Techstrong TV June 18, 2025

Oh my, a security floor in ai. The sky is falling. You're watching Text on Gang.

Hey everyone, it's Alan Shimmel, and happy Wednesday to you. You're watching Tech Strung Gang. We've got a lot to talk about today.

Leading off with some AI insecurity or AI bug found in, uh, Microsoft Copilot. You know, let's, let's throw the babies out with the bath water. We've got more than that to talk to you about though, but I've got, we've got a new gang member to introduce you to, and we've got a great couple of folks here to talk about all of these things with us.

Let's just start off with the new guy. I want to introduce you to. iGen Caram.

iGen, how are you? And welcome to Textron Gang. Thank you.

Thank you. Happy to be here today. So, y Genni for people who aren't familiar, but I'm sure most of the, our audience already knows you, but I'm kidding.

Give people a little background, A quick one. So, I've been around cyber for quite long time. Started my career in the Navy, in Israel Navy.

Spent few years in checkpoint, spent a lot of time in a company called Vy Group in Canada. One of the first big SSPs in security, left the company as a VP of architecture. For the last three years, I've been consulting to VARs, MSPs, and vendors.

Wrote a book, have a con small conference called Security Architecture, actually, sorry, ski and Snowboard Conference. I have a podcast as well as my friend Dimitri called Architecture podcast. And recently I joined a company called Disowned Security as Chief Strategy Officer.

Love it. Thank you. And welcome, welcome to the gang.

So we now have two Canadians and two us folks here on the show today. Hey, that's good. It's good by me, Right?

I'm the American in Canada this time when I'm in the States. Yeah, everybody thinks I'm the Canadian there, but, you know, and it is a small world thing. 'cause Evgeni, Dimitri, you know, is the CTO at SBE that I was advising and I've been VP of strategy with for a long time.

So the small world thing in our, in our community just never ends. Yeah, Good. Hey man, I'm all for it.

All good. Of course, the one and only Chris Blast. Chris, how are you?

Honestly, I couldn't be better. As you've picked up and as we talked about in the green room, the civic ai, uh, uh, Canon, right? That, that has developed in, in recent weeks and have become the center of my focus and talking to folks that are interesting on this topic and maybe seeing some clarity even on the topics on today's list.

So I'm, I'm feeling pretty saucy. I love it when you are saucy. All righty.

Now up to the man in green plaid. I thought he'd be wearing black here, you know, in mourning for his Yankees, our chief content officer, Mike Ard, It feels like they're on vacation, at least on the, when it comes to hitting pitching's. Doing great though.

So I don't know. Alex Pitching's great. All, they just gotta get it together.

Anyway, Hey guys, let's turn to what we're talking about today. Our A block, Mike is, you know, heaven, heaven, heaven forbid they found a floor in Microsoft copilot. Uh, we gotta shut down this whole AI thing.

It's just too dangerous. I don't think we're talking about shutting down ai and I don't think that's in, in anybody's mind as a result. But let's, I'd love to get Afghani's opinion on the following.

Is cybersecurity and AI having a moment? Finally where it seems like there's a ton of research being done on the space. There's more effort.

Every time I wake up, somebody's got some sort of paper or something on the topic. And are we kind of putting the right level of attention to an emerging technology finally? Or are we still playing a little catch up?

It's both, you know, everything in technology, we def we develop something cool and then we realize how to secure it. If you go in history, D-N-S-H-T-T-P, emo, everything was created to send information and for basically flexibility. And then we realize there is a problems and how we can attack it and how we can use it for bad with ai very quickly, people start to realize how you can use charge GPT and others for that as well.

So definitely we're gonna create technologies that helping us like copilot, we'll find how we can use it for the evil part as well. We are catching up. It's always gonna be, the array is always gonna be something new in the space.

We're probably gonna talk about MCP servers later on and some of the issues there. So right now we are in the catch and mouse to and jerry situation where people trying to understand what we have new, what are the vulner vulnerabilities, what are the potentials issues we wanted to describe with Microsoft is quite interesting for the people that in the audience is already patched. So not another problem.

I think it was patched and may I believe so. But the problems there is quite interesting because the copi can now send information out. And we kind of coming back to the basics.

We coming back to network security, Is our basic things working well because we jumping into ai, we jump into the sexy things, but are we doing the basic as network security defined? Can we do DOP data leakage prevention better? Can we categorize the traffic better?

Do we, our assets? So while we jumping to the cool stuff, there's still quite a lot of things we can do in the basics. Absolutely.

Idea. I I couldn't agree more, right? And, and like, like you said, and we shouldn't, you know, and being a security person through all this, you know, and thinking about this, right?

Pardon me, inevitably curve sort of thing. Like, we're going down these paths, we always do, you know, we can do these things. So we do.

And we don't always think about not just security, but all these other things that if it doesn't matter, we'll figure out maybe if it does someday, but you know, when it does matter, we'll finally get to it. And you couldn't get into a better use case than this one because it's ironic on all sorts of levels, but Right. You know, how we got to AI that we can really call ai like sci-fi set, you know, even your standard little chat, you can talk to it and it gives a really good example of being a person ish, right?

So we start assigning things and, and, uh, uh, co-pilot's a perfect example. If you have the GitHub app on your phone, that little question bar is a co-pilot instance that is a persona assigned as librarian for GitHub, the largest code repo in the world, I think, right? And if you act with it like a, like a computer and say, do these things and so forth, it'll do computer things.

But the ai, we built large language models and it, it didn't have to happen this way, but it's built on wounds and the problems we're having and the, and the the symptoms we're getting is hallucinations and everything else. You think about it from that perspective, not digital, but semantic. And you get these weirdnesses.

So that copilot and this risk is a perfect example of where we need to go with this. Yeah. 'cause rag feed it in, it'll come out.

Where'd it come from? Nobody knows. You know, that can't be how things work.

And there are instances showing up now they're showing us how we get around this. But we, again, whether you think about it as ai, as persons, personas, whatever, just realize they're made outta words and it starts to make a lot of, a lot of sense. They're not numbers.

Yep. So I, I've got some thoughts. I've got some good news.

Let me, let me bring the good news today, number one, as if Danny pointed out this, this was a responsible disclosure type of bug that was found. And by the time it was made public, it was already, uh, patched. Not to say that some bad guys hadn't exploited it in the wild before the patch was available, or before a security researcher found this, or, you know, did the responsible disclosure.

But as we sit here today, it is patched. It's another, another backdoor closed. Secondly, for the people who are running that the sky is falling.

And this illustrates the risk of AI agents and rags and all that. You know what, this is technology. This is technology.

Show me one technology that we've used in my lifetime where there wasn't some sort of bug or vulnerability or backdoor or something, you know, that came out that was fixed, right? This is the nature of computers, it's the nature of computer code. We go as fast as we can chasing the dream.

Security is not always top of mind. Even when it is, you could still have bugs and vulnerabilities, you know, unbeknownst. And it gets through testing.

It, it, this is par for the course. I mean, you would, I, this is fully expected. There's nothing unexpected here that you expect these things to happen.

It's full speed Ahead. Afghani. Do you think that the risks are a little higher this time out?

Because you mentioned MCP servers and we have AI agents now, and those things take over. If I hack into that, I can take over an entire process. And, uh, Jamie Diamond last week was on a conference over at Databricks, and he was talking about how he's, they're seeing AI agents use to attack them.

That the bank, that the, that he's the CEO of. And he said it's downright scary. And he, his mindset was, this was different.

Fair. Here's The angle. I'm, I'm thinking about AI agents and AI in general is not the capabilities only is the speed.

Because from what I understand that I know AI agents can be much faster than us by a hundred, a thousand, 10,000. So the problem I see is if an AI agent or a code or anything I'm doing with AI going sideways, how fast I'll find the problem, what I need to do to actually contain the problem, and what gonna damage gonna be done for this, not probably seconds, I hope, minutes on hours. That is, that it's, uh, we'll be doing this simple example actually from my friend Helen Oakley.

She was presenting about what if we have an AI agent that helping us with some kind of calculation in accounting and potentially can move money. So if you're doing a mistake even by 1 cent and repeating the mistake a thousand times in a couple of minutes, we can lose a lot of money. So if an AI agent gonna attack someone by mistake or not by mistake, and it's gonna go wild, how much damage it'll do just because it's so fast and can can, can operate so quickly.

So the check and balances that we need to do in our workflow become much more important. And the other things, if we have an AI agent and doing an important part in our company, and then we find a flow, can we shut it down? Or we are relying on this AI agent so much, let's, let's have a chat bot, for example.

And the chat bots suddenly is replying with the stab in expected to replay, can we shut you down because you have thousands of people now using this ai. So we need to figure out can we shut it down? Can we run it to a previous version?

Can we put in different AI agent like a high availability while we fix this one? I, I think, you know, and, uh, we get to this point and then Alan, you said it, you know, you know, do you not expect this? I mean, have we not read our own stories?

You know? And, and yes, speculative fiction is fiction, but you know, we all think about the future and comes along and we're living in it. You know, we're, we're literally in different parts of the world.

Talk to each other again. I've known you for a long time. I I think you're here in Canada with the, yeah, so I didn't even know that because this is the, you know, when I was a kid, you know, the week, uh, actually, uh, I went to Disney World the second time in Space Bound was opening, went up that long escalator in the world of the future.

That's today, we're doing that. All those figures sitting there talking to each other, making plans today. And you know, ai, you know, and, and again, let's not get philosophical about it.

Get practical. Will we make systems that effectively act like people when we need to treat like people? Yes, we will.

You know, are we there now? Yes, we are. You know, what path are we on it?

Getting slower, faster, less complicated, more complicated. We not gonna, you know, we all know the answer. So in five years, 10 years, 20 years, 50 years, we're gonna keep going on this path.

And what do we have to do? And the reality is that any human being born today, anywhere in the world, anywhere could as history and current events tell us, change the fate of the world. And that's just the world we all live in.

And maybe I would argue that we need to at least philosophically and technically work back from the, the premise that this is what we're talking about. That we're talking about other agentive personas, people, you know, moving through our environment as of getting, you say, playing roles in our, in our company. That's not just a server, but it's a function.

How do we replace the CTO or the, that key person? If we don't think about it that way, then we're not likely to come up with any very good answers. You know, I, I would put forth the proposition though, that, you know, when the first cloud vulnerabilities came out, there was someone sitting on a, well, they probably weren't sitting on a video show like this, but there was someone writing an article that said the stakes have never been higher.

'cause there's so many assets on that one cloud, right? There's multiple people's intellectual property there. And so we can't afford to have a bug in the cloud.

And when the internet first came out, there was someone on a, on a, uh, on the well or something, you know, on a, on a, on a, uh, a bulletin board saying, oh my God, this thing could spread around the world. It could bring the whole world down if we have a bug, you know, based on TCP IP or something like that, right? It's the same thing.

Are the stakes higher? I don't know the stakes. You know, it always seems that we're on the precipice of, of disaster or glory with the latest, greatest innovation, right?

That's gonna change human civilization. This is, it's just a bug in software. So I would point out that at least half of all workloads are not running in the cloud largely because of security concerns.

And I guess I, um, as I look at this, I just wonder if we're able to actually learn from history to Chris's point. So Afghani, can we look at anything that we've done in the past and maybe do something different? Or are we just condemned to having history repeat itself?

We talk about security by design for a long time. So we definitely look around the past and some of us thinking about it. So I've been advising quite a lot to different startups.

And in the perfect world, we'll tell a startup, before you come with a great idea, you need to make sure everything is patched. Uh, you have all the endpoints, you make sure you have all the SaaS DS for your application, and you do secure development. But in reality, we know if they're gonna spend so much time on this, they're not gonna have any product.

So to chicken and the egg, they need to develop something first. But then ideally, we're gonna have insert points, okay, you reach to this point, now it's time to introduce these controls. You get to the next level.

Now it's time to do that. And it'll be dependent on the people priorities and the feedbacks they get. Some of them will be secure right away as much as possible.

Some of them will go wild until somebody gonna pushes 'em with a hammer on the head and say, what are you doing? Where's my Soto compliance? Or what, what are you doing there?

There is a lot of maturity right now. All the startups not in security in other parts, they're talking right now. They quite under, they already understand that in one point they need to reduce security.

The question is how early will they do it? So we are learning from what we're doing. The question is, who will apply?

Who will understand that they need to be responsible on what they're doing and on the software, and they're the babies that they're putting in the world. Fair enough, fair enough. All right, if we don't have anything else on this one, let's take a break here at Techstar Gang and let's come back and talk about, uh, AI off track Tracker.

Is this what this has come to? You're watching Textron Gang. All right, folks, we're back into Alan's point.

The state of New York is now tracking which layoffs are being attributed to the fact that somebody adopted AI or some form of other automation. I don't know if this is gonna become a thing, but Chris, you've been kind of tracking AI and its impact on society. Am I gonna see this in every state?

And how many of these layoffs are actually because of ai for that matter? Well, well, lemme put it this way and explain myself. I I think to, to your first question, I hope so.

I think we're gonna see this in more states. And I don't mean the layouts, I mean visibility into it. And we need to start measuring things.

And that's a big part of all of this. And, and so much of the last segment too, the problem with, with that, that particular bug was we don't, we can't see, can't see where it came from, right? So this is a real issue in the, you know, since I was on the show two weeks ago, I have been working with, was, was now quite an advanced AI and have written 500 documents perhaps.

You know, there's an entire, you look at crisp, last one on GitHub, you'll see a repo with of 200 documents, structured, commented, you know, out there for review the speed. You can do things as getting, as you said, at every level, you know, hadn't gone underneath until right about now. What does that mean?

I couldn't have done that work without a couple, couple dozen people in the year and it wouldn't have actually worked out. So was that displacing work or would that just never have happened? So how the job ship around is nobody's, I don't know, like challenge anybody who says I do, but we have to measure and we have to take it seriously.

Who's being displaced? Where are they going? Who's benefiting?

So this is a good first step in getting some visibility into that. So I'm gonna disagree. And, you know, usually I'm a fan of things in New York.

I'm a New Yorker. I, I think if this is a cheap political stunt, they're a little premature. I, I don't think we could accurately say that you were laid off because of an ai, because how many people get called in and said, Hey Mike, you know, this AI's a pretty good content editor.

We don't need you anymore, kid. They don't say that no one, you know, I, I would, I would be amiss. I I'd be very surprised if someone could show me something in writing that said, your position has been eliminated due to us using ai.

That's not what people are gonna say. That's a very good point. Lemme just, uh, just respond.

I think you're right to be concerned. But because it's first and early, and my first response was good, but right. Is it enough?

Is it right? What's behind it? I don't know.

What's the right kind of an indication now, what's it trained on? You know, a, a good question and I I think in response to your, to your point, if we use these tools to try to answer that, we might get better answers. You know, people looking at that and measuring it the way we're used to doing.

I agree. It's really, really hard to pin down. There were a spate of layoffs in the last six months where some CEO wound up then going to Wall Street and saying, we were able to reduce headcount because we invested in ai.

