Techstrong Gang – May 13, 2025
Alan, Mike, Bonnie, JP Morgenthal and Stephen Foskett, president of the Tech Field Day arm of the The Futurum Group, delve into the rise of quantum networking before reviewing the merits of Wi-Fi versus LTE networks.
Then, the gang takes a look at the degree to which artificial intelligence (AI) represents a threat to the environment.
Transcript
Hey, everyone. You ready to get entangled with Quantum Networking? You're watching Text On Gang?
Happy Tuesday, everyone. It's Alan Shimel for Techstrong, and you're watching Textron Gang. We've got a good show here with some interesting stuff as this week starts taking off and we've got some interesting people to discuss it with you.
Uh, let me introduce you to our gang for today. First of all, with the fancy background. Mm-hmm.
LLMs ai. I, that's what he, yeah, we've, I had, that's a new one I haven't seen before. He's our friend.
JP Morgenthal. Jp, how are you man? I'm doing great.
How's everybody today? Well, we'll find out, won't we? But good to have you moving on from jp.
We'll go up to he, I think, I guess from the background, he's back in Hudson, the man from Hudson, but that was the man from Hope. Well, you're the man from Hudson, uh, or uncle the, uh, founder of Tech Field Day. Steven Bosque.
Hey, Steven. How are you doing? Well, you know, it was nice to be down there in the studio, uh, keeping Bonnie's chair warm and the, but, uh, it's nice to be home.
There's no place like home. Yes, There is. Click your heels when you say that.
Click those heels. All right. Welcome Steven.
And then also home in Harrison, New York, not named for the Presidents our Chief Content Officer, Mike Vizard. Hey, Mike. How are you?
I'll echo Steven. Sentiment. There's no place like home.
It's, and any day you're headed, home is a good Day. That's right, that's right. Speaking of home, we're home here in our book studios and joining the in studios, our resident Echo Insights analyst and, uh, editor Bonnie Schneider.
Hey, Bonnie, how are you? I'm Well. Great to be here.
Alan, great to have you in the studio. I gotta tell you the truth. It's a lot better having you here.
I mean, it's not that I have anything against Steven, but it makes for a better picture. Well, it's rain, honestly. I'd rather have her there too.
Yeah. All right. Let's, um, it is a rainy day here in Boca Ratone, but you can't have sunshine every day.
Wow. It's actually raining in Same, in two different places. Miles apart in Florida.
The same in Florida. Well, we, That's, that's, that's not a Good sign. No, but we needed the rain.
It's been, it's been pretty dry here, Here too. Yeah. We need to rank anyway, so Mike Cisco threw their hat into the quantum race Derby, whatever you wanna call it, but being Cisco, of course, they're talking about quantum networking.
True. That. So yeah, Cisco has revealed that it's working with uc, Santa Barbara, to build something that feels like a quantum networking and entanglement chip.
And I guess, you know, in my mind, a lot of us thought about quantum computers as these big kind of mainframe like things, but maybe they are gonna be a little more distributed. And every time I turn around, somebody in China seems to be building a quantum computer that's smaller and feels more like a general purpose machine. So Alan, I know you follow the space, but what's your take on what's going on here?
You know, I, I think this is the filling out of the ecosystem here, right? But this isn't networking the way you may be thinking of networking. Um, but I just wanna mention, hey, we are planning a virtual event on quantum called of All Things Quantum Leap.
Uh, and it'll be in October. And we're looking for speakers, sponsors. And if quantum's your thing, you should register for it.
I don't think the registration page will be up for another week or two, but, um, I'm looking forward to diving a little deeper into all things quantum there, from quantum to security, to quantum chips, to quantum networking algorithm, software, everything else. But when, you know, first of all, people have to be familiar with the entanglement principle of quantum, right? Which means you could have two particles or two pieces of matter that are somehow, you know, as crazy as it seems, they could be light years apart even, and still connected entangled, where if one spins, the other spins one, you know, and they, they copy each other's, uh, uh, matter kind of, uh, you know, if, if one's spinning left, the other's spinning, right?
Whatever it is. So they're intrinsically linked, they're entangled. And once you understand that and you have that, and you can manipulate that, well, that allows for all kinds of, of things.
