic sort of forced us to do some things, um, you know, to, to embrace that new modern reality.

VPNs became sort of a dinosaur because why would we wanna route everything back to headquarters and then back out again? So, you know, we really need to rethink it for this era. And, um, you know, as, as more things connect to more other things, it, it just becomes so obvious that the architecture has to change.

So, David, do you want to add to that? No, I think, um, I mean, I think John's right. I think, you know what, I think what we're seeing continued growth of, and, and importance of is, is connection speeds and connection quality.

Um, where, you know, I think years ago with dial up, it was, you got what you got, Remember, we all remember the noise of the modem, right? Yeah. You know, and, and, and then when your mom picked up the phone in the middle of your, your call, like, okay, hang up that phone.

Here goes my download. Uh, but now, you know, with, with gaming and with, with, you know, these sorts of video conferences and, and all sorts of, you know, higher bandwidth use cases, um, you know, both with download speeds and, and upload speeds are, are important. And, and not only the speeds, but the latency too, because you can have a high bandwidth connection, but if you've got terrible quality in that connection, it, it's no good.

Uh, and I think that, you know, we're seeing that becoming avail. We're seeing it become those kinds of connections becoming available more places, um, but not necessarily where they need to be as well. Agreed.

So let me put you on the spot though, guys. You know, we talked up this connectivity cloud. It's about, you know, this next stage in networking and connectivity and security, connecting all these new, new ways that we connect.

What specific, like what is the networking on it? It, it's, it's not, it's not what came before. John, as you said, we've had to kinda re-engineer what networking means.

What, and we're not giving away obviously anything proprietary, but what, what's, what's the CloudFlare special source in this new age networking? Yeah. Well, I think, I think the thing you have to understand about CloudFlare is that it is a highly distributed cloud architecture where we have literally positioned our own hardware very close to the end users wherever they are on the, on the planet.

So I think we're in something like 330 different cities, a hundred plus countries around the world, within 50 milliseconds of every internet connected person on the planet. And the idea there is to po position that hardware in a, in a way that serves a couple of purposes. One, protecting from DDoS attacks and bot attacks that, that start at the edge, but also protect the end users near where they are.

And then when you have that network with interconnects in between, right? So we are able to connect, like I'm sitting here in Texas, San Antonio, Texas. I, uh, have a, uh, connection to Cloudflare's Edge.

I think it's Dallas, probably based on where I'm at in the world. com, by the way, if you ever wanna see where you are in relation to CloudFlare. But the point is that once I get to the, the network, CloudFlare takes control over some of the decisions around routing and, and how we get to the destination.

And that's very different than just the open internet in, in the traditional sense, because you were sort of at the whim of your upstream provider. You were routing using BGP routing tables based on whatever route was the most efficient, theoretically. But it didn't, it isn't always necessarily.

And I think that that has to change, especially for business users that rely on the internet for enterprise, uh, use cases. Um, I think positioning that hardware near the end user gives us way, uh, things that we can do, um, to apply security close to where they are as well, so that we don't have to route them and introduce any new latency. And then if you start to think about what's next, um, on the horizon, we talked about already a a little bit about ai, we're gonna position, um, you know, AI near those users as well.

Um, but we can also do encryption at that, uh, the, those locations, you know, the, the open internet needs encryption. It relies on encryption because, you know, we can't trust the, the underlying networks anymore. I saw news articles the other day about the, uh, the telcos potentially breached with regards to some hackers.

Uh, so do we trust the underlying, you know, physical network? I don't know. I, I don't necessarily think we should.

And then when you start to go up the layers of, of what you're trying to protect, you've got all this data flowing between, um, you know, ais and APIs, and we've got quantum computers on the horizon that theoretically can crack most current encryption algorithms. Um, so we need to start adopting post quantum encryption. You know, David has stats on that in the radar, uh, platform as well.

So these are all dynamics that are shaping and changing the way, um, that we are building networks, the way we're securing them, the way we're using them, uh, connecting. And, and, um, I think it's gonna continue to change. It's gonna continue to evolve, and it has to take into account this, uh, extremely distributed world with, um, a lot of unpredictable characteristics.

So in, you know, in, in some ways, not always, but you said something before that. Originally the internet was a bunch of private, connected, a bunch of private networks that were connected. And in some ways, CloudFlare is a giant private network, though it's using public pipes in a lot of places, but it's, it's a private network within the internet or a network within that network.

Um, that's right. It's an interesting way of looking at it. Right.

David, you said before, if we don't learn lessons from history, right? We're bound to Repeat, I think, but CloudFlare I think is in a slightly different position there. 'cause we're in more of that middle mile.

I think we're, we're serving a different purpose than, you know, the, the, the private networks that John was referring to from, from years ago. Uh, my first job in the industry was at BN the the folks who, who built out the original a net. Oh, Well, guys, okay.

Oh, I are you, are you ex BN as well? No, and bbn in one show, that's gotta be a record. Go ahead.

