Techstrong Gang – April 9, 2025

Hey everybody. We got mainframes in ai, the death of paper checks, and well, geez, how are we gonna make that commute back to the office? More livable now that we all seem to be going back that way.

We'll be back in a minute. Hello, and we're back. I'm Mike Ard.

I'm still in Ireland, so if any of you watched yesterday's show, you'll know that that's where I am and I'll probably camped out here for at least another week or so. But we have an awesome lineup of folks joining us today in the text drawing gang from all points in the globe, starting with Lisa Martin, you were, you are not in your normal location today, so where are you? I am in my normal location.

I got a new backdrop for podcasting, so I covered up my funky wallpaper with, uh, these panels. So yeah, greetings from San Jose. Well, well, well, how will we know what mood you're in?

If the wallpaper doesn't keep changing, Know it's back there, so I can take these off if I need to, and I'm kind of missing the vibrant, bold, so we'll see. All right. Also joining us, I believe, jp, this is your second time on the show.

That's, that's right. All right. JP Morganthal.

Jp, where are you these days? I'm in, uh, the Orlando, Florida central Florida area. All right.

Right next to Disney. Not far. All right, cool.

And another normally person who's normally in Florida anyway, but is not today, is Chris Bla. Chris, where are you? I am in Hamilton, Ontario.

The, uh, snowbird cycle has begun. You know, it's like the Sparrows fraternity to Capistrano. You know, when I show up in Canada, it's a sure sign of spring.

All right. I feel like you just made it back to Florida in the boat, and now you're back up in Canada, but what do I know? All right, well, let's jump into today's topics.

'cause yesterday IBM launched a Z 1760 years young and keep going strong with that mainframe, but this particular model has a new tellem processor in it that tellem was designed to optimize the running of inference engines on the mainframe. And then they also are building something called spry, which they previewed before, which will let you run the LLM offloaded from the core mainframe. And their argument is, is that all your data's on the mainframe, so your AI should be too jp, you've been following AI closely.

What do you make of this latest move by IBM? Uh, It's, uh, it's, it's pure IBM, uh, the minute I saw it, I'm reminded of, uh, the announcement that, uh, IBM ported Java to the mainframe. And in fact, it triggers me a little because I, I thought it was a great move.

Um, actually, I'm sorry, they, it migrated. They put Java on the AS 400, and I gave them great kudos for doing that. And then Sun gave me a big backhand across the face for such deal of that.

Um, but you know, it, it, I like that they get out there and they, you know, look to, uh, see how their existing investments and integrate into the new environment. In today's world, though, you know, I, I, I, it's, it's a wait and see really how they're gonna be able to sell this, right? Because the, I think the notion of, of racks of hardware that are, you know, you know, scalable together, right?

Is become more popular. Uh, by the same token, many of them bailed out of their own data center. Um, I was with DXE for a very long time, and you know, they were one of the organizations that were selling their maintenance contracts back to IBM, right?

Uh, on, on their managed services. So it's who's buying mainframes? Who's maintaining the need frames anymore?

Um, where are they gonna be operating? It, it sounds like in a piece that's missing to the story is, you know, does IBM envision this to almost be a cloud model where you buy the mainframe, uh, from them and then they will host it, manage it, and operate it for you? Uh, and if they can prove that the economics of performance work out really well from an inference standpoint, 'cause there are many organizations that will be looking to privatize their inference capabilities and want to have very large models along the lines of what meta's creating and requires a lot of, uh, firepower to operate and run.

And if they can do that, if they can demonstrate that that box is competitive and you don't need to try to, uh, go and acquire significant infrastructure from, uh, the likes of Nvidia in order to make this work, I think they could Have a play. Yeah. So let me argue the case from their perspective, which is, uh, the whole notion of moving large amounts of data into the cloud to create, to bring the data to an AI model is passe.

They're actually going old school and saying, you know, we need to go back to the pre-cloud days where we're bringing the compute to where the data already is. And if the data's already on the mainframe, then that's where the AI model should go, and the cost of moving data. And I think jp, when you and I were much younger, somebody once said, nothing good happens when you move data.

So a lot of folks still kind of subscribe to that theory. So will people bring the models to where the data is, or are we gonna move the data to the models? So that sounds like a McCrory data gravity, uh, response.

And it, It, it, you can channel him if you are so inclined. Uh, I, it'd be hard to channel him, um, the, you know, data. I, I think it's a fair f argument to will I, you know, will, will organizations want to migrate that data, but they're already doing it.

Uh, cloud providers have made it almost, you know, free to move data into these environments. Uh, it's the moving out that ends up costing you money. Uh, Amazon has taken steps to make it very easy to transport terabytes of data, uh, even on, on hardware, right?

Right. The, the old, uh, loaded up on hard drives, put it on a truck and ship it to us, and, and we'll connect the drives and copy it off that way. Uh, there, so there are options that I, it's hard to make that argument anymore that nobody wants to take their mainframe data and move it into the cloud.

I think they've been doing it for years to do business intelligence against, uh, and this, it's not, it's not a novel concept. It's not like this is the first time and all of a sudden businesses are now gonna be faced with taking these, uh, petabytes of data and trying to make use of it, you know, in a brand new function. So they may find that a lot of the data is already in the cloud, We shall that is true.