Every, you know, from the craziness of teleport teleportation, you know, star Trek kind of beat me up, Scotty, to networking perhaps, right? And that's what Cisco's talking about here. Now, the interesting thing, it doesn't necessarily have to be networking like from a router to an endpoint or over the internet.
It could just be two quantum chips that are entangled, right? So by using entanglement, you can create sort of a, a parallel type of processing, uh, you know, sequence where you can have multiple quantum chips now working in quantum time, right? Via entanglement.
And so if, if, if a thousand qubits is the holy grail or whatever of when we could really hit quantum, you know, picture having 10, 100 qubit processors that are network really entangled and, and working like that. Yes. I, I, I just, I wanna address the, uh, Cisco announcement with regard to that, right?
I think what Cisco was pointing out is that we don't have that kind of scale on the horizon right now. We have, well, no, but we, what we have is a, a limited number of qubits, chips that can, that are created to create a quantum computer. But what the way they're saying that you can scale is by creating, leveraging quantum networking as a way to create multiple quantum computing machines, Right?
That that's what you're gonna have in, right? You're gonna put 'em in parallel the same way we did Linux machines into supercomputers, right? That that's, it's kind of the same kind of thing.
I I I'm interested in how much data can be represented by these entanglements. Because if you, the one thing we've always said is that they, you know, and I've always joked that, you know, ultimately the network owns the s for the in, for the for, for the universe of computing because it's the, it's right now the major limitation, you can only get data to, and so much data to move so fast from point to point, right? So it, it's your, it's your limitation in compute right now is how much data can you move and, and how far can you move it before you see degradation.
So if you, if this thing can carry terabytes of data instantaneously moving it from point to point, it's, that is a, then that's a huge game changer. That is a major game changer. Yeah.
But this is, I mean, the whole idea of quantum networking, I don't know if we could think of it in terms of how much data can I move over a pipe? I, because it's particle to particle. So I think the, the issue then becomes, well, can I do a hundred thousand particles to a hundred thousand particle?
But each particle has this much on it, but, you know, but a hundred thousand of those, nevertheless is a lot. I mean, jp honestly, this is, this has hurt your brain stuff, right? That it's, it's kind of so far out there and so radical to what we know and think of.
Um, because I don't think data is necessarily transported. It's kind of an instantaneous, if this knows it, that knows it. I mean, you know, you gotta, you gotta be on some medicinal marijuana or something to kind of get your head wrapped around this.
So for, So for people who, um, are, are not quantum, uh, oriented, uh, think of it like virtual mapped memory. You have your virtual map on top of your physical disc and your virtual map knows instantaneously, whatever's on the disc data. And then you move your virtual windows around to see different parts of the disc, and it becomes just boom, it's available.
So you're saying the whole universe could be on my thumbnail, I was thinking more like the cat in, uh, men in black around the, the, the, So Steven, it sounds like to me that this is the laws of physics being applied to quantum computing, or is it fundamentally different? Well let, yeah, let's, let's, uh, I, I don't know, get, get down to down to earth here in a little bit. Uh, you know, put down the, the wacky tobacco, if you can for a second.
Uh, yeah. What Cisco's announcing is pretty interesting, and I'll just point out that we actually heard a very similar thing, a similar story from others at our networking field day. In fact, in March, uh, bt the big, uh, telecommunications giant presented a whole section or a whole field day session on quantum networking.
And Cisco's describing a similar application. The idea is, well, number one, Cisco has some pretty cool technology here. They've got a quantum, uh, pro well networking processor, entanglement, chip.
It's not really, this isn't a quantum processor. This is a quantum enabled networking chip that works at room temperature, at very low power and uses the same wavelengths of light that are currently used with optical networking, which is important because otherwise it would not be practical at all. The other thing, like JP was saying, I mean, the interesting thing here is that you can instantly, and, and when I mean instantly, I don't mean just really quickly, I mean, it's there, um, transmit data in a way that cannot be snooped on that is completely, you know, synchronous on both sides.
And that's a really useful tool for things like key public key exchange and encryption, that sort of thing. Another thing Cisco has done is they are using the quantum, this quantum fluctuations as a random number generator. And, and I think that those of you in the security space, I'm sure that you all have experienced this and, and know that random number generators are not very random.