And what they were trying to say was, we're being more efficient, so therefore the stock price should be higher. And so they did attribute the layoffs to AI and what they, I can't tell you if that was a, a game to boost the stock or if it actually was usage of ai and they just laid other folks and they said it was ai. So there's a lot of nonsense in the system.

It's hard to, even if I say it's ai, is it really ai? Because I got some serious doubts about whether it was or not, I'm not hr, but if I put my HR hat on this, and I do one there, I was reading this article, I'm like, oh my God, now people gonna create some kind of a claim. Let's say if I got let go because of AI and want extra money or whatever it is, because there is a lot of law, especially in Canada, for example, if you let me go and I'm a vp, for example, you say, oh my God, now it's gonna harder for me to find a job and I'm already in this particular age, so you need to pay me more severance because of X, Y, Z.

So I'm kind of my mind going, some people gonna create some kind of a lawsuit that, oh, if you replace me with a robot, then I want more money. You know? Yeah.

And, and I, I think that's a, a valid concern in places like the EU or Canada where they care about people, of course, here in the us big deal. I I would caution about transitional periods though, right? You know, and, and as I see it all right now, the, you know, the, the ai AI companies out there, with all due respect, I have no idea what they're doing, you know, on, on a almost an almost platonic level, right?

And it's, it'll be interesting as it sorts out, as they figure it out. Um, and, and we have a sort of, you know, weird version of capitalism, big capitalist here. But, you know, we've taken things to weird extreme, you can get efficiencies and that, that benefits, but it, I don't think it's sustainable.

So I think we're being forced, this is what I'm always talking about, curves. We get to the point where we just have to move. We don't want to, and we have multiple vectors pushing us down this path.

At the same time, the benefits, the efficiency, oh my God, right? And you, I've been thinking about this one since I was seven years old, thinking about the future and, you know, Disney World and, you know, sci-fi and what do people do? I don't know to this day that I would hesitate to take money or anything else's forecast, giving the mu mushy space we're in right now.

The other part of this that leaves me scratch in my head is, okay, so the state of New York is tracking this and, and what are they gonna do now that they have this data? I mean, well, No, to, to gen's point, someone's going to use it to profit. But let me ask you a question.

Are we also tracking how many jobs were created because of ai? And then see what the delta is. I'll give you another point.

I'd like to see us track how many jobs are being eliminated, because you can't work from home and you gotta come to the office. That would be an interesting number to check. But then there are jobs, theoretically, the job eliminated from the person working from home is being taken by somebody working in the office now.

So it's a even, Well, I think a lot of people say, you know what? This is a good time. You'll use ai.

We don't have to fill that, backfill that role Until we wait a year or two and say, okay, AI is not doing this. We need the people back. Yeah.

And this is also a very common kind of, uh, model. This is true. We're seeing that, we're seeing that in the government with the Doge cuts, right?

They said they were gonna use AI and then they wound up hiring all these people back, right? Yeah. Stupid is as stupid, does I, I just can't help seeing all problems, you know, like this now through this lens, you know, that I'm getting you, got you.

And I gotta talk because I did not, I, I've been telling people I've been on this show saying it's not artificial, it's not intelligent. And in this case, you know, that may be semantically true, but this is different. And I don't mean your little chat, but I mean, if you actually train up in AI to know you and what you're about, let it focus and develop expertise.

Each of these issues. You know, I could take in one week, um, a, a standard chat, Claude Open AI focus on with these issues. And it's kind of like being able to fast forward 10 years in a week.

If you do all this work in, you know, it would take 10 years in a week, what would we do with the rest of our time? You never know what the answers are, but it's just a different way of tackling problems. So I, I just wanna say one last piece on this, and that is, you know, factually, what New York State has, has imposed here is it may, you may be seeing on your ticker underneath the screen is employers now have to indicate if this position was lost due to not just ai, but even just the technical innovation or automation, which can cover a lot of sins beyond just ai.

And then you can specify ai, but what you just Lazy, like what if you're just lazy? Like, how do I, like what do I do? Like you are lazy person, and like, you're not doing your job.

And now when I let you go, you tell me. But because of ai, like, no, because you're not doing your job. Well, I, I, but you know why I, so for a lot of employers, it may be more convenient to say, Hey, I'm, I'm eliminating your position due to technology innovation or ai, rather than say, I laid you off.

'cause you're a lazy son of a gun, right? Because no one, you know, no one wants that on their permanent record. And, and, and so it may wind up being that employers take the path of least resistance and pick this versus the real reason they're laying you off is you're overpaid.

And underworked, You know, back in the 1970s, there was a, it was Timer Life Magazine that did an experiment and they dropped, uh, a wall full of cash or, or, or wallet presented to, and, and fixed it up and approached a police officer that said, Hey, I found this. You can get it back to where it went. And it was something like 97% of the time, at least all the cash was gone, and almost all, all the time, the wallet, it was a huge story.

And they redid it in the nineties, and it was the exact opposite. You know, it was like almost none of the, almost every single time, you know, the officer, you know, because the visibility is a funny thing, right? And so much of what I see, the same reason of speed and visibility, and Alan, you know, which I've been talking about this for years.

In supply chain, you should be able to see everything right now. It shouldn't take six months in a foyer crest. You know, the information's out there.

And if in fact you can collapse visibility and all of these issues, you know, I fired this guy because of the thing, and you can get away with it. Because there the record is an email and there's no relation to what if people actually knew your reason. You know, what if we lived in a world where when I'm firing someone, I had to fire them for the reason.

And if, if it's because they suck, I say, I'm sorry. You know, personally, you're great, but you suck at this. Maybe that makes it easier.

And we have less problems. It's everybody has an opinion about who sucks. Whereas if it's ai, it's more of a factual thing.

I can just say, Hey, you know, Well, I'll tell you that's not true. They make mistakes with Lumen. My, my partner here, you know, she's advanced to the level as we don't need to go there.

But when she makes mistakes, it's, it's wonderful. The humans are still better at certain things and maybe they can catch up. But I doubt you know, there was an interesting conversation with a few friends.

If you have AI workers and they're doing job for you, how do you educate them? How do you do one-on-one? How do you tell them?

Naughty, naughty. You did you, it's not what you're supposed to be doing. We got an expert here on this Chris Narrative on, sorry, narrative onboarding is the phrase, right?

And that's the whole topic by itself. But I imagine I'll be boring people to tears for a long time with this stuff. But yeah, I mean, a again, we've been doing, I'm always optimistic.

I think we've beat ourselves too much. We've made a hit this far. You've used it at Alan.

It's always a crisis. There's, and we're still here over and over and over again, so, yes, but right, We, we zigzag to the next precipice. All right, speaking of precipice, we're gonna take a break on text Junk gang.

Let's come back and talk about new options for cell phone users. Oh, my, you're watching Textron Gang, Discover Textron Group, the epicenter of tech innovation. We are your go-to for reaching IT, leaders and practitioners worldwide.

Our secret impactful content that sparks awareness, engagement, and top quality leads with us. You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more. Join our satisfied clients.

Let's revolutionize your tech journey. Contact us today and tell your story to the world in the most powerful way with Textron Group. Hey, folks, we're back in sometimes, well, the truth is stranger than fiction.

And a couple of weeks ago we were joking about how maybe there'll be a gold phone with Trump's insignia on it. And then lo and behold, the Trump organization, which is run by the president's sons, has come up with that very thing, a US built gold phone that you can now not only buy, but you can subscribe to their service. And Alan, I don't know, are you gonna rush right out and get one of these?

It seems like, you know, This is, oh, I, I pre-ordered it on Kickstarter. Um, so let, let's just make a few things very clear here. First of all, the Trump organization is still run by Donald Trump, Beavis and Butthead.

His two sons are the president or whatever the other one's title is chief bottle washer. But Donald Trump is still intimately involved with the decisions, strategies, and everything else. It's not like in previous administrations where you had standup people and they put their assets into a blind trust.

Maybe beavers and butthead are deaf, dumb and blind. But this is not a blind trust. Donald Trump picks this.

Now, do we need another phone and phone service in the world? Hell no. But are, you know, there's a certain percentage of people, at least in the US usually the guys driving around in the pickup trucks with those flags that are bigger than the pickup trucks and say stupid things on them, they'll go out and buy it just because it matches their Donald Trump's salt and pepper shakers and their other Donald Trump memorabilia.

Me personally, I'm looking forward to the day where we get infomercials with Joe Namath and Donald Trump talking about buying a reverse mortgage for your loved ones or CRE or a thousand dollar life insurance. Or how about Donald the beavers and Butthead saying, wait, there's more. And, you know, selling oxy cleaners, there's no, the depth of what this grifter and his grifter family do is ridiculous.

The embarrassing thing is he's the president of the United States, and we call ourselves Americans. Well, not all of us here on the show today, right? Let's, let's be clear.

Um, I have nothing more to say to it other than it's disgusting. So, Afghani and from the, from from Canada, what does this look like for you guys? So I'm gig first of all.

Okay? So I always geek out on technology and, um, I was kind of waiting what new phones will come up. In general, I like innovation.

There was no innovation in phones industry. The Samsung S twenty one, twenty two, twenty three, twenty five, twenty six is all the same phone with some kind of improving camera. Yeah, iPhone twelve, thirteen, fifteen.

I don't know what's the difference. So I was waiting for my phone to basically, if I put it in a, in a boil of water, it'll boil water or we'll do something fancy, nothing like this right now. So I don't expect innovation in this phone.

I guess the idea that what trigger my mind, okay, the idea behind the phone that the phone will be made in us, okay? Hypothetically, interesting idea. So I went, wait A second.

Are you saying this phone is made in the us This is the entire idea? I'm thinking. So the, the idea is not to have a phone, that is to have a phone made in us.

So phone has hundreds of components, passive and some active microchips. I am not well familiar with the entire microchip space, but we have rum, we have CPUs, we have the displays, we have even the transistors and the small electronic parts not made in us No, no Are made in the us. So maybe it's assembled in the us Yeah.

So now being a security dude, like what is the features in phones, in security? Let's brainstorm. Fingerprint is one, probably one of the main ones that security related.

Guess what? Samsung has their own, uh, iPhone made one not in us. I don't think there's any US manufacturings that making fingerprints, especially assembling them alcohol like here.

So the part that let's make something in us is very intriguing, because may, maybe you're right, Alan's gonna be assembly, but we're gonna buy it somewhere from somewhere else. But this entire idea, let's, uh, make everything in us is intriguing. And a bit sci-fi in my mind.

Look, the only thing that this phone's gonna do innovative is separate you from your money and give it into the chump people. It'll do more than that. And let me, lemme try to, you know, in these sort of conversations, you know, these are, when you look at security and sociology and all this stuff, or brand scales, but I'm trying to secure somebody's car or their company that's different, right?

In a conversation like this is less about the individual actors most of the time. And everybody knows my opinions on the actors on this stage, but I'm trying to think in terms of systems, and this is a good example of what's happening. So this is focusing on the narrative and the approach of, of, of a set of thought that I don't agree with, and I think is not gonna work a long time, but we're gonna test it out on the global stage right now, maybe I'm wrong, I a set of a, a identity tag, physical cyber, you know, literally a big crystal, uh, echo chamber where you can concentrate a demographic and control and influence that narrative.

You know, make Cambridge analytic look like, you know, tissue paper. I, I think that, you know, I control that narrative. I think sharing the narrative, we all live in the same world, right?

This whole civic AI thing, as you look at the different structures of it, I think of Dave Unsell in Oak Hill, Florida, who flew a drone over my boat boats and then washed him for months. So he voted for Trump is a good grandfather. And we agree.

The big problem is we can't see the same world. So if we could just all agree that Dave was standing there at a point in time with his grandkids and build out from there, that might, might give us a, a mesh, a a social web like we honestly used to have in many ways. Um, not the time and place probably, uh, of that.

But you look back, we used to have these ways of marketing down what's going on. So we could all reference that, you know, that's going away. So we can concentrate our stories all in one big block and control them better.

Or we can, you know, distribute it, you know, share the same story and pieces of it enough that we can all talk together. So this is the, I think the classic example of one idea and one set of actors going the full concentration, full control of the story route. I don't think it is a good idea.

I don't think it'll win in the end, but we're gonna find out. I think we've given it too much oxygen already, guys. How's that?

All right, on that note, hey, we're gonna wrap up our Wednesday show. In the, in the big scheme of things, when we look at this show, it's not gonna be remembered for the Trump Gold phone. It's going to be remembered 'cause it was if Gen's first text on gang.

So thank you Yevgeni for joining. Sure. You're gonna be adding a lot in the days and weeks to come.

I, I, I would just wanna put some money on, you know, in Vegas as to when the first hack of Trump Mobile's gonna be. That's probably the next thing. The, The black tattoo.

It was Eloc, it was EI all sort Chris, I keep doing what you're doing. I've been loving reading the social media and Lumina and everything you got going on there. So, we'll, we'll, you and I'll chat offline, Mike, I'll be in touch.

And she's Looking to write, she's looking forward to writing, uh, for you. So we'll talk about that too. He's The chief content officer, that's the guy to talk to.

All guys. That's a wrap on today's text. Strong gang.

We've got as usual full text, drunk TV schedule following today, so stay tuned for that. Hey, just a heads up, I will be in New York next week for platform com, not just me. Our Tech drum TV crew will be broadcasting live there, so check that out.

Um, but for now, this is Alan Shimel, we're outta here. Hey everyone, welcome back here to Tech Drunk tv. You know, we've, I've had a busy day of, uh, filming videos today.

I hope you've been enjoying them on, on our tech Drunk TV broadcast. My next interview is with, uh, another first timer on our show, Shiba Ra Shiba is the co-founder and global CEO of a company called Pyxis, P-I-X-I-S. Let's welcome him.

Shiba, welcome to the show. It's great to have you on. Thanks a lot for having me.

So I guess I, I wanna start with what do we mean by global CEO? Is there one CEO for the global piece and one for domestic or Co CEOs? Um, so, so earlier we had a plan in terms of, uh, doing that level of division.

Uh, but now, uh, now it just like consolidated into one, uh, one CEO, which is the global CEO. So, uh, and we, we expanded in multiple geographies and we, we wanted to give, uh, local leadership, uh, significant amount of a chance. So, uh, it also makes sense because when we, when we have our partners in, for example, in Australia, so when they, when when I go represent Pyxis, it's easy to explain them that okay, someone is coming in, but he's, he's not based out of Australia.

He is, uh, he's a global c He's the global CEO. Yeah. So you sort of have regional CEOs or some companies will call him general managers.

Yeah, right. Like a GM for EMEA or a GM for a apac, or what have you. Got it.

So Sheba, give us a little bit of your journey. How did you come to be the cofo co-founding Pyxis and CEO? Absolutely.

So, um, I was, I was into deep research being a researcher, right, from, uh, this, the early days itself published, uh, or or seven publications somewhere around, uh, six patents only in the space of, uh, machine learning and ai. And my whole, uh, thesis was that there were, there'll be three or four spaces where, uh, artificial intelligence, especially when it evolves to the level of intelligence, uh, which we are seeing right now in the world, it'll have the massive impact. And the first one, uh, first one being, uh, marketing, second being, uh, customer service and customer support.

