Microsoft Outage, AWS AI Automation and Sea Sensors
Microsoft outage resilience leads this Techstrong Gang discussion. Mike Vizard joins Jack Poller, Tracy Ragan, Jon Swartz, Jeff Reich and Amanda Razani to examine three infrastructure stories. They consider disrupted collaboration services, an AI-built AWS development tool and new underwater sensors. Each story highlights a different dependency behind digital business, from everyday applications to the cables beneath the sea.
Outage Outrage
A widespread Microsoft outage affected Outlook, Teams and Copilot. These services support everyday communication and work across many organizations. However, their importance also complicates recovery when several familiar tools become unavailable together.
The panel considers business continuity beyond the immediate technical fix. Teams need ways to share reliable updates when email and collaboration platforms cannot carry them. Meanwhile, customers and employees need clear information about affected services. Microsoft outage planning therefore includes communication paths, service dependencies and practical recovery procedures.
AWS AI Automation
AWS Emulator gives developers another way to build and test cloud applications on local machines. The project uses AI agents to build and maintain emulated cloud-service APIs. Developers can use those interfaces locally and in continuous integration pipelines.
As a result, teams can explore faster feedback without sending every development step to a remote cloud environment. Yet local emulation has limits. A useful development result does not automatically establish production readiness. The discussion examines where this approach saves time and where testing against real cloud services remains necessary. It also considers what AI-maintained tooling changes for the people who rely on it.
Sea Sensors
New sensing approaches aim to improve visibility beneath the waves. Optical sensing, autonomous systems and AI could help detect activity near shipping routes and critical infrastructure. Subsea communications cables make that physical environment important to digital businesses.
Consequently, the panel considers what better detection could reveal and what remains difficult to observe. More information can help operators understand activity around vulnerable assets. However, detection is only one part of protecting infrastructure. The discussion keeps those practical limits in view while examining the promise of new sensor systems.
Together, these stories connect service resilience, local cloud development and underwater infrastructure awareness. The panel examines their implications without treating emerging capabilities as finished solutions.
Transcript
Hey everybody, welcome to the Techstrong Gang. It's Friday, and we have our usual assortment of well-known criminals hanging out with us today to have this conversation. But let me introduce our gang for today.
Jon Swartz, good to see you. Hi, Mike. Hi, everybody.
Good. Amanda Rosati, good to see you. Hello.
See, I can actually say that correctly. All right. Jeff Reich.
Howdy. Better being here than prison. All right, Jeff, I'm going to come back to you in a minute.
Tracy Ragan, good to see you. Thank you. And Jack Poller, of course.
So- Nice to be here. Well, let me jump through a couple of things before I get back to Jeff first. So if you've been watching the show, we've been at KubeCon all week, and now we're just back here in our regular, normal environments.
And so a couple of housekeeping notes. We did not get a chance to thank all the veterans for their service this week, so we'll just lead off with that and thank all the veterans for their service and keep them in mind as we go forward, because well, yeah, it's a strange world out there, and we may need them sooner than we think. The other thing is, it's Friday the 13th.
So, that probably doesn't bode well for the rest of this conversation, but we'll see how that goes. And, at the same time, we've also been talking about Nvidia and what's been going on over there, and that chaos continues to ensue, and I imagine it will forever, and we've kind of been talking about it on the gang this week. But I'm going to leave that there.
But, John, I'm going to start with you. I have a conspiracy theory for you. You're going to love this.
Oh, I love conspiracy theories. So, Microsoft, somebody decided to help Nvidia out by creating an entirely different news story and basically decided, "Well, I know what we'll do. " What do you say?
So it's good for the team, right? So The Magnificent Seven, they're all colluding. They work together.
Yeah. Anyway, unless you've been under a rock, you would know that there's been a major outage at Microsoft this week. So what's the latest?
Oh my God. So we talked about Friday the 13th. It came a day early for most of us.