And that's a huge problem. And that's one area that hackers and especially nation state hackers use to exploit systems. Essentially, they will, uh, clone the pseudorandom number generator and make and, and, and, and generate the same number.
Well, it's awful hard to do that in the quantum space. And Cisco has come up with a quantum random number generator. And, and also they're working with, uh, university researchers to develop future applications that would allow interconnection of quantum systems, but also quantum interconnection of classical systems, which is something that the BT presentation was talking about as well.
There's just a lot of really interesting, um, features in here, including some nuts and bolts ones, not some pie in the sky ones. I mean, honestly, a quantum run random number generator. Sounds great.
Well, you know, it's funny you bring that up, Mike. Wasn't it just yesterday's show? We talked about this issue about some researchers, researchers came out with a report that they explaining and, and being able to anticipate prime numbers, which is, you know, how all of that underlying all of our encryption and, and, and the idea of being able to generate random numbers, you know, is, is this, it's not so random as it turns out, potentially.
Anyway, there was just a research report, whether it was, you know, how true it was, we don't know yet. But, but I, I, I actually, Steven raised an interesting point that I hadn't yet considered. I in the field of intelligence, um, think about a compu two computers in two skiffs.
Those are compartmentalized rooms. Air GAed not connected to any network, not even the high side, low side of the, of the government actually Air gap computers being able to communicate with each other across skiffs, that is an insane intelligence application you would actually have, because the whole idea of the air gap computers, it's not connected to any network, right? It's a standalone computer that a, uh, performs some type of function.
And this would allow those computers to be networked And not have Well, but is that a good thing or a bad thing? Well, from an intelligence application standpoint, it's re it's a, it's critical. Depends on which side of the intelligence equation you are.
If you're trying to protect stuff, it's good. So this comes back to kind of what I'm starting to wonder about here is, so we know that the existing computer architectures we have in place are flawed, and we're using them in ways that are insecure. And to Alan's point, there's prime number issues now that people are figuring out.
It feels like all the money being poured into quantum is an effort to kind of reinvent computing. And it might take a decade or so, but, or is this gonna be some, you know, little sidebar esoteric thing that we're using for some vague computer science project? Like, you know, we're gonna figure out how to use some chemistry application differently.
But, you know, Steven, how big a deal is this ultimately? Well, the proof is gonna be in the pudding, isn't it? Uh, we'll see if they're able to productize this and bring this to market as a friend of the gang, Bob Sutor would say, and as he did say in that Textron article, you know, I mean, there's, uh, there's a lot of uncertainty here in the quantum space.
Yes. I'm sorry, I had Beautiful, yes, there's a principle about that, isn't there? Yeah, Exactly.
Yeah. The principle says that it's uncertain. Um, and we gotta listen to the principle, but, uh, no, the point is, we'll see, right?
I mean, if Cisco can productize this thing, I, I, I can envision a situation where next year, the year after Cisco brings a quantum powered random number generator for encryption tasks to market. And that becomes a very important and useful tool. I could also see a situation where they never talk about this again, and we just continue to move on about our lives.
So ultimately, it's not research that fixes things, it's products, and let's see the Product. Well, what about, what about both of those things happening in parallel universes back to the wacky tobacco, just saying, just saying. But, but seriously, You know, if, if we are going, so I'm a Star Trek Fat, I, I have to, you know, if we are going to go on these adventures where no person's gone before, if we're going to continue pushing the boundaries, right?
We, we've spent the last hundred plus years exploring basically the theory of relativity and how that affects everything that we do. And it's led to the nuclear and hydrogen bombs, fusion bombs, vision, energy, and so many other things. The next breakthroughs that will advance the human condition.
We need to understand and harness the power of quantum. And, and you know, this, um, this is one of those times you're happy to be in tech, right? Because we're kind of leading the charge here a little bit and trying to harness, harness this to recreate what we do.
And that's kinda what tech, the tech industry does, right? We don't use punch cards anymore. We don't use little floppy disc, right?
How, you know, so I'm, you know, we do it because it's hard, not because it's easy, but it's worthwhile. And, and that's my last word on quant. I'm looking forward to the Rung Gang being in two places at one time.
That'll be fine. Well, we're already yet, look, we're in Ohio, two places in Florida, in New York already. We're quantum, But then I, but theoretically I can have two Stephen Foskett arguing with each other over a quantum network.
