DevOps in the AI Era, Missing AI Jobs, and Identity Security | TSG Ep. 997

Hey, everyone. We're live on Techstar Gang. You know, what is the state of DevOps in the age of ai?

Well, what's the state of anything in the age of ai? We're gonna be talking about that and more. We've got a great gang who are live here with us today.

We've got our Colorado contingent of Mitch Ashley and Camberley Bates. They're all happy. It snowed there.

That's, that's what Colorado people like. And then, uh, he doesn't look like he's back home. I don't know Jack, but Jack Poller with us.

Jack, where are you today? I Am actually in northern Virginia today, outside of dc. Oh, very cool.

Good to have you, Jack. And of course, we've got the Dean, Mike Vizard headlining the show. Mike, welcome.

So Mike, I, as I mention, we're gonna start off with the state of DevOps in the age of ai. I don't know, is it, is it, is the state any different? Well, let's dive into that because there's a couple of things happening out there.

There's two reports this week. Um, first from Sonar are talking about how people are actually using AI within the context of DevOps. And one of the things that we're starting to see there is that, um, they're using it, almost all of them, but maybe not to write, uh, code that winds up in production.

They're using it for everything from creating experiments to, uh, working on the documentation. And it also points out that maybe the toil isn't being reduced after all, because it's just changing. It's still toil, it's just different kind of toil.

And maybe we're not exactly saving a whole lot of time at the end of the day. Um, also, there's a separate report from LogicMonitor talking about how well a lot of people are gonna change their observability approach this year and start to think about a different platform and different tools and maybe moving beyond monitoring tools. And then finally, you gotta wonder if this is gonna be in the years of mergers and acquisition, because Snowflake turned around and bought a company called Observe.

And that's an observability provider that, you know, has been trying to gain a whole footprint in that it monitoring observability space. But Mitch, as we go into 2026, it kind of looks like a year of, you know, rock and roll In the land of DevOps. What are you expecting, Mitch?

I you might be on mute. Oh, thank you. Appreciate that.

Yeah, it, 2026 is the age of AI everywhere and across the software development life cycle. And, you know, we have to be careful when we see things or hear things like 40% or 60% of our code was generated with ai. What that really means is AI was in the picture there somewhere.

Maybe it generated some of the code, maybe it was part of the, the process that that contributed to generating a writing code. Um, but the point of it is, is that organizations are very heavily using and will increasingly so use AI as part of their software development lifecycle. I mean, this, uh, the, the article on the report that you mentioned was from Sonar that does, uh, code reviews, reviewing code generated from ai, um, our own, uh, futurum intelligence data.

We just completed another round of surveys and 77% of people say that they're using AI as part of their development process, uh, on a regular basis. And it's everywhere from we're piling it to, I use it every day, you know, a real mix of that. Um, and, and I think what we're moving to is trying to experiment and understand what AI can do for us in software development.

While at the same time the industry is responding with the real more capabilities beyond just writing code. 'cause if, you know, if, if you build software, you know that actually writing the code is a pretty small part of the job. And so generating is also a bit pretty small part of it.

But I think what's interesting coming out of, of sonar is they're, they're looking at what people are doing. They're, they're not surveying people that are taking it, I assume, out of their product. Uh, the stats anonymizing that, and so kind of gives you probably a little bit more accurate picture than just a generalized survey.

Mm-hmm. Fair. Um, look, I, I think there's no doubt, Mitch, I I like what you said, you know, reminding people that when we say 60% of code is generated by, by ai, it doesn't mean that AI solely generated that code.

It means more like there's AI fingerprints on it, right? AI was involved in the process here. Um, but I, I think the bigger picture though is this, the, the plethora of AI tools that are being used to in, in the, in the generated code in the CI/CD process, right?

One of the, one of the challenges we've had in DevOps from, from day one was how many different tools can we have? You know, how many chefs are in the pot or in the kitchen, and how many tools are in the pot? And, 'cause there was a time where, look, you had Jenkins and you can have your plugins for Jenkins, but that was it.

Well then we started, well you've got your get, you got your get ops, you got this, you got that for testing, you got, you know, and, and, and very quickly we had tools sprawl, and we spent a lot of time trying to consolidate and, and make that, and the idea with AI is maybe we could actually automate and make that easier. No, the fact is, we're seeing a lot more tools now. And so if you're a DevOps engineer, what's a DevOps engineer to do?

