Nokia & AWS Unveil US AI Infrastructure Investments | Tech Field Day News Rundown: November 26, 2025

Nvidia has cloud aspirations, more like our don't, Dino NATO is googling some stuff. Amazon launches some satellite antennas. Splunk donates open telemetry injector project, sudden catcher in space.

And we're gonna dive into the new AI infrastructure investments from Nokia and AWS in this week's episode of the Tech Field Day Rundown. Hello everyone out there, and welcome to the Tech Field Day rundown. Today is the 26th of November, and if you are listening to us in the United States, I hope you're not doing it at work, because of course, it is the week of Thanksgiving, which means that most people have already decided that Wednesday and Thursday and Friday are national holidays, even though it's really just Thursday, but it is National Cake Day, no lie for everybody that celebrates.

And joining me, of course, is my wonderful co-host who doesn't celebrate Thanksgiving this week. Mr. Alistair Cook.

Al, it's good to see you again, And it's always good to be here. And it was great to see you in person last week. I'm looking forward to some of the interesting news we have coming up.

Yeah, it should be a really fun time because of course, everybody's trying to cram in their last little bits of important stuff before everybody then goes to eat, uh, gravy that has gits, you know, little bits of Turkey in it. Uh, we're gonna go ahead and start off with, uh, some big money news because Nvidia plans on spending $26 billion over the next six years to secure Cloud GPU capacity for its AI projects that would make it one of the largest cloud infrastructure buyers company is partnering with providers, the names that you've heard before, like Lambda Core Weave and Oracle. And they're gonna create a system where they're both the supplier and the customer.

The move comes among, among strong AI demand regulatory uncertainty, and frankly, limited GPU supply highlighting how even leading chip makers are gonna have to lock in compute resources to say competitive. Al, is it weird that Nvidia is not only selling them the GPUs, but also buying that capacity right back from them? It's a weird continuing trend that we've seen.

So this is, uh, very much the, the idea that the money go round of AI is, uh, Nvidia giving money and receiving it, giving money with one hand and receiving it with the other, uh, in terms of what's going on, that is a, a very big uplift. So NVIDIA's talking about this as being their need to run AI for their internal business requirements. And they're talking about it as being 1 billion in this year, rising through to 6 billion, uh, for the, the, the following couple of years, and then tapering off maybe in this particular filing.

I mean, who knows what you're gonna spend money on in five years time? It may not have been invented, uh, but there is definitely a, a lot of money being spent here, and it's being spent by Nvidia at customers of Nvidia who are buying NVIDIA's ccp, uh, GPUs. Uh, one of the vital things to see in this is that, uh, NVIDIA's GPUs are hard to get hold of if you're not buying them by the thousand.

And, uh, that's what we're seeing is these, uh, cloud providers in particular, some of the, uh, core cloud providers for ai in this case, we, we saw sort of said Core weave and, and Oracle and Lambda in here. Uh, some of this is that, uh, Nvidia themselves are stepping away from the idea that they will do a renter GPU model. That was something that was the, the DGX cloud was supposed to be this renter GPU that they're gonna be providing to others.

It looks like we were actually seeing the, the shift away from that. Instead, Nvidia has been very aggressively encouraging neo cloud providers and new people to spring up and deliver clouds that are just for ai. Um, and Nvidia has been pushing those very hard so that there isn't so much of a market dominance from the existing cloud players.

Uh, some of the spend will be on, on exactly those same Neo cloud vendors. And I do wonder whether there's an element here that they're going to be customer number one for these neo cloud vendors who allows the Neo cloud vendor to get funding to then buy more GPUs or buy their initial seeding set of GPUs from Nvidia. So it, there is a, a, a little bit of a sniff of this, of, of being kind of a, a, a tenuous money go round to make things look good.