Solar. Solar, I'm sorry, I just had to say it. All right, let's take a break and kinda sober up a bit here.
And, uh, we'll come back and Steven, you can have a report for us on Mobility Field Day. Yeah, I sure will. Fantastic.
You are watching Techstrong Gang. Well, we are, uh, hosting a lot of different field day topics, but as I said last week on The Gang, my favorite tech field day, I mean, my favorite child among, among many is Mobility Field Day. The reason is because as, uh, they were all laughing at me for saying on the gang, they are from another planet.
It is always fun to, to mix and match with people who, uh, know something that is completely alien, that, that are deep experts in a field that you aren't an expert at because you know, you can really kind of soak in it and learn so, so much, so, so quickly. Unfortunately, I wasn't at Mobility Field Day. I was in some place called Boca with a, uh, distinguished man named Shimmy.
But I was, uh, check checking it out remotely as anyone can do, watching some of the presentations. And over the weekend, I checked out some of them as well. It's interesting the takeaways that you get from this group when they do come together.
Uh, one of the things that I predicted going into it was that wifi seven was gonna be a very important topic. Well, the answer to that is a little more mixed. Um, wifi seven sure is important, but some of the key features, especially Multilink operation, are still very much a work in progress.
There are standards, but they're sort of pre standards, and the companies are still working on that technology. Another aspect that we heard quite a lot about was how to bring wifi to the masses at large public venues. If you've been to a baseball game or a football game and tried to use the wifi there, it probably works better than the cellular network.
But unfortunately, that, again, is still a, a, an area that there's a lot of development happening on. Well, fortunately, because a lot of that development is actually going to be able to leverage some of the features of wifi six and seven to really broaden access to public wifi. The other thing that's important is that we're seeing a lot more crossover between wifi and data center technologies.
So things like, uh, VXLAN and so on are actually coming into the wifi space, which again, is one of the things that I love about learning about other fields, because you can take ideas from those fields and bring them into yours. The wifi community was the first to have essentially what we now know of as software-defined networking, or software-defined computing software defined anything. This idea that you have a controller and then you have, uh, basically cattle instead of pets out there as your hardware and the controller, uh, with Zero Touch will activate and configure and, and, and integrate that hardware.
Well, that was a very widespread technology on the wifi side before we ever saw it on networking, let alone on Compute. Now it's everywhere on compute, and that has been a really powerful thing. So it's really cool to see some enterprise networking features and ideas like VXLAN come over into the wifi space as well.
So overall, that, that, that was sort of the takeaways from Mobility Field Day. I do urge you to check out some of those presentations, especially, uh, I just wanna give a little shout out here to, uh, fellow Futurum, uh, analysts, uh, Ron Westfall, who talked a lot about some of the, uh, intelligence research that we've done, uh, internally on the state of, uh, mobility, networking, quantum, those sort of things. Check out the, the Ron presentation.
All those posts are up, um, on the text on tv, but they'll also be up on YouTube very soon. So Steven, every time I turn around, some telco is trying to get me to essentially ditch wifi in favor of using LTE as my primary network. And they're pointing out things like, you know, wifi can get hacked and it's a pain to use.
And you know, you, to Alan's point last week, I think, you know, you gotta stand on one foot to get the connections, right? So, um, is there a case for LTE here or is wifi still the dominant player? But wifi is gonna be used in other things?
Well, when all you have is a hammer, everything looks like a nail. And if you're a telco with a big 5G or LTE network, uh, you might start saying, Hey, we can do that too. I think there is some crossover, but I think that, uh, those of those that I know who've deployed, um, basically LTE as a, as a broadband back haul, many of them are happy with it for the price.
Um, many of them are not happy with it, with it for the performance, especially when when things get busy. Uh, that's not a topic really that the mobility Field Day folks have talked much about. Um, but again, without casting aspersions at anyone on the call, uh, we found that it's sometimes difficult to implement wifi in a way that is reliable and high performing and meets the needs of users.
And I think that the ease of use of mobile networks has really come a long way to the point now where, I don't know about you, but I was at a hotel over the weekend. I never connected to the hotel wifi. I just used my phone, uh, on 5G the whole time and, and I didn't really suffer for it.