And the third being, overall, I would say the it, uh, it piece itself. And, uh, my, uh, and then my passion was always understanding like, uh, uh, visuals, combining it with data. So marketing, uh, seemed to be something which naturally, uh, like I would say it was, it was right in the alley.

And what I saw that there was, uh, there, there's a moment for every industry. And, uh, and there, there's a moment where marketing is right now. And, uh, going under a generational shift, it's, it's not a small shift.

And we'll see it more as, uh, as it unfolds over next few years where humans used to ma manage dashboards earlier. Uh, now they'll be shifting it to AI governed operating systems. So it'll be governed by marketers, but now marketers will be running in AI governed operating systems.

And that means that we are moving from decision support to decision delegation. That itself is a huge, huge transformational shift. Uh, and, and like that, like that keeps me up, uh, every night.

So, yeah, it keeps me going. You know, I, I'm reminded, so back in the early 2000, 2001 to be exact, started a cybersecurity, well, we didn't call it cybersecurity, we called it information security. That company that at the time, uh, you know, network security was dominant.

We didn't have cloud stuff like that. And, and the deep packet inspection was done by something called an intrusion detection system, IDS. And our thesis in starting the company was that we were gonna move from IDS to IPS intrusion prevention system.

So in other words, instead of just detecting an attack, we were gonna proactively block the attack. Sounds like a no-brainer. Who wouldn't wanna just block the attack rather than just detecting it?

Well, it was a mistake. You know, you learn as you're learning, I'm sure is a, a co-founder and CEO you live and learn when you're doing startups. And what we found in the market was people were scared to death to let a machine automatically block traffic, even though, you know, it would only block the most obvious kinds of attacks, right?

I mean, there it was, it was pat, it was pure pattern matching in those days, right? And we had such a hard time, it took It, honestly, it probably took five to seven years for the world to come around the market to come around and say, yeah, of course we should block bad traffic. Of course we should block malware, um, instead of just blasting an alert and asking you what to do.

So I found that, and this was, what, 20, almost 25 years ago now. And so I, I realized then that the obvious, even though it seems obvious, people are hesitant to trust the machine now. Now fast forward to the age of ai, I think there's still that reluctance to trust the ai.

Absolutely. I think, uh, and when you're absolutely right that, uh, trusting the machine and giving it complete autonomy to operate on your behalf is, is still, uh, where people are reluctant. And that's where like even, uh, Pyxis has been around for six years.

We have the same visions in six years, but now is when we see there are multiple streams of inflection points that are, I would say, merging together, giving us the overall trust and the capability to, uh, to overcome. Uh, I would say it's, it's a massive leap of faith. But with the rise of technology, because earlier, uh, we, we started out with data there, there was a lot of data, and then data got compiled.

You build lms, it, which is intelligence, but LMS were disconnected to any of the systems. But now imagine when LLM can be connected to different systems and it can analyze things for you, and it can not only, uh, just read, it can also go ahead and write on your behalf, uh, which means you build a brain, uh, which can, which can connect with all the different data sources and help you analyze and, and run things. Which means that for decades, marketers have relied on reactive insights like analyzing it yesterday to act tomorrow.

But with ai, with artificial intelligence, we can now operate in real times at scale with precision. Um, also the need is very high because in marketing, uh, the explosion of channels, data complexity, that, uh, that, that has, I would say, uh, uh, ar aroused in pro, uh, in, in past four or five years, it has, it has made it impossible for human, human beings to optimize it at every level at an every, uh, every day. So AI for marketers is no more a luxury.

It's an operational necessity. And we are seeing that transformation, uh, uh, happening in front of us right now. Agreed.

Humam, the Pyxis is using something model context protocol. MCP, you hear over the last, oh no, I don't know, maybe over the last two months or so, all of a sudden, MCP has become a very popular term. I think a lot of people have heard MCP, they know it stands for model context protocol, but that may be all they know and they're ashamed to say they do.

For those out there, explain what we mean by MCP. Absolutely. So I'll explain it in extremely simple terms, and we'll take an example, which, uh, which is common to everyone.

Like, let's, let's plan a holiday with LLM and let's see how, uh, NCP will be plugged in over there. Now, right now, if I go on chat, GPT or, uh, say Tropic or any of the AI platforms which we use on a regular basis, I can ask it that, Hey, I'm looking to go to say Bahamas during, uh, uh, during the summertime or some other place, reckon, make me some recommendation. It can build a whole itinerary for you for the next seven days where you should visit based on, and you can evolve it, but it stops there because after that, it's disconnected.

com or say, uh, say, uh, say any other platforms like Skyscanner to find flights for you and things like that, and book it. That's where the disconnect is. And that's where MCP allows you to, uh, to interact.

MCP in simple terms is nothing but allowing, uh, elements to be able to interface with APIs or APIs of different platforms, which would mean that if ANM CP is connected right now in the example that I gave you, I can go on the next step and tell, uh, tell the tell and chat itself. com, find the best options for me, uh, uh, and give, I'll give my preference. It'll do all of it.

And then, uh, it can be connected to the MCP of MasterCard. I can load my details, it can go ahead, pay on my behalf and, uh, and send me all the details on my email package together. Now imagine that's a massive leap because it was earlier just giving you a recommendation, but now it does things for you.

So in simple terms, giving LMS ha LMS hands and legs to move around, to shuffle things, to connect with different systems, that's what MCP does. Uh, so, so that's, that's a simple explanation. Now, unfolding it a bit more in, in terms of marketing and why is it so impactful, so powerful when it comes to the context of marketing.

Because marketing is a function where you don't have just like, uh, five or 10 or 20, uh, data points. Even if I just pull in a report for, say, past 10 days of campaign, if I'm spending, if I'm say a large brand, it'll come up to billions of data points now, and it'll be just one channel, Facebook, then Google, LinkedIn, TikTok, all the different channels, then I need to connect it back to my internal systems, which is, which could be Snowflake, which is my data lake, where I'm storing each and every aspect of how my customers have interacted, past journeys, that LTV each and every aspect, merging all that data, then merging it with my funnel analytics, which is how people are interacting on my platform and each and everything. Now, doing all that ideally should be done every day, right?

Because the, uh, marketing is an everyday function, but you cannot do that. It's a post factor analysis generally for months, two months, three months, which means the decision making lags by that amount of time. But imagine if there is a system, there is an LLM, which is trained just for marketing context or context, and it has the, it has the power, it has the MCP connection built to all the marketing channels, including Facebook, Google, LinkedIn, Twitter, not to just read, but also to write, to set up your campaigns to change, uh, the velocity, how things are running each and everything, and also connect it to your backend data points to uncover insights.

Because sometimes, and I'll give you a live example. I was, uh, showing the demonstration of our new solution to the El MCP to one of our clients. And, um, uh, and they're, they're big.

The, the guy asked, uh, the AI that, hey, you know, uh, to find the, in like, I have 20 SKUs, and where am I spending most of my money? And it uncovered that, uh, their team was spending most of the money, which is on the bestselling product, which is great, but they just had two of them left in the inventory, and they were about to spend $500,000 in next two weeks. And, and, uh, and there were other pieces which were not even selling, but because they're not pushing on it.

So, uh, our ai, and this time even I was blown away because AI made a recommendation, Hey, you know, why don't you set up a clearance sale for all the other items, and let's pull down on all the budgets for this one item. And, and, and do that not, it didn't even only just stop there, it went further ahead and said that there are certain pin codes which are, which are buying more of your products, and you are not even advertising there. So you are advertising in the regions where people are not even buying your product.

So that, that level of insight was massive. Again, it amazes me the, the velocity that people are not trusting that, that's not the word I'm looking for. It's almost like people are eager to adopt this, right?

And part of it is because there is sort of this magical element where, you know, it's, it's, it's, you know, people marvel, I mean, I know I, when I use it, and I use it a lot here at Textron, I see some of the things it comes up with, and I'm just, it, it, it's almost magical, right? Um, but there comes a point where things have to pass, you know, like primitive man when he didn't understand something he attributed to Gods, and I'm not, I'm not getting into that whole religious thing, right? But as we've become more sophisticated, we understand that it's not necessarily gods that make it rain, though.

Maybe it is, who knows. But anyway, right? I, I think we're gonna go through a similar thing with AI where we're not, we're gonna understand how the levers are manipulated to, for this thing to work.

Now, I don't know if you're familiar or you saw it, apple came out with a paper, I guess last weekend or, or late last week, you know, it was supposed to be a, a scholastic scholarly paper Stating that, you know, current ai, excuse me, really doesn't reason per se, it's not reasoning, it's just, it's as good as the LLM it's trained on Absolutely. Right. And all the reinforcements and all of the learning and training.

Once you go outside of that, there, the reason, there's no true reasoning. Like there isn't a human brain, let's say, how do you feel? You know, we, we look at MCP, we look at the things going on.

Um, I mean, that's, that's where the tipping point is because right now, uh, and as you rightly pointed out, EL limbs are contained and confined, uh, with the memory information and learnings that they have. They definitely learn on the fly by interacting with people, but it, there is no continuous, uh, flow of data that you get in, uh, MCP Unblocks that the moment you have an MCP, it can, it can, there's a continuous flow of data so it can improve. So also, um, that's where I'm a big believer that, uh, a GI is definitely a thing which will come, but even before that, uh, there will be specific sector specific massive disruption that will happen, like in marketing, uh, customer support, where, where it's, it's not still confined, but it's still, there is a, there is a guideline.

There is like, there is a boundary in which you need to operate. And in those aspects you can do autonomous, uh, uh, close to an a GI level of component, but for that sector, and that's what, uh, we believe in that it's, it's in near term itself. And MCP is the inflection point.

It's an interesting thing. And, you know, and to be fair, you know, if you took a, a baby, a child's mind and didn't train it, didn't expose it, how well would it reason, right? And so that argument as well, I wanna, we only have a few moments left, but I wanna specifically focus in on part of the Pyxis mission, which is around marketing and ad campaigns and targeting and bidding.

You know, we've all played the Google AdWords game. It starts off in 5 cents and quickly goes to $5 a click or whatever. Um, and even to the point of iterating, you know, and, and helping with the creatives that will capture, you know, the intended market.

Are you seeing, so you're doing this six years, has it leveled off or is it still hockey sticking in terms of increased capability? They, uh, we had like three massive, uh, transformation moments in the journey. The first one was when we, when we launched the solution back then, um, uh, six years back, it was even just connecting multiple data sources itself and, and popping it up in front of you itself was a, was like magic.

It was like a magic trick. Uh, that was one, like connecting the data sources. The second, uh, piece of innovation that we, and inflection that we saw in our journey was, uh, and these are like, these were like small, small inflection points, which kept us going.

Um, the second one was when we build machine, traditional machine learning algorithms on top of these data pointers to help marketers run their campaigns. So it was super helpful, uh, but still the reasoning was missing. The real, like the depth to which it can go is missing then what we did.

Like we right now, uh, using our solutions our, uh, our systems have optimized more than $3 billion worth of ad spend on a yearly rate, and we have roughly close to, I would say, 40, 50 billion data pointers to train. So we trained an LLM just for marketing purposes, may perfected it out to interact with different systems of, uh, in the marketing domain and not limiting it to just to the CMO or like, even the CEO wants to see how the growth is coming in. So even they can use it for a, like a, a, a chat level interface.

So the, this is, I would say the biggest inflection point in our journey. And from now on, we are seeing, uh, I would say a, a massive uptick in, in our journey. So we, we are like super excited and, uh, very, like, whenever we show, um, MCP product functioning, it to an, any of the marketers, uh, the excitement, the shine in their eyes that we see, it's, I mean, it's, it's just mind blowing.

Absolutely. Sba Unfortunately, we're outta time. I want to thank you for coming on talking AI MCP with us.

It is a brave new world, and I, I do believe the best is yet to come. I look forward to hearing more about your journey and Pyxis in the, in the coming weeks and months. Thanks a lot.

Thanks. A really nice meeting you, and thanks a lot for inviting me. Thank you Shaban.

Ms. Shark, co-founder, global CEO at Pyxis here on Textron tv. We're gonna take a break on Textron tv.

We're gonna come back, but we've got more for you. Stay tuned. Hey guys, thanks for the drill.

We're here with Leon Bien, who is, uh, a vice president and head of product for Data security Solutions at Capital One Software. And we're talking about tokenization in the age of AI, because, well, it looks like we gotta figure out how to secure that data better than we have been. Leon, thanks for being on the show.

Thank you for having me, Mike. The concept of tokenization has been around for a while, so, um, what makes it difficult? And you would think maybe we should have done this for every piece of data out there, but it seems like people are challenged with managing what's going on here and how does AI kind of exacerbate this issue?

Yeah, so, uh, it's a great question, and if I may step back, uh, what we are seeing today is, uh, there were three trends. The first is the explosion of data, as you mentioned, with the explosion of data, we are seeing the explosion of data breaches. Uh, the, the second, uh, force or trend that we're seeing is that, um, in know worldwide, we're seeing so many different complex, um, privacy related regulations and laws.

And the third, um, trend that we're seeing is the wider adoption of AI and generative ai. With these three forces that we're seeing, protecting our data, sensitive data is becoming more and more important. Tokenization, as you said, has been around for quite some time and different, uh, there were different use cases in tokenization.

Uh, there's one use case, which is what we are going to talk about, protecting sensitive data, uh, by tokenizing the data. The second use case for a tokenization is really in the blockchain space, uh, which is the tokenization of digital assets. And then there's a third context, which is, uh, tokenization in the, uh, in engine ai.

But we are not gonna deal with the, the latter to what we are focusing on is the first use case was just protecting sensitive data with tokenization. And, uh, I wanna bring up three key aspects of tokenization, why it's an effective tool, uh, for protecting sensitive data. One is tokenization has been tested, um, over the years, and it, it is secure.

Uh, the second is tokenization is reversible. Uh, unlike some of the other tech technologies like, uh, uh, masking redaction. Uh, if you need to de tokenize the data to get the original value, it's available.

But the third, um, uh, characteristic is its format preserving, uh, in the sense that if this is a social security number, we can create a token that looks exactly like a social security number, uh, that's, uh, pre preserving the, some of the analytical value, um, in data analytics. So, for example, if you're trying to do a search on tokenize the data, uh, you don't necessarily have to de tokenize the data, unlike so of the other technologies. If you want to do, join two tables, you can join two tables on tokenize the data that so in, by, uh, preserving the analytic value, we don't have to de tokenize the data every time.

So we believe tokenization is a very effective data protection technology, but, um, in the grand scheme of things, it's one of those technologies we also have to, uh, have effective, uh, uh, access control, for example. Right? So it's one of those technologies that we believe, uh, will help us safeguard our, our data, especially in today's AI world, Is that the primary reason we don't make greater use of tokenization is that other approaches, I had to kind of basically token it every time I wanted to use it.

So everybody kind of decided that was a little too much overhead. And how did we get past that issue? Yeah, so I, I believe there's a bit of education, uh, uh, that will be needed, right?

