Techstrong TV May 28, 2025

Hey, everyone. You know that big, beautiful bill may wind up governing how we use ai. You're watching Textron.

Hi everyone, it's Alan Shimel. Welcome back. Happy Wednesday to you.

Um, I have to apologize, I'm a little under the weather. It was a tough weekend. I had a, some kind of thing going on, but we're here and we are moving forward.

I hope your, uh, Memorial Day weekend, if you celebrated it in the US anyway, was good. And, uh, we've got a full lineup here today. Let me introduce you quickly to our gang members and we'll jump right into it.

First of all, uh, he's from Fu Invisible Impact, or he's principal analyst and co-founder, uh, and our friend Guy, Guy Currier Guy. How are you? Good, good.

Joining you all from Austin, Texas. Good to see you back home. Speaking of home, he's the man from the north now wi with it, with official, uh, uh, with official duties as well, but he's our cybersecurity expert, Chris Blask.

Hey Chris, how are you? Love. Thanks for having, being on.

And then finally he's up in the Bronx, loving this winning streak. The Yankees seemed to be on there, getting their mojo going. He's our chief content officer, Mike Vizard.

Hey, Mike, how's it going? And it, it's going great and of course it's not enough to win and he, everybody else has to lose. So, so far so good.

Yeah, but who would've thought a guy named Yarborough and a guy named Warren were going to be two of our top pitchers? I would've thought, man, minor league system. All, all this time, Alan, I thought your voice was gone 'cause you were screaming for the Knicks.

Yeah, Yeah, I wish it was, but no, I, it was a tough weekend for me. Um, I'll see a note for the, uh, Texas Stars, uh, the a HL hockey team here based in Austin, Texas that just advanced to the semis in the, uh, for the cup, whose name I don't even know in hockey. Very cool.

Very cool. Very cool. Alright, let's jump into it, Mike.

So we've got this big, beautiful bill winding its way through Congress. Of course, the Senate has got their Senate, a bunch of senators have their knives and forts out ready to chop up this bill. Um, well maybe add more to it.

You're right. What do we got? So the deal though, is that tucked into this big beautiful bill that covers a million different things is a provision that says that, um, states will be banned from having their own AI regulations for 10 years, which by the time I got around to regulating anything would mean that it was pretty much moot at that point.

It seems like, uh, a, a power grab on the part of the federal government with the aid of, uh, some folks in the House of Representatives. We'll see what the Senate has to say about this and whether it makes it through. But, um, you know, on the one hand, Alan might make sense for us to have a national AI strategy, but on the other hand, this is just enshrining in the law that we won't have an AI strategy.

What's the point? You know, look, there's so much nonsense and crap packed into this bill that I almost wanna say this AI piece of it is small potatoes compared to some of the other fascists that gender that's in there. And if we didn't, if we had a normal administration that plays by the rules and norms of society and humanity, it may not be a bad idea, but when you have an administration that is out for their own personal wealth accumulation, right?

And you've got what, what we've seen going on here with self-dealing and, you know, violations of the emo moments clause and so forth, I am hesitant to put anything into the, solely into the hands of our federal government. Um, you know, I didn't see anything in the constitution about AI rights, but you know, clearly in the Constitution, anything not taken by the federal is reserved to the states. I it'll be interesting to see if the Senate swallows this, because, you know, it's not just that all states and local administrations are forbidden from having any regulation or any policy at all.

It's also that they're socking away hundreds of millions of dollars, if not billions of dollars, hundreds for AI over these next couple years. Now, is that money gonna wind up in the Trump family, uh, library on the Qatari plane that he's getting? Is it gonna end up in the UAE AI center or some other country that agrees to have a Trump building?

Let's face it. This whole thing is, I mean, how we, I, I'm speechless right? In, in another time, in another place I'd welcome a coherent AI strategy.

But this is not a coherent AI strategy. This is everybody be quiet and let the, let the Pigs feed at the trial. So, guy, you're in Texas and are they gonna sign off on this?

Because let me pause it this way. Let's say four years from now, there's another election and somebody else is in charge of the White House, and now they got a six year run of AI policy that they can go with. And the states have no say regardless of what color they are, right?

Or both? Well, um, so I have two thoughts about this off the bat. First of all, I agree with Alan.

Let's put policy and square scare quotes. Um, if there's a policy behind these, these kinds of proposals and, and you know, a lot of the, the big beautiful B-B-B-B-B-B-B-B bill, um, that policy is, uh, allowed, you know, uh, for, you can call it grift, bribery, corruption, whatever you want to call it. Just, just enable that, um, concentrate, concentrate, uh, decision making, um, you know, among individuals, uh, whose personal loyalty to each other dictates whether they get to feed at the trough, as you say.

So that's my, that's my first thought. That's not really a policy. There's no policy here.

My, my main thought, my main reaction was that this reminds me a lot of, uh, autoregulation, um, where there was no, you know, there never has been any, um, ban on states having their own regulations. So effectively, California, which Dely turned blue, and it's now pretty, pretty, you know, democratic and democratically run, um, dictates a whole lot of the, uh, compliance that automotive manufacturers, uh, adhere to. Um, simply because it's such a big market that carving out California alone for, for example, emission standards doesn't make sense.

So they just apply emission standards everywhere. So I felt like this was just a preemptive strike against something like that, where there might be no AI regulation at all, uh, on a federal level. But, um, just as the state of Texas, and I think Florida and some other states, um, banned TikTok on their own or whatever the, I don't remember the details of that.

Um, they wanna avoid they being, you know, the grifters and the federal government want to avoid, um, states doing anything. And then why do they wanna do that? Um, well, um, I think between AI and crypto, um, the, you know, so-called oligarchs or, you know, the, the billionaires with free money to spend on whatever, um, have essentially worked what they can to buy off central decision making in the us um, federally.

And, uh, they have been making a lot of arguments and they're doing in the UK as well, um, that AI should not be regulated in any way at all and should not have to pay for the copyrighted material that it, it uses to get trained and all those sorts of things. So it's kind of the old story new again. Um, I almost feel like we're going back to the days of the Roman Republic here in the us at least temporarily, I don't think Empire guy, not the Roman Republic, the empire.

Well, okay, fair enough, fair enough. Uh, potato, potato, we can discuss, uh, that later. And your point being, my point, which is it's not what you know, it's who you know.

Um, it's not what you've been entrusted with, it's what you can get away with. So this is another example of that. I really think it's like the automotive and the emission standards.

They just don't want states to, especially big states, to, um, uh, have any influence at all so that nobody can have any influence. Well, Chris, Chris, around the world, these centralized authority typically has this, right? Anyway.

I mean, it's kinda unique to the United States and, um, in Canada for instance, if it's not in the Constitution, it's assumed to be the responsibility of the federal government here in the United States. It's the opposite. If it's not in the Constitution, it's assumed to be the responsibility of the states, but I think every other country is much more like Canada.

So our, you know, are our friends around the globe probably wondering what the heck we're debating here. Yeah. So there's a classic, uh, Disney floating around, you know, the IT space probably even, you know, it keeps getting into, you know, shooting yourself in the, in the foot with, in different programming languages.

Well, I think most people are familiar with that one, you know, and so forth. And this, this, to me, seems exactly the same. Like this is miracle shooting itself in the foot, you know, because as you sent Alan, again, I can, you know, I can understand this, um, at the federal level, you know, for, because this is, you know, highly evolving space and, and critical and ex existential threat economically and all sorts of other ways.

How when I, you know, one of the great things that the states has always had going for us is the 50 experiments, you know, state's rights, you know, let's try a bunch of different things all at the same time. We don't agree, move through it fast. It's been one of the things keeping America with the head of the pack, you know, uh, some somebody else's comments here, just, just structurally technically, you know, um, competitively, you know, to say that we are going to stop, you know, one of our big advantages, we're gonna stop states from acting like Muslim states and have them act like Canadian, Russian and Chinese provinces, you know, that they've just, you know, subsumed in the wall and we will compete.

It is, this is not, I'm throwing things away, you know, mission made those fault being strong, um, in the most important evolving area. That's okay. Yeah.

Does this renter actively somehow turn off to be a good idea? Unless that's physically possible. Um, we'll see.

Alan, do you think that this is tucked into this big, beautiful bill in the hopes that no one's gonna notice? I I think there's, as I said earlier, this is small potatoes to what's tucked into this big beautiful bill. You know, look, and now to be fair, the Republicans don't have a monopoly on this kind of nonsense.

You, you know, it's always, these bills always get packed in with, with pork barrel and all kinds of stuff. But I've seen reports of some of the things that are tucked into this bill, and it's enough to make your toes curl. I mean, it's just another step toward, towards an authoritarian government.

And quite frankly, I, I don't blame the executive. He is what he is, but I do blame the yellow belly people in Congress who are gonna sit down and lay, lay down here and, and let this happen to our country. You know, I'm sorry if that's a little too partisan for you, but that's pretty much how I feel.

We don't know that that will happen yet. We, we know that, um, well, we Know the House of Representatives, The house of, yeah, it's going to the Senate now. Um, it seems to me to do my own political analysis, it seems to me extremely likely that it will pass in some form.

Um, I, I wonder if they'll hit their deadline. They, they also have the very important matter of going on vacation to consider. And I rather wonder if, um, certain, you know, it, it takes, I think it would take four Republicans or something like that to, to, it's a little confusing, you know?

Um, 'cause normally they'd have to pass filibuster. They'd have to make certain changes in order for not to, who knows? It's a little confusing.

Also, one gets the feeling that, uh, you know, another norm will break, or if it's a norm rule, whatever will break at some point. I just think that they're gonna do anything they can to pass it. Um, and they won't change it too much, and it will pass, pass with dispatch, because otherwise they'll have nothing to say that they did, essentially.

And if they're don't pass it now, then they start pass, they pass it closer to election season, and it's gonna be, um, a, a black mark on them. So I do think you're probably right, Alan, but we, it's not for certain yet. The Senate might make some changes For the better.

Alan does goes, does this go to the Supreme Court as a test of just where the line is between what the federal government is allowed to do and what states control? I mean, who said That? Well, I, I don't believe so.

I think if the federal government legislates it, it preempts state legislation. This is, you know, it's not a state's right issue. This is Congress abdicating its responsibility to be an equal, an equal branch of government here, and putting a check and balance on the executive branch.

There's another aspect of this bill, though we haven't discussed, and I think it's also important, is that this bill, clearly FLA favors big tech. So you're gonna have the same old, same old. You're gonna have the meta Google, apple, whatever, the Max seven, right?

Tri, you know, riding triumphant here. But this real, you know, in, in a, a new thing like the, like ai, you would hope to see a million flowers sprouting a million new growth trees sprouting to harness this new advantage and look at different ways of doing it to compete. Instead, we'll look for the competition to come from China, maybe Europe, right?

Because we're not encouraging with this bill, we're not encouraging growth at a grassroots level. This, you know, if you don't have enough money to, to Greece, El Duchess palm, you don't get to play in this game. And that Well, yeah, and it's even worse.

It's even worse because, um, the Supreme Court looks ready to, um, to overturn about 90 years of, of precedent, um, some point soon. Um, the, the, I think the case is called Hubble, I'm not sure. Um, FDR area, ERA case where, uh, the Supreme Court ruled that the president can't willfully fire.