So I think that, uh, many of us may be starting, I don't know, maybe the bloom is off the Rose A. Little bit on wifi. Do you find yourselves doing that as well?
Yeah. Mm-hmm. Yes.
Well, and, and, and to the aspersions around wifi, it helps if people do plug in their wifi access points as we found out one was unplugged here, but, um, who knew? But you know what, I, real life, so at the HOA where I live, I'm, I'm the president of the HOA in Del Boca Vista. Um, but anyway, we are looking at bringing in a broadband provider for the, for the whole development, which is 69 homes.
It's not a big development, and they wanna bring five real fiber to everyone's house, not fiber, and then last mile fiber in everyone's house. And everyone would get two wifi, seven, uh, wi you know, wifi access points upstairs, downstairs, a whole bunch of other stuff, five asynchronous, 500 meg up, 500 meg down cable tv, and if you want IP phone cheap, like 80 bucks a month for all that, I surveyed the neighborhood to see, is this something we want to do? And I was surprised that there were outta 69 homes, about eight homes that are on this T-Mobile, LTE.
You know, if, if you have a T-Mobile phone, they give it to you for like $30 a month. If you a T-Mobile cell phone customer, they give you a little wifi access point, and for 30 bucks a month, they get, they're streaming, they're streaming tv, they don't have cable, they're streaming video, they're using it, you know, for basic connectivity. Now granted, these are people like all del Bulk of Vista people who are probably in their sixties, seventies, or greater, and are not, you know, I don't think they're uploading any videos or anything.
They're shooting locally, though one never knows in Florida. Um, but it works for them. It works for them.
And so in $80 now, they don't have anywhere near 500 megabits up and down, but for what they need it for, it works. And, you know, 30 something dollars to $80, they'd rather keep their 30 something dollars thing. I don't think that works in an office here, maybe with the amount of devices we'd have connected in our with needs.
Um, and I'll tell you one thing, what I do agree with Steven on is like, I believe, and maybe I'm wrong, but the, I think the LTE connection is more secure than the wifi connection, especially if I'm like bouncing into some sort of restaurant somewhere and I'm trying to use their wifi and god knows when the last time that thing was Updated. Well, Stephen out, my friend Jennifer, Jennifer jj, man, Jennifer Manilla was at your event, wasn't she? Yeah.
And Jennifer is, is fantastic. Yeah. She's the authority on wifi security.
And the truth is, uh, to your point, I don't think there's any intrinsic reason that wifi is less secure than cellular. In fact, I think that it could be more secure. The problem is that wifi is generally misconfigured everywhere.
Um, you know, I, uh, I had to reconfigure the local coffee shops wifi while sitting there, uh, because of, uh, you know, they, they just had the default outta the box configuration from Windstream, and it was just terrible. Um, you know, that sort of thing happens all the, all the time. And I think you're right.
Um, you know, rogue access points and, uh, impersonation attacks and all that stuff, it's, it's, it's not as common maybe as people might think, but it, it does happen. And that can be really bad. Uh, any, I, I dunno, if you wanna be terrified, Google wifi pineapple, and you'll learn all about that.
Um, on the other hand, uh, there are also rogue, uh, LTE cells, uh, Google Stingray, and you'll be terrified as well, and you'll shut off all your phones and throw them in the microwave and, um, and be done with it because all of this stuff, there's, there's, there's also sorts of hacks here. Yep. Steven, I wanted to talk about another aspect that was discussed, and you mentioned briefly, and that is, you know, stadium level wifi.
So I, I've been covering that basically with my friends from Extreme Networks now for a few years. And they, they have a good business around that. I think they're like the official partner of the NFL or something.
I know the Cisco Stadium, um, where the 49 ERs play in Santa Clara, I don't know if it's still called Cisco Stadium, but with Levi Stadium. Yes. They, they do that, they do a whole bunch of stadiums and, and they have made a ton of progress there.
And I, I've seen it when I go to sporting events now, you know, it used to be you can get on the network, but you wouldn't go anywhere. Right. You couldn't get out.
But, you know, most recently I, I've seen absolutely improvements there. Oh, yeah. Well, it's funny that you mentioned Levi's Stadium.