Um, I also think that things are getting a little more fluid and maybe people are getting more, uh, accomplished at moving data between different platforms. We have folks called data engineers these days, and maybe it's not as heavy a lift as it used to be. The question is what is the value of the data?

And I think that nobody really is focused too deeply on that part yet. Um, data cleansing, uh, data value, understanding what's important. Uh, you don't wanna wholesale just move everything.

Uh, we've gone down the road of data warehousing where it's lossy and they, you know, you're getting summarization and we know that doesn't work because there's always the question of, okay, but what led to that? Right? So you have a, a chain of thought, if you will, uh, a method of saying, you know, I wanna drill down deeper, and yet I've hit the, the bottom of the well, because I've only moved so, so much data over.

And I think that that might, that hybrid approach might be an answer here. It's, you know, let me move some summarization up into the cloud, and then if I get to a point where I need to drill down, well then I have ai, and then I can point my AI at my databases and start to use, uh, rag approaches to go get at that data and explore further. And I have a lot of context to bring with me from the summarized data.

And that's the piece I was missing before. So you have this up in the cloud, do this data warehouse, summarize data lossy, but I get to a point where really I'm only interested in three areas out of that, where I wanted to drill down anyway. And then I can bring it back to where the actual detailed data is and use tools like AI to go investigate that further.

Lisa, what is your take? Is IBM kinda regaining its, uh, mainframe mojo a little bit here with some AI infusion? I mean, oh, if I think back, I think it was, you know, around the turn of this decade, all five years ago that, you know, IBM was at the forefront of all this AI stuff, and then they seemed to just kinda fade away a little bit.

And the next thing you know, everybody was talking about open ai. This could definitely be that shot in the arm that they've been really working towards for, uh, the last few years. I think they're doing a good job of listening to their customers.

Customers are hearing all the time, as we know with all the conferences we go to, if you're not in ai, you're too late. And so being able to get that shot in the arm, that infusion of AI ensued the mainframe is something that I think, especially for their legacy customers, will help them modernize, become more competitive in the space. So it's something that's interesting to watch from a legacy company.

Chris, what's your take? Because one of the things that IBM is saying is that, you know, there's all this data and it needs to be secure and well, the mainframe is the most secure platform in the world according to IBMI. I just want to take a moment to appreciate that it's 2025 and we're using the term mainframe, right?

And I Looked, You know, a couple years before and after 2020, I worked with Unisys and I got, uh, I got, you know, wogan to the fact that mainframes didn't go away. You know, in, in the Unisys context, we would take truckloads of Dell servers, we would turn 'em into a mainframe, we would sell 'em to companies like Citibank, where they're running code written by Grace Haber. You know, that runs the entire world's financial existence, right?

And, and as much as I like to talk change over time, you know what, part of that is understanding how long it takes for things to go away and what doesn't go away. You know, innovation doesn't replace, you know, it's a fully distributed world and everything's gonna be yes. I mean, there's huge iron mainframes, literally mainframes, you know, running massive code and everything, you know, you and JP and Lisa are saying is, you know, there's logic to this.

You know, do I need to move all this data around, you know, mainframes, you know, as you say, from a security perspective, yes, mainframes are, as we see them today and as they've been all along, are more secure than taking a bunch of multi-use PCs and building your business on them. So there's a lot of intrinsic logic to this, right? Just because, you know, the terms have changed doesn't mean the fundamentals have, right?

How do you deal with these huge amounts of computing and inference and, and data transfer, you know, good questions. But among them is sometimes you don't move things, you know, leave it where it was. I I, I like your point, and it was in the back of my head as you said it, about using the term mainframe.

'cause let's face it, they're not mainframes. They're essentially V blocks. They are topper rack switches with high speed backbone, uh, fiber networks that have transferred gigabytes of data on the back plane attached to very large storage arrays with lots of server capacity in between.

And there were virtual machine operating system on top of it that's running Linux, windows and other operating systems. So it's essentially a V block. Um, you know, it's very big bare metal, but, and we call it a mainframe because IBM sold it, right?

But at the end of the day, I think at this point, mainframe is not a, a fair representation of what that box really is anymore. But it has a Z label on it. It must be a mainframe.

Whatcha talking about? Alright, I'm not sure how this is gonna play out, but one of the benefits of having a mainframe was the pricing model. If you, once you had the mainframe, you could run, uh, almost Linux workloads for free because you were charged based on the MIPS that normally was associated with Z-O-S-I-B-M hasn't quite said what the pricing model is for running these AI models on this platform, but might it be cheaper to run them on those platforms if I'm kind of gonna tuck that into my existing IBM licensing agreements.

Jp, what do you think? Uh, you raised a great question. I mean, you know, at, at this point, are you gonna have, you know, a fixed amount of MIPS that you're assigning, that you're signing up for as a business?

Um, if so, IBM's really going to have to put forth, uh, a very complex calculation that equates to how many tokens equates, you know, to how many mips, right? Uh, in order to understand, okay, that's great. You know, if I pass in 10,000 tokens, what's it gonna cost?