What's a DevOps person to do? You don't have to be called a DevOps engineer. I know people get upset about that, but if you're a DevOps person, what are you to do?

And this kind of bleeds into our next thing about where the jobs are in ai. The job of a DevOps engineer or a DevOps person and the DevOps team going forward is not to be the tool jockey, not to be the monkey using the tools. It's to manage what tools we're using to have a human involved that says, Hey, we've gotta consolidate down which tools, whether they be AI or not.

Can we eliminate, how can we get hold of, you know, manage this DevOps, this CI/CD practice? Okay, so let, let me ask a question around that. When I think about tools, I'm gonna walk into the, our garage and we've got gobs of tools, Me tools.

All the tools are, are tools to enable you to do something. You know, I still have to pick up the hammer with the nail or a screwdriver with the screw. I need to know how to use that in order to, you know, it's kinda like we just had a new furnace put in.

You know, I'm, I'm it up. Let me, let me explain it to you. When you walk in into my garage and say, Alan, hand me a screwdriver.

Do you want the power driver? Do you want it attached to the drills? Do you want a thin Phillips, a thick Phillips?

You want the flathead, you want the short one or the long one, which I, I've got about 50 god darn screwdriver in my garage. But that is trusting that the person that is, because in or the screwdriver is only the tool to develop the code. And so once, and you have to know, you know, what that code is, looks like, and how good that code is that's producing.

Yes. So how does that really play into all this? So I, I think AI can help us there.

I think AI can decide, you know, AI can work the tool. I think the person has to decide which tool AI is going to use. And the, the other thing about DevOps though is I'll tell you, and, and Mitch, you know, this DevOps exists in bubbles.

In an org, especially in an enterprise, it's a rare bird of an enterprise that has standardized tools across the entire development team. And from team to, you know, you may have dozens of development teams within a big enterprise working on different projects. And each team has its own screwdrivers and its own hammers and its own everything.

And, and part of the problem a DevOps at an enterprise scale, and it actually gives rise to the whole platform engineering thing, is how do we standardize on a platform, including the tools we use enterprise wide, enterprise wide, not team wide. And now with the, with the plethora of AI tools, I just feel like we're, we took two steps back and we need a human here in the loop who's gonna help us? And maybe it's the platform engineer.

Well, I some something you said, uh, didn't ring quite right with me, which is, you said AI can help make the decision on which tools to use. And I actually think that AI is actually just another tool and it's like another set of screwdrivers, or it's different than a screwdriver, but you have an engineer has a choice of tools to use. And some of those are AI based, some of them are not.

And it's really the ai, sorry, the engineers are thinking about and looking at which are going to be the most efficient and the most useful. Particularly know, I'm glad you brought that up because I said what I said, hoping someone would say that. So thank you a little something extra in your paycheck today.

'cause here, here's the thing, the engineer, so I think, I think AI is a tool that sits on top of these other tools. And I think the engineer could say, this is the outcome I desire. This is the kind of code I'm working with, this is the language, whatever, what's the right tool I should use?

AI. And I, ai I think has capabilities of telling us what it thinks Right Tool. And this is, and this is where I differ with you in that I think AI is just another tool in the toolbox.

And It's, but it, it so tough. Okay, go ahead. But for Me, the difference is it's a different way of interacting with the tools.

We use a natural language interface rather than, uh, an uh, uh, a technical code or a pseudo code interface to interact with the tool. So it changes the way we interact with the tools, but it's still a tool that is helping us generate code and building the programs and the applications that we're trying to do. So I, it's, it's a nuance, but I think it's important.

No, I get it. It's mess in that movie Fantasm where it is got of directing all the brooms and everything to clean everything up, and you got your little wand there and all those tools are being orchestrated by the wizard. Mitch, are we having a more realistic conversation about AI and app dev finally?

Because if I looked at last year, there was all these statements from CEOs about, you know, getting rid of developers, this, that, and the other thing. And it seems to me that that's all proven to be nonsense. Well, I said it was nonsense at the time and we're sure building a hell of a lot of developer tools for all these people we're getting rid of.

So I think that was a, a pretty pretty gross over statement that's made by people who don't know how software is created. Um, to, to your point, Jack, uh, yes, AI is a, is a tool, but we also introduced new tools that we're using and, and it is both working with how we work with software today, performing a lot of the same tasks. It also performs tasks in different ways that we don't do today.