Mm-hmm. So, I think we were all affected one way or another, but there were some updates posted to the official Microsoft 365 status account on X, and they confirmed the disruption that I think impacted as many as 130,000 people, was tied to an elevated load across a portion of its infrastructure, which initially affected Exchange Online mailboxes and expanded to impact services connectivity across multiple regions globally, Teams, Copilot. I know I had problems saving just a Word file, among other things.
My wife, who has to turn in grades for her school, the system, it basically erased all the changes she made, and it never saved them to begin with. So just kind of updating it, I heard from a couple of folks today, I think Mike did too, that Microsoft confirmed it was investigating Exchange Online mailbox access issues in the Americas. While they were doing that, the lack of a clear root cause or restoration timeline fueled the cyber attack speculation on X.
There were reports also that Google and AWS had disruptions. This is according to Global Data. One thing I want to throw out, maybe we can talk about it later, is this whole impact that this may have had in terms of insurance.
The insured, global insured losses from this outage are preliminarily being estimated at between 300 million and 500 million, according to Morningstar DBRS. They drew some comparisons between this and the 2024 CrowdStrike outage, which was significantly larger. But nonetheless, that's something we could talk about.
But I don't know about you all, but, yeah, just the timing was terrible. I'm still having a little few issues. Mm-hmm.
I don't know. Maybe let's open it up to the floor. Well, I'm not on Microsoft Outlook or any of those services, so first of all, I would like to thank Microsoft for reducing the volume of inbound emails to my system for the last two days.
I greatly appreciate that. And also though, from a business perspective, kind of looks like, is this like the equivalent of a modern digital snow day? Is that what this is?
And everybody kind of just got the day off, and this is kind of the new chaos and the new reality we live in, because I swear, these outages involving Microsoft are becoming more frequent. I think we had one with GitHub not too long ago, and developers were offline for the better part of the day. Tracy, I'm going to go to you after on this one, but we're not supposed to have this, right?
There isn't supposed to be these single points of failure. What's going on? This is the new normal.
And you know what? I don't care, as long as GitHub's running. Well, for me, I don't use Exchange, and I think that the world is going to help Microsoft out because a lot of these governments are moving off of Microsoft's platform because they don't want a reliance on it, which probably is a good move.
So we're facing a change. This is a change. Last Tuesday, the blue wave wasn't the only change that's going on.
I think that we're going to see some changes in how developers and how companies think about managing this single point. We have pushed developers for years not to have that, but most of our, I would say our business software is reliant on a third party, and that's going to happen. It's the new normal, and I really do believe that there'll be a shift back to having things more local.
All right. Jeff, do you agree? And I forgot.
My apologies to go with the whatever day it is celebration, and I have no idea what that's behind you. But I would say that Microsoft has positioned itself as the world's computer. I don't know if you've seen that marketing stuff.
So that thing behind you doesn't look like Crow, but what is that? Actually, whether you want to call it a new normal or not, Microsoft can call itself that, that's fine. This is not a Microsoft issue.
It manifested itself in Microsoft this week. And like you, Mike, I just got, "Oh, yeah, it's a nice slow day. " But more importantly, take a look at everything everyone does today, whether it's development or sales or managing your coil club, whatever it's going to be.
Everyone's relying on some sort of network and computing in the cloud. And for those that are really depending a lot on AI agents to determine a lot of their decisions, this should be a wake-up call that says, maybe this is why it's a new normal that Tracy referred to. It's not a Microsoft-only problem.
It was simply their turn. Mm. And I know I was an incident commander at different places I worked at for a long time, and the rule had always been, no matter how dependable it is, no matter how many redundancies your vendor says they have, find another path that's completely away from them, so that when this happens, at a minimum, you have a way of getting messages out or performing rudimentary functions.
And if we don't take this wake-up call for that, the next one's probably just going to be worse. Whether it's accidental equipment failure, bad programming, cyberattack, the end result's all the same. All right.