How much of my MIPS am I using? Uh, is there a direct correlation in that that they can make between, you know, the inference capability and the amount of MIPS you signed up for? Those are good questions.

And, and as you said, they haven't put forth any details about, you know, what their pricing will be. Mm-hmm. Lisa, I think maybe IBM sometimes takes for granted that people understand what the mainframe's all about, but in this particular use case, they need to talk to data scientists.

And there's a whole community of folks that they are not necessarily front and center with, uh, in much the same way. They were not front and center with application developers. So does IBM kind of need to reinvent itself a little bit and start telling a different story in some shape or manner?

Yes, that's a great point. They have to be able to understand who their audience is for this. Um, they probably, you know, as you said, talking to to other folks before, now they need to get what the data scientists, it's a different conversation.

So it, I think you're right. I think it is a bit of a reinvention. They need to understand what the data scientists are doing, what their pain points are, what they need, what they're looking for, to make sure that what they're designing fits that target audience.

So that's definitely worth that they're gonna have to do there. In addition to the pricing work that you talked about, is making sure the audience really is invested and understands. I would imagine they've already been doing that to develop and design this new system, but they definitely need to be talking to the right folks and understand how is what they're designing applicable, how, so what needs to be tuned?

And really listening to that audience is, is key for that. It seems to me that they're betting on the fact that, um, IT teams are gonna take more responsibility for managing the inference engines. So that part might show up on the mainframe, but then there's the whole training of the AI models.

Um, so I don't know, JP who's in charge of AI these days in these organizations is gonna be the IT organization and data science team or somebody else altogether. Right now, the, uh, it's been left up to those who are starting to build the applications and where the application's been built. Uh, I think it's a little bit of shadow IT right now that has been bringing the use of AI into the business.

And as far as a corporate or enterprise top down strategy, I, I haven't seen that, you know, enterprises have have really moved forth in this direction that they're putting it out, you know, out there as a, here's our stance on, and you have to break it down using AI for augmentation, using AI for automation. Um, how do we validate what data can be passed into the, uh, inference engines? What kind of tools are we making available or enforcing to ensure we don't have data loss, uh, that is part of this process?

Things like that are part of your AI strategy. I'm sure that companies are putting things together, but there's no guidebook. There's a, you know, I've seen a couple of things put out there by, you know, Google and others, you know, along the lines of, here's an, an AI strategy, but at the end of the day, you, it's my belief that you actually have to go and build your first application and, and decompose it all the way down to understand for your organization how you're going to make this happen from an operational standpoint.

So you need to actually do it so you understand all the pieces so that you can apply operational characteristics and, and, uh, requirements to managing that in a, in a production manner. And I don't see that that's been done. I I see a lot of point, uh, proof of concepts and a lot of minimal viable product just being thrown out there.

Um, and I think there's a lot of lessons that are gonna be learned, as we've seen with other technologies. When people try to jump on the bandwagon very early. Chris, JP is describing something you might consider rational behavior, but I'm not sure, I believe that there's rational behavior in IT organizations.

I think that sometimes we are gonna have six different teams build 12 different things and then argue with each other about who's implementation should become the quote unquote deep facto standard. 'cause mine's better than yours. So, um, does this require some more adult supervision maybe?

Well, you know, let me kind of back out a little bit, right? Security assessments, right? You know, this terrible word in, in security, which is people coming in to tell you what you're doing wrong, right?

Yeah. This, you know, I, I think again, just from a security perspective, we're slowly getting to the point where we can, where we should have been in the beginning, you know, in an ideal world where we understand who we are in the first place, what we're trying to achieve, and therefore put security in place instead of say, Hey, there's this security technology, where do I stuff it in my organization? And I think, you know, similarly, you know, to to this point, how much real gestalt organizational coordination is there and how much adoption of cool tech by business units and organizations doing things.

So I think that's the real question. How organized and considered is, is this. And JP you're saying, you know, there's, we're at this point in this topic where there are some guides around, and I think mostly people are probably sort of, you're probably understating this, but, you know, ad hoc adopting things that are working for them.

Um, unless building coordinated strategies, you know, that fit the overarching goals. There was a conversation I participated in yesterday on LinkedIn, and we, we are security people, uh, you know, people who are in actually implement security postures and, and have experience. We're discussing MCP, which is the new protocol for linking tools into LLMs.

And, and you know, the security person said, yeah, I mean, how do we entrust these things? W what, you know, we have all these servers that were created, how do we know that they were created by who they say they were? How do we know that when they called back out to the internet to some other application that they are doing so in a, you know, uh, honest in, in, you know, manner or that their, you know, intention is, is not bad.

And truth is that there isn't nothing, right? It's all experimental. Right?

Now, you're, you're taking the word you, you know, hopefully you went to Salesforce to get your MCP from Salesforce, but a lot of people are downloading this from GitHub or they're downloading it from hugging face. And the question is, well, what's the validity of the server there? How do I know it's valid?

You know? And I gave them some rational explanation, like you say I do in, in an irrational world about, you know, these things are running locally and if you downloaded it, you have the code. So, I mean, you do have an opportunity to validate there.