So members or leaders of so-called independent agencies, board, board members. Um, and it looks very cl that looks very close to being overturned, which means that, uh, among other things, the president could fire the commissioners at the FTC. So if, if, if the commissioners at the FTC serve at the will of a president, um, who, I mean, what does the FTC do among many other things?

The FTC allows companies to, to get licenses and allows companies, especially in the case of disputes, to continue trading, continue doing commerce. And one of those could be Apple, and it happened that's not given a plane or a, or, or, or a boat or a, you know, several solid gold iPhones or whatever lately, um, the president could say to the heads of the Federal Trade Commission, uh, I want you to sanction start procedures to sanction ensue apple, which then Apple would've to deal with. Or I want you to delic apple for export or whatever it might be.

And the, if, if they can be fired at any time by the president asking that, then they're more likely to go along. If they don't, they'll be fired and someone who will go along be put in their place. So it's about to get worse.

Um, if, if that does go in that direction, all we can hope for is to slow it down long enough, um, to put more proper checks and balances in, in its place. This is all part of a piece, which is this unitary executive theory, but also the essential authoritarian theory, which is that people want strong men. Usually it's men, strong men, um, in charge, because that makes everything and everybody stronger.

I don't, I don't, uh, believe in that at all, but I think that's a general belief. So ultimately, we're gonna see more AI executives doing photo ops in Mar-a-Lago standing next to the president while they drop, no, while they drop off large bags of cash in the back room. Is that where we're going?

Yeah. I told you like the Star Trek episode where they built a society based upon the Al Capone book. That's, that's where we're going.

That's where we're going. All right, let's take a break here while I packed my bags. Um, we could come back.

Let's talk about something maybe not as politically contentious. You're watching text Trump Gang. Hey, folks, we're back.

I'm not sure this is gonna be less contentious, but it probably won't be as political. How's that? But Java has turned 30, and I don't know, guy thinking back in the time when Java first came on the scene, I don't think I thought about it as kind of this language we were gonna live with forever.

But here we are. I mean, there's more programming languages than ever, but seems like everybody's still using pretty much Java. Everybody is using Java, at least in the enterprise.

Uh, I mean, um, I suppose there are good parallels with cobol, which originated in the mainframe area and mainframe era, and continues to be used in mainframes to this day. But I think better, I, I would, I would raise that, you know, by, uh, uh, you know, by one, which is that, uh, Java, thanks to the Java compilers and Java virtual machine is much more portable. It's scalable, it has security capabilities, it's resilient Java, the language.

I wouldn't, you know, I, I don't care for personally, I'm not really into c or or c derive languages at all. That's just my personal preference. Um, and the theory of, um, of, you know, writing in a particular language is not the point here.

It's the platform. That's the point. The Java virtual machine, I mean, so Mike, I'm also maybe just barely old enough to remember when the Java virtual machine originated as a way to run Java applications on your desktop.

I mean, that was the whole idea was to gain, um, uh, interactivity with portable apps that, um, you could find at a website or somewhere else and just run on your desktop. Um, it morphed, I mean, morph as needs. The, the whole thing got sort of revolutionized, um, with a recognition that, um, you could put agents into servers or full blown applications into servers using the Java virtual machine is almost like a pre, um, a a primitive, extremely primitive, uh, uh, predecessor to the container.

Um, so I celebrate Java's anniversary personally. Um, I don't care what language you are programming in, but I do care about mobility, portability and scalability and all sorts of things. And I also care about not wasting time on trying to, uh, recode or refactor or re whatever on applications that run perfectly well.

And I think that there are many, many homegrown applications or, or OpenShift based applications or what have you, using Java and the Java virtual machine. There's no there there, in other words. Um, but Java has its detractors.

And I think mostly that has to do with it not being cloud native on its own. It's being sort of old and legacy. Uh, to me, the, the, the main critique, uh, or the critiques boil down to vibes, which is go is cooler, rust is cooler, um, that's old fashioned, you know, that kind of thing.

Um, I think if there's a talent issue involved, um, that could be something. net programmers out there. If it's boring, that's not necessarily a bad thing for it.

And for the enterprise, I always appreciate those mainframe modernization conversations where it's basically, let's replace a 50-year-old language with a 30-year-old language. 'cause that's gonna make a huge amount of difference. But Chris, Well, hey, listen, listen, VMs run on mainframes via, uh, Java VMs run on mainframes, and those Java applications can port over to mainframes.

But yeah, go ahead With, with a little work in, in about a six to nine months worth of effort. Um, Chris, there are one big criticism, Java my mind is the security aspect of this. And it's a language that is 30 years old and it is prone to have a lot of security flaws in it.

And there is an argument for moving to things like rust that are fundamentally more secure. It's harder probably to program in rust, especially you're known Java for 30 years, but do you think maybe there's a time and place where our security concerns will finally maybe push us to think about a different language? Maybe we're being Touched on a couple of the, the factors thing, right?

You know, if, if it is harder to do something in the new one we're replacing go with, then, you know, does that make it less secure? Intrinsically? And I'm not saying in, in a talk show like this, you're not gonna answer that.

You actually have to think through these things really clearly to give you an actual answer. At least the fact that guy, you went there with mainframes. Because as, as I think about this topic, I loved, you know, in the, in the teens working with Unisys and realizing that mainframes are still out there, the ent entire global economy is being run by code that was written by Grace Hopper, you know, before, you know, dumb road time.

And it's running on mainframes that are really just intel, you know, merged clusters of Intel servers, um, running emulation software, you know, that's nesting all this, you know, 700 year old code. The actually works just fine. And I think that's when you get into security, you know, these obvious answers just aren't, you know, let's go to the new thing.

It is actually more secure. Will that make us more secure? Maybe not.

So, it, it's a fascinating topic, but every time this comes out, you know, there's a newer better Foster. And I can sit here and say, looking at just that component, yes, that is a more secure way to do things. Now, if you actually replace that in the real world and replaces that depends, you know, people, humans, processes, you know, economics, it bloody depends.

Will we be here for the 50th anniversary of Java? What do you say? Yeah, If be 50 twos are still flying.

So I say that, so, So, uh, I think one of the real concerns and criticisms about Java is, is Oracle's ownership of it. Um, I think that that's, I mean, there's, what is it, open, open, JDK or whatever. There's, there are some, there are some open source alternatives.

Um, as with all open source, open source, uh, you know, um, you gotta think DIY at the same time that it tends to be complex. Um, even if you have, uh, a service provider or somebody you're paying to make it less complex for you, there's this, you know, um, uh, uh, fly-by-night aspect to doing, you know, an open version of Java. So, um, your best bet is to pay the licensing fees to Oracle and continue that way.

And that is probably not a great thing for the industry as a whole. Um, I'm not sure why Oracle has an open sourced Java. Well, I could tell you why it's not Who it's, I I think it's a, some business decision to do that.

But, uh, you, you Can tell us why. Yeah, they dug the open source piece of it that my sequel, so, so here's my take on this. So from day one, almost Java's been the, the language everyone hates or loves to hate.

It was always, you know, the promise of it. And, you know, when it first came out, it was sun. And, and for many of us of a certain age, sun represented what was good in computing, right?

They, they, they were supporters of open source. They did have cutting edge technology, they moved the needle, and they seemed that it was a good company. Um, but even back then, the idea of the right, right?

Ones run everywhere, and this was gonna be the new thing that replaces everything. There were always, you know, doubters out there when, when, when it got sold to Oracle, the tide really turned against it, even though it kept being popular, even though how much It's 70, 70, 80% enterprises have it. In what way does, does that mean?

Is that indicating that a tide turning against it? Well, The, the, the tide of public opinion, it was old, it was slow, it was insecure, it wasn't open source. Oracle's an evil empire.

You can't trust them with it. You know, are they updating it? Are, you know, the, and, and all the new ones, as you mentioned, there were always, I don't wanna call 'em ankle biters, they were more than ankle biters.

There was a DA definite type of people that were anti Java for many different reasons. But here we are at 30 years old guy. As you say, 70 to 8% of organizations still use it in some form or another, in some capacity.

And it, Mike, to your point, in 50 years, they still will too. And Oracle, yes, they have an open source that they're not good stewards of open source or, you know, this is the same thing with MySQL and why we have such a strong Postgres community. They're not, you know, stewards of open source projects that, I'm sorry to say, um, but they do keep it current, right?

They do update it and, uh, you know, Java's dead long lived Java. Yeah, I would say the pace of innovation and Java's accelerated in the last two years or so. And a lot of that has to do with the open source contributors to that effort and Oracle itself.

But if you look at what they're doing, um, a lot of what they're doing essentially is borrowing the better ideas from other programming languages and moving 'em into Java faster, which is fine. We've seen that model before. It's kind of similar to what kind of what Microsoft does with Windows.

And for a certain class of customers, that works just fine. But to me, the biggest issue is just simply inertia. I don't think, you know, for all the developers who say that they will write in other programming languages, they all have a favorite.

And I think every CIO knows that it's easier for them to go find a Java developer than it is any other developer. And then the whole thing just starts to self perpetuate. 'cause now the only people I'm hiring are Java developer, so we keep using Java.

Yeah, I think that's right. I think that a lot of times, um, including on this show, we, we, we, I almost feel like it's a part of the developer culture, um, which is, uh, you know, a builder culture. And so when we see that's the main way I've always explained how many languages there are out there and how many new ones keep popping up, whether derivatives or brand new ones.

It's almost like, um, you know, a certain, you know, a a certain mindset says this language isn't quite doing everything I want it to do. It could be just as simple as, uh, memory allocations, or it could have to do with, you know, mobility or whatever it is. Something sort of more under the hood rather than the syntax of the language.

And so they just, they, instead of writing the code using what they have, they create a new language. You know? Um, and uh, Python was one of the greatest examples.

The best Python was just, it was sort of a basic successor. It was just easy, um, an easy to, to learn and use language suitable in multiple environments. And then somehow, uh, uh, almost by chance it became the language for AI development.

Why not? But you can use any language that suits you. I happen to love Pearl.

Um, I think it's elegant and cool and uh, it gives me a thrill to use it. I really think it's the platform. That's the question.

The Java platform, um, I'm not familiar with if there are ways to write, uh, write in other languages for the Java platform. The Java platform is where the security vulnerabilities are, but the, the Java platform is also where this nice integration with various kinds of enterprise infrastructure and in and enterprise platforms makes it scalable, reliable, you know, that sort of thing. So long-lived, Java, Java is dead long-lived.

Java, You know, it's funny, in the age of ai, I was talking to somebody this weekend and I said, you know, how's your job changed? And he said, well, I'm not actually writing more code. I'm reading more code.

And the code is written by the machine, but he has to understand how the code was created by the machine. But, um, I don't know if AI will result in us adopting the other programming languages because it might become easier or if the barrier to taking on something different drops. But certainly in the nature of being a developer and, and a Java developer is gonna change Like every developer.

Yeah, I agree. Absolutely. Alright, let's take a break.

