AI Workforce Decisions, Data Center Backlash and Drone Detection

Hey everybody, welcome to the Friday edition of the Techstrong Gang. And as usual, we have an assortment of folks who are going to contribute to the commentary on the news of the day. And well, it is as up and down and an insane week all week long, and this show will be no different, I suspect.

But let me welcome our guests. John, how you doing, my friend? I'm good.

Hi, everybody. It's good to see everybody. All right.

Jack Poller, how are you? Doing very good. It's Friday.

It's Friday. Happy to be here with the rest of this gang. All right, Friday.

I don't think this is your last thing of the day, though, by any stretch. No. No.

All right. Guy Currier, where are you? You're usually bouncing around the world.

I'm in San Jose, California, where we just concluded AI Infrastructure Field Day. All right. We'll be talking about that on Monday, I suspect.

Tracy Ragan, good to see you as always, and in the great state of New Mexico, and apparently determined to stay that way. Absolutely. I'm happy to be with this motley crew.

Excellent. And then it can't be a Friday without some sort of special holiday day. So Jeff Richards is here.

Jeff, what are we celebrating? Who doesn't like National Nachos Day? Ooh, well, I do.

I like good nachos. I got to say, I like Irish nachos better. They kind of have fries and a bunch of eggs and greasy stuff on top.

It's like one of those garbage plates from upstate New York. Yeah. Kids and sink nachos.

Yeah, pretty much. But if you're ever in a place called Woodlawn in New York, order the Irish nachos. You're going to like them.

Yeah. In Texas, we kind of go with standard nachos. Standard.

Well, come on. Can't every country have their own type of nachos? Come on.

What of a Frito Pie as a- Nachos- Oh, those are awesome ... yeah, they're kind of nachos. Nachos were born in Texas.

I just want to say, nachos were born in Texas, so the right place to have them. Okay. Everybody's going to Texas for nachos.

Right. I got you. That's right.

All right. Here we go. All right.

There's a new report out, and we're all talking about jobs today because the latest jobs report has us up by 160,000 jobs, and that's a good thing. It's always good when people are working. But we have this backdrop of a conversation about, well, is AI going to wipe out jobs?

And Gartner has an interesting report out saying, well, it may or may not, but if it does, a significant number of people who get laid off are going to get replaced by people that companies are going to have to hire anyway. Because basically, they're going to discover that all this AI stuff doesn't really automate as many things as they quite think or in the way that they think. And it may make the employee's life a little bit better, but it doesn't necessarily mean that there's more output for the company.

But John, go ahead, explain. Well, when I came across this, it's almost like a contrarian report, but in a sense, it kind of does make sense. So given the backdrop of a number of companies that plan these massive layoffs that are influenced in great part by AI and restructuring, among other things, there's this idea, and I think what Gartner's saying between the lines, that there are consequences of rushing to cut headcount before understanding the operational reality of AI integration, and maybe it's backfiring.

It could backfire. So Gartner essentially is forecasting that nearly one-third of workers replaced by AI will be rehired by 2029, and often at higher compensation because these organizations are going to find out that if they cut too close to the bone, they are losing institutional knowledge, execution capacity that they fundamentally need. I think they are a little bit dubious about automation as an immediate headcount reduction mechanism.

They think it's going to lead to disruption, talent deficits, inflated rehiring costs. In other words, I think companies are getting ahead of their skis. They may regret what they do now and end up rehiring humans, which is encouraging to me, but that's part of a story we've been talking about for weeks.

The idea that AI agents in and of themselves can't operate. We need some sort of human element. So the one-third number, God knows I'll probably revise this number, but I found it intriguing.

I also found it a little bit heartening. But again, this landscape just keeps shifting as we rush to enhance or embrace AI to cut costs. So let me walk through how I kind of see this at the moment, and then maybe I'll toss this to Guy because he'll probably disagree with me for grins anyway.

But it seems like I'm looking at a world like Dropbox had an announcement this week, and they were talking about how you put all your files in Dropbox, and then they will give you all the AI agents you need to resize images or whatever it is, and this will all be automatic. And this will eliminate a lot of the tools and a lot of the scut work that people do every day, and that's not necessarily a bad thing. But it doesn't necessarily equate to more productivity for the company.