A lot of this is open source, but you know, the truth of the matter is, you know, they have a re a valid point, right? Where's the app store for MCP servers would validated, uh, you know, servers because you're authenticating and you're giving them the ability to act as you impersonate you against applications. I see a great man in the middle attack opportunity for anybody who wants it.

Mm-hmm. In theory, there's gonna be like an API gateway that, uh, speaks MCP, that's gonna provide you those capabilities. But the number of people I know who equate API gateways to AI is zero.

So, well, CloudFlare actually put forth a spec that does that today. They're actually built a remote MCP, uh, protocol that works with the current design instead of creating yet a new approach. So if you have an existing CP MCP server, it'll still work with the cloud flare capability.

So there's, they, they're the only ones I've seen that's really, you know, put forward something. And then there's a lot of discussion about specification, uh, the MCP protocols. Alright.

Um, I would shout out one thing with IBM is that, uh, they seem to be embracing more standards in the mainframe. So, um, maybe to your point, it really is just a big ass server, but, um, you can use things that you use for DevOps and everywhere else now with the mainframe. So is the mainframe just another note on the distributed network?

'cause we used to separate distributed networks from mainframes, but maybe it's all just one big happy distributed network. Jp, what do you think? I'm interested.

I, I mean, this is the first I'm hearing about the tell, and two, I'm really interested in how that compares to, you know, some of the NVIDIA product that's come out, some of the r products that are coming out with regard to their ability to, uh, handle, uh, GPU calculations and, and be able to support inference capabilities. I, I think that's the secret sauce if you ultimately, right? I mean, and that's what everybody's after.

That's where N NVIDIA came out as a strong player in the AI space, is their capability to reduce these, uh, chips that support ai. And if TELLEM two has some secrets, sauce, you know, that's what we need to see. If they have a competitive stance there With these chips, that's a, that's a major play, that's a major opportunity for them.

All right, we're not entirely sure if folks, if this is your grandfather's IBM Redux or something entirely different, but stay tuned, we're about to find out. We'll be back in a second. Discover Textron group, the epicenter of tech innovation.

We are your go-to for reaching IT leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us. You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more.

Join our satisfied clients, let's revolutionize your tech journey. Contact us today and tell your story to the world in the most powerful way with Textron Group. All right, folks, we're moving on to our next topic, which is the death of the paper check.

It's been a long time forecasted and people have been working on it. And now even the federal government says that it will get rid of federal checks one day soon, as soon as I get off this couch. Gotta move.

But hey, it's a step in the right direction. Lisa, what is your take on what's going on with this? And, um, I know some folks who are probably as old as I am who probably can't imagine a world without paper checks, but has the time come?

I know, I think it has. This is not anything new. We've been seeing these initiatives for a very long time, but just what, a week or so ago, uh, the president signed, I think it was March 25th, another executive order to eliminate paper checks for all like federal disbursements, including, uh, intra governmental payments, benefits, vendor payments, tax refunds.

It goes into effect soon, I believe it's September 30th. So in about five months time, payments will be made electronically. And the White House is saying it's not just the exposure and the risk of fraud behind this decision, it would also save the federal government millions of dollars reduced delays.

I saw something in terms of delays that, um, federal checks are 16 times more likely to be reported lost or stolen or returned to undeliverable versus an electronic funds transfer. So it's, I think it's, it's about time. We do know the Trump administration has really been embracing technology.

They've obviously from supporting AI through the creation of Stargate to establishing Doge with a mission of, in theory, modernizing federal technology and software to maximize governmental efficiency and productivity. And in fact, this executive order comes with a marketing campaign. Guys, the Secretary of the Treasury, in coordination with different heads of agencies, they're gonna implement a really public, uh, awareness campaign to inform people that are getting federal payments.

I wish I was one getting a tax return. Never happens, uh, of the, of this transition. But like I said, it, it isn't anything new.

We've seen this over the last 30 years. It's gonna be interesting though to see how this impacts banking systems that might be more legacy. Is there gonna be challenges there?

Is it gonna force a lot of the banking organizations to do a massive technology upgrade? I think it could very well be, but I think it's about time we move to things to being more electronic and more secure. I'm sure Chris would agree.

Yeah, I mean, you know, I, I can't really, you know, so I worked the, at a, uh, in the early eighties at a plant in Canada that made Sears catalogs, right? And it made Sears catalogs for a hundred years. And every working there was certain that it would be making Sears catalogs forever, because that's the way the world was.

And you know, obviously, you know, they don't make Sears catalogs there anymore. And paper checks today sort of remind me of, you know, when you get a ref refund for something, you can pay for it this instant. You can be shipped tomorrow and you get a refund, refund and takes, you know, three to four weeks, six weeks.

It's like, seriously, you know, in the modern world, it it, to be clear, if you're refunding me for something, then you should be able to do that transaction in the same period of time as to make it, and it's the same sort of thing. It's a legacy holdover. There's no logic behind it.

But companies, you know, make a couple extra points by being slow on their refunds. So paper checks, yeah, they're expensive, slow, unreliable, insecure, and they exist. Uh, you know, I, here in, here in, uh, in Canada, a friend who works at a, a metal shop, uh, that's been in business, a family business for years and years, and they have paper checks because in her eighties, you know, the, the, the couple that owns it, she's been writing paper checks all her life and she's not stopping.