We'll come back and discuss our third block today. You're watching techron Gang, Discover Techron 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 and we're gonna have a little chat about will post quantum cryptography and encryption.

Uh, there's a report out from our friends over at Digi Cert, and depending on your perspective, the glass is either half full or half empty. But they noted that only 5% of folks are currently using quantum safe encryption. But we don't know when Q day is coming, which is when quantum computers might actually break the encryptions.

We have, but a lot of people feel like it's gonna be sooner than later. Chris, is this one of these kind of slow moving crisis that we can see coming and yet we won't do anything about it till it gets here because we're just too busy? Oh, I don't know that that, uh, that, that, but, uh, you know, privacy, security.

Yeah. And typically privacy and sequencing the not time. Right?

You know, one of the things that, you know, one of the, one of the ways security people like explaining this to, uh, to, to normal humans is, uh, is talking about blank faults and so forth. Like this is a, uh, a fenced little needle felted, uh, um, Pegasus, my youngest child Terry, made for me. And I can use this.

I can get through any bank fault in on earth. All I need is time I, until the middle is gone. I going through, um, all this encryption we've been using, we talked about this another episode a month or two ago on, on that show.

All this encryption we've been using since the dawn of time gives you a certain amount of time to keep your secrets. That's it. You can build pyramids, you can build chambers.

It just increases the amount of time to get in. So on this, yeah, quantum encryption is coming, quantum computing is coming, which will blow away all the encryption on all the data that's already encrypted out there. So if you have no data that's encrypted the old ways, you need to assume now that, as you say, when we get to QA, you know, shortly after that, all of that's going to be public and build your response and security around that.

Right? You know, and I know what people are meaning by, you know, quantum proof, you know, encryption that you could use today, and that's a good storm. But to be flare, that will not keep your secret secret forever.

They will keep them secret longer. So this is, I mean, I don't know how to, you know, downplay this, this is a big change. People really need to think about this, but it's not new.

You know, you know, we, we flaw is by thinking that since we encrypted it, since we put in a bank funnel, it is safe forever. It is not, you know, secrecy is on a, is on a timeframe, always. You have to think about that in advance or get ready to be surprised at times like this.

Oh guy, do we need to change the tenor of the conversation around this? 'cause I think it's too technical maybe. And maybe we need to go back to business leaders and say, Hey, you know, people are hoarding your encrypted data out there and they're gonna expose it to the world in the next, I don't know, three to five years, shall we say.

Um, and then, you know, the, do business people get more concerned then because they're, they're assuming that encrypted data is gonna be secure forever? No, they don't get more concerned. Uh, I, how long have we been having this conversation?

You know, the thing that has a low probability of happening, I would say 1% probability is a high probability. Let's go with 1% per year. Um, shoot, I could kick that can for a while.

There's, this is only gonna happen to me once every a hundred years. Why do I need to worry about this? Um, and we keep hoping or thinking that some big breach or some big problem shared data or some, I mean, possibly if a com a a company of sufficient size goes completely under as a result of a breach, that might be enough of a warning sign.

But otherwise, I don't know if the, if the, if the average human mind and, and by I average the bad, I don't know how to put this, the typical human mind, I dunno. If the typical human mind is able to process just this simple, I don't even wanna call it nuance. Chris's perspective is perfect, which is to say, um, it's not really weather you can be breach.

It's how long it will take. That's really simple, frankly, that is simple. But the mind, the human mind does not tend to work that way.

The human mind works with, it's either safe or it is unsafe. And, uh, when an expert is providing, you know, a little bit of color on the safe, but yeah, it's safe in the sense that it would take 10 years instead of take one year or take 10 seconds or whatever it is that I, I don't think business leaders in general, some do, but I think for the most part, they don't think that way. I think decisions are made that way.

I think that is the fundamental problem security. Um, which is, that is a perpetual fight, um, that you should measure in terms of, you know, a scale, not in terms of a is or is not. And that's just too difficult to rock when what you want to do is make your next bonus and have a happier Christmas.

You Know what, I think that's an actual business value to, well, you're saying guy, you know, just thinking this through, right? Because we, you know, maybe it's just those of us who've done security for a long, long time get this feeling. But like, you know, I see it and as a business person myself, you know, I'm an executive and I make business decisions.

And if I have this idea and say, look, I can lock this away forever and move on. It leaves, leads me to a certain path. If I say, alright, I can lock this down for a while and my competitors are gonna find out and they're gonna get it, and I had to be planning for that in advance, it keeps me agile.

I'm always moving forward. I'm not gonna see any laurels. So, and again, I feel very strongly about both these things.

Um, so maybe I'm downplaying I, you know, as somebody who releases a lot of people about security, you always take us whatever you wanna do, that's what you wanna do, and that's fine, and I'll help you help you do that. But I think it's, it's, it's more than just a security risk. I think it's a, it's a business risk.

If you think your business is gonna be stable, your world's gonna stable, be stable because you can keep them things as they are for a long time, you're probably just wrong. So, you know, I, I followed this since nist along with industry participation, came out with the first quantum pool algorithms for digital certificates, et cetera. I think there's a very simple explanation for this.

Most people don't believe Q days coming anytime soon. A lesson we've learned in security, and I've learned this over 30 years, is it's just in time security. No one wants to be taking security precautions prematurely, number one.

Number two, I think this problem will take care of itself because what's going on in our world is that certificates are, are being set to expire much more, uh, frequently. You know, like you don't buy a certificate and it's good for five years. As a matter of fact, Google is trying to make it that, I believe every certificate needs to be renewed every 90 days.

And when, you know, the fact of the matter is these certificates, there's, there's three or four companies that control 99% of the digital certificates out there, DigiCert's one of them, what they already are building in the quantum proof algorithms into their newer digital certificates. So when you upgrade to a new certificate, you get the quantum proof algorithm with it. So I, I think this is a great way to, to this survey is a great way for us to cover it in the media and make people aware that Q Day is coming, but this is no Y 2K.

And I say that with all seriousness, um, you know, but this is no Y 2K and I, I think it's a, a, uh, situation that'll be resolved. I should mention that DigiCert has an upcoming Quantum Readiness Day as they did last year. And, um, I am a judge for their best use of quantum, uh, hackathon type of thing that they do.

So I, you know, I am, I guess conflicted, or at least I'm disclosing that's the way it is in the collection there, right? If I say and put it out in the open, I'm allowed to take money and everything. So I'm letting everyone know.

And Mike, Mike, what, what, what do you, what do you, even you you've been writing and you and I have been writing and discussing this, you know, for a long time, 25 years, whatever it is, and it, you know, it's, it's, it, it seems to me, uh, that, um, you know, back in the day, um, we were seeing that was before a lot of the themes we see these days around like the human side of security and what, uh, you know, social engineering, that sort of, so those terms came along, but back then, more or less in the beginning it was, I mean, it's the same as today. I don't really see any difference. There's discussion out there, there's security vendors trying to raise the salience of the danger.

They're making their sales and they're making money. But the general culture is one of wait and react. And I don't, I just do not see that having changed at all.

I'm not, I think that there's two other factors at play here. One is, so if you're talking about, you know, three year window, maybe five, the probability that you are gonna upgrade your IT infrastructure to something that has these algorithms built into it. So it's crypto safe is pretty good.

So I think, you know, there'll be a lot of folks upgrading infrastructure and moving applications, and I think the building of applications and the replacement of legacy applications will accelerate in the age of ai. So they'll, I'm not saying all these apps will go away, but I, I'm saying that the level of risk might be smaller because the pace of modernization is gonna be faster. So, but I don't think that has anything to do with security as much as better security is a byproduct of those two events.

Would you calling it? Well, I, you know, since I went into applied philosophy with my last comment, you know, let me take this one to 11, because I think the very idea that you're gonna, you know, keep things secret forever, it leads to bad business processes, right? You know, and if you, you know, call me a boy scout, if you will, there was one.

Um, if you as a business and a person, whatever function we're talking about, you operate with the idea that people aren't gonna find out whatever you're doing, we say, and we'll find out, it gets harder to justify the idea that no, don't tell anybody. We'll, we'll, we'll keep this amongst us, um, because that has worked for all of human history. It works, you know, the people who take advantage of things, criminals and dictators and so forth, manage to keep secrets long enough.

If you accept the idea that, that you can yes, manage your privacy and you can keep, you know, you know, with your competitive advantages, you know, behind the veil long enough to do whatever you want, but it will all come out. It kind of just logically works. A lot of potentially corrupting influences outta business processes.

And again, you may be able to capitalize sometimes on corrupt processes and usually, you know, get the, the golden ring, but it's not a good plan. It's not repeatable, it's not consistent. So I think that the very idea, and I think in, in the digital age, you know, maybe we're just, we're we're calling out this historical, uh, flaw because we have business people saying, alright, you're the security people.

It's only how to be perfectly secure forever. It's like, we gotta keep a secret forever. You cannot stop trying to, it's bad for you.

There's a lot of people already using various messaging apps out there to have meetings they probably should not have because they're thinking that it's secure. So I'm looking forward to if Q date does come, there's gonna be a lot of disclosures of things that we thought were one way four or five years ago that it turns out are not. So, and maybe, you know, it'll affect business leaders, but it will most certainly affect government leaders when, you know, things that they thought that were state secrets are now being readily shared on the front page of your whatever favorite is.

Yeah. And I do wanna make one stupid prediction, which is that, uh, there will be several effects of Q Day that we have that we can't even see right now, that, uh, even, even doing white hat work or something that is just escaping us dangers or possibly benefits that I think there's something very unpredictable about such a huge change in, in security. However, however much we are anticipating it now, and the breaking to be easy, quick breaking of keys, there's gonna be other effects and side effects.

We, we have no idea. And, and the awesome thing about AI is the amount of time it will take to sort through all that data we decrypted, it will be, you know, now measured in minutes too. So There's a classic, uh, novel out there that that plays really moment.

This, I sort of wonder, I always wonder about this book, a shockwave writer by John Bruin, you, I did stand on Zanzibar along with their, uh, books that were lot more popular, but shockwave writer is exactly what, yeah. You know, what happens if all those secrets become public all at once? Maybe that's not a bad thing.

We may find out. Uh, guys, I, we need to pull a plug on this one. I'm sorry we're out our time, but excellent conversation, Chris.

We're looking forward to hearing more about hearing more about what you're up to maybe next time. Guy, as always, it's a pleasure to see you, Mike, go Yankees. Thank you everyone for tuning in today.

Stay tuned for the rest of our text on TV schedule. Until next time, Ms. Alan Shimel for Textron Gang.

Hey, everyone, welcome back here to another Tech Drunk tv. I'm really excited to introduce you to Matthew Shaxted, and I hope I said that right. Matthew is the co-founder and CEO of Parallel Works.

Hey, Matthew, welcome to Tech Drunk tv. Alan, how you doing? Glad to be here.

Glad to have you on. Matthew. Matthew, I want to make sure I, did I pronounce your last name right?

If not correctly. You did, yep. Okay, that was good.

I always like to, you know what, your parents gave you a name and you've got a family name and it, it deserves to be pronounced correctly. That was pretty Good. Yep.

Cool. So, so Matthew, you're the, uh, co-founder and CEO of Parallel Works. You know, I, I've co-founded a few companies myself.