You know what I did notice, and Amanda, I'd love to get your opinion on this, and I was talking to some other folks. There was a massive spike in text messages yesterday because everybody decided that, "Well, I couldn't get ahold of him. I couldn't send my email," so they wound up sending all these texts, and I was kind of getting texts from people I haven't gotten texts from in forever, and I guess they might've spent some time trying to figure out what my phone number was.
But do we need an actual plan here for failover when these things happen? Because to Tracy's point, it's going to happen to everything, and it's just the new normal. Yeah, absolutely.
Well, my husband works for a bank, and they were impacted. And they were actually being... Yeah, they got text messages telling them to go check their personal emails for communication.
" But they were trying to, how could they still communicate? And I saw this across a lot of different business environments yesterday. But something else I thought that was interesting was- Several tech leaders I saw that posted, perhaps Microsoft has just gotten so big that it's going to just be subject to all these attacks.
So will we see businesses looking for alternative options, smaller companies to switch to that aren't going to be subject of all these attacks? And I did see several other small business owners taking advantage of this happening to put little funny posts out there, like, "Who was able to keep operating? We were.
Whose email didn't go down? " So anyway, I did see that reaction. But I wanted to talk too, before we move past your whole conspiracy theory thing you said at the beginning.
Oh my gosh, the level of conspiracy theories coming out from this outage. Let me tell you. These are serious tech leaders in the industry that think that it could've been a massive AI agent attack flooding, causing this problem.
So purposely attacking. " They've been saying this for weeks. This is what it's about.
It starts here. But is it a laughing matter? I don't know.
A lot of theories come true. Well, I figured out I can live without email, but Jack, let me ask you this. There are these things called service level agreements, SLAs, and they're buried somewhere in a contract, but I also believe there's maybe verbiage in there that will say that Microsoft isn't responsible for any damages inflicted here.
Because business owners are probably hopping mad about the fact that basically they've lost an entire day of productivity for people they ought to pay. Yeah, at the cloud level, the service level agreements, the SLAs, are typically compensated through credits to your account. You paid $100 a month for this service, you lost 5% of the utility for the month, we'll give you 5% or 10% back on your payment for the month.
But more importantly, I think, is I find that this is just an incredibly weird story, that we're having this massive conversation throughout the public discourse about something that has affected so few people. Mm. John quoted a number of 130,000- Yeah ...
which I believe came from DownDetector's 130,000 reports. Yeah, DownDetector. If it was 10X the number of people, if it was a million plus people, remember, there are somewhere between 450 million and 500 million seats on Microsoft Office 365.
So we're talking about half of a percent of their entire user base being affected. Maybe it's 1%. That is a very small number.
5 million Windows devices. That- Yeah ... is significant.
When I saw the 130,000, like you, I was thinking, how can something so conceivably insignificant create so many waves? But it did. I would challenge those numbers then because- Yeah, I don't believe it.
I think it's too low, but... Yeah, because I was on a plane last night at 1:00 in the morning watching the BBC, and they're reporting out of India, where all kinds of companies and organizations are affected, and if 5% of India is affected, that's twice the number you just stated, right? Yeah.
Plus, it's not just those companies, it's who was affected because those companies were impacted. Right. And I'm just going based on the data that I have available to me.
But it seems to me now, you mentioned text, I would also like to remind people that this can be used for voice communications. And in fact, that was its original purpose. And you can pick up the phone and call your buddy and say, "Hey, Joe, I can't send you an email.
" Well, but I don't answer phone calls from people I don't know. You know, basically my- But critical conversations can continue. I think that's the point, right?
Yeah. We don't have to have emails. We just don't have to have it.
No. There's Slack, there's Discord- Yeah, that's true. There's Discord, not email ...
there's telephone. Yeah. Yeah.
Jack, can I just, just kind of stepping back, so what happened to Microsoft yesterday, I just got to say, what a bad week to be big tech. You think about the blue wave or whatever you want to call it. You think about Nvidia, Anthropic.
There are more questions about the IPO, OpenAI. I got to just throw that out there. It's...
Delta? Delta came out and said, "Hey, we're not using agents. " Yes.