Uh, you know, that's wonderful, but large organizations mailing out lots of paper for, no, this is 2025. Let's stop. Is she the lady that I'm standing behind in The supermarket When she's ready?

I was just gonna bring, I was just gonna bring that up. The only, the only, the only thing is what is the service for somebody that doesn't have a bank account, right? There's Gonna have to be exceptions.

Great point, jp. Yes. Uh, I, and I haven't seen that as corresponding information, right?

Because, you know, I'm all for efficiency and I, I I, I'm all about automation and efficiency and productivity. It's what I do for a living. But at the end of the day, um, you're a government and you have to take care of your citizens.

That's your primary, uh, goal. That's your primary responsibility. And unfortunately in this country, we have a, a significant size population of people who don't have bank accounts anymore because banks really don't wanna provide a service for people who only have, you know, live paycheck to paycheck.

There's no value in them in providing that service. So how do you used to get a check? You could go into a check cashing facility and you cash the check.

How do you cash digital money? Is it gonna be like a WIC card where you, you now have this debit card, but you know, that's still lossy, you could lose it, but you know, at least it's like American Express travelers checks, right? It's traceable.

Anybody who tries to use that WIC car, they need to pin right? You so it's safer, right? So there are options, but what, what are the options, right?

I just wanna hear, Well, lemme, you know, you're absolutely right. And from a security perspective, I, I, I honestly, I feel, uh, I, I feel, uh, like, like I've been corrected, right? Because yeah, as a technologist, yeah, let's get rid of paper checks.

But to be clear, yeah, those that pop, uh, section of the population, they're already the victims of all these money. Check cash checking scam organizations, you see them all over, you know, any, you know, any neighborhood across the country that is not middle class or middle class, you know, this is the banking system for millions and millions of Americans, and it's a ripoff to begin with. And yeah, I would have to think through this, you know, what is the actual impact on the vulnerable population of this great point.

Because, because again, they're already the, the, the least able to respond and the, and the, the least well served. I understand the president is getting ready mountains of pennies ready in the Department of Treasury that he'll just FedEx the people when he, if he needs to, right? Because that's sufficient, Super efficient.

Um, have we become, um, and I guess at least I'll ask you, but you know, yeah. Are we really on the cusp of becoming a cashless society? I mean, honestly, I can go entire week without spending cash these days, maybe even more.

Oh, same. Yeah. Yeah, I think so.

I, again, it's gonna depend on the population and of, of who we're talking about. JP brings up a great point. There's a large amount of the population that goes to check cashing places, and that's not going to change.

But for a lot of businesses that need to modernize, especially as the younger generations mature, are going to expect a cashless society. Um, I never have cash with me. I've got everything stored in my phone.

Um, I can use that. Or if I have to go old fashioned, I use a card, uh, which always feels very dated, but I think, I think we are progressing towards pockets of cashless societies. I don't think it's gonna be this blanket overall transition comes September 30th, 2025.

But I think in pockets we're going to be seeing it. And I hope that it does maximize the efficiency of the government, the productivity, reduce the risks of fraud and, and things like that to make people safer and be able to get their, their funds that they're owed in a safer, faster way. So I, but I think there's gonna be part of, it's also cultural too.

Like you talk about the lady at the grocery store who's probably, you know, a silent generation person, maybe a boomer, probably not, um, who's not gonna stop writing checks. And so they're gonna be the ones that are not going to embrace carelessness. But I think as the younger generations and we have more generations coming, I think, I think we will have no choice.

'cause I think they'll force it. Yeah. I, I'll tell you industries that well two come to mind immediately, that would really be upended by a pure cashless society.

The first is tips, because how, how do you, because there are certain areas, if you can leave a tip digitally, you know, electron like a credit card, but they don't get the full percentage. So if you leave 'em five, they get four 20. Um, whereas you leave 'em a $5 bill, they get a $5 bill.

The other one is casinos, right? When you sit down at a table, you don't have, there's no way, there's no system of getting cash, you know, from the, uh, from the window or from the casino that you can use at a table right now that gives you the pure amount. Right?

Now, if you take a credit card to the, for, to the casino window, you're getting charged a fee, right? So what, what's, what's the medium for buying chips at a table? All right, so you're pointing out two fundamental things that I might argue are scams.

Um, I'm here in the west of Ireland. I was in England last week, and you know what you didn't see since I got here tipping because everybody's got a job and it's built into their salary. And we're not trying to scam a bunch of people outta money because the owner doesn't wanna charge a full rate for something.

'cause he can't afford to pay the people a living wage. So we made up this tipping little scam. And so maybe it's time to retire tipping.

Um, and as far as electronic fees go on cards, I never understood this either. So let me get this straight. I don't go to the ba I, instead of going to the bank, I use my electronic card.

You charge me for that. So I'm not using the teller. And so now I'm paying more for the electronic transfer than I would if I'd actually used the teller.

In some cases in other places, I've noticed that I get charged more for going up to use the teller instead of the electronic transfer, which makes a little more sense to me sometimes. But there are more fees in electronic transfers than you can shake a stick at. And all of it is bogus, as far as I can tell.