It just makes my life a little bit better as an employee, and I'm less stressed, and I'm not quite clear that I'm actually going to have more output because, well, at the end of the day, there's only so many things I can cognitively manage, and we were talking about this earlier this week with the notion of software factories. But their AI agents themselves require a huge amount of supervision. So now if each person has a small army of AI agents, they're going to have to monitor these things and actually engage with them because they're kind of like children.

And left to their own devices, they will go off and do kinds of weird things here, there, and everywhere. And so Guy, is the future of work I'm just babysitting AI agents and making sure they're doing the right thing and not squabbling with each other? Well, I don't know if you're saying that for me to agree with you or you're saying that that's not the case for me to disagree with you.

But yes, I do think the future of work is babysitting agents. Take the Dropbox example. 5 hit, my initial take was that this was not going to improve productivity.

It was going to improve quality and to a certain extent, dependability of work. Because if you think about writer's block, it's a little easier to get over writer's block or creative block or do some work block when you have an agent or an AI, however stupidly, to get you started. And I think we just keep seeing that take of mine borne out.

The Dropbox example is an example. Resizing images. Sure.

People get images, they put them into Dropbox in the pre-AI world. When they need them, maybe they can find them, maybe they can't. Maybe they get them once again from somewhere else because they forgot it or they misnamed it or whatever.

But you're talking about resizing, which is to say there's some standard format that they want them in, or whoever they're working with wants them in. And you could set up an agent to resize. It's done with scut work, like you say.

But what was the situation before? Well, they provided it in the wrong format. It came out blurry.

It wasn't quite right. If people think of this productivity gain as if in the past they were slogging through eventually, manually all of those images. But what happened in the past was they weren't going through them at all.

It just wasn't there. And so I think you can extend that to pretty much all the work. Tracy, coding using AI seems like it speeds things up, but if you're doing it right, you have all the architectural and design work to do.

You have lots to supervise. You need to be strategic about it. You need to know how to use it.

You need to develop skills or prompts or whatever that do it better and better as you go along. I feel like it takes about the same amount of time, but maybe it's more repeatable and consistent, AKA quality. Frankly, I am surprised that Gartner is saying only a third will be hired back.

I think they're going to keep revising that number. I think this is just like their conservative base number. Well, Gartner's good for two...

Gartner in particular provides two, and I should note that Gartner's a competitor of Futurum, my employer. But Gartner does two wonderful things for the market, and one of them is to provide these frameworks, like the five phases of, or whatever it is, five stages of AI grief, I suppose, that they've put in their report. But the other thing they do is take really amorphous stuff and stick numbers on them.

7 and that a probability of point whatever, and it seems all vibes based, if you like. But they're really good at putting numbers on stuff that's like really hard to- Well, for context too. Yeah, sorry, Guy.

In terms of context, they made this announcement at a conference in Florida with a heavy HR influence, so they're kind of throwing red meat to folks So there's a little bit of that motivation. So, I don't want to get into a separate argument here, but I always think that the numbers that get attached to anything by a research company, if you go back in time and look to see whether or not those numbers bore out, I'm always like, "I don't think so. " But it's always fun to have some sort of number to discuss these things.

Tracy, are you prepared to manage AI agents for the rest of your days and times? Is this your idea of a fun job? You would have to pass that question to me when you know how much I hate agents.

No, it's not a fun job. And no, I don't want to have to spend my days managing AI agents and babysitting them. But there is a reality that will set in for all of us that may be where we're going in some industries.

But I feel like the mistake that the community has made is believing that AI is a replacement strategy. It's not a replacement strategy. It's a productivity strategy and maybe a quality strategy, but even in code generation, the quality isn't better.

And it gives us a lot more code then that we have to then manage. So I don't even know if right now in coding, if AI is actually a good productivity strategy, even though I believe it will eventually get there. So for example, I submit white papers and whatnot for government grants and whatnot, and generally it takes me about a week to work on one white paper, say a six to 10-page white paper.

Now using AI, I can get it done in maybe three or four days. Does three or four days mean that I can replace one person, right? Can I replace myself completely with just AI generating these white papers?

No, it's impossible to do that. So it's not a replacement strategy. It is a productivity strategy because I'm not a very good writer and I'm terrible at editing, and I can take my white paper and push it through AI and ask it to edit it and clarify any word structure or sentence structure that needs to be done, and that's how I use it.

I use it as an editor. Now, I used to hire somebody to be my editor. I would pay them hourly.

That person's job may have been replaced. Mm-hmm. But in a normal company, you're not going to see a group of developers who are being so much more productive that they can get rid of the person sitting next to them.

