NVIDIA $2B AI Play, Blue Origin TeraWave, & More | Tech Field Day News Rundown: January 28, 2026

Obsidian portal protection, VMware licensing issues. Nvidia invests even more into Core Weave. Quantum open source moving from Kuda to rock Meta is financing some fiber Andro updates for Understanding and Blue Origin catches a terror wave in this episode of the Tech Field Day rundown.

Hello everyone and welcome to the Tech Field Day rundown. It is January the 28th and we hope that you had a nice hearty breakfast that included blueberry pancakes. 'cause that's what day it is.

It's National Blueberry Pancake Day. It's also National Lego Day. And I know that all of the listeners on the rundown love Lego sets and they would love to be building one right now.

But hold on a minute, 'cause we've got some news headed your way. We also have a brand new cohost headed your way. Joining me is Mr.

Jay Coutre. Jay, welcome to the show. What is your favorite Lego set?

My favorite Lego set is the one I don't stand on or step on when I'm going through a hallway. That's my favorite Lego set. You know, that's probably a good one because as painful as stepping on Legos would be, what's even more painful is missing an episode of the tech field.

A rundown because we have some great news that we wanna share with you and get some commentary. And we're gonna start off with a story from our friends over at Obsidian Security because they have expanded their SaaS security platform to cover third party integrations, giving organizations better control over how applications connect and share data. New features include enforcing least privilege policies for who can approve integrations and improve supply chain breach notifications.

The updates aim to reduce the risk of breaches spreading across SA environments, especially as integrations and AI agents increase attack surfaces. Jay, do you think it's important for companies to start thinking about their SaaS security supply offerings? It is.

If you've ever, ever had something offered to connect something to something else, uh, as part of your user experience, now imagine that being turned up for everyone. When you roll out some of these new platforms that are seeking to connect any to any many, to many, and what you end up with is this potential for a, a security Swiss cheese or, or a fond do depending on what your perspective is. My thinking is that anything that can raise visibility and actions that can be taken on behalf of a security team that's gonna be responsible for this, uh, it's it's that many to many that that dependency mapping, this is really key for us to move forward with a de-risked, you know, uh, safe quote unquote version of adoption of things like ai.

My suspicion is that the SaaS companies will just, uh, counter that with making it even less, uh, you know, uh, cumbersome, uh, release more friction, um, from the current experience and increasingly become frictionless. So I think it'll be a little bit of a spy versus spy, uh, story as the SaaS does, you know, tries to pull friction out of that, uh, sequence and then security teams sort of try to wrestle it back into place. Um, so I'm curious to see how this plays out maybe a year from now.

Uh, but for, for this, uh, update, I think least privilege, no one's going to object to that. I do think there will probably be some tickets that will go into the, up into the right column. Especially as you put some of these controls in place, even if it's, you know, a sensible set of guardrails, I still think there's gonna be an outcome where it flows back onto some of the operations team side of security to deal with those tickets to explain why we're not gonna allow you to connect your luggage, uh, for example, to our enterprise, you know, resource management systems.

So that's, that's my perspective based on what I've been reading so far. And if you wanted to pivot from the world of all things connected to everything, uh, there are those that are part of the legacy Broadcom VMware story that are starting to learn a little bit more about what's underneath the cover of some of these contracts. So recently, uh, fidelity and, uh, Broadcom have started to settle a legal dispute.

I say started to settle because it's, I think it's settled, uh, but I also know that there can be sometimes some back and forth. But the whole idea is around access to VMware software, what that means for, uh, Fidelity's own operations for them to continue, uh, without disruption. Probably many listeners today use or have used Fidelity as part of their overall, uh, compensation, uh, within a corporation that they've worked for.

And so what this really means is there's gonna be these, again, to use that word friction. There's friction now between these prior customers of VMware that are now effectively Broadcom customers. And so each time that happens, it really highlights where happens when you bundle a certain way, then someone acquires that company that you would use that bundle with, and the acquiring company decides that that bundle shall no longer, no longer be honored.

And so I think that this is really an interesting one where there's gonna be major licensing changes that'll occur. I don't think this will be the last time we've heard of a similar story, but, uh, what are your thoughts? I think it's interesting because Broadcom really wanted to go after what I considered to be that load-bearing customer set, right?