But I think there is also a requirement for Nvidia to use their own GPUs to, to actually transform their business with AI in the same way that they're telling their customers to transform their businesses. Uh, one of the other elements to keep in mind here is that, uh, the restrictions on export of the, uh, Nvidia GPUs to, uh, less friendly states has had some impact on NVIDIA's, uh, financials. They've, uh, incurred a four and a half billion dollar charge earlier this, this year, tied to not being able to fulfill orders because of these restrictions.

Uh, and so this may also be some of the play around keeping things priced still, $26 billion is quite a lot of cloud investment. And seeing that go to established cloud players, providers who are uh, delivering good GPU services is probably good for the market. But I'm not sure that it's a, a huge net change.

Qualcomm's new terms for the Arduino have angered the maker community, uh, including maybe myself since I write a, a whole bunch of Arduino code that added sweeping rights over user content. So since Qualcomm acquired Arduino, they're basically saying anything that you produce on our platform belongs to us. Uh, and the other thing is that they've, there's a ban on reverse engineering, which goes away from the whole core open source of Arduino makers and companies like, uh, adder Fruit argue that Qualcomm is undermining what made Arduino successful and prompts many to switch to other types of, uh, CPUs for these, the same projects, things like the Raspberry Pi 2040 MCU, and my favorite, the ESP 32.

Uh, ASA says the changes, uh, were for clarity and compliance, but I dunno, Tom, uh, it seems to have angered a few people. I think it angered a lot of people. And honestly, I am not surprised because I secretly kind of knew this was coming.

Uh, remember Major League two when Charlie Sheen puts on the, the suit jacket and gets rid of the bad boy image, and everyone's like, oh, well, he's still the same pitcher. He is just, he's a little bit more grown up now. And, and that's like a whole plot line in the movie.

I, I won't spoil it for you, but that's exactly what, what, what's happened here is that Arduino basically said, well, we want to be, we wanna be underneath the umbrella of Qualcomm, uh, because we would like to get money and not go out of business. And our, the Arduino folks were happy to do that. And they, they said the things that you're supposed to say when you, when you become part of a corporate organization, right, is that, you know, nothing's gonna change.

We're still the same people. We just maybe wear suits to the office now. And Qualcomm was like, okay, cool.

You know, we, we promised that we're not gonna mess things up. We just need you to agree to this legal writer that says all of the things that everything else says, right? We're gonna collect your data.

We're going to, uh, anything you make with our stuff is technically ours because you're using our stuff. And, uh, we are also going to restrict your ability to reverse engineer a whole bunch of stuff that we don't want you to poke around in. I got news for everybody out there that's boilerplate.

That is pretty standard. If you go to work for any major organization and you invent something on their dime, it's their invention, not yours, IBM, Cisco, you name it. Any patent you file when you work for them is the patent of the company, not yours.

I agree wholeheartedly. You're, you're gonna be, I'll be the first person in line to say that this is not what the Arduino community wanted. But I think the other thing is the Arduino community did not want Arduino to go out of business.

So you've got this catch 22, right? If we wanna stay in business, we have to play by Qualcomm's rules, and they are Qualcomm's rules. I, I'm, I'm almost positive that every company that Qualcomm has acquired same kinds of things.

And yes, a lot of people in the community are gonna come right out and say exactly what we've heard from companies like Ada Fruit, I don't like this. You're, you're destroying the, the rebel spirit of, of who we are. Yeah.

When that's what happens when a company has to basically mature. And if that means that someone's gonna have to just jump out on their own and kind of spin some things out and do their own thing, okay, great. We're all better off for that.

How many startups exist in Silicon Valley? Because the company, the founders of the company, realized they couldn't make what they wanted to make under the umbrella of the corporate entity. Like that's the, the part and parcel, that's the other rebel spirit, right?

That's, uh, Steve Jobs and his team going across the Apple campus to make the Macintosh flying the jolly Roger from all the windows because they really were the gang of pirates over there. And, and now look at it. So I think that there's gonna have to be some give and take here.