Yep. Yes. That's what they stated.
Delta as in Delta Airlines, right? Yes. Uh-huh.
Yes. " So I'm not- So, you know- Sorry, I'm United. What can I say?
I'll be quick about it. Delta, the CEO Ed Bastian, has made this such a cause. We're going to have the best internet, online, website service, WiFi.
He's been saying that for years anyway. All right. Well, I will give Delta some credit, because I was on their plane last night.
Okay. So there was massive storms here in New York, so everybody got home late, and it was kind of a borderline nor'easter, and they even had tornado warnings and all this other stuff. It wasn't Delta's fault, and the people who were in charge of the plane and kind of trying to make all that work were lovely.
So, I gave them some props for being stressed out and all the things that went with it. They went out of their way to try to keep everybody comfortable, so at least the people at Delta are okay. Mike, I'm glad you got home safe.
I would just offer, I want to leave with this thought: no good crisis should go wasted. This is maybe overblown, but it is still a crisis for... It was a 100% issue for those affected, right?
" You know what, Jeff? What you said is so interesting, because I was at a press conference with the great Stephen Foskett. We were at Commvault in Nashville, and there was a panel of executives, and one of the panelists, who is a CIO at a major bank, said this.
He said, "The best thing that can happen for the AI industry or enterprise adopters is to have these crises or these events be publicized. " But that was something he pointedly brought up. Said the company that sells backup and recovery services.
No, no. I won't lie to you, typically, I'm usually skeptical about this stuff. I almost said a word.
Stuff. But this guy had had a painful experience, so he was kind of speaking from his own torture. But yes, Mike, you're right about the self-selling.
All right. Jeff, before we leave this topic, what the hell is that thing behind you? So for anyone that grew up in New England, they know what this is, especially this time of year.
It's called Indian pudding, which I know is very likely politically incorrect. But it was a version of hasty pudding brought over by the British, and when they couldn't get wheat to make it, they used corn. So it's a different version because corn was being grown by the native people, they called it Indian pudding rather than hasty pudding, and there you go.
Okay. And are you sharing a recipe for this with the rest of the country now, or how does that- I'd be happy to. It's basically corn, egg, butter, molasses, or maple syrup.
The ice cream is optional. Right. Well- And you bake it like a little- Well, if you send that email now, most people should get it in time for Thanksgiving.
Exactly. Touché. All right.
Moving along here. So this week also saw something that I have not seen before, and maybe it'll become more commonplace, but AWS announced that they had created a new tool. It's an emulation service for developers, and you can work on your local machine and emulate a bunch of AWS cloud services, and while that concept is not necessarily new, what was kind of fascinating and maybe even a little scary is that they said the whole thing was built by agents and will be maintained by agents, and they're not even going to take, it's open source, but they're not even taking poll requests from people.
They're saying that this thing is going to be standalone and fully autonomous. Tracy, is this the way of the world going forward now, or what? I think when it comes to developer workflows, agents will be the way of the world.
There's no stopping that. But I think it's a productivity win for developers, in particular, smaller teams that don't have big infrastructure budgets, and we work with a lot of those types of teams. And this is just platform engineering next level.
Platform engineers have spent years trying to streamline how developers get an environment up and running as fast as possible. Raspberry Pi, think Raspberry Pi and when you're hitting the seeing the story. But what AWS is really bringing is what used to be a cloud development environment that was something that was exposed or in a central point of control, now we have it back on our local machine.
I feel like technology's going through a salad return here because this is where we started, and I really believe that that's where it's going to return. Because developers then they can build, they can make mistakes, they can do anything that they need to do without spinning up a lot of cloud resources. And for open source communities, that is huge.
In fact, the Otillius team has spent a good portion of the last three months doing everything they can to simplify the front end, even though the back end has some complexities, so that developers can bring everything down to the local machine to do all the development they need. Because we're not going to continue providing support in an AWS environment. So for small teams like ours, this is a big, big difference.