I don't believe that's the case. But projects might get done faster. I don't see that happening either.

But Jeff, I also happen to think that, are we going to get attached to these AI agents, and each one of them is going to have an identity, and we're going to give them names, and I'm going to have- ... " Well, everyone already has an identity. And in fact, you hit a note that not enough people realize every single agent does have an identity.

And sometimes by default, you let an agent inherit yours, and that's not a safe practice. So yes, every agent already has an identity. What we need to do now, and ISA actually has a working group focused on this that's really pointing towards how do we manage both those identities and the authorities associated with them.

Most of them right now are way too broad. And by the way, to return to an earlier comment you have, I do believe in the axiom that 80% of all numbers are made up. So when it comes to these predictions, who knows where we're going to go with it.

But I also want to recognize that in the entropy of the entire supply chain of what we're doing, all we're doing is moving work. We're not necessarily reducing it. Now, it may still be beneficial, but it's going to be more beneficial to one area of the supply chain, and potentially more costly to another in the same way that with manufacturing, once you automate manufacturing and have faster conveyor belts and everything else, all of a sudden the packing and supply part of that have more work to do.

It may even be more expensive. So we need to look at the whole picture, not simply what is AI doing in a given silo and what identity does it have associated with it. All right.

Jack. " Let's go back with the real world, right? If I'm an employee and my boss assigns me something and it's not quite working out the way it should because the boss has some weird expectations, which is usually a common daily experience for every one of us, we can go have a conversation with the boss and we work this out.

In effect, if we all have AI agents, we are now bosses of our own AI agentic crew. And are we all prepared for the fact that that crew is going to message us at some point and say we're being idiots or stupid about something, and are we prepared to have that level of dialogue with AI agents that are basically going to be questioning us as we go along? Look, my wife shares a philosophy that a lot of other people do that you should be nice to the AI agents because at some time they're going to be your AI overlords and not your agents.

And they're going to be the bosses, and you're going to be the... I, on the other hand, am cognizant of things like token costs, and I like if the AI can understand it without punctuation, I don't need to spend money on it, right? And I don't need the AI agent to be nice to me.

I need it to just do stuff, right? And I treat an AI agent like an intern. I'm like, "Here's the assignment.

Go do it. Minimal instruction. Go figure it out on your own.

" And AI agents are very nice. They're designed to be nice, and worse, they're designed to be very chatty. They talk all the time.

" That's all I need. I don't need a 10-minute long conversation giving me the details of every web search you did and everything you did, right? That's one aspect of it.

Now, if you look at sort of the Gartner forecast and the rehiring, let's think about why companies laid off people in anticipation or because of AI. I think there are sort of two different layoff strategies, and we talked about this a little bit before. One is the big companies, the Oracles, the Microsofts of the world, the Googles of the world, they needed capital to go invest in the hardware to do AI, and so they had to trade off people for the capital to go for the hardware.

The rest of the world has laid off people because I think what Guy, Tracy, and Jeff all sort of touched about this a little bit, which is this anticipation of replacement of one form of AI replacing humans in one form or another. And I think that comes from thinking about AI as automating our existing workflow. And when we looked at historically manufacturing, right?

We automated everything, and we went from humans sitting in front of a table of bolts and screwing one bolt in, something comes in and they do it over and over again, very drudge work, and it was replaced by automation, by robots. And not only did humans not have to deal with this absolutely miserable work environment and miserable job, but the robots were much more repeatable, and they're designed to be repeatable. AI is designed not to be repeatable.

It's non-deterministic by default. It's designed not to do that. So you do have this sort of trade-off of if you replace a human with AI, what exactly are you getting and what are you automating and how is it going to work, and will it give you that quality that you think that Guy talks about?

And I like Guy's sort of discussion about the quality, but I think the non-deterministic outputs of non-deterministic behavior causes some questions for me about the quality. I agree. You all can all completely disagree with me about the quality angle.

I think I'm overdue to write a monograph or something about it, honestly, to just explain my- A white paper aided by your AI. I think quality also has to do with how much are you just cutting and pasting whatever's generated versus viewing it as kind of like the first draft of something and then fixing it. Bingo.

So we experts in using it are still babies in using it. For all our sophistication in using it, we have a lot of practice and a lot of learning to go. The main factor holding back the promise of AI is the human factor right now.