Companies like Fidelity can't just rip and replace software willy-nilly because they're in a highly regulated industry and their operations really can't go down. And we kind of suspected this when the announcement was made all those years ago, right? Like Htan and Broadcom said the things they had to say, we're not gonna raise prices, we're not gonna cut r and d, we're not gonna do all of that stuff.

But we also kind of knew that for what they were paying for the company, they had to raise prices. So the question is, how high can they raise prices? Uh, there's, you know, the, the, the maximum of, uh, raise the prices until somebody squeaks and then turn it down maybe 10% and that's your new norm until whatever.

And I think that the squeaking in this case was the fact that Fidelity said, that's too much. We're gonna take you to court. Or, you know, I think it might have even been VMware taking them to court to try to figure out exactly what's going on here.

But we've also heard from bigger companies, like at t they were kind of looking at this and, and basically what they're trying to say is, we know we have an agreement and we know what the agreement covers. Does the agreement require me to uninstall any patches that were installed after a certain date that I can no longer update my software? Because that was one of the things that companies said that they were doing, is they were getting a cease and desist from VMware by Broadcom saying, you can't install any more software.

I'm sure that these companies are investigating alternatives, but the problem is, is that you've got a sunk cost in the licensing versus the cost to acquire a new solution to get your people trained on it, to get it installed into make sure that operations continue and to borrow a line from Fight Club. If A plus B plus C is less than the cost of a recall, we don't do one. And that's the thing.

If the amount of money that Broadcom can get out of these customers is less than the cost to do all these other things other, then they're going to stick with it. But I promise you the investigations are ongoing and as soon as that break even point shifts, we're gonna see customers that are starting to move off of that. I also think that the next time that these contract negotiations come around, there's obviously going to be a desire for them to raise prices again.

And that's going to play into the math of how likely these large organizations are to stick around with the software. And I'm not just picking on VMware by Broadcom. Any company would be making these changes.

Nvidia has invested $2 billion into Core. Weve becoming the Neo cloud providers, second largest shareholder in deepening its role in the global AI infrastructure. Buildout.

This investment will fund massive data center expansion and tighter hardware software alignment, which reinforces NVIDIA's influence over how large scale AI workloads are being deployed. The move signals that control of AI infrastructure, not just chips, will be central to the competitive advantage as AI scales. Jay, we've seen that Nvidia has invested a lot of money into Core Weave, and likewise, core Weave is using a lot of Nvidia hardware.

Does this mean that you're kind of, have you, you've created this situation where Nvidia and Core Weave are more or less dependent on each other to continue to thrive in this market? It is interesting, Tom, because if I go back in time to when there were other large juggernauts of infrastructure like a company called Cisco, there was this manifestation of a gray market. com era, everyone bought a bunch of Cisco equipment and then allegedly some of that Cisco equipment may have made it into a secondary market.

Um, I think there's some similar parallels here with what's happening with these amazing processors that are created by Nvidia. And I think finding the right market for, here's the current current generation, here's the recently current generation, and here's the things that oddly enough, if you have enough access to it through a Neo Cloud provider, it'll get the job done. And so I think that this, this notion of where people are doing discrete amounts of training, increasingly leveraging existing models and doing greater amounts of inference, I think it'll be fascinating to see almost like a a a kind of swinging supply chain story of what was yesterday year, but still is fit for purpose to accomplish certain goals.

I think these Neo Clyde providers like Coral, weve are gonna be a part of that story. What I'm looking for is are there additional investments to other Neo cloud providers besides cor we, that also might be part of the story. And I think we've, we've already seen, there's, there's a plenty of writers out there right now that are taking kind of a contrarian view of all things related to the Oro borrows of all things in the GPU, uh, kind of, uh, value chain.

But, um, I think of it like this, I would not be shocked if there weren't more deals like this and, uh, possibly, uh, other companies, uh, that are other neo clouds that compete with if you'll, uh, with core beef, but are ultimately complimentary and lifting to the part of the Nvidia business that can't necessarily afford the latest and greatest from Nvidia, but definitely can benefit from a long tail or lineage of prior products. So, uh, with that, I'm gonna say right now, if you are thinking about Quantum and all things related to the next generation of compute away from classical now to Quantum, Microsoft's actually announced that they've got a quantum development kit. So if you can imagine, I don't know if, uh, Satya Nadal will do this, but, uh, maybe he does a developer's, developer's, developer's dance around all things quantum, quantum, quantum.