Qualcomm's gonna have to understand that they may have to relax some of these restrictions if they want to continue to let Arduino kind of be the, the leader in this space. But at the same time, the people who are fans of Arduino are gonna have to realize that, you know, maybe what you want to do with the system is not compatible with what corporate America wants. And if that's the case, there's nothing wrong with moving on mass to the next thing, but just make sure that you've got all your breadboards wired correctly.

NATO is working with Google Cloud to build a fully air gap cloud for classified workloads, letting its joint analysis, training and education center, also known as JEC, run AI analytics without using the public internet. The system combines strict security with commercial cloud tools, giving NATO both control and high performance computing. This deal, which is part of NATO's multi-vendor approach alongside AWS and Microsoft, reflects a growing trend in defense towards secure hybrid cloud architectures for sensitive data and advanced analytics.

Al do you think NATO made the right choice here by kind of spreading the resources out along with Google Cloud? You know, I think having a strategy of using different technologies for to solve different problems is pretty common. And, uh, the reality for any large organization for cloud adoption is typically hybrid multi-cloud.

But defense and NATO in this case are, is always a special case because there are always the networks that are secret of some sort, uh, whether that's top or extra top secret. Uh, and these networks are supposed to be fully disconnected. They're supposed to be separated from anything that's accessible on the internet.

And, uh, it's one of the, the fun things of working in defense is working with these area gap networks where you can carry your install media into the bunker where the network exists, but you're not carrying that install media outta that bunker ever. Uh, that's the kind of network they're talking about and cloud they're talking about. What's different here is those networks that are highly secure, usually very tightly controlled, very sort of waterfall development kind of, uh, environments.

And the suggestion of using Google Cloud platform inside this air gap highly secure network suggest to me that they're looking for more agility and the ability to do new things on these secured networks that they would previously have had to do on a public network. And there's been some interesting challenges around sovereignty recently. Some of the discussions around sovereignty have come up, and that if a US company is operating a cloud and your stuff is in that cloud, or US jurisdiction may still apply.

Well, that's us, uh, networks from the internet. I believe this is a way of avoiding any kind of oversight from outside of NATO on the data that's in this, this network. So it's not NATO turn their back on doing things on the public internet.

That's why they still have, uh, both the AWS and the Microsoft relationships. But it is a, a symptom that those secure networks are getting a lot more agility and particularly getting a lot more AI in there. And you can imagine that these secure networks are involved in things like, well, maybe running surveillance, uh, drones and systems, and that AI could be very beneficial to lightning the load on the human operators who are involved.

Uh, I think this is definitely gonna be an interesting project. Uh, I would love to be working on this project just to see what's going on inside it. Although I would also hate to be working on any defense project that requires security clearance, because then you can't talk about it.

Amazon Leo formerly Project Cooper has launched its new ultra antennas offering up to one gigabit per second down and 400 megabits up along with an enterprise, uh, preview for select customers more than 150 satellites in orbit. The service targets business and governments, uh, agencies needing fast and secure connectivity in remote areas. The new ultra terminal uses Amazon design silicon and supports private networking options like direct to AWS and is built for demanding environments.

Early partners include JetBlue and Hunt Energy. They'll be testing this new antenna before broader rollout in the next year. Tom, this sounds like an incredible piece of mobility that maybe is targeted against some other provider of cite, uh, internet, Who knew that the two people who were racing to be the first, uh, CEOs in space would come out with competing projects that were designed to provide connectivity to space.

Yeah, I would, I I I, I totally knew this, uh, this is the outgrowth of what we saw as Project Kuper, right? Is they're gonna launch all these communication satellites into lower Earth orbit, and now they have an antenna that will allow you to talk to them. I will say it is nice that it is a gigabit downlink in 400 megabit uplink.

So, you know, that's, that's competing with most of your home internet plans. And I, that's like the, the ultra tier, so I'm sure that's gonna be like the ultra expensive tier. Uh, there is a smaller tier that runs nano, which is a smaller antenna.

So naturally the, the, the bandwidth is a little bit more restricted. There's also a pro tier, which I'm sure probably just means that you can cut in front of everybody else in line. Uh, they're testing it with partners like JetBlue and JetBlue's gonna be using it to augment their existing, uh, plain wifi.