But I think the real story, and I don't want this to get missed, is that we're moving back towards local machines. There was an article, and I know I sent this article to you, Mike, I have to refer to it, about a legal firm who built a private AI server using eight secondhand NVIDIA GPUs. Their goal is to keep all their confidential files inside the office, and so the system runs an AI model locally, allowing them to review documents and do their work without sending sensitive information to an external provider.
And they claim that they're saving about 50% to 60% of their AI workloads now can run locally. Mm-hmm. Tokens, right?
And reduce their need for a commercial AI service. And they say it cost them about $4,800 to do it. Yeah.
Interesting. I have been predicting this, and I was happy to see it. So we're just returning to local.
Yeah. We're just returning to the local machine. We mentioned something similar at KubeCon this week.
Red Hat was talking about how they have moved to an on-premise or a local cloud instance that they run for open models that they're using and exposing to their employees, and they cut their AI costs nearly in half. So it's not just small companies. I think this is good for larger companies as well.
But does anybody besides me appreciate the irony of the fact that a cloud service provider is now telling you to use your local machine? Because Mike, there's an economic component. I think they change.
There's an economic component here we have to consider, right? " Right? " At some point, that costs AWS real money to support these.
There's real bandwidth costs, real virtual machine costs, et cetera. They can then have you push that to your local machine and you own that cost instead of AWS owns that cost. They still get you to use their environment, but they're not paying for your development time.
And so that to them is a big win. And a win for small companies, right? Absolutely.
It's going to be expensive to do on your own. And open source contributors. Yeah.
And John, I think this got covered a little bit more on the financial side of the world, but maybe it wasn't really noted as much. But AWS was telling the Wall Street folks that they're sold out for 2027 pretty much. They don't have additional computing capacity to give to folks, and maybe this is part of their way of pushing some folks off their machines- Yeah ...
to more production environments. That's interesting. Yeah, that's probably part of why they said that.
One thing, when Tracy was talking, I wanted to ask Tracy a question, and beyond the immediate utility for developers, does this project serve as kind of a real-world proof point for agentic software engineering? I think so. I'm not sure that this is the most complex piece of software in the world- Oh, for sure.
Yeah ... so maybe this is a place for them to do that. But I am wondering, am I going to see more and more of this?
I think I noted in the story, Anthropic built a C compiler the same way, and this is what gave AWS the idea to go do this. But there could be a wave of these. What do you think, Tracy?
Yeah. Well, the CD Foundation has a new AI SIG that is all about having these developer workflows and DevOps pipelines pushed using agentic. So what I'm hoping is that that team, and this is a call to that team, is to take a look at that, start testing it, and really take it for a test drive to understand it.
Because when it comes to just setting up a developer's workstation, the faster we can do it, the better we are. It's crazy how long it can take at a large company to make this happen. And setting up a Jenkins server anymore and building all the plugins is not something anybody should be doing.
So this is the next step. I really do believe that, and I'm looking forward to playing on it. All right.
Just one thing quick, Mike. I wonder also in the messaging from Amazon, I'm wondering if they're saying that AWS is simultaneously lowering the barriers to local cloud development, right? Yeah, in theory, I believe they are, and I think that's part of Tracy's point.
I got to tell you, if you go read the announcement, there's two paragraphs about what the thing actually does, and then there's 20 graphs about how they used AI to build it and what went into that. So maybe they want other people to emulate that whole motion as well. But Amanda, is the world fundamentally changing here?
What's going on from your... What do you think? " Yeah, it certainly lowered the barrier, but I think there's still a trust and security issue here.
It said that it was going to be maintained and updated completely with AI agents. I think there probably should still be some human in the loop verifying. I would ask all the more subject matter experts on this.
Can we trust it? I don't know if we're at that point to just leave it up to AI agents. Right.
Is it human in the loop, human on the loop, or human out of the loop? Which one is it? A human sitting on a loop.
All right. A human writing it. Yeah.