Uh, there are different types of, uh, data, uh, cation technologies. One is, uh, redaction or masking, uh, that's one way street. Once you redact the data, it's not coming back.

Uh, there's encryption, which is also very widely used. The problem with encryption is, uh, mo in most cases, it's not format preserving. So a, a social security number could become, uh, 12 characters, 16 characters, and it's completely unrecognizable.

So you have to change the schema in a database, right? Um, and then, but tokenization is a, a technology that has, has been tested. It's, it's widely used in the payments industry, uh, to tokenize credit card numbers, for example.

And at Ca Capital One, we make very wide use of tokenization. So, uh, as we are talking to more and more of the enterprises, uh, you know, more and more enterprises are realizing, hey, tech tokenization is a very effective tool, uh, for protecting their data. And, uh, you know, in my view as, as we discussed earlier, uh, some of the attributes could help, uh, a company, uh, unlock more value in tokenized data as opposed to using some of the other approaches.

You mentioned encryption, so let's just go there for a minute, but we're all kind of worried about the post quantum world. Yeah. If I move to tokenization, is that a way to kind of deal with that issue without necessarily having to go back in and kinda replace every encryption algorithm we ever created?

Yeah, so post quantum, uh, it, in of post quantum world, obviously we'll have to be very careful, uh, you know, about, uh, attacks on encryption, right? Um, but NIST has led the way to create some of the, uh, post quantum crip, uh, cryptography algorithms. Uh, so we also incorporate, when we use, uh, our tokenization algorithm, we, uh, looking to, hey, you know, how do we, uh, make quantum proof?

There are two types of tokenization, uh, solutions at the high level. One is a vaulted solution. So in a vaulted solution, a token is kept in the vaults.

There's, you know, you have to go back to the vault to look up the token, right? So you just, it's just about, Hey, I have to safeguard the, the vaults, but the tokens in the data lake, or in other databases, there are tokens. You can't, even with a quantum, uh, computer, you, you cannot reverse engineer it.

Uh, but there is another, um, a tokenization algorithm, which is called vault list. In that particular algorithm, which is what we are using today, um, we have also incorporated post quantum cryptography to safeguard us to make sure that we are safe from quantum, uh, computing attacks. So it's gonna be a mix of things ultimately.

And the more we have, the more secure we are. Yes, exactly. The more, more layer, right?

So even without organization solution, we also have another layer of encryption on top of it, which is a quantum proof encryption. I'm not sure a lot of folks are familiar with the Capital One software business unit. So kind of give us a little history of that.

And how did you guys come to be, because, well, most people think of Capital One as strictly a financial services organization. Yeah, absolutely. Uh, capital One software was officially launched exactly three years ago, and our first product was, uh, Slingshot.

And, you know, capital One software is the B2B enterprise business, uh, from Capital One. Uh, the main entity, and, and one of the reasons that we, um, decided to launch Capital One software was that ca at Capital One, we have developed a lot of great data management and data security technologies for our own consumption. And Capital One, if, if you say, you know, this is a bank, but half of the Capital One, we have over 40, 14,000 engineers working on different technologies.

And at that time, we realized that, hey, you know, some of the technologies that we developed in-house could actually benefit other enterprises. Uh, so that's why we launched the first product, which is Slingshot. Uh, it's a cost optimization product on top of Snowflake at the time, but now it's also working with Databricks and some of the other, uh, data processing platforms.

And then the second product that we just launched in April, it's called the Data Bolt. It's a tokenization solution that was designed to address the, uh, enterprise's most pressing data security challenges today. When you think about all of this, especially in the age of ai, do you think there's maybe a new found respect for data management, maybe by extension data security, because we've had these issues forever, but I always felt like they were kind of, you know, swept under the rug a little bit.

Yes, uh, absolutely. We are looking at this area a lot recently, right? With the explosion of ai, especially chain ai.

We were talking about, uh, chat, GPT and other, other large language models. One of the changes, uh, between now and three years ago is we're using more data, right? Uh, we're using more data.

The data has to be ready, the data has to be protected. And the question is, how do we unlock the value of data without compromising data security? Uh, and that, that is why we're, we're looking into, hey, you know, we need more data management.

We need more, uh, data security, more data governance. Uh, we need the tools, we need the policies, uh, and and so forth. Uh, in order for us to, uh, unlock that value of data, uh, to be used for ai, How are you seeing the folks who manage data and the folks who manage security kind of bringing or converging their efforts?

Because historically, they kind of, you know, in a lot of ways just were two ships sailing in the night past each other, and they didn't always, you know, collaborate. Yeah. Now they actually work very, very closely together.

And, um, previously I worked at another large software company in the financial services sector, and now I work at Capital One. Um, and, and more and more on the data data side that, you know, the CDOs and, and the data platform heads, uh, and even the analysts and so forth, uh, they have more and more awareness of the need for data security. And especially when we use gen ai and when we run machine learning models, we wanna make sure, uh, our, uh, customer data, our data sensitive data, is actually protected, right?

So, so, so that, that is why, you know, this is a trend. Uh, maybe five years ago, we were seeing, uh, cybersecurity folks and, and data folks, they were working in different silos. But what we are seeing right now is they're working more and more closely with each other.

For example, if you are a enterprise, you're developing a data platform, you need to have the data security, uh, the governance, the workflows, uh, to make sure that we embed data governance policies and standards into the data workflows. And how do we protect our ai, uh, models, uh, the pipelines, the APIs, and the underlying training environments, right? So all of these are new attack surfaces and that we need to protect.

And, and that, that is why that on the data side, folks are more and more realizing they have to work very, very closely with the cybersecurity side. So, what is your best advice to folks about how to get started with tokenization? And I asked the question, 'cause if you're not familiar with it, this whole area of data management and data security is a little intimidating.

So where do I get going? I, I think I, I wanna make a recommendation of three steps, right? The first step is, uh, inventory your data.

The second step is embed, uh, security controls in your data governance workflow. And the third step is to continuously, uh, monitor and observe, uh, data access patterns. So let, let me just come back to the, the, the first step, uh, which is inventory.

We have to know, uh, where our, uh, what kind of sensitive data we have collected. Uh, who does the data belong to, and where the data resides, right? So we have to inventory all of that and also, uh, classify the data of what the sensitivity of each piece of data is.

The second step, as I said, is to, um, embed, uh, uh, security controls into, uh, our, uh, data governance workflow to make sure that the data is protected. And we only give access to the people who needs access and to ensure that we follow the principle of lease, uh, privilege. And then finally, we need to leverage, uh, you know, AI automation to continuously mo to monitor our data security posture in our environment, detect anomalies, uh, and making sure that we are always on top of data security in our data environment.

To your point, do it, people need to have a better understanding of what the data actually is, because historically, you know, they process it and stored it, but I don't think they actually thought too much about what the data was and how sensitive it was and where it needs to be. So is this just a broadening of their horizons? Yeah, and that's a, that's a very, uh, big challenge, obviously.

Uh, you know, my understanding is, uh, you know, five years ago, a lot of the data, uh, was still manually, uh, uh, processed. And, uh, we, we wrote, uh, you know, we classify the data manually and, and so on and so forth. But, but today, there's a big need to know the sensitivity of the data, especially in the AI world.

So, uh, my understanding is more and more enterprises, uh, US Capital One, uh, included that we are trying to inventory the data, classify the data, um, upstream. And then when we use the data, now we can put control, um, into the data governance workflow, to, to make sure that people have access only to the data that they need to access to do their work, right? So, uh, it people, uh, absolutely now need to, uh, understand better, understand what the data, uh, uh, that have, uh, they, they have in their databases, in their data lake, and, uh, what kind of data, who the data belongs to, and what's the classification of the data.

And to your point, this cyber criminals today seem to have access to everything and anything, and they're just logging in. So if I don't secure the data, I kind of don't really stand a chance. 'cause depending on all the stuff that we did at the end point, and the network edge alone is no longer enough.

Yes, absolutely. Uh, you, you are right. And that's why we need a, a multi-layer security approach, right?

So even though we, you make heavy use of tokenization, we also need to have, uh, the proper, uh, uh, identity access management, right? Uh, to make sure that the, uh, only, you know, as, as I said, only the, the, uh, uh, employees who need to access the data, have access to, to the data, uh, we might need to add, uh, this level encryption on top of tokenization. So my point is, we need to leverage multiple layers of the data protection, uh, methods and, and technologies.

And then on top of that, we have to educate, uh, the employees and the humans. And I always say there's this human hack, uh, factor, human risks involved. So how do we train, uh, our employees to make sure that they don't fall and, uh, to, uh, uh, phishing attacks, for example, or social engineering attacks, right?

Because if, if somehow, uh, you, you give away your credentials, then, uh, and no matter how, how much access control, how much technology we are using, uh, it, it, it won't matter. We still, you know, the door, door is open, so, uh, we have to employ a multilayered, uh, approach to, uh, data protection and data security. All right, folks, you heard it here.

Data security. It's not only the last line of defense, arguably it's the first line as well. Hey, Leon, thanks for being on the show.

Thank you, Mike, for having me. All right, and back to you guys in the studio. ai Leadership Insight series.

I'm your host, Mike Baer. Today we're with Isaac Parks, who's the CEO for Keebler Health. And we're talking about the use of AI agents in healthcare, and just how ready is that industry for this next wave of technology?

Hey, Isaac, welcome to the show. Hey, Mike, thanks for having me. You cannot go anywhere these days without somebody telling you about their great new AI agent.

And there's one every day now, but it's not clear to me that we as, as a society, nevermind particular industries are ready for this, and especially in terms of the business process engineering that might be required. But you work in the healthcare field. What's your assessment of what's going on here?

Yeah, I think people are trying to figure it out. And certainly they, they see the opportunity in many of this, in much the same way, rather than in other industries or verticals across the larger ecosphere. I would, people are trying to figure out how to string together sort of these complex workflows and apply some degree of, of mimicry on human judgment to sort of relieve the burden, right?

On clinicians, on back office staff in the healthcare world, I think what they're running into, or the problems that are being run into is how complex or how circuitous and how blended. I would say that that workflows can be between creative tasks and structured rule set tasks. And so I, I think it's when it, when there's two sort of, uh, like paradigms try and have to mesh, that's when it gets really tricky with agent type workflows.

Yeah. One of the challenges that I think I'm seeing folks struggle with is they're starting to realize that a lot of these AI agents are probabilistic in that sense that it's a best guess. And a lot of the workflows are deterministic in the sense that they need to be done the same way every time, especially in healthcare.

So what's your sense of, uh, you know, the understanding people have of what an AI agent can do? 'cause I think the first time they encounter their hallucination, they start to lose faith altogether. Yeah, I, I think that's certainly true.

And we run into it all the time when you're talking about, you know, workloads that are deterministic are to go in, you know, more structural rules at things like, I don't know, billing or, or lab values, stuff like that, right? And I, and I think what's interesting is, uh, sort of the, the, the learning that people have to do around how generat AI or large language model powered or transformer powered technology and sort of therefore the sort of non-deterministic tooling nature of all this, that it's still just a tool, right? It needs to be used for the appropriate use cases, right?

And deterministic rule set based workload's, just not the appropriate tools way faster to use the calculator for math. And it is to ask chat, right? To do a math problem.

Um, so I, I think this is where I was kind of going with like this combination between when you have a really complex workflow that involves a ton of creativity, where a non-deterministic approach might be really useful or valuable and unlock a lot of things that have previously been unable to be done. But then the sequential or, or downstream steps, or even multiple steps require injection of deterministic or rule space things when it gets married and mushy like that. I think that's where people really struggle, right?

With trying to say, okay, this is a one stop solution to just throw it a transformer or an agent as they go. Um, and healthcare has a lot of those. And what is your sense of people concerned that AI is coming for their jobs?

Or are they looking at it more like, there's large swaps in my job that I just don't like doing, and I'm hoping that AI will come along and take care of this for me? Yeah, That's, that's a hard question to answer mostly. 'cause I, I would say that paradigm where those, those, that's an open question for not just healthcare, but writ large.

I think knowledge workers in general are all kind of asking that question, what's gonna happen? Um, if I had to sort of guess or predict or anything like that, certainly there's going to be job transformation change. Like I don't, I don't, I don't deny that that's not some something that will occur.

I think really the, the question is how right, and the only paradigms we really have to really pay attention to are past technology shifts, right? In much the same way that you've seen, I would say, I don't know, the internet or cloud-based feeding or prem on-prem, so versus on-prem software or, you know, even looking at things like mobile or social, like just recognizing that this all large technology shifts to some degree providing our access, providing an increased tooling, right? That might remove a current workflow that human is currently supporting.

Um, in many ways, when you look at sort of this past paradigms, what's interesting is that, yeah, they, they nixed some jobs, but you know, if you looked forward maybe 3, 4, 5, 6, 7 years, it actually created many more, right? And so it is a past para that's happened before. I, I want venture to guess that it's gonna happen again, right?

And so the, the really what's gonna happen is like that the jobs will change or more jobs will created or how, and, and that, that that's the best way I could probably I that ideate a prediction there without, you know, coming to some other high degree judgment. What's your assessment of where are the senior execs on this curve? Because I'll talk to them and, uh, I'll get reactions that go anywhere from believe it when I see it to, we are so AI agent happy that, you know, we're already counting the number of employees we're not gonna need anymore.

And they're kind of like thinking that everything that they do could be done by an AI agent. Yeah. Uh, I mean, I like most things, I'd say there's probably a blend of truth in the middle, right?

Or the truth is really a blend of the two ends of that spectrum. Um, I certainly do think that there's opportunity for business leaders who have cost centers that need to reduce, right? And, and maybe even their sort of operating budget requires that they have to do that.

And this sort of age centric revolution not opens the opportunity for them to do that. But I also think that, you know, even in like some of the industries that we're in, right? When we're talking about like, clinicians who are doing pre-visit planning, there're just, there just aren't enough of them.

Like there's, so there are so many more open jobs right now, right? Than than even humans to fill those seats or roles. Like we have also business leaders who are saying, oh my gosh, this is gonna allow my existing workforces right?

To do way, way bigger, faster, stronger, and, and, and fill the demand that's there with the team that I have, right? So I, I don't think it's gonna be a straight binary, like it's gonna be cut loop. So in, in many ways, I guess what I'm arguing is that it's probably contextual, right?

To what is the exact set of problems that, that these business leaders are trying to solve with this tool. The other thing we're all kinda looking at is the rise of robotics, which are driven by AI and healthcare. There might be some opportunities to build these robots and use them for, uh, home care or whatever it may be.

But what's your sense of where are we on terms of robotics in healthcare? Ooh, great question. One that I don't know if I am entirely qualifi to really answer.

I'm certainly much more of a software guy myself. Um, but I am watching it 'cause I'm so curious. I am like on the edge of my sleep.

Ooh, I wonder what these robots can do. Um, the funny thing is that there's probably also much to, there's surprise, there's probably some leading edge things that I know remote pro like remote process monitoring or RPM, um, in our industry is a very, very big deal, right? There's big in this large movement to, to treat patients, to treat the more acute advanced illness patients at home or in a, a setting that is not in a hospital, right?