And, and this is a trend that I've actually seen in a lot of other places because starlink is becoming the defacto communication system on a lot of like, hypermobile things. All of the billionaires that are building brand new luxury yachts, they all have starlink terminals integrated into the yacht. Uh, that's so that they can pretend to be in the office when they're actually floating somewhere off of a coast and they're doing work and they, they, they spend all of their money kind of refurbishing these offices so that it looks like they're actually working when in fact we know that they're probably not even wearing shoes at that point.

And that, I think, is one of the places that we're gonna start seeing some of this kind of, uh, flushing out a little bit, is remote access terminals. We, we've seen that with starlink quite a bit. In fact, I know a couple people who have starlink terminals who use them for things like, uh, being a digital mo nomad or, uh, providing ultra high speed connectivity in places where it's really impossible to get a cell signal, like the back country.

And I think that the value here is that with the smaller nano antenna, you're gonna be able to basically carry that like a backpacker, right? Uh, but for those kind of fixed and placements think oil rigs, uh, think remote construction sites, uh, this is a great way to provide a, a deployable network without needing to like, you know, run cables and get access points up and running and deploy, like a huge infrastructure. You can just drop this antenna down and it works.

And the key advantage here over starlink is that this is built by Amazon and it's optimized to use Amazon services. So you can get a direct Amazon link, which I'm sure will in, you know, speed up any kind of AWS uh, up link down link type stuff, which is what you would want if all your stuff is located in AWS. But also by optimizing it for things like video conferencing and other applications that you use, it means that you either A, don't use nearly as much bandwidth, or b, it will give you a better experience overall in case there's some kind of weird outage, sunspot, solar flare, you name it.

I, I think that the other advantage of doing, of having this come out of Amazon is that it will create a competitive market so that then it will force other providers, cough, cough, starlink, cough, cough to get better at what they're doing and to provide better service, to provide better connectivity, to not have weird restrictions like, oh, you can only buy this smaller antenna if you're a customer on the bigger antenna and that kind of thing like that. So this should be good overall. I just hope that we don't end up populating low earth orbit with a constellation of satellites so thick that we can't fly through it.

Um, 'cause you never know who's gonna do something crazy, like decide to build even more stuff in orbit. Wait, hold that thought. Splunk has contributed its open telemetry injector library to the CNCF open source project, making it easier to monitor legacy non-con containerized applications.

The tool automates instrumentation for production apps, helping DevOps teams collect telemetry with minimal effort. Splunk also added a schema for tracking AI workloads and LLM performance, supporting consistent metrics for cost, performance, and bias. This donation highlights the growing importance of observability for improving operation security and AI readiness beyond traditional DevOps.

Al, do you think that this is going to encourage people to wanna adopt more CNCF, uh, telemetry tools, or is this Splunk just basically saying, we're getting rid of this because we're moving on to better things? Well, I did see, uh, coupon this year a huge amount of open telemetry, um, adoption or observability really coming back with open telemetry at its core as a way of getting data, uh, telemetry data from your applications to somewhere else, be that, to Splunk or to any other, uh, open telemetry compatible, uh, receiver. Uh, in terms of what's what's announced here, I like that what, uh, Splunk has donated is a, a really simple tool to install on a legacy, uh, virtual machine or physical machine that is running a legacy application.

A non-con containerized, I'm sorry, I should be calling that a heritage application because legacy seems like a bad thing. Uh, heritage application that is still running in, in physical machines or virtual machines. Uh, you know, those of you living in the cloud native world may not realize that there are still an awful lot of applications that just run in virtual machines because there's no value to shifting those into containers immediately breaking them into a tiny set of microservices.

There's a huge amount of work there. So being able to instrument those applications without having to rewrite the applications, uh, the, the injector that's been, uh, donated here is just a piece of software that gets ins installed on a machine and sends what is local metrics into hotel and into open telemetry. And very much as an enabler for adding, uh, your legacy components of your application to your modernized components of your application.