For all of my friends on red team testing in companies, this is a whole new branch of service that needs to come out. And Amanda, to your point, can we trust it? It doesn't matter what the environment is.
The answer is no. No matter who's getting it, who's offering a service, you can't trust it. The question is, what level of testing do you need to do before you can have a level of reliability that you're ready to invest in?
Mm-hmm. I also think, by the way, this is a huge PR coup for AWS, the first one to do it. They're talking about, "Hey, we make it all local.
We had developers in mind. " Mark my words, that will be a PR coming up at some point in the near future. That's an interesting- And I just want to say one last thing about the fact that they claim it to be open source and it's going to be managed by these agents.
They need to take the next step to make sure that they have community involvement and governance around this open source project. It can't just be an open source project that the AWS says, "Hey, look, this is open source. " It has to have governance, it has to have input, and it has to have community support.
Does it? I don't know. Yes, it does.
Absolutely it does. And the more that they embrace the community around it, the faster it'll move forward. 0?
I don't know what else to call it. We already are. It's already happening.
Every day we go faster and faster in how we're using AI, and AI's getting so much better at coding. So we are. It's happening already.
And all that AI is happening on our local machines, except for a couple of calls out to a cloud service somewhere. Or, in some cases, not at all, right? In terms of the agent discussion, if we can create boundaries with it, if it's securely bounded, I'm excited to play with it.
All right. Jack? But from an architecture point of view, in software development and software in general, everything old is new again, right?
0. 0. We've been going centralized, distributed, centralized, distributed compute for a very long time, and it's every three to five years at this point where we cycle back and forth.
So we've been doing this for a while. All right. So this is kind of like my basketball moves, they're so old, they're new again.
Is that what you're saying? Yes. Kind of similar.
All right. We need to let that die right there. All right.
Let me move to this third topic because it's Friday, so we like to get something that's a little more esoteric going. So, the United Kingdom has revealed that it's spending, I think it's £900 million, which I guess rounds up to, I don't know. Does that round up or down these days?
I don't know what the dollar value is. It's up. It's up.
All right. So, I decided that it must be real because if it was unreal, it would've been like a Trump number where it was perfectly rounded to a particular made up number, but that's a whole other issue. Yeah.
But it's interesting because what they're saying is that they need sensors and other collateral, I guess, or information to protect these undersea cables that we're so dependent upon now for networking, the internet services, shipping, and everything that goes with it. And we have seen, over the years, multiple reports of submarines acting suspiciously around these cables. So Jack, as you look at this, how big an issue is this going to be?
I don't know if it's a big issue. It's an interesting issue. We can geek out for a little bit about this, if you don't mind.
There's some cool technology here that, if you think about this, back in the 1950s, the Cold War era, we had Russian submarines chasing American submarines, American submarines chasing Russian submarines. And as we got better and better at the technology, we made them quieter and quieter. So to find these things in the middle of the deep ocean, we buried sensor networks.
And these sensor networks were originally a standard microphone like one of these, that you'd literally make waterproof, drop in the middle of the ocean, and listen for a submarine passing by. And we had long, long lines of this with something called SOSUS. And then we moved from regular old microphones to piezoelectric, which is basically a crystal that as you bend it or vibrate it, which is what sound is going to do, it gives an electrical signal, and you can detect smaller sounds or sounds from farther away.
So we got pretty good at this stuff. And now this company that the UK is dealing with, I believe it's a company called Optics-11. It's a Dutch company that has a variation of this technology.
Rather than using the physical bending of the crystal to detect sound, you detect the changes in the phase of light as it moves through a fiber optic cable. So that's pretty cool technology, and it means that it's much more sensitive, which means you can detect things farther and farther away. So the worry here now is twofold.
One part of it is, as you mentioned, there are not only submarines, but ships that we believe have been attacking fiber optic cables. And they attack the fiber optic cables in two ways. One is it's believed that they can basically put a tap on the cable and either physically cut it and put a tap there, or in some other way, get some of the data out of the fiber optic cables that are running across the sea floor.