Uh, unfortunately we weren't able to film it because of confidentiality reasons, but we actually had the company responsible for the wifi at Levi's Stadium lead the mobility Field Day delegates through Levi's Stadium to show exactly how they implemented that. And it is really well done. That's an example of, um, you know, the, the, the, the good aspect of the wifi deployments as opposed to the bad ones that you find, um, mostly everywhere else.
Even there though, um, we've caught up with those people over time. They've had all sorts of interesting edge case kind of failures that they've had to resolve. And again, happily that group has come back to us and spoken with us and told us about the lessons learned and the things that they've done to improve access there.
It is incredible what happens, for example, when 50,000 people arrive within a half an hour and connect to your wifi. Uh, that's not something that most people experience, but it is something that they've experienced. So I think that it's one of those, um, one of those things where expertise really matters.
You know, when you go in the office in the morning in those big office buildings and you show up around nine o'clock, you experience the same phenomenon where everybody else logs in at the same time. Right? Well, you know, it's true.
It's true. And, uh, and you know, it's the same when you have like high volume applications. Your Dropbox is synchronizing, your time machine is backing up, your email is downloading, uh, yeah.
And then you're on the meeting on the third floor, but your office is on the ninth floor and they didn't route you back to the right connection point. So you're trying to access your access points six floors above. Yeah.
You're just, he's just making excuses so he could work from home. Come on. Um, speaking of home though, an interesting thing that came out of this project I'm doing with my HOA is the average house has more than 30 IP addresses right now.
30 connect, 30, 30 connected devices. You think about it, You know, I mean, US nerds, well, especially with the IOT space, um, IOT is causing a rapid proliferation. 'cause a lot of these devices are using IP networks now.
And, um, and that's actually one of the benefits of some of these things, you know, matter and thread and ZigBee and Z-Wave and all those things because they're not on your wifi network. Yeah. We're seeing a, a huge proliferation.
4 gigahertz wifi, which is increasingly crowded. 4 gigahertz is usually 30, 40, even 50%, which means that basically only about half that bandwidth is even available because there are so many devices on that space. Yeah, on on that point, I, I was configuring a router in the clubhouse of the, of the HOA and um, it's a six E router, and for the first time there were three networks.
4, the five and the six. Yep. Yeah.
And six E, that's another benefit of six E is opening up that six gigahertz wifi because it's, uh, you know, there's a lot less contention, a lot faster too. Six E is great If you have a device that can support it. Yep.
Which I have. There you go. All right.
Hey, let's take a break here. We're gonna come back and Bonnie has an Echo Insights report on AI carbon footprint. Yes.
Alright. You're watching Textron Gang. Join Cruise Con Virtual on May 22nd, 2025 for breakthrough strategies to address advanced threat intelligence, proactive incident response, exclusive bonus material and regulatory adaptation here from our keynote speaker, Admiral Michael S.
Rogers, former director of the National Security Agency, and an outstanding lineup of industry experts as they navigate emerging threats. The core principles of crisis management and the evolution of CISO leadership. Register now for free.
Welcome back to the Textron Gang. Well, there's a lot of Talk about AI's carbon footprint. So a carbon accounting firm called Greenley based in Paris.
I've actually interviewed them before they came out with, uh, new research, kind of breaking down the carbon footprint of ai. Not just the training, which we talk a lot about, but the queries everyday use. And they had some interesting findings.
I took a closer look at their findings and their research in this video. AI's carbon cost doesn't stop at training. Every prompt, every task fuels a growing footprint, and it's accelerating fast.
That's the conclusion of a new study from Carbon accounting firm greenly, which shows that emissions accumulate from development to deployments. Building a model like GPT-3 can emit as much carbon dioxide as 112 gas powered cars do in a year. And that's just the beginning.
Envision endless queries used billions of times, it all adds up. Greenleaf's research also shows that not all AI tasks are created equal. 05 kilograms a staggering 60 fold difference.
The gap widens further with added complexity. Supporting additional languages, for instance, typically requires separate rounds of training, compounding the carbon footprint with every expansion. Fortunately though, the study does point to tangible ways to reduce AI's impact, such as training, less often, reusing models, improving hardware and data centers, plus tapping into renewable energy.
Combine that with efficient code and emissions could fall sharply. Reducing AI's carbon footprint isn't about choosing between training and usage. It's about optimizing both.