So we're moving to, and already vendors who are introducing what they call intent-based development, where you're really prescriptively describing what you want, but even more so you have to, when you develop software, you have to do things like describe the environment that you want, the technology stack that you want, the testing strategy, all of those things. Now you can use AI as a source to say, which model would be best for my, uh, for my planning and, and design phase. What would you recommend and why would you recommend it?

I'll make the choice, right? Maybe someday that'll, that'll, uh, be a choice that it makes for me. It may even make the choice for me based on finops, you know, what's more economical to, to develop with for, for this particular tool.

But what's happening is we're elevating, there's kinda two phases to it. One is how do we use AI to develop software, kind of like the way we do now, just using AI to, to augment it or do it a little bit differently. Then there's the place of, we don't, we don't mess with code anymore.

And we really work with and AI to create what we want it to create. And we have ways of verifying and validating that we're getting what we want and securing it. And that's what we're evolving to.

We're not there, but I call that AI centered, uh, de development really, where AI is at the center of it and we're orchestrating driving it. We're being that, um, control point, if you will, uh, directing what we have, what we desire to happen. And that's where the development role is, is certainly turning this, this survey is really just talking about how people are using it today, where they're using it, what they're seeing for success or not, which is all kind of road signs along the way to where we're headed.

Mitch, I agree with you. I'm sorry. Go ahead, Kim.

Mitch, okay. So you're talking about the DevOps, we're not getting rid of them. Um, but it's kind, what, what's happening with the folks that are just coming outta college?

There, there's this issue right now that the, all the, the junior roles have disappeared. And, um, is it that we have to retool our schedule or education for them so that they're more val, they're have a different type of value when they enter in? What are those roles gonna be?

Cam, Mitch, don't answer that right now. That's your next statement. Can you hold that for the next segment?

'cause that's the, that's the heart of it right there. Now is good time for the next segment then. Well, no, but we, there's something we haven't touched on in this segment and we still got another minute or two.

And that's the observability Piece of it. Yes. Right.

I, I am gonna plant a flag right here that says we are entering the age of continuous observability co. Mitch, I'd love to see you do research on co You read my pred observability and it's important, it's important for security Jack, right? It's, it's, it's kind of a sea change for us, right?

Because observability, are we talking about observability after the deployment event horizon or, or shift left pre-deployment horizon? I think we needed it both, but we need to, we're moving to continuous observability. It throws off a ton of data.

It it takes a lot of overhead to manage and make actionable intelligence out of it. But AI, I think is, it's another great job for ai. Another great task for the tool of ai Jack, if that'll make you happy.

Um, right, that we could use this tool to, to, to really empower continuous observability. And so this is why we see Snowflake buying, uh, uh, Observer. And we see, you know, people switching observability platforms.

'cause it's taking an outsized role in the process and we're moving to continuous observability. So to jump on that bandwagon plug for predict 2026, which is next week on the 15th, and I'll be talking about this as one of my predictions shift left, is giving way to continuous guardrails, continuous security, continuous observability. And by that I don't mean some utopia, but I mean you can see it in the market, uh, where Dynatrace is part of AWS's announcement for their DevOps agent, um, or their security agent.

You see it with the Snow Snowflake acquisition. While we're building kind of the scaffolding and the tooling for AI based systems for agent-based systems, the rest of the market is saying, Hey, I'm not gonna get left out of this. I don't, I don't know what's gonna happen, but I wanna make sure that my stuff's being used early.

So you, that's why you see the sonars of the world talking about code reviews and you see the observability organizations being part of these announcements. So I think the point is, 'cause those things are gonna be gentrified, if you will. Uh, also they're, they're gonna be turned into automated processes even beyond what they do today.

And if they're ingested into, they're embedded into the development process and the testing, instead of it being a point in time function, we do this in this phase of development or we do this when it hits in production and they use this tool to monitor it. No, that's gonna be baked in. Um, and I'm not, again, I'm not describing Nirvana.

That's the progression we're moving towards, uh, because we can do it now and we have the technology, we didn't have the autonomous ability to, to bake it in through the development process. I agree. Hey guys, we do need to jump to our next block though.

com. You can go register, you can see the whole lineup. Actually, we're still adding one or two speakers, but most of the lineup is all there.