I, I always like to say, as the founder or co-founder, you kind of breathe life into your creation every day, right? You can have a lot of employees, you can have great employees, you can be working with friends or strangers that you just met, but basically it's your passion that carries the day. It's your passion that invigorates this company.

So give us an idea. Where's your passion? Where's it come from?

How did it wind up being at Parallel Works? Yeah, that's great. The, the passion stemmed really early on.

I mean, we, this, we're a 10-year-old company. We've been building this thing for 10 years. It has the same mission as it had the day one and even before when I was working.

It's, it's make, trying to make computing systems easier for people to use. That's kind of the whole goal of what we're doing. Uh, and we're doing that through software.

We try to make, you know, increasingly complex computing architectures and things are now distributed all over the world or country, depending on who you're talking to. And how do you make it easier for the people that actually need to use these things for building machine learning pipelines or, you know, creating, you know, RR and D activities or, you know, uh, translation activities or whatever the workloads may be, how to get it into their hands in an easier way to consume and make them more productive and ultimately innovate faster. That's what it's all about for, for me personally.

Absolutely. You know, I, I had a discussion this morning. I was doing a podcast called the PO Platform Engineering Show.

We're talking about platform engineering and, and one of the, and you know, we're talking about why, why do you want platform engineering? Why do you need platform engineering? Is it just for the cool kids who are doing DevOps and Cloud native?

What about, you know, if you're on a highly regulated industry, you're in a data center, you've got a monolithic architecture, do you still need platform engineering? And mm-hmm. You know, the discussion drifted into it just seems that as time goes by, right?

If we have that space time continuum cone, that you always see things grow more complex, right? Mm-hmm. You could, systems drift, systems grow, you know, state of technology grows, and with all of that, it grows to more complexity, not less complexity.

Right? I would agree with that. Yeah.

And you as it, that it, it's like, you know, does the universe eventually just fade into as a result of, you know, uh, continuing to Guy Currier out and they say that, you know, the stars go out. Does that complexity, does it ever stop? Or what, how do we, how do we aren't or get ahead of that complexity, right?

Because Yeah, I mean, it, I think that, that's a good question. I mean, from, just from like the, the world that I live in, which is kind of computing infrastructure perspective, 10 years ago, it was much more common for people we talk to organizations that we, we serve and our customers and things to have the monolithic computing environments, or they were, you know, just moving into AWS for example. And they were, you know, happy there.

And, you know, that that was kind of the status quo at the time. Monolithic, you know, single data center, single colo, maybe just, you know, a single cloud. And really over even these 10 years, I've watched that transition from those type of environments from a complexity perspective to ones where now they want to connect those with several clouds.

And you've seen these hundreds of neo clouds and GPU as a service providers coming online and each with different offerings and different availability and pricing. Organizations need to bring those in. In many cases, they're buying these new architecture systems, NVIDIA systems, you know, and a and B systems in some cases, all with different interfaces on top of them.

Kubernetes moving from virtualized environments to containerization. I don't see that trend really going away anytime soon. In fact, in fact, I think it's gonna get even more complex when quantum becomes valuable for organizations, which, you know, there's all debating timelines of what's gonna Happen.

But I was just say, Doug, what, where, where do you wanna throw a dart at the board and tell me when you think that happens. I, I am not going to, no, but sometime in the near future, let's say, there's gonna be definitely applications that there is value there. And similar to how we've seen accelerator systems come, come in line so fast and Enterprise is starting to adopt 'em, that will happen.

The complexity will continue to grow. And so for me, how do I get in front of it? You know, I focus on user experience of the computing systems, right?

It's delivering software tools that make it so that the people that are using these things, regardless of the shape of 'em, regardless of the interaction, can get to them easily and productively. That's kind of, that's how I would say we get ahead of it through kind of user experience and interaction. Excellent.

You know, another had a recent conversation on this as well. The whole user experience, uh, is, is at the heart of things. Look, let me before we jump in further, I, I feel like we didn't do a good job though of telling people, pa the parallel work story.

You, you're doing this 10 years trying to make the complex manageable for people. How does Parallel Works do that? Exactly?

Yeah, great. Great question. So we, yeah, again, 10 years ago we spun out of Argonne National Lab.

Mm-hmm. Actually, and we were, my partner and I created the company as essentially a user interface on top of batch scheduler, supercomputing Systems. That's kind of the world.

We, we came from. I was a engineer, my partner was at Argonne. We, we started building a, a user interface and a piece of software to make it easier to submit jobs, computing jobs on these big, you know, generally building size buildings and the difference in these type of systems, these computing systems, they all work together to solve one problem from a networking and storage perspective.

But that's where we, we started Parallel Works, you know, creates a piece of software that aims to connect in wide varieties of computing infrastructure, both architecturally and distance wise, uh, into a single interface. And once they're all connected in, then you can serve those computing interfaces to your users within an organization. You know, the teams that are actually consuming these things, uh, as a really kind of treating it as a single system because all of the details about how you interact with the systems, whether it's Kubernetes or Batch scheduler systems, or VMware or OpenStack or any of the clouds, we make all those things look and feel the same.

So that's kind of the goal of Parallel Works. We have a, a software tool, um, called Activate. So our actual platform is called Activate.

We have a managed SaaS version for commercial side, uh, users. And then as of three weeks ago, we now have a FedRAMP PI IL five SaaS as well, where specifically DOD defense users and contractors can come into this turnkey managed IL five FedRAMP high environment, click a button, provision resources, and run their export control workload. So mission critical, export control.

Um, so that's, that's what we build. It's this Activate platform. Um, organizations can also take that and put it inside their own cloud or on-prem boundary and do the same thing, connect all the resources together and deliver it to their end user teams.

Excellent. That's a little bit about, about what we do. Yep.

Absolutely. I'm glad. com easy All one word.

Very simple. Yes. All right.

Alright. The name actually came about, 'cause we, when we started, right? You're, you're running on these super computing systems that run parallel computing jobs and we kind of played off the parallel, parallel works, uh, story for that.

Yep. And you know, now, now kind of as we evolve into this complex world of diverse architectures that are geographically distributed, you know, we like to think of ourselves as a type of distributed operating system, so to speak, for hybrid computing. That's, that's kind of how I position it.

People always have different opinions on what does hybrid mean. And I, I try to define it, but Yeah. You know, so it's funny you mentioned that, right?

I always got the hybrid piece. I, I'm gonna have a little bit in the public cloud a little bit, maybe in my colo center or maybe some in my, you got it. Data center closet, whatever.

I, I'll be honest with you, I didn't see multi-cloud catching on the way it has in retrospect. It's easy to see, right? 'cause each one of these clouds has something they do a little better than the other guy.

And, and not just the big three either. I mean, whether you're going to Oracle or IBM or, or some of the ones like these GPU as a services and stuff like that, you know, they all have their own special source. And you know, as I've gotten older, I've learned it's all about having the right tool for the job.

And, and so you pick the right tool for you, whatever particular job you're doing. Um, mm-hmm. But that being said, you know, then we have this whole other world, Matthew of, of the federal government space, right?

You guys recently had some news out in here. Um, I, I have a background in one of the companies I co-founded was about 65, 60%, maybe more of our business was in the fed space, mostly on the DOD side agencies rather than civilian government. Though we did do, did do some of that as well.

It's a world unto itself. Yes, it is. Talk to us.

So far, I'm, I'm still in the early stages of it, but Yeah, talk to us about your recent news there. Yeah, sure. So, um, really three years ago we won a competitive solicitation from what's called the Defense Innovation Unit, DIU.

And they're, they serve as their transition organization in the DOD to help industry get into the defense sector, really. And they, what they've done is they've innovated on some contract vehicles, really, um, that they call these commercial service offerings, CSO, that makes it more approachable for DOD organizations after you go through one of these to adopt your solution without having to go through like the big primes or the far, you know, far contracts, all that type of thing. Uh, they call it like a follow on authority 'cause it's already been competed.

So we won one of those three years ago. Uh, it was really fun. Part of it was working with some researchers on machine learning image detect, image detection pipelines from essentially satellite imagery that can see through cloud SAR SAR data and detecting objects in that data.

And, you know, putting that into our platform, running it on, uh, cloud providers as well as the big on-prem, you know, computing machines, super competing machines and that was great. And it worked. And, you know, it was, it was fun collaborating with some of those groups and solving real, you know, machine learning challenges specifically.

The other part of the, the engagement was bringing our activate software to a, what's called FedRAMP High Impact Level five authorization. And really what that means, it's the highest classification level before secret. So it's specifically for US citizen export controlled or ITAR software and data.

And, you know, we're like, when we started like, oh, sure we can do that. And we didn't really realize the depth that that goes into to get a software solution like ours authorized through all the proper channels, uh, three weeks ago that we had the conclusion, the, you know, the culmination of that three year effort by getting the authorization through the DOD authorizing body called dsa. Sure.

And now what that means is we're open for, you know, this platform. Our software is open for business for anybody in the defense industrial base to, you know, apply this tool for their own multi-cloud hybrid computing needs for their defense engagements at that, you know, IAR level security and they don't need to go through what we went through is really what it means. Right.

And spend all the money, The time. I'm just gonna ask you that because I've been through this, right. We had to, we had to, I had to go through an information assurance certification process where there was some, there was some encryption involved.

So we had to be FIPs compliant. Got it. And, You know, it was, we got information Sharon certified in the Army, but then we had some Navy business and some other agencies.

And it, you know, the, the one smart thing we did, The look on your face is showing Yeah. No telling Send money. We had to go to one of those labs where they do the certification and Yep.

We, we were blessed. They sent me an angel from Heaven, this guy named Mike Leman, who had about, Mike had maybe 40 plus years in the federal space. And so he knew how to navigate.

But even with an expert guide like that, much like you, it took us three years. And I'm telling you, tens of thousands, if not hundreds of thousands of dollars. So this, this problem that we're talking about, it's, it's called like the accreditation barrier.

Some people call like a valley of death. Mm-hmm. And if you ask like Amazon or Google or Azure or other clouds, it's really impeded cloud adoption in the, in the defense sector.

And there's definitely a pockets of it going on There is cloud computing. There's these big cloud contracts that let you get cloud accounts easily. Like I can get an Amazon account in their GovCloud pretty easily with all the defense discounts worked in through things like JWCC and these other contracts.

But this accreditation step, you have to go through this process of certifying everything. We're trying to basically remove that. And there's other groups that are kind of taking angles on that.

Palantir, uh, second front have their own angle that kind of own the accreditation barrier. We're trying to do it in a SaaS mode where now anybody can come in and provision these IL five resources through a system, like through our system without having to go through that themselves. So it's a little slightly different angle.

That's, That is Scott's work you you're doing though. I hope it wear, I mean, 'cause we, you know, you hear the stories about the thousand dollar toilet or whatever the heck it is. Right.

Or they have the hammer. It, it's not that the hammer itself costs that much, it's all of these other accreditation and that This is the several million dollar hammer Yeah. To basically prove to actually, you know, you have to implement all the controls to show that you're secure and then you gotta prove it to a bunch of people.