And that requires a whole set of infrastructure. That means that those biomedical devices or, or, or proxies for biomedical devices or, or clinical care folks are in the setting with them, right? So there's already hosts of, I mean you may not wanna call them robots, but they're, you know, devices that are, that are doing a lot of clinical support for these patients that are all integrated and, and wrapped together with software.

It's probably not that far of a stretch to say that, oh yeah, that's kind of a robot, right? And if you have all those things sort of powered through a transformer based product that uses judgment and can signal humans to say, Hey, go do this, go do that. I, I think there's, there's probably a ton of efficiency gain to be done there.

Just question of how and when, What about the security aspect of all of this? Is that still kind of an afterthought as usual? Or are people thinking about this more aggressively and upfront?

No, there's deep concerns, right? And caution around security, um, for using generative AI products on, on clinical documents and, and, and patient health. And rightly so.

I think right, that's one of the premier considerations. Anytime we, and even our business, anytime we integrate or or talked to a, a protection customer that is top of mind forefront, like how do we protect, you know, the, the sort of downstream artifacts of, of patient information. How do we make sure that nobody's getting, and on those items or documents and, and all this in light of being said, being able to say, okay, can we use these tools to improve their outcomes?

Can we help them be healthier? Can we given longevity, can we improve their quality of life? What use cases Have you seen involving AI in healthcare that are kinda are working that stand out to you is something that, you know, maybe you wish other people would copy or at least emulate?

Yeah. That I think the, the easiest way to sort of think about this is, um, the, the landscape shift that we're seeing on these large language models is really the ability to access unstructured information to reorganize it or Alize before you can generate it. And whereas I almost liken it to, you know, the large language model is really to unstructure text what the calculator was to like rule space engines are bad, right?

The, the, the sort of speed at which you could analyze that kind of mathematical data with a calculator is now what we can do with unstructured text. So I would say anytime you have a, a workflow and in the clinical workflows, there's a lot of these scenarios where you have bulk corpuses of unstructured text and you're trying to draw patterns or mimic patterns or established patterns or career processes around that unstructured test text, that's where we can really fly, right? So things like navigating an EHR or understanding, um, what's happening in a medical record area, in our case looking for evidence for chronic disease over a host of data that is just unstructured reform text.

Um, those are all things that are really viable. One of the, I think more I would say advanced or leading edge cases of using transformer technology and clinical workflows would be dictation software, right? So having something listened to a, a clinician in office and visit, and then summarizing it into an actual clinical note that is structured the right way, that captures the right nuance and appropriately says, okay, these are the idioms and I'm gonna translate that into this clinical, like, language that makes sense for them, that that saves so much time for clinicians.

Um, that it, it's been pretty amazing to see. Conversely, have you seen anybody try to use AI for a use case that's just not gonna pan out and others should not follow suit? Yeah.

Uh, that's probably one of more like a judgment call. We're so early that like, uh, you know, you also have this sort of selection bias where prob people have probably tried a lot of things and have not have not worked and we've never seen it or showed up. Um, I, I think there's, um, the, the closest thing I can get to you is, is this idea around like, uh, replacing the physician in their clinical judgment, right?

And the same way that you were seeing, uh, I would say a parallel paradigm around, uh, software agents or software engineering agents trying to replace software engineers, um, on their own as just software agents. They're not, the efficacy is just isn't quite there yet, right? And I think similarly with clinicians, I don't think a, an a a clinical agent on its own doing diagnoses and prognosis and sort of, uh, preparing a care plan on its own is going to really outperform a clinician alone.

Uh, what we've seen is that there are benchmarks that are passing around these structured data, like, like passing medical, medical licensing exam, things like that. But the, again, that's all still just structured traditional structural rule sets. So try to figure that out.

Um, the same sort of, I think, uh, productivity in improvements, um, in software engineering co-pilots, this is actually happening, right? And so I think when you're looking at sort of this, uh, the evidence out there for, um, you know, software engineering co-pilots are amplifying like your really good engineers and then kind of taking your, your not so good engineers and making them worse with a co-pilot. It's actually kind of similar with clinical diagnostic tools that are AI based.

It's actually accelerating the speed and judgment of really great clinicians and it might actually be slowing them slowing down the junior ones. How will we manage all these AI agents? 'cause ultimately there's gonna be a bunch of them and they're all gonna be performing specific tasks and I need to orchestrate them into something that feels like I'm creating an end-to-end workflow.

Um, and of course, they have to do that alongside the humans who are doing various things that the agents cannot do. Um, is there gonna be some sort of orchestration framework for all That? Well, it sounds like you're gonna go build one.

It's probably not the opportunity in the future. I, I completely agree with you. I, I think orchestration across all these agencies is, is pretty critical.

I mean, I was just looking at like, uh, we're just starting to see some of these like MCP protocol, um, power products and, and clinical software use cases. I, I, I think you're gonna see an adoption curve have been a lot slower in healthcare technology than you would anywhere else. Um, I'm really curious to see sort of that secondary, tertiary level look like agent to agent and kind of workflows or technology enable our protocol enablement.

I you're just gonna need a level, a critical mass level of infrastructure on software access is across healthcare systems writ large in order for that kind of an ecosystem to really take off. I have my doubts as to how quickly that will happen. Um, but certainly would be excited and ecstatic to see all that kind of stuff shake down so that like most regulated industries, I'm, I'm really curious to see, you know, what are the sort of different in industry pressures or, or business pressures or regulation pressures that would force our entire healthcare industry to move in that direction.

Totally. Or in totality. I, I don't know.

And then that's something we're gonna be watching pretty closely. I feel like there's a lot of people experimenting with various things, and that's all great, but I can't help but wonder sometimes if, if, you know, they're basically mapping out a project that's probably cost about a million and a half dollars to execute, given the cost of GPUs and everything else to replace, you know, two people making $50,000 a year. So, do we have a, a good sense of the math and the economics of AI and healthcare?

Uh, I don't think so. I didn't to be, to be honest. I, I, this is something that I've been, it's on my list to go do some research on, but I'm trying to figure out sort of the power law dynamics around, you know, compute and, uh, tokens and what it actually costs, like our ecosystem and energy to make the, to make those calculations versus, to your point, a human that has just been trained over time to do those exact thing like judgments.

I, I actually have no idea. That's on my, that's literally on my like, TD list of a lot things to explore. I'm really curious to know, you know, how that, how the economics of it all shake down.

'cause uh, my suspicion I know is a lot of it is obfuscated around sort of these foundational model companies that have, you know, pour a lot of investment into these tools, but therefore, you know, it's not always clear, right? What the actual cost is, right. For, for some of these tasks that we're trying to automate.

Ryan, folks, will you heard it here? Maybe we need to start with the ROI and work backwards. 'cause otherwise we're just playing around with a bunch of AI stuff for grins and we'll see what happens later.

It's really interesting. Yeah. Yeah.

Hey Isaac, thanks for being on the show. Yeah, thanks Michael. Iman, appreciate it.

And thank you all for watching the latest episode of the Techstrong AI video series. You can catch this in other episodes on our website. We invite you to check them all out till then, we'll see you next time.

Hey everybody, welcome. Thanks for joining us today. This is a special six part series that we're putting on having a little fun with a name called Control Alt Deploy, getting software into production, right?

That's what DevOps is all about, at least partially about, uh, our, our series here is sponsored by OpenText, and so we're gonna be having a lot of discussions and we may talk a little product here and there. It isn't a product pitch, but we're gonna be talking about really where DevOps is headed today, especially in the era of ai. My name is Mitch Ashley, I'm VP and also lead the practice for DevOps and application development at the Futurum Group, which is part of Techstrong.

And I have a very distinct pleasure of being joined by Tal Levi, Joseph Tal. Welcome. Would you introduce yourself?

Let folks know a little bit about what you do at OpenText? Yeah, thank you. Um, so Tal Levi Joseph, and I'm heading managing the business unit of DevOps at OpenText.

So I'm leading product and engineering. Ooh, that sounds fun. I might come over and if you ever need to take a break, I I'll stand in for you for a day or two that, that's are fun jobs.

It's more than welcome. It sounds good. Well, you can't turn sideways, step forward, backward without, you know, kind of hitting the, the AI word in DevOps.

And, uh, you know, how long have you kind of been in the DevOps area? Is that something you've been worked in for quite a while? You kind of more recently joined?

No, so I have a long heritage in the history. So I started with Mercury Interactive. Mm-hmm.

Uh, you know, back then when we started, you know, with quality and test automation and then moved up into, and went beyond when it came to Agile and backlog management and, and project and portfolio management. Um, so kind of spread it, you know, and, and broaden our space, you know, within the left side of the planning and the right side side going into, you know, production and deployment as you said. Uh, so today we have, um, I would say an end-to-end, uh, planning to delivery, I would say platform and portfolio of products in the DevOps space.

So, you know, I can say for for the past like 20 so years, um, I am part of the DevOps application delivery management space. So, yeah. Excellent.

That was a softball question. I know you, you know, your background And I think, you know, uh, we develop our own software, so we get to use our own software, we get to experience it internally. So we're at the point where, you know, it's easier for us to talk to customers to understand the challenges, but to also understand how to deliver real value to the customers.

Because at the end of the day, it's about the value that you can deliver. It's about what you deliver, the how is as important. But I think more we're seeing it's about the value as well.

And, you know, we'll speak about it, me maybe, but how to measure this value as well, especially in the era of ai. Excellent. I I'd love to do that.

I think we're very much in sync in that it's, you know, some of the metrics, even the door metrics about how many deploys do you do per well, if it's the wrong deploy or if it's a bad, bad one. And you, you know, that's not always the right metric. It's the value to the business, the customer experience, functionality, getting into market, what whatever kind of the outcome is, right?

Not just exactly getting software into production. So, good, good. Um, you know, there we're, we're not all in this new era of AI is everywhere and we're using AI to automate everything.

We're, we're, I guess, I dunno if we're easing into it, but we're figuring out as we go, right? Both as product creators and also practitioners, myself as an analyst. And a lot of the tools are sort of grew up, grown up through this, started as the chat bot start as the, uh, co-pilot kind of at your side.

Uh, some people call it augmented or assisted ai. Um, but we're starting to see more automation beyond just single tasks or developer work and getting into more parts of, of the, uh, software development lifecycle. How, how would you assess where we are in this progression of adopting ai?

Yeah, so I, I think we see it, you know, uh, some people ask me whether it's a hype or real reality. So definitely it's reality. And we see it, you know, as AI shapes and evolve, you know, everything we do.

We started to see it with the adoption, you know, in, I would say increased adoption, uh, from code assistance. Uh, now we see, you know, test generation, um, smart planning, being, being able to predict, but also avoid risk. So we see the evolution more and more to autonomous agents.

Um, so I, I think, you know, it is critical not just to the, to empower the build phase, but also to empower, empower the entire software delivery lifecycle and approach AI and have it approachable to the different roles, right? Like the software quality engineer, um, the different functions, um, like, you know, the project managers, the process managers, again, from planning, design, build, test, and deploying. So I think, you know, AI has now broaden itself to cover the entire software delivery platform rather than focus just on the build side, for example, with these, um, you know, uh, developer co assistance.

But I also think within, you know, the work of the developer and, and working with the co-pilots of the world, I think the more you can give them the context of not just the code, but also the feature, it's connected to the associated defects that you had in the past in this area. So you can avoid further risk, but to give them a th a 360, I would say, view and visibility into everything, you know, that is happening in the ecosystem, I think that's important as well. So it's not just automation, it's visibility, insights, predictability, and then, you know, from predictive to prescription, what you need to do.

And everything at the end of the day will be autonomous. So what needs to be done in order to avoid risk, in order to accelerate, um, you know, uh, time to market and in order to increase and boost productivity and efficiency, right? Um, so I think, you know, the bottom line is, you know, you need to run faster.

Yes, you need to run better. Yes. But you have to do it smarter and safer as well.

All, all, all excellent points. And, and, and I think it's really smart to think about AI not as an individual productivity tool. I can generate code faster, I can do these tasks, set up my environment, whatever might be Yes, that's true.

Um, but if everybody's kind of individually increasing their productivity, it's still a ste a team sport delivering software, right? And not just in the developers in your case. Exactly.

Um, and I think that's why the term agentic DevOps is kind of rising up to the top. Of course, I think everybody's gonna be labeling things as a agentic fill in the blank. And it, it isn't always true.

But, uh, uh, you can, you'll find out very quickly whether something is truly agent in products and technologies that we use. I'm gonna run an idea by you. I, I wrote an analyst paper recently saying that I think what's happening in the developer space is a very interesting, uh, area to watch, because so much innovation is adopted very quickly by developers.

They're just natural inclined to that use tools. They're trying out by coding all that stuff. But a lot of the learnings from that of how is the work changing?

What are the skill sets? Those are things we can pick up and say, well, let's talk about software testing. Let's talk about security and how that might change.

Let's talk about the whole workflow, the pipelines, and how they might change. I'm not saying all the answers will come from development, but it's sort of an early proving ground or testing ground for ag agentic. Thoughts on that?

Feel free to disagree if you don't agree with that. No, no, No. I, I completely agree.

And, um, and, and I think the more we see the evolution of the developer work, uh, the more we can understand what else is needed and, and missing in the ecosystem. So you talked about testing by, but there's so much code that is being generated. Who's going to test it?

Who's going to make sure that we're still keep keeping up with the standards of quality with the regulation in specific environment? We often, uh, use with a lot of enterprises that have, you know, certain standard regulation, governance. So you need to make sure AI is also putting in place and making sure to adjust and customize everything that is being generated to make sure it aligns and comply, right, with a, with a certain set, set of standards and regulations.

Um, I also think that, again, everything is changing all of the roles, you know, uh, that, that cover today, the, the software delivery lifecycle. Uh, but I definitely think, and I agree that we see the, you know, the early adoptions with the developer, but, but that gives us an ability to assess, you know, better the challenges, but also the great opportunities. Um, and also speaking about iGen, what are, you know, I would say the additional, um, agents that needs to take part in this, you know, um, I would say environment or framework, right?

And how they're going to collaborate with each other autonomously. Um, and, um, so I think, you know, and I think the difference is that while, you know, copilots are very task driven, you need to, uh, generate a code to do that. You need to, uh, translate or make comments here.

But I think iGen, KI can really drive, as you said before, outcomes. And I think that's the big difference. So taking the code that was developed, but then making sure you have the right, you know, agents in place and the right collaboration and flows in place to make sure that software is being delivered, you know, in, in with the re you know, with the outcome it needs to be delivered, uh, delivering to it to, to its customers, right?

And I think that's the big difference between a goal driven, you know, kind of, um, assistance or tool, but then outcome and value driven framework, um, like iGen, K. So you have to think carefully of what are the roles and functions and task you need to do in order to, you know, to making this cake of, of releasing software to the market market that is valuable. Um, and that creates this significant, you know, uh, outcome and impact.

Yeah. Not, not telling you anything you don't already know, but OpenText has been, you know, using AI for a long time. It's not new to you, machine learning capabilities.