This is absolutely vital where companies are developing new capabilities that still leverage older data sources or that are following the Strangler model, where you start building microservices probably in containers, uh, on top of Kubernetes to build the new features around your existing legacy application. This lets us instrument that legacy application. I like it.

I also like that they, uh, announced more open telemetry and, uh, large language models, being able to able gather more information out of your, uh, l LM based applications, particularly as we see that move towards age agentic applications, getting more telemetry data out of what the actual LLM is doing. Not just the performance information, but is it actually doing the thing it's supposed to do? Is it doing things safely?

Uh, are we having sort of compromises in in the, uh, LLM? These are things that the companies are very much concerned about. And I think it's, it's great to see more visibility of what's actually going on inside, inside these modern applications.

So both really good things. I think Splunk has done a great thing here. Uh, I don't think that it's entirely altruistic.

I think they definitely want to show some leadership in the world of open telemetry because sometimes is talked about as a way of replacing your, uh, heritage, uh, logged tools like Splunk that charge by volume of data moved through them. So it's, it's not entirely altruistic, but it's a great thing. So I told you to hold an idea.

Recall that idea now, because Google's project SunCatcher is testing the idea of moving AI data centers into low earth orbit using solar powered satellites with high speed laser links. The system aims to provide more energy and faster connections than earth based data centers. Hmm.

Early tests show promising speeds for connections and prototypes. Satellites are planned, uh, by Planet Labs in 2027, uh, full scale orbital data centers. Yeah, probably a little further away.

And the project reflects a growing effort to meet the massive demand for computing and power for AI applications. Does lead us to some concerns about filling the night's skywood data centers and making the whole place glow in the wrong place and dark in the wrong place. Tom, are you gonna service these AI data centers in the, in the, uh, low earth orbit?

Yeah. Let me just, uh, break out one of those space shuttles and, uh, and oh wait, we, we retired those, uh, dragon capsules. Yeah, maybe, I guess, I don't know.

I, we, we, we heard about this from Amazon, remember? Like, they, they wanted to, they wanted to launch things into space, and now Google's like, no, no, no, we, we want to do it too. Uh, can I see the tests that, that, that these things are promising?

Because I have a lot of questions about how this is supposed to work. Uh, number one, how are you gonna cool all this stuff in space? And before everybody out there says, well, space is really cold.

No, no, it's not. Space is empty. There's a difference.

One of the things that we've learned over the years is that things in space can actually get really, really hot for two reasons. One, when there's no particles around, you can't radiate heat away. And two, there's this thing that's about 93 million miles that way that, uh, will really heat stuff up.

Because again, there's nothing to block the, the solar radiation, uh, when you're in space, uh, we, we've learned that a lot from astronauts. Like it's, it's, it's a little bit different out there. Like literally the sides of your spacesuit can be two radically different temperatures because one side is being hit by the sun and one isn't.

Uh, the other thing is, like you said, what what's gonna happen when something breaks? Uh, you think it's hard to do a truck roll to a data center in the middle of prior Oklahoma? Uh, wait until that, uh, truck is gonna have to roll, you know, a few a hundred kilometers north, uh, and galactic North, by the way.

Just, you know, this straight up, that that's gonna be a huge problem. I, I wonder though, why, why are we lo launching these things into orbit? Because one, if you want to use solar power to power data center, cool, build it down here.

Like, like there's nothing stopping you from building a solar array down here. Uh, we turns out we're really good at building those things, those things. Is it a space congestion problem?

I don't know. Uh, there's a lot of space, but here's the thing that most people don't understand. For things to work down here, they have to be in relatively close space.

That's why they're in low earth orbit. Uh, GPS satellites are actually almost too far away, uh, for the amount of data transfer that they're gonna have to do. Uh, something like the James Webb Space Telescope doesn't even orbit Earth.

It orbits a LaGrange point, uh, closer to the sun. Uh, if you wanna look that up, that is not the song by ZZ Top by the way. Uh, but the, the ultimate problem that I have here is there's no reason to be building these things in space other than somebody said that they wanted to do it.

