Episode 18: The New Frontiers of Telecoms: AI Traffic, Drone Detection, eSIMs & Mobile Money

Peter Jarich (00:00)
Hi everybody, and welcome to Intelligence and Noise, the GSMA Intelligence podcast, where we look at the week's news, give you an idea of what's important, maybe point out some things that you didn't notice, and try and leave you with some intelligence, sift through some noise, as the name implies. We do it through the lens of the GSMA CEO newsletter. I say this all the time: if you're on that distro, congrats. It means you're important. If not, don't worry.

It's the GSMA's opportunity every week to give you an idea of some things that are important from the desk of our Director General. And Tim and I. Hello, Tim. Good to have you as always.

Tim Hatt (00:41)
Hey Peter, how are you?

Peter Jarich (00:42)
Good, good, good, good. Tim and I and the rest of the GSMA Intelligence team are lucky enough to be part of the team that helps to sort through the various stories. So we're here to talk about some of those stories, give you an idea of some things that maybe didn't make it in, but probably could have.

So I'm Peter Jarich, Head of GSMA Intelligence. Just gave you a clue. This is Tim, Tim Hatt. Tim heads up consulting and a lot of the work we do on the content front. He's here with me as often as he can, as long as our travel schedules allow.

If you've been listening to the podcast, you know that we've got this week the Nova Summit in Napa, California. That's why you see this wonderful, beautiful background with stone and greenery behind me and not my usual home office. If you're lucky enough, some wild turkeys will make an appearance. I mean, wild turkeys, the animal.

When you're in wine country, not whiskey country. But no real wild turkeys. It's a beautiful setting and it's going to be an exciting time here. And I recognize I've talked a bit about it, and I try to kick off these every time with a little bit of, as my wife calls it, a story, but I think it's worthwhile just making again another connection here to why we are here in California, why we're here on the West Coast with this Nova event.

And, you know, to some extent it doesn't necessarily seem like a natural fit, right? And you see this at an event like MWC, where you've got the telco world and you've got the tech world, and the tech world's worried about things like AI and scaling AI and building data centres, and the telco world is worried about what they're doing with 5G and are we going to make money from 5G and the basics of running what perhaps some folks in the tech world see increasingly as a commodity or as a utility.

But if you take a step back, the reality is a lot of what takes place out here, a lot of what is being driven increasingly from the tech sector, increasingly from the folks in Silicon Valley, if we think about things like AI in particular, are driving what takes place in the telco world, right? All the advances that are driving the industry forward have an impact on how telcos need to build their networks, how telcos need to run their businesses, how they need to serve customers, but also vice versa.

And this goes back to we've all talked about the telco for AI, AI for telco side of things. How the telcos build their networks will also play a large part in how some of those new services get supported, right? How AI gets supported across those networks.

I'm lucky enough I'll be talking to Verizon here on a fireside chat on the Tuesday. That's the topic of conversation. It is how do we think about our networks? How do we think about what we are doing from a technology angle? Yes. From a strategy angle, yes, but from a chief product officer angle of how do we roll these things out in a way that supports everything that we're trying to do here on the West Coast.

At the same time, earlier in that day, I'll be part of, I'll be running a 6G conference session. They call them boardroom sessions. And there again the question is how do we need to think about 6G, not just as the next G, not just as an evolution of what's going on in 5G, but as a technology that can benefit from what's going on, benefit from advances in AI, in cloud technologies, benefit from all the amazing innovations that come out of places like Silicon Valley, but also is ready to support them and, in the process of supporting them, maybe help the telcos earn a little bit more of a return.

And maybe this is all obvious to all of you of why we're doing this. I think sometimes I've gotten some questions of what is the point, but it's those synergies, kind of bringing those perspectives together, that I think will be really useful if this thing works out.

It's why I'm here. It's why I've got my Golden State Warriors hat on, it's why I've got my Cliff Family Winery jersey on. I'm trying to play the part. I've even got my glow-up kombucha, because, you know, it's California. So, you know, I'm trying to blend in as much as a kid from Wisconsin who's been living in the UK for almost a decade can do.