And it's so easily misunderstood. So I'm going to mention something that, so it was at a Cisco event. And it wasn't just Cisco.

I think there was another company. I went to some event. It was a ServiceNow conference.

And back a year ago, a year and a half ago, when they were first foisting AI agents upon us, they were convinced, they were trying to sell me on this idea that We're going to have super AI agents. That was the big phrase they kept using. Who is going to manage all the other AI agents?

There will be maybe minimal human interaction, and they're trying to sell that idea, and now they're backing away. I've got to move us along, but I am going to say that the new AI boss is arguably the same as the old boss. Gloom, and doom.

All right. Shifting gear a little bit, but we're still in the post-election cycle for the week, and we are starting to see some changes in the tenor of conversation around AI and data centers, and OpenAI is promising more transparency. And of course, folks in Congress are already saying that they're going to be chairing a couple of investigations looking into some of this stuff as well.

And I'm sure everybody's going to show up and play nice. But Tracy, I know you've been tracking this whole election thing for a while, and so what's your overall take on what's going on here, and can we believe any of these people? Well, I think it's a clear referendum on administration and an industry that has not been transparent or inclusive to the communities that they are going to live in.

It reminds me a lot of oil and gas, to be honest, because I live in a state that we don't own our mineral rights. And there was a period of time, probably 15 years ago, that oil companies wanted to come in, and they were drilling holes within 15 yards of people's front doors. And these are dirty, nasty kinds of industries.

Data centers may not be dirty or nasty, but they do come with a concern about the usage of water and more expensive utilities. And there was a period of time that nobody would say anything if a big manufacturer came in and brought a bunch of jobs, and the state gave them huge tax breaks and incentives for being there. And they used this for data centers, and it was the wrong strategy.

The requirement for any of these infrastructure projects at scale should be community involvement. You have to have from the very beginning because it's a way for the community to express their concerns and work with these companies to make it fair. And in this case, in a lot of these data centers, I know the one that's going down in New Mexico, they're looking at giving them way too many benefits.

And what about the people who are having to... They get discounts on electricity and water. Well, their neighbors don't, and it's a supply and demand, right?

If electricity usage goes up, which means it's going to be more expensive, it's going to be more expensive to live in a place that was historically a very inexpensive place to be because it's in the middle of nowhere. So community has to be involved, and it's across the board in everything we're talking about now. Everybody's voice counts, and state governments, local governments now have to consider that, and there's no other direction to go.

Mm-hmm. Guy, will big tech have to learn how to have these conversations? Because I think these are the types of things, they're soft skills, and they're not very good at it because that's why they went to the dark smoke-filled room to have these deals in the first place.

Yeah. It would be great if they did. Will they have to?

I think yes, but I just don't think that they will. But here's the counterpoint. I think we're a little bit lucky that the leading personalities in tech are multi-billionaire jerks, because I think that is the leading reason for hostility towards data centers.

I do think, for sure, that the backroom deals and the surprise initiatives and all that other sort of stuff, as well as the illegality like, "Come and take my gas turbines if you like," that sort of thing, I think that those are contributing factors. But I'm really more of the opinion that this is sort of a political movement in response to billionaires and trillionaires and, like, five people globally wandering around and apparently dictating how we're going to live and work. And I think that's lucky because the dangers of over-utilization of energy capacity and environmental impacts and all that other sort of stuff are real and problematic.

And I also think, as I've said a lot of times, including this week at AI Infrastructure Field Day, that AI is currently probably the most wasteful technology that we've ever had. The most inefficient and wasteful. I think that will change for sure, but the way it's built and run and used is extremely wasteful right now.

So, will the set of jerks I just mentioned, or however many of them there are, will they get any better at reaching out and talking about these things? I think OpenAI's grand announcement that they're not data centers anymore, they're AI compute campuses, I think that's an indication of how good they're going to be. But I also think that if that political problem is addressed, then we're going to start to see capacity built out in a quieter and more community-engaged way, just through a natural process.

Jack, do you think that the data center lobby, let's call it what it is, can paint a better picture than it has in the past, and will there be an effort, and will it be earnest? Yes. I think they can, and I think they will.

I think all of lobbies go through... AI is in its infancy. It's going through its growth curves.

Maybe you could call it the early tween years right now for AI, and I think their lobbying efforts are at the same spot, but they will learn very, very quickly how to do PR and lobbying and get what they want from the politicians. I agree 100% with Guy. This is very much a political backlash driven by a lot of hate and anti-sentiment, anti-this, anti-that sentiment.