I highly, highly encourage you to attend next week on the 15th, as Mitch mentioned, in virtually the entire, uh, FU analyst team will be presenting, including Daniel Newman, Tiffany Ver, and more so, and Mitch. So check that out. Let's, um, I also wanna acknowledge, you know, we are live and we give people a chance to comment and I saw some comments pop up on the LinkedIn board here.

We appreciate the comments hello to you to Jordan from Jordan as well. Um, and, and we, you know, we're still, this is still an alpha, not even beta, uh, we're showing a preview release of the gang, but we're gonna figure out how to incorporate more of these comments in here. But we do, we do appreciate them and thank you, uh, Mike, let's move to the next segment, which is around AI jobs.

Kimberly already kind of threw the bandaid off. Well you, You have a post up on, uh, text drawing AI talking about, well, just where are all these new jobs that AI is supposed to be generated, where are they gonna manifest? And we're also seeing the new jobs report is out today, and I think there's 8,000 or so, but no one's quite clear what kind of jobs those are just yet.

I think we'll sort that out a little bit. But Kimberly, we've been talking about the impact AI has had on jobs all last year, but are you seeing new roles and new opportunities here? Or what do you think is gonna happen?

So I haven't really seen entirely new roles at all. Um, more what I have seen is evolution of roles that are already in place. I mean, one of the other things that you guys posted on was, um, what Mark Beni Benioff said, um, was that he got in front of the, kind of the, the horse in terms of thinking that it was gonna change the di you know, the people that in his company in terms of his layoffs or whatever, and he has found, that hasn't turned out.

And we've heard that from a number of companies saying the same thing, especially around customer service operations, et cetera. It just didn't change that. But what is changing is the operational level of how they operate, how they integrate systems together, um, and produce new outcomes or new pieces of it.

In fact, I went up there and, uh, went up to the Salesforce thing that you guys had sent a guide on, you know, a link to a Salesforce guide and I went and downloaded that thing wasn't that good, so don't worry about it. But anyway, and I downloaded that thing, but within like, I think, you know, two minutes, I have an email from the salesperson already and it's kind of like, okay, so was that AI generated or what? Yeah, whatever it was, it maybe, maybe not, or, or whatever.

So I'll respond to them and have a conversation with them to see how that came through. Because I'll, I wanna know what's going on. But the other thing that, you know, I read this morning, um, and I'm gonna quote, pull a quote from there.

This, um, something that was on, on Substack and they were talking about the anxiety of the different anxiety around the world. China is not an anxious about AI US is very anxious about ai. And he was asking the question why.

And part of this is because we're, we are so concerned about what we do for work. And, um, I love the quote that he had in there is this one path of AI people, people that are driving technology looks at it as human augmentation, which is I think is what we're seeing. AI really is.

The other one treats it as profit extraction. And I think that as we, what the, initially I thought it was more of a profit based kind of thing. Um, I was thinking revamping all of these operations, but what it really is, is augmentation.

It's looking at brand new, I believe it's looking at brand new ways of looking at things. But that, you know, it does get back to the first, the thing that I'm, I asked Mitch about, which is what about the folks that are getting straight outta college and can't get hired right now and are struggling with it? Well, I need to go to whatever ski resort you were at this weekend, and, uh, because you got some really good thinking going on.

It was at ER park, we had six inches and they're getting more, They're getting more. Well, you know, thi things evolve and let me give you an analogy, um, in a galaxy far, far away, a long time ago, et cetera, I was in the construction business and you know, all of a sudden people started showing up with rafters that were pre-built. They don't really, we used to build them by hand, right?

We'd go up there and, and every board we put up there until we built the roof, suddenly they're showing up with a crane, uh, loading them up on the roof. Next you get prefab homes, right? You didn't hire the most experienced, uh, contractors or, or in our case developers to go build the prefab houses.

They built the process of how that gets built. Mm-hmm. Um, and other people ran those machines, ran those systems.

That's very much what we're going through now, right? And, and I think the entry level job problem is a direct consequence of this overreaction over rotation of great AI's gonna do all this for us. Nobody stopped ask of is it ready to do all this for us yet?

So, so, because one tech bro, let's go of a bunch of people, they all have to do that. And everybody pauses and says, well, we don't need to hire. Unfortunately, he has a massive ripple effect on, uh, our entry level, uh, community of workers.