And, uh, that was, that was our customer base. It was, it was those enterprises, it was those, you know, schools and government entities and whatever. They had their, like you said, they had their, their network, they had their firewall and then they had their internet connection.

And I think now a lot, like, like John said, a lot of that is blurring around the edges. And, and CloudFlare Yes. Is its own sort of ostensibly private network, but we're, we're, we're in that middle to, to help facilitate the interconnection of everything else that's become more blurry around the edges.

Fair enough. Just as we're kind of coming up on, towards the end of our conversation, I'm curious what kind of questions are customers are asking you, or you were thinking about was you work with customers going into 2025 that might be, might be newer or trending, right? We didn't talk as much about this a year or two ago.

Now this is much more of a topic, uh, like I may throw out there, like portability of cloud workloads, right? We necessarily talk about that three, four years ago, not like we do today. What are some of those trends or topics that maybe people should be thinking about you're having conversations about?

Well, I, you know, some of them are sort of longstanding trends that just continue to persist in terms of, um, the challenge for, if you talk about enterprise customers specifically, like, you know, multi-cloud, everyone, I, I just did a panel, uh, round table with a bunch of CISOs and, and, uh, in Philadelphia, a bunch of, you know, highly regarded chief security officers from various companies. And they were, they were all challenged and continue to be with sort of how to protect this multi-cloud architecture that they find themselves having to manage. You know, they're in multiple hyperscalers, they've got a foot in, you know, all three or four, however you, you wanna count 'em if, depending on who, who you're, who you're counting.

And, um, you know, trying to wrap their arms around it is hard because that means they have to have expertise in lots of different platforms. It means they may have to define WAF rules and different platforms differently and try to keep them, uh, you know, sort of coordinated. Uh, I, you know, one of the anecdotes back to these early days of network that I brought up was, you know, when you're trying to build resilience, resilience is a topic that comes up constantly building resilience from cyber building, availability of resilience as well.

But you know, the more complexity you put into the equation, sometimes the less resilience you have because you've got to, you know, replicate, uh, rules in different places. You've gotta keep things in sync. I always remember the failover firewalls that were supposed to protect our network.

And when they failed over and they weren't in sync, you had a bad day, right? That you had a big outage, nothing worked. Uh, and, and so that was a, a degree of complexity that was supposed to, uh, build resilience.

And, and in some cases it led to, uh, brittle, uh, brittle nature of, of networks rather than resilience. And so I think we we're struggling with a number of, of ongoing trends that, um, that we, we aim to solve at cloud. So we're trying to take complexity out.

We're trying to build resilience in, we're trying to solve for some of these connectivity challenges that companies are struggling with as they try to wrap their arms around the multi-cloud. We're introducing zero trust into the picture. So we're protecting the end users from the threats that land in their inboxes and on their desktops.

So all of that is, is really the topics that I end up talking a lot about with, with, uh, cybersecurity leaders. How about you, David, from a kind of observer of the network and data, data, we'll call you the data wonka the network, like what's happening? Okay.

What are your thought On this? Um, I mean, I, unlike John, I, I'm not having many customer conversations and that's probably something I do need to remedy. Um, but I think, you know, as all this stuff is changing, it, it makes my life I think more interesting because there's just more and more to look at and, uh, you know, more things to consider.

And, uh, you know, I had a, a conversation this morning, in fact with, with one of our, um, application security PMs, and he was, you know, briefing me on, on some of the new stuff they're rolling out. It's like, okay, this is gonna be really interesting and actually improve our ability to, to bring insight into some of those areas. Um, so I think as these, as these things evolve, um, you know, there's, there's more more metrics to, to collect and, and aggregate, you know, draw trends from David shines a light on things that we should all be paying attention to.

You know, the adoption of those pon post quantum encryption algorithms, the adoption of modern protocols that help make the network more resilient, that, you know, a lot of these things are trends that, you know, IPV six, another one that, that we track. Um, that One's been dragging on for 15. I've been tracking that for 15 years.

It's Yeah. Yeah. You and me both, But I, yeah, Well I, the post quantum stuff, so I, I was a judge for DigiCert's Quantum Readiness Hackathon this year, so I happen to be somewhat familiar with some of these, uh, you know, post quantum algorithms we're using.

Um, some of the Futurum guys are at the, and gals are at the SC 24 show this year, and there were two, not one, but two quantum computers on display there. Wow. So I don't think it's going to shape our, you know, in the next year or two, but 5, 7, 10 years on, I, I, I do think that that's something we need to start putting on our, It's on, it's on the radar, no pun intended.

I mean, it's certainly out there and, uh, you know, we need to start adopting some of those things now to protect the data that we're exchanging now from future, um, potential, you know, threats. The, this, these threats haven't even manifested yet, but they, they have the potential to, because of the trajectory of the, the technology evolving. And I think this is one where we're getting out ahead of it, or at least we're, we're making it possible to get out ahead of it, at least in some respects.