" Somebody's going to see an opportunity here that the rest of the country is not looking at, and that state is going to reap huge benefits for it when it happens. Jeff, do you live in that state? I currently live in that state.

We've been living in for the past few years, and I believe that had something to do with the complete flip that happened in this state, because the now governor-elect ran on, that was one of her key tenets. So I really think that has a lot to do with it because it has been pretty much, Tracy mentioned because you're in the middle of nowhere, we have a whole lot of nowhere here, and all they need to do is find a way to either generate local electricity or pipe it in and find a water source and bam, they have a huge campus, Guy. So that's already happened, and there is a big bipartisan backlash to it for two reasons.

One, people feel they're losing control of the rural environment that they live in on purpose. And two, the jerks. You add that together, and why would people like this?

And they were never asked. They need to be asked. They just talk to the community.

Right? Work with the community to come up with solutions, because in my experience, living in a rural area with the oil and gas example, we came up with some solutions. We came up with some agreements of how it could be done.

So, the community may want it as well, but they don't want to get charged more for their electricity, right? " So he was arguing for reopening a nuclear plant that we had shut down, like, two decades ago. Not a popular thing and highly...

Well, OpenAI, they're trying to say the right things right now. They claim they're going to retreat from these municipal non-disclosure agreements, but I don't think they have a choice. Meta's saying something similar to that, but they don't have a choice because We just did a story this morning about the Democratic majorities.

When they take over in Congress, they're going to introduce measures aimed at shielding households from surging utility bills tied to data facilities. They're going to mandate independent safety audits for frontier models, and they're going to strip legal shields that protect AI developers from product liability. So these guys don't have a choice at this point, and I think, Jeff, I think it was like 27% of the data center activity is based there, and they had the ultimate villain pumping in $250 odd million in those races.

Can I go to my resident government senate, Jack? As you think about this for a minute, will this be fundamental change, or am I just going to see data center money move from one party to the next party and they'll put a happier face on it, but basically, what comes out the other end is just going to be a slightly heavier touch than what the president has been doing, but hardly what everybody voted for. I've used this expression before and I'll say it again, there are multiple Scrooge McDuck swimming pool piles of money floating around in this one, and there is way too much money for the politicians to ignore.

And as I said, the people who are trying to do more data centers are going to get very good at spreading the money and the love and the PR and the lobbying very quickly, because there's trillions of dollars here, right? One business is looking at $800 million in cash that they need in the next three or four years. That money's going to get spent on people as well as on digging holes in the ground for power and cooling in data centers, and it's going to influence a lot.

Here's my killer political app, and it's probably out there and I haven't seen it yet, but it seems to me with AI, I should be able to have an app that lets me track what my local politicians are saying about any particular issue, and then I can send them a thumbs up or thumbs down that says, "Here's my sentiment on this particular topic," so maybe they know what this is before they go and say something stupid that I'm just going to get the rest of the electorate riled up about. Jeff, yes, no? Yeah.

What makes you think an AI agent's going to be able to make sense of what a politician says? I think all of our politicians should just be AI agents from now on. Let's just do that.

See, now, that's the kind of dumb intern I'm looking for. Yes. I like this idea.

For sure. But yeah, it's true. I mean, at some point, maybe we collectively elect an AI agent that's going to represent the community that we all agreed on, that is trained in a certain way.

Who knows? Maybe we could have a Congress full of AI agents. Guy, what do you think?

I just want to jump on Jack's highly cynical, everybody does it, and the money rules take on the blue wave that we've experienced, and the fact that that doesn't matter because it didn't really work for crypto, and that was under red rule of the past few years. I think that the money sources actually are suspect right now. But yeah, sure, it's way more than enough to buy a bunch of people off.

But the situation just overall is just way too uncertain to know. And I think that there's two ways this market can go, because we've talked about the imbalance in investment levels versus return quite a few times on this podcast. Is it a podcast?

I don't know. Anyway, we've talked about that a lot, and my take has been that we don't know how much return we're getting on all of the spending. In this case, we're talking about data center build-outs and data center spending, but there's a lot of other huge areas like boundaries, and such, that need build-outs.