It's middleman, it's brokers, it's a broker industry with design as a broker industry, and people need to get paid for running the service, right? Everything started out as a distributed network and everybody who runs a node in the network wants to get paid. So yeah, to, in order to get rid of that fee, you'd have to create a point, which we can today, but you'd have to get back to a point to point, you know, architecture where an ATM machine could, you know, doesn't cost anything, uh, you know, along the way, you know, for the transaction and it can go directly to your bank, right?

And that's what's missing, right? If I, if I go to a generic ATM machine hosted, let's say again, buy a casino, right? Uh, somebody has to pay for that machine, the maintenance on that machine, getting that machine filled, and it doesn't belong to any one particular bank, okay?

But the casino's offering it as a service, so it should be free, but they have to pay all the brokers. Now, if there was a protocol that said, oh, you are with Chase, great. I have the ability to communicate directly to Chase, I'm just gonna die, you know, send my transaction over to Chase, it's not gonna cost anybody anything.

'cause it's point to point. Now I can Chase says, yep, here's the money, you can release it. Uh, and it drops it in my account.

That gets rid of that whole network. And that's what Zelle was initially supposed to be, which is why all the banks jumped on board. But it didn't turn into that.

I just had this issue yesterday, so let me elaborate. Um, so, and again, you guys seem to think that paper ch checks are less safe than electronic transfers. And yet my son in college is calling me up because somebody rated his bank for $300 and now he's waiting for the bank to re to put that money back into his account, which won't happen till Friday.

So guess who gets a text saying, dear old dad, can you help me out? So sure enough, I'm like, first thing I go to, well, it's my bank, I'll go to Zelle. Zelle, um, wants me to therefore enter into this transfer.

Great. I've transferred money to him a thousand times. And then they're like, well, here comes the text, you gotta wait eight minutes for the text.

But I'm in Europe, so I don't know where that text went, but it never arrived. 5 seconds. So I gotta say something about our little electronic banking system.

It's, it feels like the online ticketing scourge we were talking about before and a couple of shows ago where everybody's got its hand out now, that it costs too months to do things electronically unless you get the right service and you know what you're doing. Which in itself is crazy. Whi which raises the whole question so many, uh, so much of Wall Street has, which is why is PayPal stock so low when it ha when it provides such a critical service and it's leveraged, you know, it's, is seriously undervalued, you know, stock.

I, I just don't understand where PayPal, what happened to PayPal? Uh, I guess it just went out of fashion. I think, you know, it's kind of like the mainframe of electronic payments, But they also own Venmo.

See, here's the thing, they're also Venmo, they're PayPal and Venmo. So I mean they, how many, how, what percentage of overall non-banking transactions do they own? You know, it's very high.

Mm-hmm. So anyway, Chris, are you sure that electronic payments are more secure than paper checks? 'cause there are some people who would beg to differ, but thoughts No.

In short, right? You, that's The wrong, you Asked the wrong question, question of a security person. Sure.

Certain, a hundred percent, no, there's no such thing, right? It is, it's all in context. And I can't, can't help think we we're thinking about this, right?

You, you know, the first half of the last century, you know, if you're into technology, you watch airplanes right? From canvas to the B 52, which is still flying today because a basic airframe is a basic airframe and there's only so much you can do with that. And you know, similarly, you know, the last 50 years we've done this technology thing and everything you talk about in a show like this, right?

You know, this is new, this is new, this is changing the electronics. Nothing has settled into a pattern, right? You know, long enough as it has in other disciplines like aviation where you can put predictability on it and say, this system is put, is well put together or not.

Um, and again, that's, that's sort of the world we're in, I think, you know, several more decades. You know, we still have mainframes, you know, we are establishing certain patterns, but we keep doing entirely new things with emergent properties that we can't predict in advance. So yeah, I, generally speaking, yeah, electronic transfer is traceable.

We know how that works. A paper check, you know, can get on the world and do weird things, but it depends. Make sure your son puts two factor authentication on his bank account.

I guess that might help, but I heard that they figured out how to get around that too. So who knows what's going on. But, and you know, I have no way of actually verifying any of those statements.

'cause you know, kids in college, s**t happens. Yes, It does often. All right, folks, we'll be back in a minute.

We're gonna move on to our next subject, which is, how do we get from A to B without driving ourselves crazy back in a minute? com is the leading resource for news analysis and education on challenges facing the cybersecurity industry. com covers all aspects of cybersecurity, including data security, DevSecOps, cloud security, application security, network security, security threats, and more.

com has the largest selection of security content featuring breaking news, blog posts, podcasts, and more. com to learn more. com.

Home of security bloggers network. Hey folks, we're back with one of my pet peeves out there. So we are now convincing or at least ordering everybody back to the office.

So we have returned to office scenarios. There are more people on the road. There are more people using mass transit.

Mass transit doesn't appear to be any cheaper in New York. In fact, it might even be more expensive than ever. And even though more people are using it, um, here's my issue of the day, right?

So if we have people go back to the office, in effect, they are now shelling out money to drive. They gotta pay for gas, they gotta, um, get a bus fare or a train ticket or whatever it may be. And that adds up.