And that that's the hand, that's what brings the cost To the handle. Like, what's a story about the plumber who fixed the pipe by just bang, he hit, hits the pipe one time with a monkey wrench, boom. And the water starts flowing and he tells the lady, Hey, $250, she said $250, all you did was bang the wrench on the pipe.

He said, no, no, that was free. It was knowing where to bang the wrench on the pipe. There you go.

That's 250 bucks. That's exactly like that. Yeah.

So it, there is that, let me ask you a question though, Matthew. And you know, as a co-founder of a company, as I said, went through this, my, my, my in retrospect and hindsight's always 2020, right? How do you stop it from eating you up?

Like where that becomes the focus of your whole company and you forget that, hey, I've got a business that's 10 years old with real customers who pay us real money and I owe it to them to keep, keep on. Totally. A hundred percent.

And we're, you know, we're a team of like 28 people now or so or so. We're fairly small and cashflow profitable. We're not institutional backed.

So it's, you know, we gotta, we have to keep bringing money, money in the bank to pay the bills, right? Yeah. I mean, that's kind of how we're operating.

And it is a good point. 'cause it becomes an all intensive thing or a can. And we've had to be, we were, we were fortunate with some of our DOD sponsors to, to cost share a big aspect of it.

So they provide some funding and it's like NRE funding non-recurring to help us achieve some of these goals. And those are just starting to convert now into more of our recurring subscription model. And we got a a, a big demand after it.

But yeah, I mean, you said it right. We gotta kind of focus our attention in our other engagements and Fed side and in the enterprise side. And that's kind of where we focus and make, make sure they're very happy while like this DOD what we think is a behemoth of business opportunity for us makes it, makes its way through.

'cause it's a, it's a long tail for sure. Yep. But, you know, as I said, it's a worthy goal.

And, and it's, and quite frankly, look, you know, we've talked about Doge and everything that goes on in the Fed space, and I'm not here to, to get political with it, but this is something we really need. Forget just firing people randomly and shutting stuff down, you know, to cut money. We need to make it easier for companies to do business with the government and for the government not to have to pay a million bucks for the hammer.

You got it. Exactly. We are, and we, and we've been talking to a lot of startups and other groups that need to go through this accreditation process, but they don't wanna really spend two, you know, whatever x million dollars in, you know, two, three years to go through it.

They can work inside of our boundary, deploy their software, and now they're accredited and they can roll this out to the, you know, so that, that is the idea that we're trying to kind of, uh, bring out. com. com?

We have a contact form you can reach out there. com. So pretty straightforward.

I'm sure that my name will be somewhere. Yeah, no, it's on the bottom shirt. People are seeing it.

Yeah, we have an info at Great. Okay. I am not seeing that, but we have an info at Parallel Works.

So any of those ways we, you know, get in touch with us, we can definitely have a, a more in depth discussion. All Right. Hey Matthew, I appreciate you coming on Tech Drunk TV today.

Good luck with this. I, you know, I've been there. I will tell you, you know, people are always looking for the pot of gold at the end of the rainbow.

I don't know if it's a pot of gold, but there is success at the end of the rainbow when you get to the other side of this. So, um, exactly. Yeah.

And that's why in the meantime We'll be, we'll be Busy with, you know, our other engagements. Another second. Absolutely.

Yeah. Matt, Matthew, thanks for having me on. Thank you.

Matthew Shaxted did co-founder CEO Parallel works here on Tech Drunk tv. We'll be back in a minute. ai video series.

I'm your host, Mike Vizard today with Efrain Ruh, who is CTO for Europe for Digitate. And we're gonna be talking about, well, the rise of domain specific large language models, which are a little bit different than your general purpose ones. Re welcome to the show.

Thank you. Uh, uh, nice, uh, nice to be here with you, Mike. Pretty excited for today.

I think we've all seen or experienced by now in some form a large language model. Um, but there appear to be general purpose. They're jacks of all trade and maybe not optimized for specific tasks.

And now we're talking about the rise of Agen ai. So my question to you is, are we about to see a whole new generation of these things that are really optimized for specific domains? Yes.

Um, and definitely Mark, um, we see a big trend, uh, as you know, they, these generic models are being, basically, they're getting better and better over time, but there are specific constraints. Uh, then, um, like, um, enterprises, they're looking a way of embedded own basic asset knowledge of their industry. They also want to have a little more control about these models.

And there's another important factor is to make it more cost effective, uh, to make the business case. So those two, three parameters are actually driving this, basically the evolution of these domain specific LLMs. So to your point about the cost, are they gonna be smaller and more efficient and maybe less costly to train?

'cause I don't need as much data. I mean, how do they become something more affordable? Well, um, if you think about the very large language models, I mean the biggest more parameters you have, basically, you have either or you host it.

So, which basically would have a very constant cost around hosting or basically you taking care of you pro getting it provided by, you know, um, uh, companies like OpenAI. And then, uh, basically every time you make a call, you getting in charge. So ideally is to make this call, I mean, utilize, depending on the use case, you may be able to utilize a smaller model, which is more cost effective, has a smaller footprint, you're able control more than then you're able to run by yourself and not depending on, on a third party vendor for run for you.

Is there any trade off to be made in terms of the reasoning engines that are embedded in those things? 'cause every time you hear from the large language model folks, they talk about their next generation model and it has more memory and consumes more power. So, and they say, well, that's needed to drive these reasoning capabilities, but can I still have the advanced reasoning capabilities with smaller models?

Yeah, sure, sure, sure. I mean, there's, uh, there are, uh, there's, uh, so many models available on the market today, and if you look at, uh, what the, the cloud vendors, so today are providing, uh, companies like Microsoft Azure or AWS are giving you the possibility of selecting the model that better fits for your use case. And then, uh, so you're able to do your a specific training and then, um, and then areas where you have to basically differentiate whether you're gonna go with more, let's say, a specific TAT training or you're gonna basically, um, bring on some, some rag retrieval augmented generation into your model.

So, um, um, what we have right now, uh, in the market is a's a very large variety options that, uh, companies can leverage. And instead of basically going to this very large and very complex, um, um, um, LLMs, which are being, um, you know, offered in the market, yet We also hear about the rise of AI agents. So, um, will these agents essentially be able to navigate multiple domain specific LLMs to drive a process?

Or will it be more like multiple agents talking to each other and each one is trained on a specific domain? Yeah, so, correct. So we've been working on this AI business for quite some time.

I mean, we, we've been in the market for the last 10 years. We've been started investing, uh, a lot on traditional AI like, uh, machine learning, and we move into gen ai, especially from the it we are working, especially with these AI agents to deliver value on our area, our expertise, which is the automation of IT operations. And basically, and it's more or less the same approach that, um, that the, the industry is taking.

So you select certain agents to, given a very specific task. And for example, you wanna have an agent, which is very good on, on understanding data, then, then you might have another agent, which would basically be in tuned for more reasoning and decision making. And then you would have another agents which are more focused on the execution, basically taking, um, remote connections, collecting, uh, data, or taking even, even actions.

And then all together in an architecture where you have a kind agent that helps orchestrate everything. So it is, it is like, um, like you were mentioning, the trend is to, uh, have multiple agents, which each specific agent being like very good or other specific domain or a specific task that at hand, How disposable are these domain specific, large language models gonna become? Because I can imagine a world where they're constantly being, um, updated.

So will my agent kind of be loosely coupled to that LLM and then when something else comes along, I'll just whip out the API? Yeah, that's, that's true. That's one of the things that, uh, uh, when we talking about this very distributed architecture where we have like, say multi agent would allow us to basically swap sim certain agent, maybe in an area where they have an an, uh, AI agent, which is very good, for example, of, um, applying changes or fixes on an inter IT infrastructure, maybe have an another agent which is more efficient or more, more, uh, faster or even has more knowledge, then I should be able to swap them, basically maintaining that, uh, we are keeping the same API, um, communication within the agents, so that, uh, that is already happening, is happens on, on these very distributed architectures where we have established, um, um, very standardized communication protocol within these agents.

So then be able to make those changes when we, when required, because remember these agents then, some of these agents we may have to, uh, keep updating them or, or, or, or doing some fine tuning on, on, on retrain. And the whole ecosystem is just evolving so fast that the agents are getting more and more efficient. They've been, uh, requiring less computing power, uh, and even reducing costs.

And then, uh, we have to keep these architectures flexible enough to be able to, to do this swap. Yeah. How will we orchestrate all these AI agents then, and how will we kind of connect them to these various LLMs that they may be using?

Because sometimes it may, may even need to access something that's a little less domain specific, but how fluid is all this gonna be? Well, um, it, it, it already depends on your use case or where are you gonna make this from? And, and, and currently there are, um, a very large amount of different patents that been been deployed from everything.

It could be a single engine that, uh, that is pretty good to basically take care of what you need to do or a complete multi-agent approach. Um, so it, it really depends of, of what kind kind of use case, uh, we're gonna make. If, if we gonna make, we make, for example, where, like I was talking about in, in the IT environment where we trying to make automation across operations, not making agents autonomically, uh, take care of it faults or, uh, outages and downtime.

We need to have those, those agents, which are basically, um, able to take, um, these remote connections, uh, and take actions in the IT systems that we have deployed into our environment. Mm-hmm. So as we kind of put all this together, um, who has the understanding of the processes that's gonna be required to stitch all these agents together to automate something on an end-to-end basis?

Well, that's, that's a very good question. So one of the cha big challenges, uh, right now in the industry is the, the lack of skills. Um, we have, um, lms as you know, they're being out in market, uh, I think less than two years.

We think about chat, GBT was launched, uh, just by the end of 2022. Um, so there is a lack of, let's say, expertise, uh, in, in, in, in our IT teams, the people that have to, to deploy them. So we have to make, um, very strong investments on our, our teams.

Um, we need to let the people, um, let's say try and experiment with, with different models. And that's where the, the, the hypervisors, like I said, you know, uh, like, uh, Azure and, and AWS already provide, um, you know, like a safe, uh, sandboxes to, to start, um, um, improvising or, or start, uh, innovating with this, with this ls. So it's, it's definitely something that, uh, the companies need to invest internally.

Of course, they can get help from, um, external providers, you know, very large, uh, companies or, you know, the typical SOS to help them accelerate this, this, uh, this process. But, um, um, definitely it's a skill that needs to be grown up, uh, within the, the enterprise. Um, will, I need multiple AI agents to be trained on the same domain, specific LLM?

Because one, I need an AI agent to kind of check the work of the other AI agent. 'cause all these things are somewhat probabilistic and I'm trying to maybe use them in a process that's more deterministic. So I need it to be done the same way every time.

But these LMS don't do anything the same way every time. Yeah. So that's, that's what, um, we need to, um, bring, um, another capabilities.

I mean, one is of course, uh, the, the correct guardrails, um, to make sure that, um, you know, our answers are within, um, the, the scope. Uh, that's also very important. There is also the idea of bringing, um, things like, uh, uh, I mean very standard, uh, reasoning into the system.

Um, and then, um, and, and the knowledge know, I think one of the big challenges companies has is that how do, how do I collect all this knowledge, how to make it normalize it? How do I, I treat this data, how do I clean it, make sure the data is correct for, for training, um, do my my model training, and then, uh, then what happens when this data becomes obsolete? How do I get it updated?