But this looks like it's gonna be an open source set of tools that are gonna allow for anyone that's currently using kind of like the integrated development environment experience of Microsoft today vs. Code using go, uh, GitHub copilot to truly come in and harness and make use of all things related to development in the quantum realm of computing. So this is gonna be about a workflow story.

I think it's more than just saying like, we have, you know, hardware. I think this is more of enabling that developer audience to participate in what is arguably a fairly recent development in alternatives to the classical computing we had talked about earlier with GPUs or cpu. So I think this also means that if we look at the near term impact, there's probably going to be some amount of hardware that's available for Quantum.

But much like the initial scarcity of all things related to GPUs, when everything became, oh, it's beyond machine learning, we're moving to ai and suddenly you had supply chain, uh, constraints or supply chain constraints, I think there will probably be a modicum of hardware available, but getting developers easier, more, uh, fast, uh, ready access to the ability to touch that hardware is very interesting. So if you go back and look at the original D-Wave and some of the other TTI stories around there where they've launched with a completely amazing website and kind of get people into it, Tom, what's your perspective on like kind of crossing over from the scarcity of hardware into now enabled developers to maybe create even more scarcity around it? There comes a point in every technology lifecycle where the early ideas about what it's capable of doing are sky is the limit.

There's no, uh, you know, there's no idea what this could happen to do as we move along. And then usually around the series A funding round, uh, some investor steps in and goes, yeah, but how are we gonna sell it? And the problem becomes those pie in the sky ideas have to have a return on investment, otherwise you're not gonna get a series B round.

Quantum computing has been around for 30 plus years, even in the cultural zeitgeist as we know it. Um, I, again, I I referenced sneakers. All you gotta do is go back and watch that, that's quantum computer.

The problem is that programming on those things is not traditionally straightforward. You have to have a different kind of methodology for thinking about how it's gonna work. Well, normally if I'm gonna learn how to program on something, like when I learned how to program CI had a C compiler that I installed on my computer and I went in there and I messed around in a c compiler and I wrote code.

And if it compiled, great, and if it didn't, I had to figure out what's going on with it. And that was on my 4 86. I can't do that on a quantum computer because those things are expensive per hour of runtime.

What I need is a development environment that will kind of desk check my code and make sure that what I write is runnable on those systems. And that's what Microsoft is offering to do. And they have taken a page out of the Nvidia playbook.

'cause if you remember years and years and years ago, Nvidia was like, oh, well, if you wanna do things on our graphics accelerators that are not texture mapping, we have this language called Cuda. And if you wanna learn how to write Cuda, and we have these software models that'll allow you to build programs, you know, they'll run on GPUs. Fast forward to today, almost everybody's writing their stuff in Cuda because that's the kind of become the, the standard for how to write on GPUs.

So if Microsoft can really open source these tools so that everybody has access to them, and they are training people how to write their code to run on quantum computers, then in a few years when quantum computing is cheap enough to be more, uh, dispersed throughout the industry, people are gonna write the way that Microsoft taught them to do it. It in a way, it's kind of the freemium model, right? We're gonna give this away because we know it'll get more business down the road.

I I hope that, you know, we, this is a, a, a noble goal, but we'll see if people latch on to using VS code and GitHub copilot to do all these things. Or if they kind of reject it in favor of some other approach as soon as they had more ready access to the hardware. Speaking of Cuda, a developer used anthropics cloud code to port Nvidia Cuda code to AMD's Rock M platform in about 30 minutes, which raises a lot of questions about NVIDIA's longstanding Kuda lock-in.

While the demo suggests that AI assisted tools could lower the cost and risk of switching GPU platforms, experts caution that complex performance critical code may still require deep manual optimization even. So, this example shows how agentic coding tools could begin reshaping GPU software ecosystems. Jay, I know I just said that kind of kuda is the way that people want to go, but what about people who are looking to shift that off onto a MD or, or other options?

Do you think AI and specifically Claude can help them rewrite their stuff and be like, ah, yeah, it mostly works. Tom, when it comes to commitment to Cuda, I think they're gonna be people that'll stay all in, uh, on the platform native with Nvidia. I also think they're gonna be those that will take somewhat contrarian positions trying to either, uh, find an alternative, uh, like this example of someone doing unique porting for an A MD platform.