And, we might be set at the beginning of a correction now, where the money sources try to hedge against risk. And in that kind of environment, I think the politics are, at least for a year or two, are going to be a lot more significant than the money. I'm going to leave this here, but I would ask everybody who voted, first of all, I want to thank you for participating in this experiment called democracy.

But please, stay engaged. This conversation is far from over, and you need to keep an eye on what all these politicians are really saying and doing in your local community because, as Jeff pointed out, what they say and what they do is not always one and the same. So pay attention.

All right. We're going to move to something kind of geeky and fun, but I hope it's fun anyway, we'll see, but it's also maybe scary. But Ericsson is talking about figuring out a way to use networks and wireless networks and all those things that we depend upon to detect drones.

And this is important because we are starting to get concerned about not only drones in war, but drones that maybe shouldn't be in a particular location, and maybe even malicious actors using drones for terrorism attacks and things of that nature. Jack, I know you've been following this drone conversation for a little bit, but what do you make of this approach? And I kind of like the idea, but here's the thing I am a little worried about it.

Am I going to have to put an anti-aircraft cannon on the corner of every street because we've got to watch out for drones that's going to be connected to this network, or how's that going to play out? Well, I don't think we're there yet here, but the motivation for this is this guy in Russia who has sparked a drive to reestablish the Soviet hegemony and launched a war into Ukraine where he thought he'd steamroll Ukraine in a couple of weeks. And the Ukrainians said, "Wait a minute, we've got a whole bunch of these cheap drones here.

We can go buy them on the open market, fly them out there, and we can see where the tanks are. " And wow, that works really well. And the way war works today has completely changed in a period of months to the point where drones have completely reshaped the battlefield.

And what's sparking Ericsson is Russia is now, both Russia and Ukraine, but Russia has developed enough long-range drones that they're flying past Ukraine and around other countries into NATO block countries in Eastern Europe. And there have been a couple of unclaimed, but downed Russian drones that fell out of the sky or were downed in Poland and some other areas. And the Europeans are suddenly scared that what Russia is doing in Ukraine will now come to the rest of Europe.

So Ericsson said, "Well, instead of trying to build our own radar, anti-aircraft type things and all of this, we have a whole bunch of radio systems sitting on cell towers all over the world. What happens if we use the existing radios that we have and say, 'When something flies in the middle of the radio, we can see a reflection off of that. ' So why don't we apply AI, our friendly AI, to analyze all the radio signals from all these towers, and we can start seeing things that are moving through at 100 feet, 200 feet off the ground," which is where drones fly, and traditional radars, even both military and civilian radars for air traffic control don't see that low and can't see planes or drones, things that small.

So I think this is a really cool utilization of existing technology, maybe a software patch, additional data analysis, where we then enable this. And as you said, Mike, that it's not just Russia that we're concerned about. These things are a couple of hundred bucks and they can be pretty dangerous at $1,000.

And a terrorist can do something with this pretty easily. And so having this technology available in the United States would be, I think, just one more level of assurance that something bad, like another 9/11, isn't going to happen. All right.

Tracy, a couple of months ago, we were talking about drone shooting parties and basically all these drones that were flying around without permission and people would be taking potshots at them. Is this a way to maybe distinguish friend and foe, and what's a legitimate potshot target and what should maybe belongs to Amazon and not be shot down? I don't know.

I still think if somebody's delivering a- Their food via a drone, and flies over somebody's swimming pool next door- Mm ... there's going to be issues. But I agree, this is crazy cool, innovative use of our existing infrastructure, and certainly an argument against having the need of the giant arch in the middle of DC because that's one of their arguments, is they need it so that they can shoot down drones, which is ridiculous, to be quite honest.

But the truth of the matter is, it's great that we're seeing this come from the private sector, because the public sector right now, the DOW is not focused on drones. I tell you right now, they are focused on LEO, low Earth orbit. They are thinking about the golden dome.

They're thinking about- ... missiles and satellites, and missile detection and satellite sensing. There is not a big focus on the management of potential risk around drone technologies.

And as we've learned from Ukraine, they can be a very serious security challenge. So I love this idea. " And using our existing infrastructure is exactly the way we should go.

Yeah. But there's certainly problems with it. There's lots of discussion that needs to be had around it.

Precisely about what you're just talking about, how do you detect what's a good drone and what's a bad drone? See, Jack, can you black hat this a little bit, though? Because that's what always escapes me in these scenarios, is it's all clever and cool and everything, and then some bad actor somewhere uses the exact same thing to do something not so good.