Et cetera, et cetera, et cetera. So that's, that's I think one of the biggest challenges on, on, on tuning this, these, uh, models for, uh, specific, specific domain tasks. And that's what I, I kind of make emphasis that I think we have to keep these models relatively small and simple, uh, not a lot of complexity and try to divide, you know, it's very, when we talk about a specific domain, um, we need to basically constrain it so that it doesn't become a very complex problem, uh, you know, this process of training and, and, and keeping up to date.

So I'm with you. Um, so what's your best advice to folks about how to go about doing all this? 'cause it does feel like it, it's gonna take a journey, it's gonna require some expertise that has to be aggregated from all across the organization, but I think a lot of folks kind of look at this and it's a little overwhelming.

So how do I get started? Yeah, yeah, definitely. So especially on, on, on automation.

So we need to make a first a good assessment when it makes sense to automate, because, uh, um, it's like you said, it's, it's a lot of effort. And then, uh, you know, the cost of implementations can, can really explode. Uh, sometimes it's very hard to, to estimate the overall, you know, TCO of this kinda implementation.

So one thing is just to make sure that we solving a real problem and we are able to, to, um, estimate very clearly what is gonna be the impact. I mean, how much, how much time my team is spending on this, how important is for me to reduce those effort? How important for me, for example, to accelerate, I mean, there's, there's two things.

One is the effort I'm able to reduce, which would be money, um, but also how fast amm actually getting a response. So that actually give me an edge in the business area. For example, when I'm, um, I'm quicker on giving you, uh, the best quote, or I'm quicker in giving you the best solution because I have a complete, let's say, adjunctive solution for, you know, responding, for example, from A to an RSP.

Uh, so that will give me a, a good ad advantage. So I think it's best to make sure that we pick up the, the correct use cases. Uh, we define, um, very clear KPIs that we're gonna track, uh, to make sure that, uh, we heating, we hitting those als.

Then of course, uh, keep an eye on, on, you know, the, you know, the, the, the overall distance case around it. Well, folks, you heard it here, there's a lot of math involved in AI and LLMs, and it doesn't all have to do with science. It's to do with the return on investment.

So look twice before you jump. Hey, Efran, thanks for being on the show. Thank you.

Thanks. Uh, it was, uh, a really a pleasure to be with you here, Mike. Uh, thanks very much and I hope to we see you again.

ai series. You can find this and others on our website. We invite you to check them all out.

Until then, we'll see you next time. Samina snaps up ZT Systems. Arista has their head in the clouds.

Microsoft Open source Linux, Palo Alto gets mobile Salesforce converges charter cashes out on Cox Communications. HPE makes Morpheus a little more green. And we're gonna take a closer look at the Take It Down act in this episode of the Tech Field Day rundown.

Hello everyone, it's time for the Tech Field Day rundown. Today is May the 21st, and I hope that you're all being very nice to your wait staff because it is national wait staff day. In fact, you should probably send a memo around the office to do that because it's also National Memo Day.

One thing you don't need a memo for is to check out my awesome co-host this episode, Mr. A Cook Out. Welcome to the show.

Thanks, Tom. It's always a pleasure to be here. And I think National Juice slash Day might be one of my favorite days today.

Well, it is the best time to go get one of those, but not right this second because we have some great news stories that we wanted to take a look at. There's some acquisitions brewing, there's some acquisitions that are closing, and then there are some legal things that we're gonna get into. But first, let's talk about one of those acquisitions because A MD acquired ZT Systems last year, you may recall that we covered that on the rundown, and now they're gonna be selling most of what they acquired to Sanmina A MD states that this will help them manufacture more chips on US soil and will also help mitigate the impact of those pesky tariffs that we keep talking about.

A MD didn't sell all of ZT systems to Sanmina, but I have a question. Al, how does this help them save on tariffs? You know, I'm not really sure how this saves on tariffs.

The plans to produce chips using the TSMC manufacturing facilities in the US aren't kind of closely linked to this, as far as I can tell. Maybe ZT systems, uh, does manufacturing overseas, and therefore divesting of that is gonna help a MD be importing less, less systems. Uh, it's important to see that this divestment was planned from the very time that a MD purchased zt, uh, ZT systems.

It's important to see that this divestment was planned from the very time that A MDA acquired CT Systems, and it's the manufacturing part that has been sold on to mia. The AI design piece is staying inside a MD, and that's what a MD actually wanted. They didn't want to be a manufacturer of systems.

They're manufacturer of chips. And, and so the design part is staying within a MD and the actual manufacturing is the piece that's being moved on to sunna. The AI design is absolutely a, a critical part of, uh, a MD's future.

We know that huge amounts of infrastructure are being purchased and, uh, be this infrastructure adoption is being driven by AI requirements, generative AI applications, and IMD wants a chunk of that, and they want a larger and larger chunk over time, obviously, in order to be more profitable. So having a, a design solution, a design practice around building infrastructure for AI seems like a very good thing for AMD to do. And sanmina with their part of the Z two systems, the manufacturing part, they were a first partner for a MD.

So it's not simply with getting rid of the bits that, that we didn't want outta the business, we're also continuing to deliver some of that value through from the piece we're keeping to the piece we're divesting in. So it's a, a bit of a closer relationship then between a MD and Samina with ZT Systems. I think it's, it's a good move to, for a MD not to be competing with its its own customers in delivering these systems.

That's the part that was a little puzzling at the acquisition time, but now it's become clear that it's the AI design piece that a MD wanted, and they had to get the whole business in order to, to actually have that part. Six months later, they're now allowed to divest of that manufacturing. So I hadn't stopped competing with partners.

Of course, now they've got a, another partner that is more equal than maybe some of the other partners, but that's how partnership works. A blockbuster bit of reporting last week when it was rumored that Arista Networks would be acquiring VeloCloud from Broadcom fellow Cloud, cloud was formerly the SD WAN business. Uh, that was a startup acquired by VMware back in 2017, about four 50 million solution, then got bundled into the networking and NSX component of the VMware product suite and seemed to do pretty well.

I I quite liked the product when I did some work with the, uh, SD WAN team at VMware. Of course, Broadcom buys VMware and everything changes, and Broadcom wants to get back to the core products in VMware and has been progressively divesting of the non-core assets. Reporting's not yet been confirmed by either side.

Community feedback seems to be pretty positive. Tom, what's going on? I wish I had a really clear answer because when I first heard about this last week, I, I, uh, uh, I was a little bit shocked to be honest with you, not shocked in the way that, you know, VeloCloud was getting sold off, because I think we, we see that that's kind of VMware's, sorry, Broadcoms trajectory for the brand known as VMware.

Now that anything that's not core to VMware Cloud Foundation is gonna be sent away. Uh, you think about their end user compute that got sold off to KKR, uh, that, you know, that was not really a, a surprise to anyone. Uh, the fact that nobody wanted to buy carbon black off of them actually was kind of surprising to me.

And that's one of the reasons why Carbon Black was kind of brought back in-house. It's not because it was critical to the operation of, uh, VMware Cloud Foundation. It was more than nobody could afford to buy it.

I think this though, has the potential to be a really good deal. If you look at what Arista has been doing for, you know, the last few years they've been making quiet acquisitions. You know, you may think of awake security or you may think of big switch networks, or honestly, you may think to yourself, I don't even know who those companies are.

That's because they don't go out and make splashy deals that are huge dollar values and then struggle to integrate those components. What they do is they buy smaller companies where they can take the assets for those companies and kind of roll 'em into their existing product lines. And I think that that's what we're gonna see here, provided this deal actually goes through.

Because you know how it is, right? Um, where there's smoke, there's fire, but sometimes there's not anything at all. So we're working off of the assumption that this is actually going to happen if it does.

This gives Arista honestly, a play that they've needed for a while, which is SD wan. Uh, if you look at the market, there's been people who've been speculating for the last couple of years that the SD WAN market is down because they have seen some pull through, right? Like everybody needed an SD WAN appliance when the pandemic started, and then the people who were gonna buy it kind of bought it.

So where does that leave them now? Well, when you think about the fact that Arista is very active in cloud computing, they are, they're considered to be one of the networking funders of choice for people in the cloud, whether it's Azure or AWS or a lot of other companies being able to offer SD-WAN on the on-prem devices as well as the, in the cloud devices, it's a huge win for cloud providers because they can provide secure data tunnels using SD WAN between those two locations. It also gives Arista a place to compete against Cisco, because I mean, in that market, Cisco is the king that everybody's coming for, and Arista wants to make sure that they're not gonna miss.

VeloCloud has really good technology, and I know if you watched our Mobility Field Day coverage a couple weeks ago, you probably heard from CTO Ken Duda, who is probably the best proponent in the industry of code quality. And I think that that's one of the things here is he's going to be scrutinizing the code quality of the VeloCloud engineers, which is high by the way. But that ultimately means that if we're gonna see, uh, VeloCloud SD WAN technology integrated into Arista e os operating system, it's gonna have to pass muster.

And that could take a little bit of time to integrate. So probably what you'll see is they'll, if this, again, if this goes through, they'll start selling VeloCloud appliances on the edge, very similar to what Cisco did with Vitello. And then as the code quality requirements get to where they need to be, you'll probably start seeing that integration piece where maybe those hardware devices become end of life, end of support, and the software itself runs in Arista boxes on top of EOS, and it just becomes transparent.

That typically is the way that Arista handles these things. I think that's a good thing. Overall, the question is whether or not there's actually anything here, and we're gonna have to wait a few days, weeks to find out, because the, the one thing I'll say about these acquisitions, especially if it's not material to the bottom line of the company, is that we often find out about them after they've already closed.

So make sure you stay tuned to the rundown, because I'm sure when it does close, we will definitely be covering it in more depth. You may have heard of the Windows subsystem for Linux. It's something that enables Windows users to leverage running Linux and Linux commands on their computer.

It's a favorite among system administrators, developers, and other power users. Since it was released in 2016, Microsoft finally opened, sourced the Windows subsystem for Linux, by the way, that request had been in for nine years. This was first a GitHub issue for WSL entitled, will This Be Open Source?

And they finally closed it nine years later. Is Microsoft now truly an open source company? Al I, I mean, I'm not gonna lie to you.

I think I already know the answer, but I also never thought that when a subsystem for Linux would ever be open sourced, what do you think? Well, I think we both know the answer that, uh, Microsoft is in the business of making money, and if open sourcing part of their, their product, uh, helps 'em along that way, decreases their cost or increases their profitability, they're all over it. And so for Microsoft Open Source is a, um, tactical move rather than a strategic move.

It's not like they're a, uh, open source first company. I do still remember the quote of Bill Gates. Um, or actually it was not a quote net.

I do still remember the attribution to Bill Bill Gates of a accusation that somebody was a communist because they were going to be giving away software for free. However, Microsoft has embraced quite a bit of open source over time and has also recognized that the operating system that they have in the Windows operating system doesn't fit the needs for every user and every, every need for every user. And so the Windows system for Linux has, has been very popular amongst particularly, you know, system admin people that I know of and, and developers who, who are working in a mixed environment that is both Linux and Windows environment.