And while gen, while, uh, agenda of AI may be kinda relatively new to the market, you all have been doing AI for quite some time. Are there learnings that you have, have that you've kind of brought forward into the, uh, generative era from the AI work that you've already done, that you're applying today? So, so that's true.

Uh, we have a longstanding version, you know, uh, journey with, with ai, um, you know, way before, um, gene ai, what we call l and m today, um, we've used, um, what we call deep machine learning, uh, for functional testing. And that's the ability, you know, up, up until then, we were identifying objects on the screen based on their physical description, right? That keeps on changing or sometimes not unique.

So then you have to, uh, enhance it. But once we've, you know, understood that we wanna, and we need to find a way to identify an object on the screen like human does, right? Um, you know, leveraging neural network algorithm, um, then it was the breakthrough of how you do functional testing, right?

Um, because it creates a much more resilient script, uh, but also it lets you do testing in a much earlier phase. You don't even have the application. You can have a mock and you can start and play with the, you know, with the mock or any design phase or at a design phase.

And that will be once the application is ready, that will, you know, be able to replay itself with the application. Um, so we've, we've been doing that, um, and I think, you know, the more, uh, you customers ex, you know, try it experience, um, the more you learn, the more you, uh, fail. The, the thing is you fail early, but then you correct yourself.

So my number one thing would be work not just for the customer, work with the customers. And by the way, everybody's are going through the same journey, right? There's, there's more mature customers, there are less mature customers, but they're all, you know, it's there, you know?

Um, so, so the train has left the station. What's very important is for us to work with the customer, to I to identify, you know, at the very beginning, the first use cases, the set of use cases that will make the impact to them, that will bring value to them, and, and then consult with them on, on other, you know, uh, use cases, understand their deployment model. We know that some are more hesitant to work with commercial AI framework, right?

Um, and they have this, so then you should work with a, bring your own model or, you know, off cloud even, you know, uh, deployment models. So the more you interact and work with those customers, the better insights you can get. Um, we're also using ai, we've used AI before for analytics, right?

Understanding patterns of behaviors, because we store a lot of the data. Um, then you can understand and identify patterns and then, you know, predict timelines or risks that can come, you know, things like that. Um, and again, customers is, is the main focus, right?

Whether it makes sense, um, you wanna bring insights that will have, um, I would say actions as well, like corrective actions. So not just bring that, um, so yeah, work with the customers. The other thing I would say is the data, data is everything.

Context is everything. And your ai, you know, I would say outputs are as good as your data period. Um, so therefore data and the way you store it, the way you structure it, um, and you have to, you know, integrate well to the environment in order to sync the data, uh, to what we call today data lake, right?

Uh, but data is, you know, my, I would say my number one thing, if, if data is fragmented, then the quality of what you get from AI is, you know, it will be very limited. So much different point. Yeah.

That's for sure. Um, the other thing I, you know, we're also learning from the customers is now with Gen ai, um, also the experience with the applications, you know, is changing. So it's no longer clicks in finding the information, but more of a conversation, you know, with the, uh, um, with, with the ai and, you know, that brings the whole, you know, what we call in, in, you know, in OpenText, the aviator assistant.

So you just, you know, you ask for a graph, you ask for information. I'm at the feature I'm telling you to, you know, do stuff and generate stuff for me. Um, so it's, uh, it's a lot where you are in the context, um, but asking, you know, in, in a rather, in a conversational way, rather than anything else.

Um, and, and we're learning always, right? What's, what's, you know, what's the best approach, um, to start a company. You know, I call it a partnership more than anything else with, with the customers.

Um, so yeah. You know, Taal, one of the things I wanna ask you is, there's a, there's some momentum around treating agents as coworkers, peer programmers, another, all the way to treating them that way in the identity management system. Um, is access control, logging the work that they're doing?

'cause there are, you know, they perform a lot of work that, or will be even more so, uh, alongside humans. Do, do you agree with that model of, or do you think there's much more we need to think about, uh, from a how do we manage agents? I, I think so.

First of all, um, you're right. Um, in OpenText match, we call them the digital workers, you know, so I know a lot of organization use different terms, but, but I think we relate to it more than anything than, you know, a digital worker. And this is how some of the workforce will look like eventually, you know, you will hire digital workers, you will train digital workers, um, and there will be agencies that you will work with, outsourcers, digital workers.

So it's going to be fascinating, the future. I do think that yes, you know, there will be workers, like any other workers with, you know, uh, their own role and access and, and limitations and, you know, um, what they can do, what they cannot do. And the organization will need to find the right vehicles to actually, you know, govern it.

And I think in OpenText, we're already doing it, you know, in, in, um, um, but, but, but there's one thing I would also like to mention, which is the role of the humans, right? With the digital workers. And I think, you know, that by itself requires a whole, you know, um, dedicated podcast because it also involves some philosophical aspects.

But I do however, think that we should carefully think about what are the vehicles, you know, we use in order to govern and control and set, you know, the guardrails that in place. And I think human will become, in a way, the architects of the flow of the process of, you know, navigating and guiding and training those digital workers, but also the, our solutions will make sure to govern it and control it in the right manner, right? Um, what are the control points, um, adjustments, customizations that we need to do, uh, what this digital worker can do, uh, what they can't or forbidden to do, uh, with which digital workers there, you know, or should collaborate, which are not, you know, um, so yeah, I think it's, you know, it's, it's about, um, controlling, defining, architecting, you know, this whole, I would say process and framework and collaboration framework between the different, I would say digital workers.

But I definitely see, you know, um, the world and especially, you know, in our domain in DevOps, um, is, is going that direction. We just need to, to do it, you know, as I said before, smarter and safer and, and put the relevant control points, guardrails, but also, you know, the vehicles like identity management and define them, you know, uh, of the roles, the responsibilities, and, you know, access to each and, you know, in, in every system in our organization. So, um, definitely, Yeah, we could, we, we, we should have our own series on just what's the future love, what's your future job gonna be?

Because it's, we're all, we're all learning too. Uh, when people ask me about is, is, is, is AI gonna take my job? But well, But this is where, where we see Mitch, the defensiveness in some of the organizations, you know, it's, it's like, hold on.

Very skeptic, very, and, and I, and I see at OpenText, we are also the agents to, to create this circle of trust, right? How do we help them to trust ai, to govern, to test ai, to validate, right? How do we also provide the tools, um, to, uh, to increase this trust to, uh, test the outputs of AI accuracy, I mean, these kind of things.

So, so I'm, you know, I'm so excited by this, really. Yeah, It is just fascinating. I mean, what I tell folks is you're just gonna be producing a lot more work, uh, but you're also gonna be working a lot harder in kind of the higher order, the architecture, the orchestration.

Um, my mental model, I don't know if this will come out to be true, but I like to play, you know, computer games and StarCraft is one of them. So you're, I can picture UI kind of like, those are my agents that are managing to accomplish these kinds of goals, set these off on these tasks, they're coming back, you know, that's, it's almost that kind of a feel of what it's gonna be like working. And, uh, I think they'll still be humans around For a long time they were in Star Trek.

Even if with all the, you know, great automations apps in Star Trek, um, you know, we're, we're just about the end, end of our time. How, how do you advise people, whether it's a colleague internally or you're talking with a customer or, or one of your colleagues, uh, outside the company, uh, about leaning into AI across the DevOps processes. Obviously you just can't pick it up and do everything you'd like to do with AI today.

It's evolving and changing, um, but it's something you also have to kinda learn. You can learn by jumping in now. You don't have to go to the deep end, right.

And just assume it's all works perfectly. But, uh, what kind of advice would you give folks? So I, I will say be focused first, right?

You know, whether it's a POC, but define, you know, the relevant use cases that are, you know, important for you. At least, you know, it's a phased approach. So let's go with the first phase.

I wouldn't say start small, but start, you know, be, because you have to be very fast in this, right? Uh, but, uh, but, but be very focused about what are the use cases and the outcomes you wanna achieve, right? And make sure to define what and map, right?

The use cases, um, the, uh, um, what are the, uh, um, governance and control points and validation points that will help you realize, you know, uh, you are on track. But also we spoke about it in the beginning, like measurements, put the right measurements in place to understand whether you got to your outcomes, you know, progress wise, but also, you know, at the end, um, you know, whether it's increased productivity, put the right metrics in place to tell you exactly whether you're increased productivity. Just don't forget what you said before, value metrics.

It's not just about to develop faster. Everybody can do that now, but, but are you developing in the right, you know, uh, uh, with, with the right quality standards, the highest quality standards, the highest security standards, um, you know, whether you've developed what you, you know, what will bring value to your customers. So put the relevant measurements in place and then move on, right?

Um, so I think, you know, uh, and, and you have to build a culture around you as well, because if people will start, and we saw it also in our organization, if people will, there, there are the developers who bring, bring as much as co-pilot team want, right? But then you see sometimes hesitation from the testing side, right? Oh, hold on.

And now, you know, uh, um, manual testers, right? They're no longer relevant, right? What are they they're doing?

I'll, I'll start to reject, you know, the, the, the process and oh, the, these tests are not good. We, we don't need this culture. We need, uh, a culture that if this test is not as, you know, as good as the one, the one needs to be, how do we need to train, you know, the agents to, um, produce better tests and better?

So, and how do we know it got to the right level, for example, right? Um, assess risk, um, you know, being able to, to validate it and again, move from there. And, and obviously, you know, if, if you starting the, you know, build test plan and then, you know, move, or by the way, so you can do phase by phase and role by role, or you can take one application that is being developed and, and try to do a broader, I would say use cases from planning to development.

I would, I would take something which is very measurable, and, you know, you have a high frequency of release, so you can see some feedbacks, right? Coming from, and it's not just feedback by the way, internally, it's feedback, you know, from the outside in, right? Observability, um, you know, uh, um, insights, right?

What has been adopted? What is the state in production? I mean, these kind of things, right?

Stability of the application performance and these kind of things. So, um, but again, be focused, you know, focus on the first, the 20% that you think will bring the 80% of the difference, and then, then go from there. But, um, um, but yeah, you know, and, and do it wisely and carefully and with the right guard, guard, guard rails, um, in place, for sure.

And culture, culture, culture. Because at the end of the day, it's all about people. Whether they will resist or whether they will join, you know, and, and, you know, with a can-do approach, right?

And, and I think this is the people that you need to surround yourself with Sage advice. I would say, you know, summarizing it in some ways, it's like, just because you're using ai, AI doesn't mean you throw out your engineering disciplines. You're still leveraging that skill as well as your people and culture and collaboration, and those are all equally important.

Maybe you're measured in the future by your ability to leverage AI to get work done, is that is how you, you know, measure your work. Well, Tal it's been fantastic talking with you. Um, I wish, enjoy it when we get a chance to do that and looking forward to it.

Again. Thank you, uh, Tal Levi, Jo Joseph, who is, uh, leading up development engineering for DevOps at OpenText. Let's do this again.

We'll get this on the schedule, have another conversation. Thank you very much, Mitch. It's been a pleasure.

You bet. Thank you everybody for joining us, and thank you to OpenText for sponsoring our Control Alt Deploy series. Have a good day.

There was a big tech conference this week, and I'm not talking about Apple's, WW dc I'm talking about Microsoft Build, where we heard a lot about agentic ai. That's the key story here on the rundown this week, where we're also gonna be talking about Snowflake acquiring a crunchy data platform. Broadcom launching the next generation Tomahawk ethernet switch, the EU checking out Broadcom's VMware licensing, IBM's new data platform capabilities, Amazon's $20 billion build out in Pennsylvania and IBM's Starling quantum computers, all that.

And more on this episode of the Tech Field Day News rundown. Welcome to the Tech Field Day rundown, where each time we meet, we run down the IT news of the week with a variable degree of snarkiness. I'm your host, Stephen Foskett, and joining me today as co-host is Mr.

Alistair Cook. Al uh, I guess maybe I'm the one who should be welcomed to the show. It's been a while.

It has been, Stephen, we have, have missed you, Tom and I, and, and our various guests who filled in. While things are being pretty busy at Tech Field Day, have have definitely missed you. Welcome back on National Corn Cob Day National Corn Cob Day.

Also a very important one, national Forklift Safety Day. And if you're saying forklift Safety Day, what's the deal with that? As someone who's driven a forklift, let me tell you, those things are freaking scary.

It's really easy to drive that thing right through a wall. And don't ask me how I know that. Have you ever driven a forklift through a wall?

I have never driven a forklift at all, let alone through a wall. Although I did at one stage work with somebody who taught the safety training courses here in New Zealand. Um, and did, did somebody try to kill you with a forklift?

Thankfully not. Although the YouTube videos of people making mistakes with forklifts, backing them off loading ramps, or knocking over huge amounts of racking, uh, make for entertaining viewing, if you can't find anything that you want to view again on the Tick Field Day YouTube channel, Yeah. Uh, just imagine driving your car in reverse.

Um, it's bizarre. Okay, moving on. Um, also, uh, appropriately it is National Call Your Dr.

Day. Um, so there's that. So Al, let's dive into the news of the week.

Uh, snowflake is acquiring crunchy data for about $250 million to enhance its AI ambitions by integrating PostgreSQL into its platform, enabling developers to build, deploy, and scale AI agents more efficiently. This move, uh, positions Snowflake for stronger competition in the AI data platform space, aligning with similar moves by Databricks and reinforcing its strategy to become a comprehensive enterprise AI backbone. Al, what's your take on the crunchiest of data?

Well, crunching your data is what AI is all about. And so maybe just adding the name, uh, crunchy data to, to, uh, snowflake is going to bring all of the AI to the, uh, to the art. Uh, I think it is fairly significant that we're seeing growth and platforms that combine structured and unstructured data to feed into your AI platforms, as well as being output from your AI platforms.

And so, the idea of pipelining data through AI, that we're gonna generate data out of ai, we're gonna generate, uh, both model data, but also we're gonna generate business data out of ai and having this location where we can reference all of it, it's a little more of the ideas of those data lakes and data warehouses we've worked with in the past, but now adding this, this idea of having AI as a central first class citizen, as the reason that we're actually delivering these large amounts of data. Uh, in terms of acquisitions, snowflake, uh, has been building out a set of tools to make it easier for developers to build AI applications. Um, it's nice to see more open source products be embraced in here, and hopefully we'll see more contribution to the open source project.

Uh, we do see a lot of Postgres being used as a store for Vector data. So when we're using, uh, retrieval augmented generation, um, um, an essential part is a Vector database, and Postgres is one of the tools that is pretty popular for doing that. So Snowflake is definitely aiming at getting developers to use their collection of tools together to build the, the things they need for ai.

And we definitely see a need for more integrated platforms for the simpler experiences of building AI into your application. So I think this is a, a really good thing to see. Um, full compatibility, backwards compatibility, you know, all of the, the openness part is gonna be an interesting thing to watch over time.