So I, I think what's what's gonna happen is the people coming into the job market, maybe they've been trade on traditional development, but they're gonna be part of this journey of, uh, as we automated and do more with ai, the people in school need to be retooled and start to learn AI in school and developing software with that. That's a little tougher. 'cause educational institutions don't turn around that quickly.

What that means is you, you, you have to educate yourself. You may have a traditional job skill of writing code and testing code and whatever you do to, to develop software pre ai, now's the time to get engaged and start using cursor or whatever Kira, or whatever tool you want to use as your entry point to do that. So it's, it's, I think it's unfortunately it's the productivity measure that the US uses, which is the profit taking measure You're talking about Kimberly?

Mm-hmm. I think there's, there's some something else there that Kimberly mentioned earlier on, which is that yes, the, the ENT people coming into the entry level have to figure out and get trained and train themselves on what they need to enter the workforce. But there's also a problem in the tech industry in general, and that views anybody between the ages of 28 and 45 as extremely valuable and productive to the company.

And anybody outside those ages is not, and it's on both ends of the spectrum where new college graduates and young people entering the industry are changing careers, have a very hard time because they don't have quote unquote experience. And when, if, if you don't hire those inexperienced people, you end up not developing the experience and end up without 28 to 45 year olds too hire at some point. So the large tech companies folks, yeah, right.

They, they, they need to, they need to accept the fact that they still have to hire entry level people, otherwise they'll have no mid-level people to hire. By the same token, they also have to hire the older people because that experience that they have is extremely valuable, even if they're in their fifties or 16. Then, and This gets back to internships, yes, Germany and other, I think European countries are very into internships and apprenticeships.

And a Are you worried that we're gonna see a, sorry, Alan, are you gonna be, are you worried that we're gonna see a lost generation here of people because there's all these new kids out there who didn't get trained on AI in school and they're gonna wind up not being in the industries that they trained for, and we're gonna have to wait for the next generation of graduates. You know, I'm glad you asked that, Mike. No, not worried at all.

And I'm gonna tell you why. I actually did a shimmy says on this yesterday, and if you really, really wanna get into a lot of the minutia, I spent about 15 minutes talking about this stuff. Here's the issue, you know, this isn't the depression and all of these displaced poor people coming outta college who can't find jobs, let's not give them a shovel and have them built digging ditches to make busy work, right?

That doesn't do us any good. It, it gives them the dignity of having a job, but it's kind of on the government doll. You know, that's how we handled things in the depression when there weren't jobs for people.

What, you know, we all hear AI is going to maybe threaten my job. Your job may be eliminated due to ai, blah, blah, blah, blah, blah. Um, and it, but almost like knee jerk, we also say, but AI's gonna create all these new jobs for us.

And my, I asked ai 'cause I got tired of asking people and not getting a straight answer. I asked the AI itself, I said, what are these new jobs? What are the new jobs?

These AI enabled empowered jobs? And I, and it came up with some really interesting things. Let me break it into three categories.

First of all, in terms of tech, don't tell me you're gonna be a prompt engineer. Every single one of us is gonna be a prompt engineer because every single one of us will be prompting our AIS for work product. What we're gonna be is whatever we are now, but enhanced, enhanced, we're going to be, instead of a developer who develops code, we're going to be a developer who takes requirements and, and does a better job of telling our AI of that what we want in that code, what functionality, and we're gonna manage it and make sure the other AI that tests it, test it.

Well, we're going manage code development, not do code development, perhaps security, same exact thing. We're gonna have continuous observability with AI providing the actionable intelligence, but it's still gonna take a security person to put the, the brains behind it to, to make the edge. AI is terrible on edge cases, and the world lives on edge cases.

You're going to need people to do those edge cases. You know, IIII, not to let the cat outta the bag, Daniel Newman's keynote, noting predict 2026 I, him and I did it together, his session. And we spoke about, hey, Daniel's not a coder.

I'm not a coder. Mitchell's a coder. I know Mitch, 25 years, he could tell you, I don't know Diddley about co I mean, I know something about coder, but I know how to code and either does Daniel, but you know what, we both dug into AI and taught ourselves how to use AI pretty damn good.