And the other thing is literally just dragging an anchor or something across the sea floor and destroying the cable, and we've had outages that we believe were caused by that. So that's one fear, and this is to detect that type of stuff. The other thing that goes into this is the United States pushing the NATO organization and the NATO countries to actually start investing in their own defense rather than relying 100% on the United States.
And as a result of this, now originally NATO countries were required to spend 3% of their GDP on their own defense, and they have not to date, since the beginning of NATO, ever done that. So they are now starting to actually invest in more defense and specifically in areas where it is new technology against new types of threats. " And to understand how critical this is, there's been a lot of news recently in the last month or so about how reliant Ukraine is on the Starlink satellite communications network for the defense of Ukraine and what happens when they don't get access to that.
And so, the UK and the NATO countries are trying to invest in protecting their communication networks with some really cool technology. Oh, and Jeff, I saw you raise your hand, so I'm going to toss this to you. But we're also starting to see underwater drones.
It's not just submarines. These things are getting smaller, and they're more potentially just as lethal. Absolutely.
And I think to Jack's point, and Jack, you did a great job of giving us 50, actually 70 years of history with that. " Come on. When he started talking about quieter and quieter subs, that's exactly what came to my mind.
But to your question, Mike, about drones, that is really the next step. We are at the point now where it has to be more than sound because now you can get nearly silent vehicles, in many cases very small, that can get near these cables and potentially create damage. So, the technology Jack's talking about can also take a look at really a combination of what is the equivalent of a new sonar.
So even if you're not getting sound, you're getting detection of non-biologicals that are in the area. And even then, they can filter that. So, I think it is a great step forward.
It's clearly what we need to be doing, and I think you're going to see even more really cool advancements in this area over the next five to 10 years, without question. Because those cables are critical to running the world, and without them, it's not an email outage, it's a severe issue that takes more than a day or two to repair. So the proactive protection of them is vital, and more and more of these tools, I think, are going to start showing up, and they really are going to be cool.
And then you're going to be doing the same thing in space, by the way, too. That's next. Sure.
We like to play on Friday something that we've discussed that will keep John up at night. So Jack, are these cables going to be the first thing to go if the war gets hot, and is this the thing that this is critical infrastructure and probably the very first thing that somebody hits? Yes.
Very likely these are very prime targets because whether or not, as I mentioned, Ukraine is of course depending on this, but we just had a conversation about a small number of people who can't communicate when Microsoft Outlook goes down. Think about what's going to happen when the entire European continent can't tell us about how to cook coq au vin in the United States. If we can't get our TikTok and our Netflix, and the world is going to fall apart now.
Like I said It's a phone, but it's also a supercomputer, and people need their TikTok. That's an ominous sign. Jack, that's an ominous scenario that I was thinking about going back to the Microsoft conversation.
We live in scary times. So Amanda, let me pull you in here for this because there's a certain irony in this conversation as well in my mind. So we, the United States, have spent the better part of the last two years telling Europeans to spend more on defense and carry their own weight.
But I think the assumption underneath that conversation was that they were going to be buying more US stuff to accomplish that goal. But as far as I can tell, every time I look around, the Europeans are doing something with their own tech, and whether it's from the UK to Dutch. And I almost think we're going to see some sort of technology renaissance coming out of Europe because all that money that they are going to spend is going to be on their own industries and their own plays.
What do you think? Well, I have seen a lot about them breaking away. I can't foresee the future, but it does seem like they're not doing as much investing, and they're talking about their own things.
I really don't know what to say to that. Honestly, I'll kick it to somebody else. All right.
John, you want to take a shot? Well, there was something that was kind of interesting, kind of a sidelight, is tech companies and the governments are looking at ways to move computing infrastructure offshore. I think there are these tidal-powered underwater data centers off the coast of Maine and Alaska.
And I think it brings to mind, I think Frank who's already wrote the story, like this idea of the ocean floor is becoming an even more critical and maybe even exposed hub for global data and power. That was the takeaway I had. Speaking of ominous.