But it will be, it should be a good couple of days. Tiring, a lot going on, but excited for it.

Well, with that out of the way, let's get to the news.

So this week, two stories made it into the CEO newsletter. Again, congrats if you know that already. Number one was about the SA roaming offer that was launched between T-Mobile US and Reliance Jio, supported by our good friends at Syniverse. The other was the support for ISAC, integrated sensing and communications, really sensing over 5G in Finland by Ericsson. And the one we're going to talk about is the latter.

But before we talk about the latter, on the ISAC side of things, I just wanted to give a nod to the SA roaming piece, in part because it was definitely an internal GSMA fan favourite. A lot of people from across the GSMA sending notes going, "Ooh, ooh, this is a story. Have you seen this? Have you seen this? This should really get in the newsletter. This is really big."

You can imagine it's very inside baseball GSMA, standalone roaming. And I think it's important, right? Because it is the first of its kind, as they call out. It means that you don't have to fall back to non-standalone. We're going to make it quick.

I think it's great that we're finally seeing this take place. And of course, we all know SA, and I talked about this. I'm going to try and make this connection as much as possible.

SA is really supposed to be there about supporting those enterprise IoT applications, enterprise use cases. Those could include, obviously, going forward as we think about things like AI and slicing. So it fits into what we're trying to do. Critical for 5G monetisation, but we're talking two countries, right?

So it's great. Folks who are going from one operator in the US to one operator in India. Great. So for those enterprise use cases where you've rolled it across the US and you're using T-Mobile US and you've got applications and you're on the Jio network in India, you're set, right?

That leaves hundreds of other operators, hundreds of other markets yet to be uncovered. And the reality is, you know, our numbers show, I say, still 20% of operators, thereabouts, which means the agreements to set SA roaming are even rarer.

It's great. It's a sign of what needs to take place. It's also a reminder we need to make much more progress.

Tim, before we move on to the bigger picture, anything? Are you going to sign up for your backup SIM on T-Mobile US and your backup SIM for Jio for taking advantage of that?

Tim Hatt (08:09)
Yeah.

Peter Jarich (08:10)
You are, you are, you are?

Tim Hatt (08:13)
I think you laid it out well. I mean, it's not a game-changing move, but it's an important move to validate the international, you know, cross-border activity that comes through roaming.

But ultimately this is a scale game here where, you know, only around, as you called out, 20% of operators who have launched 5G do have an SA network. So it is still a minority.

I think the interesting thing here, Peter, will be whether we start to see other operators launch these types of roaming deals that can catalyse a virtuous circle. Because ultimately you need the network effect to make this work, where you could say, "I'm going to take advantage of a stadium use case. I'm coming from Washington State to go to see some cricket in India. Or I'm an enterprise customer."

But unless you have the scale across the networks, it's not going to make a huge dent. So something to watch.

Peter Jarich (09:02)
Yeah, no, that's, yeah, I think that is the hope, right? This provides a reminder, it's a catalyst.

Obviously, this is being done with the folks at Syniverse, and I was talking to them last week, partly just to catch up, partly as a congrats on this. And, you know, it's obviously something that other folks can take advantage of.

So it's not like it's, you know, it's not like it's brain surgery or something that would be impossible. I just realised, though, maybe I should take advantage of this. I mean, I'm going direct from Nova here to our board meetings and working group meetings in India.

So, in fact, if I could find myself a T-Mobile SIM, I could possibly take advantage of it. So I'm not really sure what benefit I would get, but it's maybe just breaking rice.

So let's move to the story we do dig into. Do you want to dig into it a little bit? It is Ericsson's demo with Finland's mission-critical comms operator. I'm not going to even try and pronounce the name. Why not? Erillisverkot. Someone, if you're out there and speak Finnish, let me know how bad I got that.

So they rolled it out in Helsinki, and it was basically showing as many of the ISAC use cases and demos now around tracking, locating, positioning, detecting drones. Not drones that were connected to the mobile network, but specifically drones that weren't connected to the network.