So it's definitely been a popular tool to enable you to use a single computer, not have to run a discrete virtual machine that is all isolated from your Windows machine in order on Linux commands, uh, support for a variety of different Linux distributions, uh, to appear under Windows subsystem for Linux has been great and continued development and particularly community driven development. So Windows system for Linux, yes, open sourced. Great.

We've only been waiting nine years for it. Is it a big deal? Probably not really.

Uh, in the end, open sourcing doesn't change the functionality. It does get a little more visibility of what's going on in the product and Windows subsystem. Linux is just a small part of the massive estate of software that, uh, that, uh, Microsoft is ma making good money from.

They did shift a little while ago from focusing on making their money by protecting the, the back office and Microsoft office space towards being Azure as the, the crown jewels at Microsoft. So things that are on the desktop is no longer quite the crown jewels that it used to be 15 years ago. And the cloud things are really where Microsoft is making their big money.

That's the place to look for them doing things that'll move the industry to T-Mobile. And Palo Alto Networks have partnered to launch a managed solution that combines T-Mobile's 5G Advanced Network with Palo Alto Network's Prisma, uh, SAS E 5G platform. This offering integrates T-Mobile's clientless authentication system and dedicated virtual network segmentation to deliver advanced security and performance by embedding security directly into the network infrastructure.

The solution aims to simplify IT management and provide robust protection for a wided range of devices, including those that cannot support traditional sassis software such as IOT devices in 5G routers. This sounds like implementing 8 0 2 11 x in a much larger scale. Tom, can Palo Alto help T-Mobile break into The business connectivity market?

1 x, lest you summon the Eldridge horrors that were, well, pretty much anybody who ever tried to implement that, that solution. But you're absolutely right. Security works best when it is absolutely invisible to anybody who's using it.

And that's one of the reasons why T-Mobile has developed this, uh, you know, architecture that allows you to effectively sign on without having any kind of need for like, you know, challenges and, and things like that. What they needed was a way to project that in a wider area, and who better to do it than Palo Alto Networks. So Palo Alto on their hand gets a provider to start offering this service.

And if you think to yourself, well, this just basically sounds like a sassy device or sassy setup at the edge. Bingo, you're right. But the key there is that it is aimed at businesses and it is aimed at IOT devices because this is a thing that a lot of people are finally starting to realize as we add more and more of these, um, sensors and collectors to our environment, I guess would be the best way to put it, is that the data that they produce, they don't care.

It's, it's, it's out there available for use and whatever. Okay, well, we'll just secure it. Have you ever tried to do that to an IOT device?

Have you ever tried to install a certificate on an iot device that isn't centrally managed somewhere? That's why you turn into an elder horror is because you don't survive that process well, very well. That is the value of having a solution that is completely transparent underneath you have T-Mobile offering this solution on top of what Palo Alto is, you know, well known for, is that ability to analyze application traffic, is that ability to provide SSE services so that it doesn't matter whether it's a milling machine or the operator of that milling machine getting on the internet, all of that is protected.

So I think that this is valuable for T-Mobile to be seen in the business space, but I also think that it is absolutely critical to get more Palo Alto, uh, technology out to the edge. Finding skilled people to build your agentic AI is a challenge for many businesses. And a recent survey by Asana suggested that half of IT leaders are regretting implementing AI without providing staff training.

Who would've thought Salesforce took another approach? By acquiring Convergence ai, gaining both AI technology and team of skilled AI developers, Salesforce is betting the farm on Agent Force, which is not Perry the platypus ass secret partner. But I have to ask, is Aqua hiring the best way to get ahead when it comes to ai?

Yeah, it seems that It's a dual purchase here. Definitely Convergence AI has technologies that Salesforce wants, but they have a skilled dev team. And I think I saw a couple of stories this week around just how difficult it is to build that team and that study from Asana suggesting that there was some regret at having tried to build the applications without actually upskilling your team first.

I think Acquihire is, is definitely an element of this, this deal as well as the tech technology acquisition. So there is technology that will be used from the, uh, acquisition that to improve the agents that are currently behind Agent Force going beyond the, uh, relatively simple least what's described as relatively simple Einstein models that are used with, uh, with Agent Force at the moment towards, uh, more context rich and aware AI agent for Salesforce. Uh, this is Salesforce, making sure that they're the one, not the ones whose lunch is being eaten by another company that uses AI and agentic AI to do management of interfaces with with customers, yes.

Customer relationship management. The question for me is also, is this a valid strategy for anybody else to do acquihire? Are there enough AI startups with skilled teams going around looking for an exit that can be purchased by larger companies that can't grow their own AI teams?

And I think the short answer is no. There's still a shortage of skills. There aren't that many startups that you can acquire hire to get a large AI dev team.

It is gonna come back to continue to develop the staff that you have and acquire new staff. We're seeing this throughout the industry. We see rounds of layoffs of, uh, reductions in workforce, and then they are usually followed or aligned with acquisitions of the hiring of new staff, uh, in, in different areas, right?

The, the rounds of layoffs are usually about organizations making a shift in focus rather than a, uh, reduction in the size of the business. Definitely businesses should be continuing to develop their teams and AI skills seems to be a very important part, both infrastructure and software development side. So yeah, I don't think there's enough startups to acquihire all over the place to solve your, your shortage of AI skills, developing your, your own staff and acquire hiring staff that you can develop seems to be a, a much more extensible method.

The Cable giant charter is making a bid to buy one of the other Titans Cox Communications, uh, for almost $22 billion. The deal would create the largest cable TV and broadband provider in the US with 38 million subscribers. And let's say the movers designed to combat the intrusion of 5G providers like Verizon and T-Mobile into the end residential broadband market, which has historically been dominated by cable t TV providers in the us.

The companies had looked at a merger back in 2013. That plan didn't happen. Tom, is Charter and Cox getting together a good deal for users?

I, I don't know if it's good, but I also don't know that it's bad. And I will say this in complete transparent disclosure, I use Cox Communications as my internet service provider because I don't really have much of another choice right now. I mean, there's, there's a few options, but, but it's, you know, familiarity kind of breeds contempt in this case.

I am curious to see if this is gonna make it through the regulations, because you're effectively creating the largest cable company in, in America. And, and I know that there's already been some discussion with the, the DC lawyers, technically Charter and Cox operate in different sections of the us so there's not any overlap, there's not any reduction of competition, but I don't necessarily know that this makes it any better for consumers either. Uh, the the thing though, and this is probably where the, where they will use it to get through the regulations is that Cox and Charter are not competing against each other.

They're not, uh, you competing against any former cable companies that you might have heard of from years ago. They are competing against the residential service providers. So I mentioned that one of the reasons why I use Cox Communications is because there aren't really a whole lot of other options.

Well, one of the other options around here is at and t, but when you look at the fact that Verizon and T-Mobile are offering small cell service for the house and essentially saying, rather than running cables to your house or fiber to your house, why don't you just use one of our small 5G cells and, you know, we'll put a bandwidth cap on it and do whatever we want and you'll be able to use it and everything will be hunky dory. And oh, hey, by the way, that means that we get to increase our capabilities all over the place, uh, because now we have a small site cell signal kind of in every house. And I mean, that's one of the things that Cox has actually done as well, is if you run a Cox modem, then you're using Cox wifi and Cox subscribers can hop on your wifi at your house and use it because it's transparent to all users.

But I think where we're gonna start hitting, you know, the, the, the ceiling of this is if we get down to a couple of providers, is that valuable for the customer? I mean, yes, owning the last Mile is really expensive, so expensive in fact, that Verizon sold off all their last mile Fios to Frontier several years ago. They're in the process of buying it all back after some bankruptcies and some debt consolidation and some financial trickery.

But we're starting to see that consolidation phase. And, and everybody who's ever been in this knows that there's the divestiture phase 'cause they need to re meet regulations or they've loaded, loaded it down with debt. And then there's the, you know, pro the period where, you know, revenue's extracted somehow and then there's the reconsolidation because it's actually better if it's back under our, our building.

So I think that that's where Charter and Cox are right now, is that they realize that they're never gonna compete against each other, but together they have an opportunity to compete at a nationwide, um, stage against the T-Mobile's, the Verizons, the at and t's, or honestly, a lot of the local companies, so like for example, the other provider that's literally just over the street, there is a local electric company that's been running a lot of fiber through the area, which I would switch to in a heartbeat. OECC, please come into my neighborhood. But more importantly, that's the way that they're gonna survive because the value that Charter has been bringing for so long has been the bundling of the internet and the tv.

And we all know that cable TV is effectively dying with all the streaming services that we got and phone and, and all these other services. You run outta subscribers sooner or later. So how do you continue to operate, operationalize that and, and gain new customers?

Well, if you're region bounded, you have to buy companies that have new customers that you haven't reached yet. And I think that's kind of where charter's at and it's obviously worth, well, $22 billion to 'em. But again, you've gotta clear the regulatory hurdles and we don't actually know how that's gonna shake out.

So I expect to be talking about this well into next year. You may recall that our friends over at HPE acquired one of our other friends Morpheus data last year. And the apparent target of that was to be able to integrate HPE GreenLake private cloud management with public cloud management, which is the holy grail for pretty much everything they want to do at Cloud Field Day.

Right before the acquisition, we got a glimpse of the Morpheus private cloud platform based on the KVM hypervisor. Guess what? HPE now offers, HPE Morpheus VM Essentials as a private cloud platform along with HPE Morpheus Enterprise software for multi-cloud management.

Hmm. Where have I heard this before? I have to ask Al, does this open a whole new world for hybrid multi-cloud as a service, which is kind of what the target of what GreenLake is?

Yeah, I think think that is very much what the, uh, integration and the rapid integration of Morpheus into the, uh, HPE portfolio does mean. Um, at Cloud Field Day last year when Morpheus presented to their not yet released, uh, Linux, KVM based, um, private cloud platform, uh, it looked very nice and was really nicely integrated into their console. The Morpheus Management Console is really a cloud management platform console for all of your clouds, be they on premises or, uh, in public cloud or potentially even in a, a colocation center.

So it's good to see that very rapidly integrated into approach of the Morpheus, uh, virtualization platform in the business essentials. Uh, nice to see, really nice, and I think it's a reflection that Morpheus had built really good technology, that it's very rapidly seems to be replacing the original private cloud op offering that, uh, HP had in GreenLake. Uh, so definitely interesting that there's two products in here, the one that provides the virtualization platform for your private cloud, which is the business essential side, and then the core Morpheus piece, the piece that really is, is what Morpheus has been about for a long time, which is integrating the management of your hybrid cloud environment across multiple locations and hybrid multi-cloud in particular, if you've been following Cloud Field Day over the last few episodes, hybrid multi-cloud is definitely a focus that I, I see in large enterprises and the need to try and unify that management across the, the very different platforms that we see and the different cloud providers.

Um, Morpheus does this and I think that's exactly why they're acquired by HPE and this, this product should be really nice. Uh, hopefully we can, as usual with GreenLake choose which bits we want and choose to only have maybe the, uh, HP Morpheus enterprise software, the hybrid multicloud management, if that's what we'd like. And obviously if we're running a an on-premises cloud, maybe we also want the VM essentials.