And, and hopefully we'll learn from, from, from that, you know, we'll learn from Y 2K and say, actually maybe we should some resources in, you know, shifting to post quantum and making sure everything is, we got A trifecta, A y2k, re Y2K on top of bvn and, and, and I'm Not even that old. And I, I Sort of, yeah, David, you're really, come on. You're killing us.

You're killing us. Prehistoric. Um, So one last, one last topic that I think is shaping us.

We mentioned ai, but for a lot of us we're talking about gen ai. Gen ai 2024 is the year of Gen ai. 2025 might be the year of ag agentic ai.

We're all gonna have an army of agents, digital workers doing our work for us that's gonna shape how we, it's gonna shape our cloud. What is Cloudflare's response to that? How are you gonna make sure that all works?

What agents are you going to give us to use? I mean, you know, there's all kinds of implications here. Well, this is back to that thing that I think I, I alluded to is sort of the, um, idea that machines are going to be consumers.

You know, bots can come in various flavors, good bots, bat bots. Some are crawlers and scrapers, but some are, are our agents, right? If, if I have an agent that's going out to, to book an airline ticket for me, it's going to try to exhibit the same behaviors as a human would in terms of going out and, uh, hitting those websites.

Especially if there's no API present, right? The API being available makes it sort of easy, but when there is no a API, then a then a AI is going to mimic a human. And so I think Cloudflare's big challenge in 2025 and beyond is starting to figure out what is a bot and how do we describe the behaviors of, of a good bot and a bad bot.

How do we classify and AI is going to be the mechanism by which we use, uh, to, to, uh, machine learning specifically. Those are gonna be some of the, the tricks and, and, you know, tactics that we use under the, under the hood to help discern what's good, what's bad, how should we filter, how should we block? Uh, but it's gonna get more complicated before, before, uh, it gets easier.

So, Absolutely, absolutely. But you know what, let me end this on a good note. What a great time to be alive with all this stuff, right?

I mean, think back to Novell and Y 2K and, and the, and getting online with a modem, right? And your mom picking up the phone and yeah, it goes, your download and all that good stuff compared to where we are now and where we're going, talking about having AI agents do things like this, for us, it's, it's a great time to be live. It's an exciting time to be in tech.

It's an exciting time to be at CloudFlare, frankly, right there. There's a lot of good stuff going on, and I, I, I want us to end with that. Um, just really a quick shout out to Vincent here.

It says, mainstream media seems to realize finally that social media is taking a big bite of its subscriber base. What's the potential of social media on the internet? What's, what's the effect of social media on Cloudflare's operations?

David tracks the social media platforms in, in some of the statistics, so you can go see directly the, the trends up, down and yeah. And sideways in terms of social media. Um, I, I think certainly it's playing a bigger role in our lives.

You know, a lot of people share news. A lot of young people especially get their news via social media. It played a big role in the elections this year in the United States.

So we can see that, um, being a part of elections going forward, uh, it is, it is a way for us to, um, have our voices heard. If you're in certain parts of the world and nobody, you know, knows about you or knows you exist, it's a way to get the word out on something, right? And so we need to protect that back to that open internet, you know, having the ability to, to share, uh, obviously can't break the law, but we, you know, know to the degree possible, we have to make sure that we keep the lines of communication open because social has become such an important part of the fabric of, of all of our lives.

And, and to that point, we see social media sites and, and platforms as one of the key targets. A lot of times when countries are looking to block and filter and, and sensor, uh, you know, the internet, uh, a lot of times the first thing they'll do is they'll go and they'll, they'll block access to X or Instagram or Facebook or, or whatever. Yeah, absolutely.

Guys, we're about outta time. I gotta tell you. It was a little bit of a walk down memory lane, a little bit of a walk into the unknown, But I could oblige What a great panel this was.

What a great discussion, huh? Um, I, I enjoyed it tremendously. Probably more than the people watching even, but, you know, we'll, we'll bring it back again.

What I always love to do with these look aheads, look, you know, retrospectives is, give it a little time and then come back and see were we foolish? Like what did we miss, right? Or what were we thinking?

Or maybe we were dead on, on some of these things, right? We saw it, we were prees it, it's always interesting to do. But, um, David, John, Mitch, thanks for joining is always, Hey, all you guys.

Radio Shack. Yes, Corp. That was a lot of our first computers.

But anyway, I, I should point out, hey, and it was been pinned here the whole time and I didn't even say it, but this radar report, the Year and a View was right there for you to click on if you didn't guess that from the link. So do click on that. Folks, thank you so much for listening in, thanks to CloudFare Cloud Flare for sponsoring.

Thanks, David and John, Mitch again, until next time. Actually, this is the last one before Christmas and New Year's. So on behalf of the last great cloud transformation team here, thank you for being with us.

We've got more next year coming up right after the first of the year, so stay tuned for that. We're out. Everyone.

Have a Merry Christmas, happy new year. Thanks.