And my challenge to the, to the young people coming out of school right now who didn't take AI in school, ah, poppycock, go get yourself a good book or get online and dig in and learn to use AI because your life and your livelihood's gonna depend on it. And you don't need to be on the government doll. You don't need to go back to school necessarily.

We're at an age right now where any one of us, including Mitch, including Tam and Jack and maybe even you, Mike, can go learn and become an AI whiz today to make your, make you an enhanced version of yourself. Make yourself more valuable than we, than we, you are now that's for tech people specifically, but it extends beyond tech. Beyond that, AI is gonna open huge new avenues where jobs will be needed.

No, they're not in data centers. If you are looking to be a data center employee, bad news, that's over, right? But AI enabled autonomous vehicles are gonna open space to mining asteroids, permanent bases on the moon, uh, space, space in general.

The final frontier, it's here for all of my Star Trek and Star Wars fans, right? There's gonna be a ton of jobs. I'm not the one saying this, you know, who's saying this?

The, the architects of ai, right? Elon and Sam Altman and, and Jeff Bezos. Well, they all own space companies, but they're all saying that space is, is going to be a place where we could get jobs.

I'll tell you something else. I I brought it up on my shimmy says, creativity for raw musicians and artists and those left brain people out there, man, what AI empowers you to do and what you can make and what you can compose, it's gonna create whole classes. You know, we talk about the billion dollar one or two person company.

I'm telling you, it is going to be possible. May, maybe not a billion dollars with one or two people, but you are gonna be able to do so much more with less with this stuff all. Well, I'm, I'm bullish, right?

Thanks, thanks for the sermon. We appreciate that. I'll step, I'll step down.

No, you can, I'll pass the hat around. Everybody could put in a couple dollars And frankly, I'm not sure that the world is ready for an enhanced Mike Baard, but that's another person. Kimberly, last word to you.

So, and, and actually Alan, I'm gonna play off a little bit. I was with some friends this week, earlier this week, and they were talking about their kids, kids, 30-year-old kids that were having trouble with jobs. And one of them in particular was a graphic artist that had been working in LA and she's really having a tough time and doesn't, you know, and I'm like going, well, how do I help her?

Because we talked about ai and one of the things I think I could say to her is say, okay, here I am, I'm 30 years old, these aren't, go into chat GPT, Gemini and gr all of them and ask them kind of the same prompt. The prompt is, here I am, this is how old I am, this is my skills, these are the jobs that I've done. I am struggling getting the next position.

How do I use my skills and what do I need to learn from about AI from you? And can you give me a training program that I can execute? Because I was thinking that she was gonna completely have to go back to school and retool.

And what you just said to me is like, no, we've got a big tool here that can help them personally, The greatest learning tool we've ever had else. Yeah, you are Kimberly, you know, do that and come back. Let us know what happened because I'd love to hear what she comes up with because I, I that's, that's it right there.

Yeah, that's it right there, man. Yeah. Anyway.

All right. The, uh, the sermon portion of our service today is over. We're gonna be moving along now.

Um, Mike, uh, number, the, uh, segment three is an identity crisis. We, we've got, we've got some, uh, you know, work going m and a news and other things. What's it about?

So this week, CrowdStrike acquired an outfit called sg and l to bolster their identity management capabilities. And we've been talking about now kind of like identity is kind of like the new perimeter. Yes.

No, I don't know. That's debatable, but Jack has a great article up on Security Boulevard talking about how, you know, well, maybe the cure for all of this is plastics right out of the graduate. But Jack, please explore.

That's a yes. I was thinking about the plastic flying from the graduate. I wasn't sure if my audience would recognize it.

Um, let's, let's first talk about CrowdStrike and SGNL or signal. The acquisition is to buy an identity company, and it's really the, the culmination of the platformization of security plat of security, really. And so now you have a number of consolidation and big vendors building, building security platforms, and everybody coming to the realization that identity is at the center.

And I look at it as at the center of everything rather than the perimeter. Uh, people think about identity as the perimeter because it is the first thing you interact with when you try to do something on a computer. However, as you move from our old way of thinking of castle and mode architectures to zero trust, we wanna go from a one-time identity check to a continuous authentication and authorization.

And that's part of what SGNL is bringing to CrowdStrike. And it's where the industry is moving. As we move into ai, more and more ai, you have more instantaneous and repetitive actions by more actors, most of which are not human actors, but are actually machines.