Uh, but I also think those hyperscale cloud service providers have, uh, made significant investments. Specifically I'm thinking of Google, uh, with their, uh, you know, now seventh generation version of Ironwood on their TPUs. Uh, there's obviously the, uh, graviton family of, uh, processors from AWS such as, uh, traum, infra, et cetera, uh, that offer an alternative, uh, to the NVIDIA processors and, and even Microsoft.

Uh, if you think of what they may or may not be doing related to Maya and Cobalt, which are probably some older stories people could look back on, eventually they're going to get there. And I think that what'll happen is, um, in addition to other players, you know, arguably open ai, um, grok, you know, coming up with their own, uh, hardware platforms, there will be some competitive pull. But I believe today, CUDA is, uh, effectively defacto standard.

And so these kinds of, um, enablers, uh, it's possible. I go to my, my, my ables enables, you know, there's what's possible. We've shown it's possible, then there's what's permissible.

You know, are we really gonna set fire to capital to maintain this, you know, fork if you'll, of how we might do this on an alternative platform? Or is this now sustainable? Oh, now is it repeatable?

Oh, have we reached advisable? And I, I think that'll be really interesting to see if there is any kind of a bifurcation or even possibly a splitting of different, uh, you know, if you, if you will, different, uh, types of development stacks. But today, I, I do say that, um, you know, with the entirety of all the GPUs that are in market from Nvidia, the secondary markets, the neo clouds, it's gonna be easier, I would say, to get access to some of this traditional Nvidia, uh, uh, approach, which I think does support Cuda long term.

They were first, um, we could argue whether it's best or not, but ultimately getting a more, uh, vibrant ecosystem will raise all boats. Um, what's the competition to Cuda? And I think that's a, a really reasonable question to always be asking, is there something that's coming around that's more viable?

And I would, again, look back to, uh, the stories of open source. Um, there wasn't always an open source alternative to the everything that is Windows and Wintel, so maybe we'll see. But for right now, I, I do think that Cuda has a first mover advantage and an absolutely dominant amount of silicon that's been put out into the field.

Uh, as you mentioned earlier, the current scarcity of hardware doesn't necessarily mean there will always be that scarcity of hardware. And as these other competing platforms come in from the cloud service providers and some of the actual, uh, truly vertically integrated suppliers of AI services, uh, they'll go from probably, you know, uh, silicon to software to services, um, all the way up through the asses. And my suspicion is, is that, that that competition will be good for everyone.

But, uh, today I would still place a bet pretty heavily on kuda related to some of the stories we've heard about building ai. It's interesting that Meta has now agreed to pay Corning apparently up to 8 billion, uh, $6 billion, sorry, maybe it's gonna go to eight, but apparently it's 6 billion through 2030. And this is related to fiber optic cabling within their data centers that are doing these AI workloads.

What we've learned is that the soaring demand for fiber is apparent as these AI workloads require far greater amounts of connectivity than the traditional cloud computing that was being used to fill those, uh, data centers previously. So if you think about this where, uh, Corning stock is, um, way up, uh, arguably, uh, around 75% over where it was last year, this expansion of production and development into ever, ever denser, um, AI specific fiber solutions means that all of the hyperscalers, whether that's meta, uh, Nvidia, Google, Microsoft, open ai, and others, it looks like this could just continue to grow. Now what's it displacing, uh, copper, most likely.

The question is, is is the partnership really underscoring how the AI infrastructure is spending or reshaping a backbone for global data centers? Or is this just a one point in time, much like a pendulum? Okay, we're swinging over to fiber, maybe we're gonna come back to some other kind of a alternative interconnect.

What are your thoughts on this, Tom? Is this, is this seasonal generational revolutionary or evolutionary? I think this is generational because, uh, for those of you, uh, that watch the run who may not be familiar with the Pentagon Pizza Index, uh, it is this thing that somebody came up with that said, Hey, we noticed that when the pizza places around the Pentagon get super busy late at night, there might be an indicator that something is going on that we might not know about right now, but there's a secondary indicator that says something is afoot.

To me, this deal is a secondary indicator of the plans of these build outs because one of the things that people think about is like, oh yeah, you know, uh, you're gonna build a new data center, you're gonna need, uh, walls, and you're gonna need a roof and you're gonna need floor tiles. You're also gonna need a lot of power cables. You're gonna need a lot of fiber runs, you're gonna need a lot of air handlers.