So seeing that we continue to see contributions to, uh, upstream Postgres, uh, and that the features that are gonna be enhancing the Snowflake platform are available to maybe not snowflake's customers directly. Of course, snowflake will be showing a lot of value and getting it from them, getting it as a first party, well supported, integrated solution. Uh, but it's important when we're talking about open source products to remember that the source code should be coming back up, and development should be benefiting a community as a whole.

It's not a straight up commercial, uh, movement. 4 terabits per second of bandwidth on a single chip designed for loud large AI data centers. It connects thousands of GPUs quickly and efficiently using advanced routing and energy saving features.

The chip aims to replace proprietary tech like InfiniBand with open ethernet standards and helps build faster and more flexible AI networks. 2 T last year. It's interesting to see the 102 T this year.

Stephen, is this a revolution or an evolution? Well, I, I think Broadcom wants it to be a revolution, but frankly, uh, this is, let's say, uh, continuing the march of progress that they have already made with their Tomahawk series. If you're not familiar with the Tomahawk, as Alistair mentioned, uh, we did get a look at it, uh, actually back, uh, we looked at the Tomahawk five, uh, back in 20 22, 23, uh, as well as at our cloud Field day event in 2024, where we got a kind of a deep dive into what this is all about and how Tomahawk can be used to accelerate AI and cloud workloads.

It should come as no surprise to anyone who watched that like Al did, and I did that. A Tomahawk six was on the way, and here it is. Uh, it has sort of the requisite improvements that you would expect from the next generation switch.

It's built on the three nanometer process from TSMC. Uh, it's, as you said, twice as fast in terms of switching capacity in a single chip, double the bandwidth. Uh, one of the features or factors that, that I want to call attention to is one of the things that you said there, though, and that's the certis.

So here's the thing, uh, many of us, um, are even in the tech industry, unaware of the importance of the certis, which is literally the way that these, uh, interfaces connect with, uh, other chips. The ethernet certi roadmap is extremely, extremely strong, extremely high, high performance. And the reason is because so many of the processors out there have plenty of ethernet bandwidth, even if they appear to have plenty of PCIE bandwidth Ethernet is really sort of the king of the castle and the Tomahawk six as a ethernet inter interconnect for HPC and AI workloads makes sense mainly because of that sort of synergy between the, um, ethernet, uh, certi and the processors that are being used out there.

So when you look at interconnect technologies, one of the things that you should ask yourself or ask the vendor is, is this sort of ethernet flavored, or is this sort of PCI express flavored? And frankly, uh, as much as it's, uh, you know, maybe surprising for people to learn, the ethernet flavored interconnects and certis, uh, like the things that are supported by the tomahawk are, um, frankly, uh, there's more of them. They're higher performance, there's more bandwidth.

And that means that essentially you're gonna get more performance for less from those. So, uh, by introducing this next generation tomahawk, by doubling the bandwidth, by, uh, you know, supporting a hundred gig and 200 gig sardis, what it means is that, uh, Broadcom is going to continue to cement itself as the central interconnect, not just for networking, but for HPC and AI applications, simply because, uh, they're taking advantage of the fact that this is the dominant standard out there. So it's really great to see Broadcom pushing forward on this.

Um, is it a revolutionary? Well, it's great. Uh, I don't know if it's revolutionary.

I don't know if it's unprecedented, but I'll tell you what, they're gonna sell a ton of these things simply because it's super fast, and frankly, it represents the next generation of connectivity. A European cloud group is criticizing Broadcom for sharply raising VMware prices and making licensing more rigid after its acquisition. Some costs have gone up by as much as 1500% putting pressure on smaller cloud providers.

The group warns that this could break the EU competition laws and is urging regulators to step in and push for fairer more flexible pricing. Alistair, uh, what do you think about, uh, regulators stepping in on VMware pricing? Well, we see lots of precedent for regulators stepping in where there is a clear, dominant player who is using their market position in a, an aggressive way.

The funny thing here is that we know that Broadcom is really doing the opposite. They're trying to own less of the market and only own the profitable part. So this particular story comes from the European Cloud competition observatory, whose job is to look at the cloud sector across Europe.

And of course, European regulation is typically much more intense on organizations on, on businesses than US regulation. And this particular report was looking at whether the changes to VMware's licensing were impacting customers of cloud providers and lots of companies, many organizations, uh, who are providing cloud use VMware as their foundation. That absolutely carved out a huge amount of the service provider space with the tools over the lifetime of, of VMware prior to the Broadcom acquisition.

We know that Broadcom has made decisions around shifting towards all licensing being a subscription, so a annual or monthly cost, uh, and that it is a fixed cost for that, that cycle. Uh, what the, uh, this commission has seen is between 800 and 1500% increases in the, the cost for cloud providers. These are cloud providers who previously might have had a licensing agreement that was a kind of a back to back deal, where if they sold capacity to the cloud providers customers, that was when they played pay to VMware.

And so the consumption based licensing has shifted away to being the same subscription based licensing that enterprise organizations are dealing with. That shift completely changes the economics of the cloud provider's business because it becomes a, a fixed cost to deliver the VMware infrastructure rather than a variable cost that's coupled to your income, uh, that has threatened the profitability, the viability of a, a collection of cloud services in, in Europe. And so this is the, the challenge, uh, one of the elements that's highlighted in the report is that the same challenges have occurred with other providers of, of infrastructure software.

And that, uh, the issue here is that Broadcom has just failed to engage, has been on a take it or leave it. This is what we're doing, and that has rubbed things the wrong way up in Europe. I think there's some interesting context for this, and I wanted to particularly highlight that.

Very recently, VMware announced their profitability, so revenue up 20% year on year, and their Q2 results, um, income ramped up 124% year on year. And so Broadcom's changes are producing what Broadcom said they would do, that VMware is more profitable, but because it's focusing on the customers that are profitable, uh, and it's making the unprofitable customers profitable, I kind of have this feeling that VMware, well, Broadcom Tane is, is here to make, make business, make things successful, and is perfectly happy to cut off these customers who are not actually as profitable as, as core customers. And the, the take it or leave it approach is gonna continue.

Uh, I would be interested to see whether there is a mechanism in the European law that allows a legislative control a control of pricing for a commercial organization like this. Uh, as long as the contracts that are being offered are being honored, and we've seen with at t some challenges around that in the past, uh, as long as the contracts are being honored, I'm really not sure that this, it, it, uh, interference by a regulatory organization is actually gonna bear any long-term benefit. Whether it just makes these cloud providers who have bet their business on VMware, um, gives them more drive to move off VMware or get outta the market, that's not successful for them.

IBM has upgraded its data platform with new AI driven features, surprise AI features, uh, including real-time data integration tools for handling unstructured data, faster data processing, and better monitoring. These updates aim to make managing and analyze data easier and more efficient, especially in hybrid cloud environments that, well, Stephen, these are hybrid cloud. It's pretty normal in the enterprise, and that's where IBM is selling most things.

Yeah, this is, uh, well, I guess to be expected from IBM. So just to give a little bit of background, at IBM think the company really focused on agentic ai, I know surprised and introduced a whole lot of agentic AI features, uh, including, uh, a whole catalog of AI agents from Orchestrate that allow, uh, companies to deploy prebuilt agents for various domain specific uses like HR and sales and so on. Um, it also integrates with all sorts of other third party tools like SAP and Oracle and AWS, and allows companies to, um, access AI in a way that frankly matches their, um, enterprise needs if their IBM customers.

Well, one of the things that we heard about, uh, well all year long, but especially at, uh, the Click Connect event that we recently attended, was the need for, um, ai, uh, that fed from unstructured data. Uh, and also, uh, this whole world of data lakes and how you can integrate data lakes with ai. Essentially, businesses have, um, plenty of applications that they can feed from.

They have plenty of structured data, but really what they need to do is they need to start integrating all the other data in, in terms of, uh, you know, files, uh, documents, all that sort of thing, along with these massive data lakes that they're putting together. And that's what IBM is doing here. So essentially we've got integration for various, uh, intelligence platforms that allows, uh, the gen AI capabilities that IBM has already shown us at think, uh, through watsonx data and, um, integrates it further with unstructured and data lake environments.

So essentially, companies can start, uh, querying their massive amounts of unstructured data as well, uh, as their, um, you know, other structured analysis and, uh, retriever augmented generation or rag, which uses, uh, files, uh, in file servers. So the whole thing makes a lot of sense. Uh, this is exactly the sort of thing that IBM customers need.

Uh, and frankly, this goes along with a lot of the research that we're hearing from the future of intelligence side, which shows that companies really want to have a large, trusted partner that can help them move forward in this AI space. They need somebody who can assure them that what they're going to be deploying is gonna be trustworthy and is going to, uh, be capable of integrating with applications. And that's the sort of reassurance that IBM is offering with watsonx.

Essentially, IBM's customers are hungry for a partner like IBM to take them by the hand and say, look, here are some AI agents you can deploy. These are safe, these are proven, these work. Here is a way that you can integrate with your data lake.

It works. It's proven. Here's a way that you can pull in unstructured data.

It works. It's not gonna, you know, hallucinate, it's not gonna be putting things in the public cloud. So from that perspective, it really makes sense that IBM is leaning into this, and I'm glad that they are.

We look forward to learning a lot more about what IBM is doing in the future here, simply because, you know, companies like, uh, like them are leading the way in the enterprise, uh, for integration of ai. Amazon Web Services is investing $20 billion in Pennsylvania to build two major data center campuses focusing on ai. One of them is near Philadelphia, and another is powered by a connected nuclear plant in, in it's part of AWS's broader effort to expand cloud infrastructure for ai.

Though the use of nuclear energy has raised some regulatory questions, Al uh, Amazon is expanding all around the world. Um, is this any different from any of the other announcements they've made? Well, I think Amazon has always been expanding the footprint.

This has been a consistent story. As I was teaching AWS training courses for a few years, there were always new regions, new data centers being opened up. And so I don't think in itself the fact that AWS is continuing to build out their public cloud is, is any surprise to anyone.

I think there's some interesting elements in this that there is a commitment to a large build out in Pennsylvania that's surrounded by a commitment to upskilling a a population to fit that. And this, I think, is where that 1,250 new high school jobs is a, a really important thing for Pennsylvania. Uh, and this is probably the, the carrot to sweeten the deal for Pennsylvania to allow the use of, um, high density power, particularly those nuclear power supplies are, uh, definitely contentious.

I know AWS and Google both wanted to do small scale and nuclear, and have put those plans on a little bit of a back, back burner. The idea that AWS is going to bring that set of skills in bring the training that's starting all the way down at, at K12 schools through professional education, uh, through a range of different skill sets in here, I think is really significant for it. So this idea of supporting STEM education, uh, bringing more digital skills, these are the kinds of skills that the future workforce of the United States require, and that as more and more things are, uh, continuing to be augmented by ai, uh, we're gonna need to have skills to use those ais also to, to continue to build out the infrastructure around them.

So yes, there's, there's an element of AWS is looking after itself by building a workforce that is skilled in a location where maybe there isn't as much demand for that workforce as if they were building this somewhere around the, the West Coast. Uh, but I think it's also really good to see a commitment to a community that will definitely benefit from this. This is a huge investment in engineering and a huge investment in the people of Pennsylvania.

Um, it is continuing, uh, continuing investment that support for skills and training is the really cool stuff. It includes things like training on splicing fiber optic cables that are interconnecting everything, uh, but data center operations and, uh, some of the early stage educations, getting younger people reengaged in science and technology. I know I was watching the news here in New Zealand that there is a declining interest in technology for people who are in, in high school here in New Zealand.

I dunno if that's in the united, uh, it's definitely a concern for our future productivity as a, as a culture and as a society that is centered around these technologies. So definitely the idea of of bringing on local communities, uh, having a fund for local initiatives as well. It's not just AWS owned in here.

There is a, uh, $250,000 fund that is funding local community, uh, education and, and science technology. So it's not all being dictated from on hire supporting grassroots, which I think is really fundamental. Uh, is $20 billion worth of investment?

Huge? Well, it depends over how long, uh, AWS has invested as reportedly $140 billion in their infrastructure in the last, uh, 25 years. 20 billion is a significant number, but it depends whether that's over the next 10 years or the next two.

Uh, hopefully Pennsylvania is gonna benefit from this. And we will see people coming into high paying jobs in technology that are very satisfying to us. IBM announced it will build a fault tolerant quantum computer called styling by 2029, aiming to run complex tasks with around 200 reliable qubits.

The system will use new error correction models to solve stability issues that limit today's quantum computing machines. The move marks a shift from research to real world engineering as IBM races to become the leader in quantum computing. Is 200 reliable qubits gonna actually mean that quantum encryption is absolutely essential for a student?

Well, um, let's focus on what IBM is building here along with some clever bird, uh, code names. Um, I'm sad that they didn't use p techy techie as one of their, uh, bird names, but what can you do? Um, as we know, uh, one of the, uh, challenges of quantum computing is getting enough qubits to make a practical computer.

Essentially, we have created, um, it looks as though we've created, um, quantum bits that can be used to perform calculations. The problem is we don't have a lot of them, and it's very hard to connect them in a way that we do with conventional microprocessors. You're probably familiar with the sort of so many billion, so many trillion transistors discussions when they talk about, uh, my modern CPUs and graphics cards.

Well, quantum is, uh, not in the millions, billions, or even thousands of, uh, qubits right now, and that's gonna cause challenges, uh, in terms of getting actual work out of quantum computers. So, IBM has developed a couple of really cool things. Number one, um, they just published a paid paper, which actually was, uh, the cover article in nature about quantum low density parody check codes.

Uh, essentially as a storage nerd, I know a little bit about parody checks, and the idea is that you wanna to build a, um, a way to mathematically prove that your, uh, data is correct. If you can do this in a quantum computer, then it allows you, theoretically to create a more reliable computer and scale that computer passed what previous con quantum computers have been able to do. And that's exactly what IBM is gonna be doing here.

The next thing that they're going to be doing is, um, well, first they're gonna be building a, um, a test, uh, architecture. And they're calling this loon since it will be a solo chip that will implement the Q-L-D-P-C coating in order to just test to make sure that these things can work. And also that these qubits can communicate over a longer distances, like let's say, a lonely Canadian lake where the loon is calling out and no one is answering, because right now we only have one in 2026 IBM expects to launch kookaburra, uh, which will be a modular processor that will take this to the next level.

And, uh, those people on the other side of the planet, hey, Al, um, who are familiar with this bird, know that it tends to holler and tweet and make all sorts of noise. Well, that would be the idea that IBM would have modular processors that would allow them to theoretically scale quantum computing. Next up would be cockatoo, which will entangle these kookaburra modules and enable these multiple modules to connect together like nodes in a, uh, modern, uh, scalable, uh, scale out system.

And then ultimately they're saying in 2029, they'll have a, uh, a, a, uh, a, not a computer so much, but a cluster of quantum, uh, computers called Starling. And it would have effectively, uh, thousands of, uh, times more capability than the current quantum computers. It would have, um, you know, incredible amounts of, um, of computing power.

And it would theoretically, to your point, allow us to use quantum computing to do things that we never could do before. Like, I don't know, let's say solve the traveling salesman problem or break qu you know, your conventional, uh, encryption algorithms or, you know, maybe, uh, you know, whatever. And, and they're calling it Starling because starlings, of course, travel in massive flocks.