There's still plenty of space down here to build those data centers. I think that this is a, uh, marketing tactic, a ploy to get people to maybe give more favorable terms to have it built somewhere. Uh, you know, 'cause I guess their argument is you can't tax things in space.

Uh, unless Google or Amazon or Microsoft or Oracle or whomever, uh, developed their own fleet of space vehicles last week, they're still gonna pay somebody to launch this stuff. And I promise you, if you thought getting overnight delivery on something from Amazon not on Prime was expensive, wait until you see what it costs to lift that stuff into orbit. It is not cheap at all.

So I I, I watch this with bated breath only because I'm hoping that somebody in this weird fever dream is gonna wake up and go, why are we putting things in space? Uh, if we, if we can't maintain the ISS and we're worried about having to de-orbit that thing pretty soon. The last thing we need to be doing is putting a whole bunch of data centers up there, because eventually those suckers are gonna come back down to, we hinted at it in the opening of the video, but we wanted to take a closer look at some big investments that have been going on this week.

It seems like right before the holiday, everybody wanted to announce the fact they're investing a lot of money into AI infrastructure. I wanna start off with a company that we just recently covered on networking Field Day, and that's Nokia because they've agreed to commit $4 billion to US research development and manufacturing to speed up AI ready network infrastructure. Most of the funding will support r and d at Bell Labs in New Jersey, and the rest is gonna be going to facilities in Texas and Pennsylvania as well.

The initiative, which is backed by the current presidential administration aims to boost domestic AI capabilities, improve connectivity and strengthen national security, while also meeting growing demand from AI and cloud customers with and advancing Nokia's partnership with Nvidia. Al I'm gonna jump in here and talk a little bit about this because Nokia is really trying to make moves to be a valuable partner in the networking space. They are building out infrastructure that will allow them to build AI ready networks.

And one of the things that we learned at networking Field A is that's not easy to do. So you may probably look at this and say, $4 billion doesn't sound like a whole lot. Well, when you're looking at the amount of money that Nvidia and Microsoft and open AI and cloud core, we and all these other companies are throwing back and forth at each other.

Yeah, 4 billion doesn't sound like a lot, but $4 billion for a company like Nokia to invest in AI networking. That is a ton. And I think it's super important because that's the way that the data gets moved from where it is resting currently to where it needs to be processed and run on these inferencing tools and things like that.

I'm glad to see that they're, they're doing this. And the, the discussion we had at Bell Labs, uh, it was actually something that was discussed by Scotton from AL during the Nokia presentation at Networking Field day. You know, they have a lot of good data points on this.

Like they've done a lot of work out there. They're really, you know, hitting this pretty hard. I think it's important to understand that, that just because we're not seeing huge dollar values flying back and forth outside of GPUs and compute clusters doesn't mean that there's not a lot of research going on.

Al what's your take on this? I think it is important to put context on, on what $4 billion of networking, uh, and particular networking research looks like compared to the tens of billions of dollars that we talk about for buying GPUs. And, uh, GPUs are ridiculously expensive.

Uh, a single GPU can cost as much as a new car. Uh, but you don't need a, a new network switch for each of those, uh, GPUs, you, you hang those GP multiple GPS off a switch. So in terms of what things cost and, and the scale of investment, yeah, $4 billion is a big deal here.

Uh, and as you say, the, the network is often one of the challenges in building an infrastructure for ai. And we'll look at this, uh, again in January with AI infrastructure field day, uh, moving the data to feed those very expensive GPUs is a, a vital part of getting the most value out of them. This is why network design for AI is a critical part of any AI data center infrastructure design and this commitment to, to do r and d, uh, onshore in the us really great thing.

It's very helpful for, uh, the US sovereignty to here to not be looking outside to network vendors that may be Chinese, but, you know, having Huawei and, uh, the likes being on on the outside these days, um, due to some, uh, concerns around sovereignty. So seeing Nokia committing in here, supported by the, the current government, uh, hopefully supported by future governments as well. Uh, really good thing to see.