And one of the things that I learned right after I started doing this job when Steven FoST and I were doing some, uh, work in New York City, was that a lot of analysts on the supply chain side of things can suss out that something's going on based on orders for components that people wouldn't normally be able to put together. So like for example, if a company, oh, I don't know, went out and bought the entire manufacturing capacity for, uh, crucial, uh, high bandwidth memory systems, that might be an indicator that they're gonna do something with that, like build a lot of servers that require HBM, but in and of itself, that might not necessarily be something because there are such things as strategic acquisitions where I'm not buying it for me to use, I'm buying it so somebody else can't use it, and maybe I can sell it back to them at an inflated rate. But when you have an additional secondary indicators like I'm buying more fiber because I have to run this through my data center, that tends to tell me that these people are, are wanting to build up.

And the fact that you brought up copper is a very important point. These AI data centers really cannot run on modern copper infrastructure. I mean, they can do it over very short distances like server to server, but the majority of the communications that are happening, whether you are using ultra ethernet env link, um, spectrum X, these have to be copper.

I'm sorry, these have to be fiber. They can't be copper because the 800 gigabit interconnects, I mean, they run over fiber for a very specific reason. Not only that, but like the number of DSPs in these modules are insane.

I actually wrote something about that on my blog this week where I was talking about the fact that you wouldn't believe the amount of heat that gets generated from a modern switch simply because of the, the processing power that has to be stuck inside these fiber modules. But that also means that you have to have high quality fiber runs in order to make sure that your light is passing and it doesn't eat up extra DSP resources. And who better to go to than Corning?

I mean, they developed Gorilla Glass. The reason why the modern mobile infrastructure looks the way that it does thanks to the iPhone and the the Samsung phones is because of Corning's ability to develop things like Gorilla Glass, it's hardened against all kinds of, uh, impacts and things like that. 6 terabits, and you don't wanna buy that plant infrastructure and then have to rip it out in two years because you need to move up to a, you know, different laser wavelength or something like that.

So I think that this is meta basically saying, we're putting our money where our mouth is. We really are gonna build out these data centers. Let's give it another year before you know, when all these things start shipping to make sure that it's actually gonna get built out.

Moving over to another AI provider, Andro has released a new constitution for its AI assistant claw, which is shifting from strict rules to teaching the AI to understand why it should behave safely and ethically. The update emphasizes helpfulness clear limits on harmful actions, and even considers Claude's potential consciousness made public under a Creative Commons license. The Constitution highlights transparencies and, uh, positions claw as a safer, smarter AI for businesses, no word yet on if they've incorporated the three laws.

But I, Jay I have a question. Can we teach an AI how to be ethical or is it going to assume the ethics of the people who have created it? Oh, geez, ah, AI and, um, I, I'll, I'll go back to one of my p prior, uh, uh, uh, prior writings.

I, I, I know people talk about stem, you know, the science, the technology, uh, engineering, mathematics, sometimes they put an A in make it steam for the arts. Um, I did something, uh, called a steam, which is where I'm injecting ethics and empathy, um, into stem. So, uh, what by doing, so you, you get the clever acronym, obviously, but, um, what I'm kind of wondering here is if the goal is to create or publish a policy, and if that policy is to be adopted, uh, you know, this, this notion of, you know, coming under a Creative Commons license, is it, is it really, is it really effective at making its way into an organization where there may be particularly, uh, uh, proprietary considerations that might run afoul of, uh, of a creative, uh, commons license?

Um, the other question I would have is, does it really mean that the ai, uh, in this instance around what is the appropriate use of ai? Is it, is it inextricably linked to anthropic use? Meaning like if I'm using, you know, a code or, or other, uh, anthropic uh, models to come, is is it designed in mind to align to that policy?

Or will this be like a lot of early initiatives where let's launch a fancy website, everyone feels good about it, let's all kumbaya for 15 minutes and then immediately go to the next adventure in, uh, you know, how we grow the next market? So I, I do wanna understand how durable, uh, this would be. I, I think there is a cliche of simply open sourcing something or putting a, you know, a a a clever license on top of it.

Everyone feels good about it, but I do wonder what's the, a appetite or the passion for adoption and putting this into actual practice? I am very, very concerned as a technologist about what will happen as smaller language models make their way into computer vision systems, which may be connected to other things that have real kinetic impact for alive things like humans, for example. And in a, in a prior update I was watching earlier this week, I think it was, uh, uh, Steven Foskett had mentioned like, how, how it's happening, you know, as these, as you go towards the edge.