Uh, I think in New York, the cost of, um, Metro North ticket is about 200 bucks a month. Um, if you think about this for a minute, sending everybody back to the office essentially is the taking back of what they perceive to be a raise. Uh, they got during covid when they worked from home, and that helped them navigate some of the costs and expenses.

Theoretically, all these people are gonna be a lot more productive in the office, but as far as I can tell, they're more p****d off than ever because they just fought traffic to come into the office. So, uh, they're not exactly thrilled to be there and it's not a great experience for all concerns sometimes. Chris, I'm gonna dump this one in your lap 'cause you're like the world's biggest commuter all the way from Canada to Florida and back again a couple of times a year.

But what makes sense here and can tech help us here? Uh, what makes sense is, is what works and tech can help us. So in 1990, you know, the South Carolina General Electric, um, Jack Walsh had was putting in a, a global video conference center, and I just happened to be there and, uh, when that was installed and took it over and did silly things with it, and the efficiencies, you know, being able to crawl under a machine and point a camera at something to a, a vendor engineer in Germany, as opposed to getting that person on the plane to fly across the ocean.

And so obvious and thinking about that then looking outta the future, what does this mean when everyone gets sent, right? And, and as you say, since then, I, I have virtually always been working remote, you know, into the, the, the nineties. Uh, in fact, the day my, my first child was born, uh, first day of summer, 1995, I called in the office, as you know, was deep into the, uh, uh, the firewall market.

And I said, I'm not coming in the office for a month. I'm staying here at home with the telephone doing the same things I'm doing, but I'm not moving from this spot. And people in the office couldn't believe it.

I was like, you're seriously? Oh yeah. And you know, 90 moving forward a little bit, 9, 19 99, the great Stuart Phillips, uh, I'm, uh, working at Cisco.

I'm running the firewall product lines, which are a real dog's breakfast mess at the time, and we're competing the whole firewall space at the time with my encryption or my thing, or better than your technical whatever. And Stuart Phillips redid that messaging. And I, you know, set of slides I carry to this day that said things like, well, my favorite slide to this purpose is that virtual private networks, IPSec was a big deal at the time, allows you to have, uh, remote workers and remote workers cost less money and work more hours making you more profitable and competitive.

So if you'd like to not lower your costs and increase your profitability and be more competitive, then don't buy a Cisco picks or an iOS firewall and use VPNs and get remote workers. And it, you know, the irony that, that in 1999, in 2000, there weren't a lot of actual remote worker operations, right? You know, I can say it now, but that actual economic benefit for most of those customers on those days, they did not realize that then, because, you know, it takes more than that to change things.

And it was, you know, not just the pandemic. We slowly over the last couple of decades have moved to remote work. The pandemic exercised that a lot.

And to the topic of this segment, yeah, is there certain valid and invalid, you know, thoughts about bringing people back to the office? You know, there is more to being productive than just a camera and a microphone. Um, you actually have, have a business process and get the value in the relationships.

Um, and on the flip side of that is some folks, you know, particularly older folks and executives are just used to seeing people buy the paper shredder and wanna bring 'em back there even though it's not a bloody good idea. So we're gonna mix that right now. You know, it's remote work is overhyped and not used properly in many, many cases.

Going back to work is nostalgia for a pass that didn't really exist and won't give you the business value. You think either JP, for as long as I can remember, somebody's been telling me that they've finally figured out how to make that routing algorithm work so that the traffic lights will be aligned and we can get that last mile of delivery optimized and we'll be able to, uh, make everything optimally run across any kind of networking thing. Include which traffic really is, is a giant exercise in networking.

And 30 years later, I've still not seen it working anywhere. So is this just a problem that tech can't solve or, you know, maybe, I don't know. Is there some secret thing in the reasoning model of the LLM that's gonna figure this out finally?

Uh, I don't think it's, it's a problem that's designed yet for ai. Um, because it, you know, it, it goes first and foremost to the sensor network. Uh, and you know, you're dealing with physical motion sensors, identifying and being able to perceive, um, the, the traffic pattern that's currently, uh, you know, what you, where you are currently in, in any point in time.

And then being able to analyze that and then retrofit the network to support that. Now, I mean, they do have built in, uh, safeguards. So even though your horror stories in New York, they do have the ability to enable an ambulance to go down and turn all the lights green or police to, you know, uh, to, to have all green lights so that they can get through, uh, in an emergency situation.

Uh, I don't know how easy it is to enable that. There's probably a lot of bureaucracy and a red tape, and it's that bureaucracy and red tape. I wonder that holds up the ability to allow them to turn over responsibility to a sensor network to actually measure this.

It's almost like we must have a human in the loop, you know, a air traffic controller, so to speak. Well, once you get a human in the loop making decisions, they can't operate nearly as fast as the need, uh, for, for the, you know, the, the traffic network patterns that are emerging and dissipating. Um, and that's where I think they get into trouble, right?