Well, John, if I might, there's another data center stuff that's about to be built in space, and that relies on, A, easily intercepted, and B, easily interrupted communications by radios. So that's another thing. You're going to put a big data center out there and you talk about downtime and lack of ability to repair it instantaneously when something goes wrong.
Yeah, they have about 10 minutes every hour, I think, to be able to make any updates or fix something. Yeah, I'm sorry, Tracy, just really quickly, is this a consideration given the state of data centers of where they may go or maybe they've displaced or there's a diversity of where they're based? I'm not sure I believe in data centers in space just yet.
I believe that there'll be processors on satellites, but I'm not buying the whole data centers flying around in orbit just yet. But I think there'll be more of them under the ocean than up in the space, but we'll see. Well, then, the amount of money spent on this particular project was just 90 million pounds, which is about $140 million.
And it's really critical, right? If it's that critical, there should be more money being put into it. And when I read the article, I said, "Okay, so if they sense something, so what?
" Somebody's messing around out there or an anchor got dropped. Now we know it's broken, but we already knew it was broken. So the sensors have to be able to do more than just report.
It's kind of like knowing a vulnerability exists in your software, but you don't know where exactly it is. Oh, that's right. So it's hard to fix it.
Right. So my disappointment in reading that, and I understand why it would be coming from the European Union. I believe that when we had Trump up there with oligarchs from all of the big data companies, the world got nervous, and they should have.
Yeah. And the stuff with Elon Musk and Ukraine, they got even more nervous. So how do you go protect what is the majority of our communication, which is not satellite, but it's the cable laying on the bottom of the ocean?
I just didn't think it was a big enough spend. I think part of that is the unique dynamics of the UK. So I agree that they recognize the fact that they have issues with countries like Russia that are coming, buzzing them by with airplanes and all this other stuff and ships.
" And then they start screaming about it. So I think the UK always underplays how much money they're actually going to wind up spending. It's the anti-Trump number fixation, right?
They always lowball. They tend to lowball. Well, we also need to remember that putting a huge investment early on, you run the risk of are you backing a loser or not?
By moderating your investment, you're moderating your risk. And if you go back to computer networks, intrusion detection existed before intrusion prevention, and it had to, and I think this is the first step. And it shouldn't be that difficult if you think about it.
It could be expensive. But if you can coordinate with every tail number supposedly is recognized as it's up in the air for any aircraft. The same with any ship that's on a registry and their satellite imagery.
" I think we're in a much better position to do that now. I think- I think- Go ahead. I was just going to say, well, Jeff and Tracy, I think both are right to sort of drag in the concepts of cybersecurity here.
And we always say in cybersecurity that the very first problem you have is visibility, right? You can't protect what you don't know about. You can't defend it if you don't know about it.
And you can't defend against attackers you can't see or find or don't know what they're doing. I think this is the first step. Right now, does the EU need to invest in their own defensive and offensive capabilities in the sea?
Absolutely. The United States Navy members something like 450 combat ships, and the British Navy has 60 combat ships. So there's a lot of investment to go.
But investing 90 million pounds in a sensor network is a lot different than investing 90 billion pounds in a single ship. I also think that they are playing the president, and it'll go something like this. " And then the British are going to sit there and say, "Okay, great.
" I'm sure it's crossed their minds because they think that way. It wouldn't be the first time he was played. Right.
So there we go. Hey, it's been an interesting week, and I hope you all survived it. And if you couldn't work or you couldn't send emails, I hope you did something enjoyable with that extra time that was handed out to you this week.
I'm sure that we will all pay for it next week when we get those inboxes, and there's 3,000 messages in there from one day or whatever it is. So, no good deed goes unpunished, as they say. But in the meantime, enjoy your weekend and don't open your email is what I'm going to do.
There you go. John, Amanda, Jack, Tracy, Jeff, thanks for being on the show as always. " Please stay tuned for the replay.
And of course, we have a word from our favorite sponsor, otherwise known as us.