Of course, at the same time, AWS is spending lots of money as well. Uh, AWS is spending the, the kind of money for building entire data centers, so $50 billion for government AI infrastructure. 3 gigawatts of computing capacity across, uh, the government cloud regions that, that AWS operates, that's pretty significant.

We know the government cloud regions are usually slow to adopt new technologies. It takes quite a while for the technologies to get certified. And I think what we're seeing is US government customers saying to AWS, we, we need this certified, we need these AI capabilities, and we're going to take them from the cloud.

So this, uh, $50 invest, uh, 50, $50 doesn't get you very far. It doesn't even power the GPU for an hour. Uh, $50 billion is, uh, quite an investment in the GovCloud side and bringing AI to it.

Tom, uh, do you see the same thing going on with more and more AI being used in, in government use cases? Yeah, actually I do. And I think that one of the things that you have to understand about why this is happening is that the government is kind of running behind here, but that's the government's job, right?

The government doesn't jump out on the, the bleeding edge and do all of this stuff. They want tried and true, secure, capable systems because when they buy, they buy for, well, a decade or more. And, and by announcing this huge investment, because we've also got the project target stuff hanging out there and a bunch of other things, this is specific to government agencies.

This is kind of, you know, remember Project Jedi, we've talked about that a lot. Uh, it feels almost kind of antiquated at this point to think about, oh, well, you know, the, the pentagon's gonna be moving into the cloud. Oh my man, that was, so six years ago now we're, we're talking about getting up and running on AI tools and, and figuring out how to do all this stuff.

And it's important to note that this is something that was laid out in the previous presidential administration under their action plan. And the idea is, is that we need to start adopting this stuff. We need to start moving along.

But of course, as you mentioned, these things are not sudden. And so I think that the value here is that any advances that are made will probably end up trickling down into other offerings from Amazon. We're not gonna get access specifically to the stuff that's gonna get built out through this big investment, right?

Like that's, that's government is government and we don't touch government, but the lessons that they learn will be kind of cross pollinated everywhere else. And so maybe we see better connectivity, maybe we see reduced resource utilizations. Maybe the government decides to build a couple of new nuclear power plants to run these things, and we all kind of benefit from that.

Who knows? Maybe the AI will figure out the secret to cold fusion and then we won't have to worry about power ever again. But I'm not holding my breath on that one.

Well, Al that's, uh, a good look at what's going on in the news. And I know that we're kind of bumping up against a holiday here in the us but what's the excitement that we've got on tap for 2026 already? Well, 2026 is gonna tick off with a big event, uh, at least for tech field.

That would be a big event with the AI infrastructure field. Day number four, uh, we're looking at spending three days out in California hearing from a whole collection of people, including people we've mentioned on this podcast in particular. We'll have NAIA there.

Uh, we'll also have a bunch of other interesting companies talking about how to build infrastructure to support your AI requirements. And particularly I think we're gonna see lots of support for production, ai, you know, inference, we get value from your area. The next event I've got locked in is Cloud Field Day 25.

I'll be back to California March 11th and 12th. And again, we'll be looking at some fun technologies in your hybrid multi-cloud world. Uh, some of the topics we often visit here.

Uh, check out all of the upcoming events on the Tech Field Day website. And you may find that events get added. We've got some interesting things still in the works, maybe in that first quarter as well.

Thanks for watching this episode of The Tech Field Day Rundown. You can catch new episodes every Wednesday, uh, as a YouTube video or on your favorite podcast application. Remember to give us a review and maybe a little like in there, the rundowns also streamed on Textron tv, so you can catch us on other Textron properties, and you'll often find Tom and I at RUM group events and other fun places.

We'll be back next Wednesday to talk about all of the IT news for the week. That was until then for myself and for Tom Hollingsworth and for all of us here at Tech Field Day. Wishing you and yours a great day and a great Thanksgiving.