And one of the questions I asked is, how much Ram is gonna be allocated on those devices to think about that constitution? Is the constitution loaded into, do we believe that the Constitution fits in enough to justify the unit cost of that thing, thing that will be very much a part of the kinetic world around us? So, um, I appreciate the, the, the forethought, but I am absolutely, um, wondering what the uptake and adoption of this is going to be.

So I hopefully don't sound too sour on it, uh, that it was created, but I just wonder if it's just gonna be abandoned by the roadside. Like so many of their great ideas, what does they say about the road to hell paved with good intentions? So let's hope it's not that.

And on a, uh, slightly more, uh, light note, blue Origin is actually going to expand, uh, beyond rockets and invoke something called Terra Wave, which is gonna be a new satellite network designed for high capacity enterprise grade connectivity. Does this sound familiar? It should, because this is not something we're, we're talking about starlink, we're not talking about AWS Leo, formerly known as hyper.

This is a new hybrid constellation of over 5,400 satellites to start that will provide multi terabit connectivity to various, uh, supplements of other earth-based terrestrial networks. This means that you could have redundancy for your data centers. If any of you ever remember why there were satellite dishes on tops of data centers, welcome back.

The difference is now it's gonna be a lot more performant. And so when you think of this for cloud to cloud links, uh, everything related to business continuity, disaster recovery, imagine the new diagrams that would be available to us when we start drawing these new architectures of what's possible with infrastructure going from just purely terrestrial going beyond radio. Now thinking in terms of extraterrestrial connectivity.

So by targeting business with these high bandwidth, strict security and avail, uh, reliability requirements, um, what you're gonna see is Terra Wave is definitely gonna compete with starlink. Um, it's probably gonna compete, uh, while in this mission critical focused area around enterprise applications, those kinds of workloads. So to to, to, to afford terra wave is to, uh, is to mean that you are most likely a very large, uh, corporate interest of some kind.

And so I would, I would say this is very different than like, I'm gonna put a little, you know, starlink mini in my backpack and go out in the woods and be, you know, ever connected to the world around me, um, while I'm in the woods, which is always kind of odd to me. Why would you go into the woods and then wanna have a starlink mini with you? But point being is that this should be deployed sometime around late 2027.

It it means that promising much, much faster than current geo latency solutions today, and more importantly, highly symmetrical as opposed to asymmetric. So symmetrical being really important for these enterprise applications where again, you're, you're thinking about new topologies. So having known this, and, and by the way, I'm someone that actually has, uh, taken the CNCS landscape too, um, uh, taken a fork of it and I've loaded the entire space value chain that I've been interested in and I've been tracking all these layers.

I can't wait to add Terra Wave as a logo into that near Blue Origin for all these other, uh, low earth orbit and geosynchronous style, you know, providers of connectivity. But Tom, do you think this is like, is it getting too busy up there? Do we, do we have room for another, I mean, space is very large, but do we have room for a tear wave?

Do we have, do we have room for more competitors in the space or is it really gonna be like three companies that dominate much like in the US we have with the at t you have the Verizon, maybe you have Sprint Legacy. What do you think is gonna happen once we're up in orbit? I guess it depends on whether or not the guys that run Google decide they want to get into the space thing.

But I believe you mispronounced Terra Wave, oh, I believe in Silicon Valley it is pronounced Chaching Chaching Chaching. Ah, Because I I missed that. No, well, when you think about it, that Blue Origin does not have the government contracts that SpaceX does.

So what they've gotta do is they've gotta figure out a way to pay for all of those rocket launches. And for, you know, people at SpaceEx, it's real easy. You throw a constellation of those little mini SATs up there and then you sell, uh, starlink to people who hike in the woods with a starlink mini or my brother who put a starlink dish on the top of his car so that when he's driving across the country, he has internet connectivity, but that doesn't pay the bills all the time because the people who are on the consumer side of it want the best deal for the money.

And they're not going to overpay for things they don't care about. I don't care if the starlink constellation is unreachable for 18 minutes a day when I am, you know, on the side of a mountain in Colorado, I'll just go do something else for those 18 minutes. But for businesses that need that kind of critical reliability that are using this as a backup link for, uh, you know, mission critical things like think oil drilling, uh, think, uh, mobile hospitals and things like that, they have budgets for these kinds of things, which is why it's pronounced Chaching Chaching Chaching with a comma and some more cha chings after it.