HP Morpheus, uh, VM Essentials hybrid multicloud are very strongly thinkers with, we we're landing as a realistic place for almost every enterprise organization. So really good to see, um, rapid movement of the, the Morpheus product into HP's line up. Of course, HPE, like all large companies, normally moves fairly slowly.

So, uh, this has been under a year from acquisition to ship product. That's gotta be pretty good. It is time for us to take a little bit of a closer look and this time we're gonna take a look at the brand new Take It Down Act.

The US government signs a new act into law that has opponents concerned about the ramification of what it can be used to do. The Take It Down Act was proposed to allow the rapid takedown of non-consensual intimate images, or NCII, uh, from the internet. The law requires social media platforms remove content within 48 hours, uh, and that people could be convicted of violating the law.

Also, this, uh, the same penalty applies if, uh, other copies of the media are known to be around by the provider. Uh, what's the catch though? Seems like a good thing to, to be getting rid of these non-consensual images, uh, that are often used for blackmail on, on those kinds of elements.

Uh, according to the Electronic Frontier Foundation, uh, and another Freedom organization, the law has been written too broadly and a content and doesn't seem like an intimate image. The law could also be used to pressure companies into weakening encryption and to allow for easier detection of, to allow easier detection of that content. Uh, the element in there of the weakening of encryption is that this law, as it's written, makes the company that is carrying the content liable for it, irrespective of whether they actually know that there is objectionable in this case, uh, non-consensual intimate imaging, uh, on, on the platform.

So the ability to break that encryption to view whether this is actually NCII that's being carried is gonna be vital for compliance. And of course, the transport people, the people transporting your data, being able to break the encryption is rather a concern. So is this a Trojan horse?

Is this in fact, uh, the NCII, the non-consensual part that everybody agrees we should be, uh, putting measures in place to restrict, limit remove? Uh, is this just a, a cover for bringing in some laws that are gonna allow, uh, stifling of free speech, uh, suppression of, uh, dissenting organizations, particularly media organizations? Uh, is that a track that maybe has been heard a little bit in the politics in the us I'm not particularly well qualified to comment on how the US politics works, but if this law has not been written well, or if it's been written with an agenda that is not the stated agenda, this could be a significant problem for the us.

Tom, you are over there in the us what's going on? There's a problem here and, and you've hit on most of the issues that, that people have with this. 'cause there's two things that are problematic here.

I want to tackle the second one first because it's the one that's most important to our audience. This could be the lever with which law enforcement organizations are finally able to force providers to compromise their security mechanisms to ensure that they can get access to, to devices, right? Because you need to know who created the image, whether it was one that was taken of someone that was shared without their consent, whether it was an AI generated image.

But in order to do that, you have to prove possession, right? You have to prove that the the picture was sent to somebody, you have to prove that it was stolen. You have to prove that this was the device that generated it.

Well, let's just say we're, we'll say texting for example. Well, if you're using iMessage or you know, Google RCS, those are encrypted end to end. So that means that there's no way for you to know that unless you get the device and you have a warrant and you get the codes to get into that device.

But getting warrants is messy, right? Like, we need to take this down right now, it has to go down within 48 hours. So wouldn't it be better Apple if you would just provide us with a key to get into anyone's iMessage that we wanted to, and, and we could, you know, obviously we, we need to make sure that we're not abusing this, so there has to be a reason for it.

And, uh, we'll just, we'll get into somebody's text messages and we'll make sure that that really is the, the NCII that they did. And oh man, I hope we don't find any other, you know, crimes that they may have committed along the way. That would be terrible and we'd be forced to arrest them.

And not just for this, but for other things too. And yeah, you don't have a problem with that apple 'cause you're, you're aiding and abetting criminals if you don't let us do that. And yet every time that that's happened, it, that power has been abused.

There's a reason why companies like Apple say that the way that they have things configured, not even they can get into the system, it is on purpose that they have no access to the device. Because if they don't have access, then no one is tempted to go looking. And I would say that if the barrier to getting a warrant to search a device is high, it's high on purpose.

That's why we can't just stop people and take their devices and say, you have to unlock this for me and show me everything that's on it, because that's an illegal search. When you get a warrant, you have to tell the judge what you are looking for. You can't just say, I would like a warrant to search this house for anything illegal going on in there.

No judge in their right mind would sign that warrant. That's a fishing expedition. The warrant should say, I am looking for evidence of this particular crime that was committed on this date.

So, for example, if you're looking for a baseball bat that was used in the commission of an assault, you can't go looking around for guns or golf clubs or chainsaws and you say, well, these were obviously used for crimes too. Mm-hmm. Warrant says you're looking for this.

So that is one of the reasons why we have such a high bar to clear in the US and what they're effectively saying is, we wanna go in and we wanna be able to just look whenever we want and take whatever we want. 'cause you know, no one has ever been accused of stealing images like this and keeping them for themselves. Not that I would think that any law enforcement agents would do that, but you know, other people might.

Because once those keys leak that then the, the game is up. And we saw that 10 years ago, more than 10 years ago, when, uh, a lot of those CIA hacking tools were leaked to the greater internet, and then they became the foundation for a new wave of malware. So let's go back to that first problem, you know, the one where, uh, the way that the, the law is written, it could be used to stifle political speech.

So lemme give you an example of something, um, I don't know if you've seen this or not, but there have been images that have been shared over the years of, um, we'll just, we'll, we'll take a prototypical example, uh, of females who are clad in revealing swimwear. And, um, they're, there's nothing wrong with that. That's absolutely a way to make money to share yourself on the internet.

But what if I put a sticker across it with political speech? Maybe you're supporting a particular candidate, maybe I am putting a sticker over it of you supporting a candidate that you don't actually support. I've generated an image that you did not consensually share with me now, I as the image rights holder, because it's my photo, call the provider and say, you need to take that down because I don't support that candidate.

Now, I have violated the take it down act by posting an image non consensually, and for the simple fact that all I did was take one of your images and put a banner on it saying you support somebody. You don't. I could go to jail for up to three years.

Do you see how this could be used in a false flag operation to take collections of those pictures, attribute them to other people, and then force them to be sent to jail by manipulation? If you don't see how that could be done, then you believe in the nobility of what this act is being used for. Yes, we need to be able to take down when someone, uh, has revenge porn of someone and they share it online, and, and that's what this was originally designed to be.

It was designed to be a revenge porn restriction act. The problem is, is that nobody stops at the limits of the law. They're gonna keep pushing the boundaries until they find a spot where they're gonna get away with as much as they can before they get caught.

And Al I know that you remarked that, well, you know, this is probably a US thing right now, but I promise you that if it goes into law and it gets used effectively, there aren't people in the world that won't jump at the opportunity to implement a similar law in their country to affect similar change. Yeah, I we definitely see law changes in one country, and particularly law changes in the United States, uh, set a precedent for other countries. We have seen definitely, uh, follow-ons of changes in legal stance in the US drive changes in legal stance here in New Zealand as well.

Uh, historically we've been very British oriented because we've still have, have the, the king on our currency. Uh, but we definitely see the, the trends that come out of the US driving law changes. And I, I think there's also some interesting elements around if, if iMessage, uh, message gets sort of legal backdoored by, uh, law enforcement, then where are people actually gonna communicate?

It's gonna more likely cause suffering to people who aren't actually doing anything, uh, illegal. Uh, the people who are truly out to, to take illegal action will very rapidly move to platforms that aren't being, uh, knowingly backdoored by US law enforcement. We may see the rise of tools, more tools like Signal and Telegram where, again, end to end encryption, but not necessarily us based a little harder for, uh, US companies to, uh, apply the leverage or us law enforcement to get access to offshore, uh, resources.

Uh, yeah, there's, there's all kinds of, of dangers sitting behind this very noble idea of preventing the spread of revenge porn. There's been some suggestions that, uh, this might even be the bad for the victims of that revenge porn, although I didn't quite understand how, um, the, the takedowns were, were bad for the victims or at least worse than the content being up for the victims. Unfortunately, there's no telling how these laws can be interpreted and applied to other things.

But what I can tell you is, is that the intersection of where these policies come into contact with enterprise, it is honestly one of the places that we discuss quite a bit at our Tech Field Day events. And the next three weeks are extremely busy. So I'm gonna be gone next week because I'm gonna be at Security Field Day where we're talking about problems just like this.

We have a great lineup of companies. com and check those out, um, they can, you know, you can see the lineup of all the people. We're gonna be talking about AI security, we're gonna be talking about network monitoring.

We're even gonna be hearing a report from our forensic Futureum research about what the future of cybersecurity looks like right now. That is May 29th and 30th. com for more details.

Then I'm flying back, but then I get to see you in the US because you've got the next event. I do, I have Cloud Field Day on June 4th and fifth and uh, you may have heard me talk about hybrid multi-cloud. Well, we'll definitely be talking about hybrid multi-cloud at Cloud Field Day 23.

I have a great lineup of presenting companies as well as delegates flying in from all around the world. People traveling further than I am in order to get to, uh, the uh, San Francisco area that week. Of course then I hand the baton back to you and you head back to the West coast the following week.

Yes, only this time I'm headed down to San Diego because we are gonna be at Tech Field Extra at Cisco live US 2025. So I thought about this, this is my 19th Cisco live in a row. I've gone to everyone since 2006.

Uh, I've been to San Diego. This will be my third time. And I can honestly say even though it's a pain in the neck to get there, I love it to death because it is a great place to be and it's gonna be a great place for you to all check out.

Uh, that's gonna be June 10th and 11th. We're gonna be having presentations from Open Gear, Avi and Cisco. And I know for a fact that they have some exciting stuff that they can't wait to share with you.

I can't tell you what it is yet, but you should definitely tune in. We'll have the schedule coming up very soon. You're not gonna wanna miss it.

You're also not going to wanna miss what we've got coming up after that right now all I can tell you is, is that we'll be back with another edition of Networking Fill down July 9th and the 10th. So after you get done setting off the fireworks and having all the picnics settle in and get ready for more networking content, there might also be some things that get added to the calendar between now and then. com and click on the link for future events.

Trust me, you're gonna want to do that just like you do when you tune in for the Tech Field Day Rundown every Wednesday. That's right. We're gonna be uploading the rundown to YouTube.

We're gonna be checking out the audio portion of this podcast and your favorite podcast application of choice or head over to Techstrong it and check out the show notes and all the articles that we linked for all of the fun snarkiness that we leave you in, our coverage of the news. Make sure that you check out all the other Techstrong and Futurum group, uh, shows and properties and reports while you're there. And I also wanna make sure that I shout out Futurum research people because they're doing amazing work.

We got to hear some great conversations about that at Mobility Field Day a couple weeks ago. We're gonna hear about it at Security Field day next week. So make sure that you are, uh, tuned in for that because you're not gonna wanna miss it.

Al and Steven will be back next week with more great news content and I'll be back the week after that. Uh, hopefully fun things happen between now and then. But until then, for myself, Tom Hollingsworth, for the wonderful Alastair Cooke and everybody else who works very hard to bring you the rundown every week.

Thank you very much for tuning in. Go have some strawberries and cream 'cause it's always a dessert day around here, and we will see you next week.