So being able to do real time and continuous authentication and authorization checks becomes more and more critical. Then you and the plastics part come in when you bring the human in and you start talking about less interaction with computers and more about physical security. Uh, in this case we're talking about passwords and, sorry, passports and forging passports and other identity documents.

So in the olden days, we had a paper document that says, I am who I am that was interested in, uh, issued by a government entity and it had a, maybe had a photo on it, and that photo was actually glued onto your document and you could ue the photo and replace that photo. And that was how a forger did it. And you could change the type and change your identity, uh, passports now include, and other identity documents include a computer chip and RFID reader.

And so is that chip manipulable? And it is. And so this particular company came out with a type of plastic technology that you put on the computer chip that basically can tell if you've physically manipulated that chip, if, right?

So one scenario that people can do is walk up to the passport, uh, uh, control checkpoint, and the machine tries to read the passport, the computer chip and the passport, and it doesn't read. And they say, oh, well my, I dropped my, my passport in the toilet and the chip is damaged, right? So you'll have to just do manual verification.

Well, they didn't do that. What they actually did is they put it in the microwave and zap the chip to destroy it and then altered the text on the passport to try to assume somebody else's identity. So this particular PLAs, uh, plastic would show different behaviors, different colors, different types of things, if it's got burned microwaves manipulated by a magnet or other things.

So it helps the passport control officers detect physical manipulation. And all of this is centered around really people trying to break in somewhere or do something bad using an identity, whether it's a computer identity or physical identity. We need to shift back to some sort of hardware based or physical security in, in to get to what we ultimately wanna be.

Because, you know, it seems like digital security solutions just aren't working. I don't know. Mitch, what do you think?

Yeah, it, it, it, we're, we're always in this race of, right. We create another physical thing and then we need something better than that. You know, we create, you know, timed passcodes and now we need, uh, pass keys and we just like the chip that, that, uh, that Jack is talking about.

There's t tramper proof technologies, and we've had this for, you know, regular integrated circuit chips for a while too, so you can't get access to it and look at the code or manipulate the code. And that's essentially what we're talking about. So it's a bit of cat and mouse game, but I think it's, it, it's good.

In your article, Jack, I like also that you talked about exposing, showing everybody the chip that's there, as opposed to just embedding it within the document, um, which is another way of validating what it is looking at it, who the manufacturer is, all that kind of thing. Um, so I think there's good reasons for that, but yeah, I think it's, you know, until we have chips in our heads and we get, we scanned as we walk by, I don't know if that's really gonna happen. Maybe those will be manipulated.

It makes you think we're not doing it already. Is that what It is? I just wonder, I I kept doing these things.

I don't know that I'm, why I'm doing it, so that must be good. There was A movie like that. So let's, let's, Let's just say for argument's sake that I have some sort of thing on me that allows me to be authenticated by everything.

But how do I make sure that same thing isn't being used to track me and invade my privacy? And what's the trade off there? Well, right now you do Mike, and it's called a cell phone, right?

And so right now the cell phone is used as the primary authentication factor in 80 90% of authentication interactions. It tells people who you are, uh, why we trust the cell phone. I don't know.

But that cell phone also tracks you, right? Google and Apple and Facebook and a whole bunch of other companies have trackers on that phone that figure out exactly where you are at all times. So look, I've, I've said for a long time that as far as I'm concerned, identity and access management are the killer app of cloud security because we traded our perimeter defenses a while ago with the move to the cloud in 2006 or whatever it was.

And, and you know, so we don't have those perimeter defenses when we log in anymore, but we, we do have IAM and you know, it pains me that we're still stuck in the password conundrum. It, it, it's broke, it doesn't work. I think we all agree.

Um, I was always, I always felt we were gonna be moving more to biometrics, whether it be, it's something as simple as fingerprint or, or iris or, or something like that. Um, but identity is personal and, and we've gotta have a, I almost don't like using the chips and these things because they all can be cloned and used in a fraudulent way. I'd rather it be something inherent to our DNA pattern, which is unique, right?

And can we get, can we get to something like that? But then there's another issue, which is identity beyond human identity. We live in a world where machine identity is 10 XA hundred x the numbers of human identities, every container has a unique id.