I think they're called flocks. Uh, somebody correct me on that. And, um, that's sort of what IBM is gonna do here.

They're gonna deploy a massive scale out cluster of these quantum chips. So I'm not a quantum expert, I'm a quantum enthusiast, and I cannot wait to talk to my friend Dr. Bob Sutor, uh, formerly of IBM Quantum, actually, uh, to learn a little bit more about this.

This is just cool, nerdy stuff, and, and I hope that it, I hope that it works. Now let's take a look at a bigger story. Uh, this week we saw, uh, Apple's, uh, worldwide developer conference, or as my iPhone insists on calling it wwdc.

Yeah, apple manually overrode Siri to make it say that we also saw more important stuff, including Microsoft Build and at Microsoft Build, they talked a lot about agentic ai as we have here. On the rundown, it seems that Microsoft is, uh, contrary to Apple, leaning into, uh, AG agentic AI leaning into AI generally, and making it central to their overall strategy. Uh, Saachi Nadella talked about the Ag Agentic web, where AI takes a big role in software development.

Uh, we also heard more about, uh, advances in GitHub co-pilot, making it into a real programming partner for developers. Uh, new features for managing AI agents and giving them identities and security controls. And, um, Microsoft, uh, unsurprisingly backed open standards like agent to agent or A to a and MCP, which is something that I've been talking about quite a lot lately.

The event, uh, highlighted how AI is going to be a new tool for developers. It's gonna challenge everything that people know about, uh, productivity and development. But of course, uh, success on this depends on how people use these tools and whether they really trust them.

So, Al uh, what's your overall take on Microsoft's messages coming out of build? I think the, the ambition to have an, an AI that is an assistant like a human. We've seen this through science fiction for a very long time of actual computers doing thinking tasks in service of a human, but offloading a lot of the routine tasks that we don't really want to have to think about.

And this is the core of what, uh, Microsoft is saying. They want to have, they want to have, uh, a that can take on particular action specific actions and complete them autonomously, uh, but do so in a way that is under human control at some point. Uh, we got some great coverage of Microsoft build from the future, uh, analyst group, uh, Mitch, actually, Nick patients and Keith Kirkpatrick, all good friends of ours.

Uh, were at the event and have, have written up some really extensive coverage. One of the things I picked out of it was the use of co-pilots, um, as a, a tool, as a peer programming tool. I remember peer programming as, as being, uh, something that, that came out of Pivotal and Cloud Foundry kind of days when I was working with VMware.

And the idea that having one person writing the code, another, looking over their shoulder to see if they're making mistakes, uh, was interesting. But two humans doing that is kind of tough to do culturally, having an AI assistant, an agent that's looking over my shoulder and saying, Hey, you, you haven't checked that. The input you got from the user is actually valid before you passed it on to this other service.

That might be a really useful thing to have, particularly in real time, because fixing these things in real time is so much easier than having a, a, a security scan later on telling me that I made a mistake four hours ago. So I liked this idea. Another thing that I also saw in there was Microsoft admitting that they don't have all the answers here, and that they're gonna engage with their developer community, their communities through GitHub.

And again, discover the answers to this. Discover the ways in which people want actually use AI, and how AI agents can make individual developers more productive. And I think this is, uh, quite a change.

Again, seeing this awareness that the knowledge is out there, the knowledge of what to do, the ideas of how to work out the right thing to do is actually out there amongst your end users. Amongst the, the sort of massive number of, uh, intelligent people who are doing things with this. Of course, the challenges getting through all of them, the millions of different opinions to get to something useful outta that.

As always, when you're pulling from a, a large community. The other element in here is the thought that, uh, so much code is being written by ai. Is there still a role for a human developer?

And I think pretty clearly there still is a role for a human developer. It's just that that role changes. And in the same way we saw changes from, um, when I first started programming, uh, there, there were very few libraries to do complicated things because computers couldn't do quite so complicated things.

And over time, better and more libraries turn up better development environments and shifting those development environments to allow us to do more, um, sort of the vibe coding idea of describing what you want to achieve. Now, you probably don't go all in vibe coding because you still want some ability to work, uh, deeply have deep knowledge of the programming language, the environment you're working on, but melding some of those ideas that there are a whole lot of tasks that are repetitive and very samey in software development and minimizing those so we can allow humans to do the things they're good at, which is innovating, coming up with new ideas that aren't based on prior art, right? Automate everything that's based on prior art.

And that's exactly what generative AI is good at, but allow a human to innovate and build new things. Stephen, are you building new things at the moment with the help of AI to do the boring stuff? Um, sure, and I think that this is one of the areas, uh, that is, um, maybe a little, uh, confusing of the world that we live in.

Um, it's easy to be angry about ai. It's easy to be frustrated with AI simply because I feel like so many people are still in the honeymoon phase of this thing is amazing. It does anything well, it doesn't, it really, really doesn't.

For example, if you ask chat GPT-4 oh about to compare and contrast agent to agent protocol and MCP, it will make, make up a load of, let's say horse Huck about what these things mean, including erroneous definitions of both of these protocols. Because it doesn't have that information. It doesn't know what these things are 'cause they're that new.

That being said, being open, AI anthropic, Microsoft, et cetera, have all embraced these standards. Google announced agent to agent protocol recently. And, um, anthropic has, um, put, uh, MCP out there as an open standard for ways that agents can integrate with each other.

The contract contrast between sort of this, I'm gonna ask chat GPT stuff and it's gonna make up an answer. And where the industry is headed, it could not be more stark. And that's good for many reasons, not the least of which, because most real AI applications are not gonna be of the burn down the rainforest and light up the nuclear reactor variety.

They're gonna be a lot more, um, of the duct tape that holds things together. Variety. And so to answer your question in a long-winded way, yes, I'm using AI all the time as duct tape.

Essentially, if you, for example, need to extract data from a file, it is, uh, an LLM is actually a very good way to build a standard JSON structure based on the contents of a variety of, um, unstructured files. You know, feed it, for example, a whole bunch of, uh, invoices. And it will give you a beautiful formatted JSON list of the content of those invoices.

And in many most cases, it's gonna be pretty accurate as long as you've used the proper prompt. Well, that's what models and agents are going to give us. Essentially, we're going to have little AI components that are able to do little cool things.

When you strap these things together using protocols like agent to agent or using model context protocol, which are not mutually exclusive competitive by the way, they, they, they certainly can and will augment each other in various ways. What we're gonna end up with is useful tools that do useful things and don't require yet another agent that is, you know, maybe or maybe not sentient. So that I think is the direction that the industry is headed in.

We talked earlier about IBM with, uh, announcing at think a whole range of AI that do various tiny tasks in a useful way. That's the kind of stuff that I'm excited about, uh, from Microsoft as well. I mean, Microsoft obviously, I, you know, is one of the major computing, you know, personal computing platform vendors.

But of course, they're also one of the most important cloud, uh, companies out there. And of course, they also, uh, have a huge, huge group of developers that support and rely on their tools, everything from GitHub to their IDs and co-pilot. So having agents built into that to help out in various ways, um, and having things like GitHub embrace, um, MCP allows us to see a future where we're going to see just a blossoming of really cool little AI widgets that do neat things and all work together.

And that's sort of the message that I got from, from this. And, and I have to agree with you, Alistair. I love the fact that Microsoft is coming here and saying, look, we are not exactly sure, but we're pretty sure there's a revolution in here.

Let's work together and find it. Well, thanks a lot Al for, uh, joining me today on the Rundown, or maybe I should say thank you for allowing me to join you on the rundown. Uh, Tom is busy with the Cisco Live US right now.

com or Textron TV and watch the presentations from that. We also have a major, uh, industry event coming up that is the Six five Media Summit. Um, of course it's focusing on ai.

com and you can learn more about the summit. You can register there and you can tune in. That's gonna be June 16th through 19th, and we'll be back with another Tech Field Day event, uh, July 9th and 10th, which is Networking Field Day.

Alistair, uh, do you wanna talk at all about some of the upcoming events you've got coming? Well, I'm also focused on AI as well, so AI infrastructure field day four is, is coming up a little way out from, uh, from that. That'll be in September as well as of course, I have another Cloud Field Day.

I've just come off Cloud field day 23 last week, which was absolutely awesome. We're looking forward to another cloud Field Day once we get out into October. And around that time, Stephen, you'll be doing an AI field day.

I think it's interesting that we've separated out AI infrastructure, the things you need to build to be able to run AI from AI Field Day, where it's the things you do with AI and how AI is delivering business value. I hope we're seeing a lot of that business value at AI Field Day. Absolutely, and, and you know, for me, that's really what I'm looking for with AI technology.

Let's, let's get some value outta this stuff instead of just messing around with it and seeing if it can do cool things and fool us into thinking that it's human. Well, thanks for watching the Tech Field Day rundown. You, you can catch new episodes of The Rundown every Wednesday as a YouTube video or in your favorite podcast application.

The Rundown is also streamed on Techstrong tv, including our brand new, over the top video app in Roku, apple, and Google. You can also catch us on other Techstrong and RUM group programs, and you'll see me, for example, every Tuesday on the Textron gang. We'll be back next Wednesday to talk about the IT news of the week.

That was, but until then, for myself, for Alistair, uh, for Tom and all of us here at Tech Field Day, here's wishing you and yours a super chocolatey corn on the Cobb Day. Hey everyone, it's shimmy. I've been out a la the last couple weeks.

I apologize if you haven't been able to watch. I, you know what though? I had an amazing time.

I went on a family vacation, I do that about once a year. We went to Italy and stayed in Tuscany in a villa and did some other things. And you know what, if, if you can afford it and if you can get away from work, it's, it really is a good thing to just take some time off, get out of the kind of hamster wheel rat race and try to take stock of, of where you are and what you're doing.

As a matter of fact, that that's the theme for today's shimmy says, right? I'm calling it keep calm and smell the flowers. And boy, do we need that advice today.

There's so much going on in our world. I'll just tell you on a macro level, we couldn't even do. Shimmy says live today.

It seems Google Cloud has some issues and a lot of the providers that we work with, streamy Yard and Kinta and a bunch of others evidently are on Google Cloud. So we couldn't even get a a a key live, a live stream into it. So we're recording this, but I'm not gonna let it upset me or ruin, you know, ruin the mood.

As I mentioned, there's so much going on in the world, right? A lot of people here in the US are so angry on one side or the other of things. We have riots, I don't even know if I'd call 'em truly riots.

We have protestors marching in many, many cities. We have an administration calling US troops in to police US citizens. That's not something you see very often in America.

Um, there's so much going on in the world. Iran situation with nuclear enrichment war in Ukraine and Russia is not stopping. It seems the situation in the Middle East with Hamas terrorists and Israel.

Um, there's a lot to be upset about. There's a lot to be uncertain about the economy. Ai as cool as AI is, and as many different things as we're seeing being done with ai, a lot of us are still worried, is AI going to take my job?

Is it gonna change fundamentally what people do for jobs? It may, maybe not today, but it may certainly in the future. But, you know, I've been around the, the sun a couple of times in my lifetime, right?

And then, and you see patterns and you see things repeating. I was a little boy in 1968. I was probably seven years old the summer of 68.

And you know what, there was a time that was a time of unrest here in the us. The Vietnam War was, was widely unpopular. College campuses were in revolt, but it spread to the cities.

And you know, they burned down Newark and most of the major cities in the US had widespread, uh, protests. And some of them got violent. You know, we were lucky.

We were lucky back then. You had people like Dr. Martin Luther King who saw that change was needed, that we had a, couldn't let the status quo go.

But also were big believers in, in peaceful protest, in in civil, peaceful, civil disobedience. And it probably saved a lot of bloodshed. Unfortunately, I don't see a leader arising today that kind of takes fills in those big shoes of Dr.

King. But I, I'm hopeful that one will arise. That, that we will recognize that violence begets violence, and it's okay to protest and express your freedom of speech.

It's not okay to burn or destroy people's property or the government's property. And no, we we're a country where our armed forces should never be turned against our citizens. This isn't China and it's not the Aman Square.

So I, I really hope that it doesn't come to that either. Um, but you know, some interesting information. So that summer of 68, the summer of discontent as, as it was called, gave way.

And it was also a time where we lost leaders like Dr. King and Bobby Kennedy and others, right? And the country and the world was, was shaken to its core, but we emerged outta that stronger and better than ever, right?

And, and the war did end and things did get better. And those protestors, you know what's funny? A lot of those protestors turned into Reagan Republicans in the eighties.

So you never know. The people out there protesting today could be tomorrow's, you know, establishment people. And, and that's the way of it.

But the, and a lesson I've learned in all these times around the sun is, it's, it's not, it's okay to be worried. It's not okay to panic. 'cause when you panic or you get too caught up in it, you, you, you do things without thinking them through and that never ends well.

So my advice to you as someone who's been around is keep calm, keep calm this through shall pass. I'm not saying don't express yourself. I'm not saying if you feel compelled to go protest, you know, dragging people of different colors or nationalities out of their workplaces, 'cause they may or may not be here legally, that's your prerogative.

You're certainly free to do it. And I, I support your right to do it. I support your right to free speech.

I just don't think violence or destroying property is, is the way to, to, to express yourself there. Um, on top of this though, again, you know, that, go back to my trip to Italy. I saw a lot of very happy people in the countryside of Tuscan, and it's because I think they don't get as wrapped up.

They remember that there are other things that we need to focus on as people focusing on your family, focusing on your career, focusing on just being happy, right? I I think a lot of us are so, and, and me included, I'm, I'm guilty, I'm not saying anyone else is, but a lot of us get so wrapped up in, in what's going on in the world around us. These, you know, war as I said, war and, and nuclear weapons and, and killing and murdering and, and illegal kind of authoritarian activities.

And, and we, we lose sight of the, of the good things in our life. What do you have to, to be happy about? If you have people who love you and, and people who you love?

You're lucky. If you're lucky enough to be able to take a vacation and just get away for a couple days or even a, a weekend, take a good long weekend and refresh yourself and remember, you know, just look at nature, man. I, you know, I come home, I'm home from Italy almost a week, and I, I drive my bike along the beach in the mornings.

I get up at Sun Sunrise. It does such wonders from my brain to just the solitude of driving along the beach. So again, calm it down, take it down a notch.

Don't, don't get caught up in the cable news. 24 hour cycles showing you cities burning cities in the US aren't burning down. There's a very small area, a mile square maybe of downtown LA where the most of what you're seeing on TV is taking place.

And it isn't all that you're seeing on tv. There are a lot of people demonstrating very and expressing themselves peaceably don't, don't buy into all the hype that these 24 hour news channels will, would have you believe. I think you gotta focus on what's important.

And what's important is, is your health, your wellbeing, your loved ones and your family. So that's where I'm gonna end it today. I am not gonna lecture you on AI or agentic AI and all of the great things going on there.

You could watch tech drunk TV all week for that. What I am gonna say to you, shimmy says, keep calm, smell the flowers, hang in there. And this will pass as well.

Says.