Well, as I mentioned, Greenly had some suggestions for reducing the carbon footprint of a IU, such as training on 70%, uh, use of the model. That was one of the things that they had listed and many others like leaning into renewable energy, which is something that we've been talking about for a while that all of big tech is doing. But it was an interesting report because it wasn't just from a regular large research firm, it was from people that are actually doing the carbon accounting for, uh, these types of organizations.
Yeah. You know, look, I'm a big believer in reducing carbon footprint, so I don't want this to be taken the wrong way, but I think we'll reduce the carbon footprint footprint when it's economically, um, incentivized to reduce our carbon footprint. So rather than saying do this to reduce your carbon footprint, we could say, do this to reduce the cost of your compute.
I think at least in today's world, it's it's a better selling point than it is. Do it for the right reasons. Reduce your carbon footprint for Regulation In the case of GreenLake, 'cause they're working with lot of European, right?
That's true. Yeah. Eu Yeah.
So at the risk of, you know, confirming that Alan might have been right about something if you were looking over at, uh, open compute project, I think two weeks ago now, Google, Microsoft, and, uh, AWS, they were all over there talking about ways to build more energy efficient data centers. And they had like whole new approaches where they're pulling out the, uh, the electrical conduits. There's gonna be in a separate cabinet from the compute.
They had all kinds of interesting designs that would be essentially more energy efficient. It might take four or five years to get there in terms of redesigning data centers. But it was interesting that there was a lot of work in this area.
So maybe we will get to the point where we are more efficient in the consumption of the energy to drive these AI models. And then the other thing I would just point out is, um, you know, the folks over at OpenAI we're reminding everybody that you should be rude to your AI model and not type thank you all the time. 'cause it sucks up energy.
No problem. That's a good point. You know, With regard to the video, what I found interesting or a couple of things.
One, the comparison of what it takes, uh, energy-wise to produce images versus text. Uh, in the inference side, not even the model side. Uh, the other thing I think is really interesting, uh, didn't get, uh, pointed out because it wasn't directed to any one particular model provider was the impact of deep seq once again on the industry, right?
Because what they recommended, or what Bonnie said was recommended about the reuse of models. I mean, this is Deeps Seq 1 0 1, right? It's like that's a, I mean, they actually now can put themselves in yet another camp, which is with, with a green AI camp, right?
Because we did exactly what Greenly is recommending with regard to how to produce a model and make it available in a economical and, uh, also, uh, ESG efficient manner, right? So really interesting that it, uh, it, it wasn't the place for it to get mentioned, but in another report of similar ilk, you would probably hear deep seek getting mentioned again for its, uh, innovation. But, but Isn't that the way of developing markets and developing technology is as you, as they develop, you bring in efficiencies you would hope.
I mean, if, if there's a reason to, right? I mean, why does, why did automobile manufacturers stop creating, you know, uh, babe from Supernatural, right? The, uh, the, because you know, where are the muscle cars?
The muscle cars are today, right? Not, uh, ecologically friendly, so to speak. Uh, and, uh, you know, they, they become a hobbyists, uh, in invention, but not really practical for every day.
Now, when did that happen? When we hit peak oil, when the cost of fuel and the cost to manufacture gasoline reached a certain point, it became, you know, and then all, and there was some support at one time in this country for, you know, not producing a larger carbon footprint. Of course, now we're probably reversing that because you know, what the heck billionaires run the planet.
But, you know, you're, your analogy is a smart one, I think because, you know, if you look at cars, I'm a car nerd. Um, one of the reasons I think that we've gotten so much more energy efficient as well is because the technology has improved to such an extent. I was looking at, uh, a recent, uh, just a mainstream car model that a normal person would buy.
3 liter, three cylinder turbo engine. 3 liter, three cylinder turbo powered SUV, they would think I was crazy. But the truth is, the technology has gotten so incredibly good that you can deliver 32 miles per gallon from the Buick, whatever it was.
And, um, that can be a perfectly satisfactory car from somebody who doesn't know anything or care anything about cars. And I think that that's kind of the analogy for deep seek, right? The deepsea was able to basically do the same thing with a lot less.