Yes. So Ericsson's claim is that reusing existing infrastructure means you don't have to invest in additional infrastructure or develop new technologies, like a radar system that would detect this. It's leveraging a private organization's resources.

They are the cell towers. Ericsson's not running cell towers, they're building the equipment that are on the cell towers. If a bad actor got access to it, they would see what is flying, assuming the technology matures to a point that you can actually detect a bunch of drones and stuff.

They could see that and potentially not only black hat that, but blind the system to the bad drones that they are controlling. The malicious actor not only is flying drones to do bad stuff, but is also blinding the detection system. Mm-hmm.

Like anything else on a military side defense, and like we do in cybersecurity, we believe in defense in depth, and so we do not rely just on this technology as a sensor network. But it's one- Or how about spoofing? of many sensor networks.

Yeah, spoofing is the same thing. Spoofing the detector system to think that there's something there that's not, and drawing attention away from- Well, and we have these challenges today, and we're trying to figure out how to deal with them on GPS. If you think about it now, everybody's got a GPS in their cellphone, in their car, and everything else, and they use that to navigate and go about their daily business.

We land aircraft using GPS. We are now learning, particularly again from the war in Ukraine, about what to do when you're in a GPS-denied environment. And GPS is a very weak signal, and it is very easy to spoof or deny the signal.

And so we are now having to figure out we can't be 100% reliant on one system, because single points of failure fail, and you have a problem. So if they were jamming the signal, wouldn't I be able to at least know that the signal was being jammed, and then I can make some assumptions about what's happening in that particular area where that radio is? Yes and no.

So from a military strategy point of view, we do a lot of rubadub, is I think the expression. Which is, I make a whole bunch of noise over here and say I blind a whole bunch of things over here, and you think the attack is over here, but it's really on this side over here where we haven't done something like that. And I'm not a military strategist, but I've just read about this a bunch because I follow this.

But there's a lot of things that go into this. These are very complex topics, as is the technology itself. But from a perspective, the guy's question is any technology that we create for peaceful purposes can be reused and manipulated for non-peaceful.

Anything we create for defensive purposes, we can almost always repurpose for offensive use as well. It's very, very hard to not create that scenario. Mm-hmm.

So I don't know, but Jeff, will there be some small town in Texas that creates a... And I'm just picking on Texas for grins, but a no-drone-fly zone, and if they detect them, they're just automatically going to take them down. So yeah, go ahead and pick on us.

First of all, I do think this is a cool technology. And to Jack's point, two of them really, all new technology becomes hack technology before some given amount of time. And when you rely on just one technology, it will fail, for a number of reasons.

And what the scenario Jack just described is just like a DDoS attack on the internet. A website gets blasted with a whole bunch of traffic, gets overwhelmed, and then there's a low attack that's exfiltrating data. Same sort of thing could happen with that.

I envision we're going to end up with tail numbers for every drone. Mm-hmm. And when you look at FAA controls right now over planes, I think you're going to see somewhat of a parallel going on with that.

Yeah. I also think we should arm every cell tower with a laser. Yes, that's what I think.

I don't think this is going to be anti-aircraft guns on the street corner. I think it's going to be lasers on the cell towers that will take out these drones as they're going by. Because that would be the closest, most logical way to do that.

Jack? Yes, no? Well, that just sounds so great.

Let's get machines on them too. Yes. The amazing thing is lasers are now no longer science fiction as a defensive weapon.

So Israel has successfully used lasers to shoot down incoming missiles. And a drone is going to be much easier to deal with, and it's very conceivable that you could do that. The problem always is, and this is the first rule of firearms, is not only do you shoot at your target, but know what's behind your target.

Because when you miss, that laser keeps on going until it stops and hits something. And that's going to be the challenge. But then again, Tracy's got an army full of rednecks in the back of New Mexico swimming pools with their shotguns ready to go.

So... Yes, we do. I'm going to leave this here, but I'm optimistic that maybe the world will be safe from drones.

I don't think we maybe have the complete solution, but this notion that we were going to be overwhelmed by drones that are being used for malicious purposes, maybe we do have an answer, or at least smart minds are working on it. I want to thank our panelists for sharing their insights. As always, happy Friday.

Thank you for being here, and please stay tuned for the rest of the Techstrong TV replay that's coming up right behind us. And as always, stay tuned from a word from our sponsor because it's our favorite one, us. Everybody, have a great weekend.