What's going to happen is, is that Terra Wave will be positioned as the Cadillac of, uh, connectivity. No, your Apple Watch is not going to be on Tarawa. No, your iPhone will not be using Terror Wave to send videos of you scaling Mount Everest to your buddies back home.

What it is going to be used for though is the medical personnel at Base Camp that are trying to do remote consultations with doctors for severe altitude sickness and things like that, where the amount of money that you poured into the expedition paid for that. And so this is how they're going to justify launching, you know, 5,400 satellites. You're right, space is big and we've seen that there's already a lot of stuff up there, but one of the things that we've learned from the Starling experiment is that those things go out quite frequently because it turns out not only is space big, it's also very hostile to all forms of technology.

And so what'll eventually happen is those satellites will start dying off, they'll fall back into orbit, they'll burn up. And so there's gonna have to be this constant refresh cycle of putting more of them up there. And the idea that starlink had was we can offset those launch costs by carrying other payloads from a commercial space exploration problem when you're Jeff Bezos is that you are the second person to the dance and now you've gotta figure out how to make that work.

And I think, think that what you're gonna see is that once this becomes a big deal, you're gonna see him kind of moving into the, the yacht people, right? Like this was a thing that I learned late last year, is that all of those really awesome luxury yachts that you're seeing being built all have a starlink terminal on board because the people who live on them, not just visit that live on them, are literally working out of them. Like they have a cabin on the boat that is their office and they all have star links and they are all able to communicate and do voice and video chat and all that other stuff.

I think you'll see terror waves starting to take over in the highest end of those because they are always on high throughput. I don't think that this is going to be the start of them using it for things like orbital data centers, because I still don't believe that orbit or orbital data centers should be a thing or will be a thing. But rest assured that if anybody is gonna figure that out, it's gonna be the guy who is bald and owns a lot of stuff and probably has a thing against Ians.

Oh wait, no, that's the other guy. Eh, one thing though that I think that you will be excited to hear about is all of the cool stuff that we have coming up here at Tech Field Day, because the first thing is happening right now, as a matter of fact, we are doing AI infrastructure Field Day this week. That's where my normal co-host Alistair Cook is.

He's out in sunny Silicon Valley. He's on the right side of the earth for once, but he has got some great presentations that you're gonna wanna check out. com and you can see a link to all of the presentations with the schedules and the presenters and probably some cool pictures and things that are going on with the delegates.

And make sure you're tuning in and use the hashtag AI A-I-I-F-D on Twitter X wherever you go, uh, to uh, learn a little bit more about that. And then coming up, uh, in March, March 11th and 12th, Al will be back in the US once again for Cloud Field Day 25. He will, uh, have a great lineup of presenters.

com, uh, to, uh, en entertain you after the long winter is over. And, uh, March has come in like a lion, uh, before it goes out like a lamb though. I'm gonna be in San Francisco for our SAC this year.

Uh, that is the RSA Community Conference. Uh, we're doing Tech Field Day Extra there for the very first time. And we have great presenters.

We have Veeam, we have Object First and we have Commvault, and we're gonna have more very soon. com to check that out because we'll also be listing some delegates that we'll be attending there and, uh, some more fun stuff going on. Speaking of fun stuff, Jay, if people wanna check out some of the stuff that you make, where can they go to learn that?

It's pretty simple. org and if you're really interested in space, click on the right logo with a little satellite and it'll take you to the space value chain. Awesome.

And if you're interested in space, may I suggest that you go to YouTube and look for all of these Space Spheres, dialogue from portal two. There's like 45 minutes of it and it involves space. We wanna thank you very much for listening to this episode of the Tech Field Day rundown.

Remember, we publish all of our episodes on Wednesdays on YouTube or in your favorite podcast application. The rundown is also being streamed on Techstrong tv, and you can catch me and lots of other people around here on other Techstrong and future and group programs, including all the podcasts that we record on a regular basis. We're gonna be back next Wednesday to talk about all the IT news that was in the week, that would technically be this week.

But until then, for myself, Tom Hollingsworth, Alistair Cook, Jay Rell, and everybody else who does a, an amazing job to make the rundown happen. Thank you very much for tuning in. We'll see you all next week.