Every API, every, every AI agent is going to need a unique id. And we can't do biometrics on AI agents. I think so I, I don't know if we, I don't even think we don't know what we don't know about this bigger problem, which is how do we apply identity control to non-human identities because they're just as important or, or they're, they're as much of a security risk as human identities are.

Then. And then finally, you know, what did Ben Franklin say? If you trade something for something else, you have neither I, I forgot to quote, but Mike, if we, if you know, if trading a better identity means giving up personal freedom, If you train liberty for security, you have neither is the Quote.

Exactly. And, and I'll say that, I'll leave it at that, Ellen, I think they're, we're in the continuous trap of looking for the next right answer. This is what's gonna solve it RFID chips, this is what's gonna solve two factor.

And, and, and Frank, and frankly Jack said continuous, or really when you're just kind of kicking off this topic, you know, imagine that security's checked continuously. It's not one thing we do when we log in, it's, it's checking because we have the device with us, but it's also checking other mechanisms, right? It's checking our past history.

It's looking at, um, the, the data that we have on our devices, but also the activities that we've had in, in the applications that we're using. It's almost like GPS, right? Isn't one satellite that tells us where it's, it's it's correlation of what three or four that that actually tell you where you are.

And you have to have at least those, those signals. And it's the timing between those that get verified, uh, to make sure that you're in the location that you think it, that you think it is. I think we need to think of security as it is this continuous, uh, conveyor belt of technology's gonna continue to prove.

And sometimes it's because bad guys figure out how to get around it, and other times it's because this makes it better, makes it, uh, less intrusive for the end user, et cetera. But it's not the next great, you know, chip layer that's gonna solve this problem for us. Kimberly?

Yeah, so when you were talking about the technology, talking to the technology, I would think that one of the, we have a, the technology there, it's called blockchain that does identity management across, um, those kind of environments. And it's probably time that we really implemented the more of the blockchain technology in those areas. Yeah, no, you know what happened?

Blockchain got sucked into the whole crypto thing and got a black eye. But, But I think that, I think there's areas now that, um, it's coming out. I mean, I was with, um, gosh, can't call their name out, but that what they are working on is, you know, those kind implementing blockchain more into enterprise type application environments.

Um, and, and using it for probably the first, the areas that was probably, you know, first to identify for. So blockchain is too slow for at the level of scale that we need, though it's okay for, you know, a couple of crypto transactions, but it isn't gonna work for, you know, an AWS Amazon e-commerce app. I don't know, But So like an Opportunity, make it faster.

There. There was a comment that mentioned CHIP and bio and CHIP bio, right? Biometrics is, and there's a lot of different biometrics technologies.

Uh, there's a lot of chip technologies, there's a lot of blockchain technologies. But Alan brought up something that I'd like to highlight, which is, if we think about it, and I mentioned this earlier on about our old castle and moat environment versus the way we do things. One of the things we used to do versus what we do today was zero trust.

And one of the things we used to do in Castle Mote Technologies is we had this concept of implicit trust is if you're inside the ring, inside the moat, in the castle boundaries, then you have implicit trust. And that led us to create and still exist today in the cloud environment. Anonymous identities, right?

We have a whole bunch of things that actually don't have identities, and that's where our greatest risk comes in. So all these cloud services, all these APIs, all of the AI agents don't have the identities built in today and don't have the identity controls and the continuous access and authentication checks that they need to. And the first part that we need to solve is that part, eliminating anonymous identities and then making continuous access checks.

And then all of the other technologies come into play and will help us solve these problems. It's Alan's point though. I think what he is trying to say is that we're all gonna have to get outta swab and then kind of run it around our cheeks and then touch that to our computer to get authenticated.

Come on my computer, ever. That's not personal information is it? Right.

So why, and when you do that, they say, Hey, you know what? You should be looking at breast cancer, skin cancer, clic cancer, or something else. Guys, we gotta end this show though 'cause we're over time.

You know, I, I love this live format. Kimberly, thanks for joining us. I hope you love the live format.

I know it was your first foray, Jack. Always great to have you, Mitch. I'll see you in a little bit.

Sounds good. Hey, for those of you who commented, thank you, we appreciate it. We're figuring out how to incorporate these.

We'll keep working. It's a work in progress. But have a great weekend everyone.

We've got Techstrong tv. com. Go register.

We're gonna talk about all this stuff and more this Alan Shimel and we're out.