You have to put some trust into the system, and I don't think that they've done that yet. Lisa, let me tell you a story about what is going on in New York here with, um, there's all these motorbikes now there are little minibikes and whatever they are up and down the roads and, um, they go really fast and the city's starting to figure out that they're running into people and people are getting seriously hurt. So the solution to this problem is to, um, increase the frequency at which the lights turn red so that it forces those motorbikes to slow down so that they can't get up ahead of steam, which sounds great on paper, except for one thing.

Now everybody who drives is p****d off because the light changes every two blocks. So Yep. That would be very frustrating.

So is there, um, you know, how would you market that as a plus? Or how does somebody who's in charge of a city that's trying to provide these services kind, um, come up with something that people will kind of understand, appreciate, or maybe they should just come up with a different plan? I don't know.

Yeah, it, it, it's a marketing challenge because you're gonna have some people that are gonna see ostensible value. Um, if there are statistics in terms of the reduction that these red lights frequency would make in traffic accidents or even fatalities, you can use that data, uh, from a campaign standpoint to, to try to really reinforce that there is value here. But you're also gonna have the other folks on the other side of the spectrum who are going to get so annoyed because the lights are turning red too frequently.

I I have a few red lights streetlights in San Jose here that are awful, and if I hit a yellow, I know I'm gonna be here for three or four minutes. And it's frustrating, especially if you're trying to get somewhere you got an appointment and you're looking at your watch. But I think the challenge from a marketing standpoint is trying to, to bring forth the positives.

And I think data collection will help to do that, can help to do that. It's a matter of how the, the governments actually use the message and how they're delivering that on a consistent basis to their constituents. San Jose, my city where I'm sitting here right now is actually investing in cloud-based transit signals at it's almost 200 intersections.

Um, and it's saying, Hey, we're gonna re reduce delays for numerous bus routes. Um, and so I think that decision, that technology decision can demonstrate how data can be used to improve efficiency and attractiveness of public transit, but we have to have that message really crisp and clear and the value has got to be there. Mm-hmm.

Chris, there's this light in my town. Um, it's notorious for being slow. So what does everybody do?

They deliberately don't go that way. And as a result, they are now more traffic than ever going on all these other side streets, just so everybody can skip that light. And probably we're expending more gas and it's taken us twice as long to get where we're going.

Um, is this some sort of like fault in our human condition that we can kind of figure out cause and effect when we change one thing, we don't understand the implications, or is just somebody out there kind of looking at a two dimensional spreadsheet and trying to flip a button and doesn't think it through, Uh, the, the, the, the former, you know, if, you know, thinking through, you know, immersion properties, he's just hard. And most of us are focused on one individual, uh, task at a time. But I'm thinking, you know, a good analogy, I I I would say that this is my litmus strip, my most, my most important sort of measure of where we are with this ability to work remotely and so forth.

And that's the virtual conference, right? You know, I want the ability to go and interact with people as if I was there in person. And I can't be more clear about how far away we are from that, right?

This is wonderful. We can do this right here. We can have meetings and, and we, you know, almost all of my business, you know, my work has done this way.

It's wonderful, but it's not the same every time. Even in our industries, in our technology industries, we say, all right, we're gonna have a mixed hybrid conference. We'll be part in person and some people will be, be remote.

It's, you know, it's like the nineties, you know, anybody working remotely during those days, remember being the, the voice in the speakerphone in the middle of the room, there's eight or 12 people in the room, or we're like, Charlie's angels, we're just this little voice trying, trying to speak up over it. Yeah. So, so we're really, really, really quite a long way from the real remote interaction that gets the kind of value and business value, you know, being in person with someone and being able to talk for five minutes.

You can't replace that. We can replace that electronically, but not today and not next year and not this decade. So y you know, as I sort of said in the beginning, I, I understand, you know, I, I am, I am skeptical of anyone saying we need to get people back to work, but I understand there may be very good reasons, but it's that sociology.

How do we get value out of human interaction that needs to be studied? Not the technology. I don't know.

I just think we have to adapt. And speaking of adapting, I want to congratulate Chris on turning 50. That's 50 years of adapting, right?

60 pet 60, yes. 60. Oh jeez.

Come back to that. The big six. Oh geez.

Now we're getting serious. Oh yeah. It's bringing, it's all bonus rounds at this point.

com, all these little short lines, you know, we can complain about getting older, but as far as I'm concerned, you know what, it's, it's bonus rounds and defender now, so it's all good. All right. I just, I just want all these traffic issues solved before Chris turns 70.

Think we can get the, you better get on that then. I do think we have an accelerator and, uh, that, that's what I think the biggest thing that we'll get out of ai, uh, innovation is gonna be right now, is the, you're gonna see a high acceleration of answers to things that before just we're taking, not that we didn't necessarily have an answer, but it's gonna speed up now. Mm-hmm.

Significantly. Mm-hmm. Yeah.

Yep. That's a great point. Yeah, because I'm, 'cause I'm gonna ask the car, what's the fastest damn way around that annoying traffic.

Yeah, exactly. And it better have an answer. There you go.

All right. Hey guys, thanks for being on the show today. It was great insights.

Were awesome. And wanna thank everybody else for watching the show with us, and please stay tuned. There's an awesome amount of content coming up right behind us and the rest of the tech strong TV lineup.

Until then, we'll see you tomorrow.