And if it's good enough, then nobody cares about having a V eight anymore because it's good enough. But to Alan's point about, uh, cost as well, I think one of the challenges, one of the biggest problems with AI right now is that effectively the cost is zero because these things are so incredibly subsidized by people who want to see AI succeed. And so you have these companies, um, raking in VC money and using all that money to build these models and to deploy these systems and, and, and they never ever have to stop and say, wait a second, we're using how much power?
If that costs me how much dollars? They don't care about the dollars. If you've got a billion dollars in VC investment, who do you what?
You know, what are you gonna do? You're gonna fret about A few watts. But, But I worry about it.
But again, this is the natural way of things. This is the natural way of things. They will start carrying, the VCs will start carrying the VCs.
I mean, it's the market at work. They'll start rewarding companies that are doing it more efficiently, Especially when they have those, um, you know, conflicts with the local power companies and things like that. And that gets into the news and that's, you Know, yeah, no, it, you know, positive.
Look, I re I've said this before. I remember interviewing the two CEOs of two widely successful, no early on no SQL databases. And I said, there's no sql, Stanford, no security.
He said, we'll have security. When our customers demand, we have security. AI will become efficient when customers and investors demand, it's become more efficient.
Well, it actually may be, again, the laws of physics that forced them to do so. Here, there is a cap on how much energy you can actually suck, right? And, uh, and without nuclear reactors now, you know, as a widely available source of energy, um, you can, I mean, shoot, uh, if I was Texas, if I were a company, I would not look at going near Texas with my, with my data center, right?
Because the last two years winner, they've lost power for how much, how long. But yet they building, you know, uh, Amanda, our managing editor says that San Angelo, Texas is on its way to becoming, you know, the data center capital of the world. My my youngest son's out in Abilene, Texas, they're building huge data center out there, but they're using wind energy because I was gonna say, they're not on the, they're not on The Texas grid.
No. In tech, well, the biggest, fastest growing energy in Texas is not oil or gas. It's wind and solar in spite of what the root and toot and cowboys may want to tell you.
So, Which makes perfect sense. 'cause it's hot, it's sunny, it's dry, and, and It's windy out and it's plenty out on the planes. It's wide open.
So I mean, you go out to Abilene, Texas, you see wind turbines as far as the eye could see. And, um, it'll be interesting. And I, I, I'm not counting out nuclear either, jp.
I, I think this may be a catalyst to, to revisit that. And, and you know, we, Steven, we've discussed it on our show before. There's, there's new generations, there's these mini ones, the thorium, there's all different kinds of technologies we could bring to bear here.
I don't think that, I don't think the fat lady is sung yet on that one. So it was fantastic. It takes up space.
That's the only downside to it, is the amount of real estate it requires, but it, I mean, it can generate entire cities worth of energy. Sure. You know, I saw an ad the other day, someone came out now I think what they were Korea from Korea, transparent solar panels.
You could have every building become self-sustaining with that. I didn't watch the video to, to find out if it was powering the train or not, but they were putting solar panels in between the tracks on trains And, uh, yeah. And they're putting 'em above, uh, parking lots in, in, in France above canals and rivers.
Um, all sorts of incredible ideas. Um, yeah, I don't think solar is the end all be all, but I think that it's just such a promising technology. And, and I agree with you.
I mean, I'm gonna shock you, Alan. I do agree that, that nuclear, Wait a second way. I nuclear you both agree with me in one show.
Wow. I know, I know you're gonna, you're not gonna invite me back. No.
I'm wondering if something's gonna fall on me or something. All right. Must have had a good night's sleep.
Anyway, Bonnie, thanks for a great episode on that. As usual, with all of Bonnie's, uh, videos, they are at Echo Tech, echo tech insights, and you can get them yes on text, TV text, that's right. Or our text drum tv OTT channel.
If you have nothing to do, you wanna watch that laying in bed at night? How about it? Um, but I think that's gonna call a wrap on today's text.
Strong gang, thank you for watching. Steven, thanks for a great report in participating. Jp, my friend, always a pleasure seeing you.
Hopefully maybe this week in Orlando, Mike, we'll be in touch. Bonnie, thank you. Thank you.
Remember, we've got a full text trunk TV lineup immediately following this, so stay tuned for that. And as Steven mentioned, if that mobility field day review, uh, wet your whistle, you can go check them out on Textron tv or the OTT app right now. Until tomorrow, this Allen Schmo for Textron Gang.
We're outta here.