It leveraged a whole bunch of technologies, right? You want to get decent granularity of the signals. We're talking about massive MIMO, beam steering. There was some edge compute in there, obviously the ISAC sensing software, and then detection and classification of things like AI.

Ericsson positioned this, as is often the case for any of this ISAC stuff, for airports, borders, any critical infrastructure.

Tim, this isn't the first time we've seen this. Your first thoughts when you heard this? Just another, just another, just another.

Tim Hatt (11:06)
No, I don't think it's just another, just another. I think this is actually pretty significant as a proof point.

And, you know, you called it a what's interesting here is it's a demonstration of a public safety use case. That is changing how we think about the value of a telecom network.

I do think, Peter, I'd call it a couple of points of significance just to build on what you said. You know, one of these is that when we look at ISAC, first point is this happened on a 5G network. It didn't need to wait for a 6G network.

So why does that matter? You can use existing investments and you can preview how you might monetise the 6G infrastructure.

Peter Jarich (11:43)
And just taking a pause on that, I think it's important to recognise, right, that we're talking about this now, but in February, Ericsson demonstrated ISAC at their headquarters in Plano.

In July, we saw Cohere get a $28 million grant. Grant might not be the right word. They got money from the US Department of Defense to build a multi-waveform ISAC RAN prototype with the idea of commercialising quickly, not necessarily waiting for several years down the road.

Tom Rondeau with that FutureG division of the Department of Defense will be here at Nova. So it'll be interesting to talk with him.

Not long after that, Ericsson worked with AT&T to roll out or demo this at AT&T Stadium. Earlier in September, Verizon did some demos. They did it with Ericsson, they did it with Samsung, did it on Qualcomm's campus.

So just to your point, we're seeing this, and the whole point is we talk about ISAC as a 6G technology, but you're right. I think all this, the key message is people aren't waiting.

Tim Hatt (13:02)
And it will happen on 6G, but the point is, to your point, these deployments show it doesn't need to wait for 6G, and that's important for the industry as a testing ground.

I think it's also significant, the fact it was Ericsson here. They have positioned themselves as an ISAC leader, really a trailblazer. We're starting to see this in action.

And then thirdly, the fact, and I'll be interested to get your reaction on this, that we are seeing, as much as this is a public safety use case, you can see ISAC expanding into a number of verticals.

Whether that is through drone detection, whether it's through precision agriculture, whether it's through smart city-style applications to gauge population densities.

I think the key here, Peter, will be, as ever when we look at this through a business mindset, as it comes time to scale, can a vendor like Ericsson and a telco partner look at a monetisation model over the long term? How do you see that?

Peter Jarich (13:56)
Yeah, no, I think both of those points, right? On the Ericsson front, it was probably telling that I listed out all those things that we've seen in the not too recent, not too distant past from ISAC demos. Ericsson clearly part of a lot of those, right? So they're making a lot of this.

We had some folks out at their recent analyst event. ISAC was a big part of what they talked about, particularly in terms of federal communications, in terms of Mission Critical Communications.

So they've got their whole messaging there in terms of a timeline: POCs, demos this year, small-scale deployments in 2027 and large-scale deployments in 2028. Right. So they see this in a couple of years going large scale, and then 6G coming in 2031-plus, which, if you believe it's 2031 and not 2030-plus, you really see the need to roll this out today.

Roll this out and support it. So they really have this vision.

It also includes, I think we'll see this more and more, this idea that ISAC isn't just leveraging the network. Yes, it's leveraging the network, but as they talked about, you know, you're going to be using AI and ML to actually do some of the detection based off the sensing.

But the data feeds are going to be pulled from not just the mobile network, but the mobile network along with cameras, along with radars, along with a whole bunch of things.

And so Ericsson is clearly making that noise. I think you're right in terms of we could go down the road and see some other use cases.

What I see happening with a lot of this, and why I think it's particularly important, is that the number of times, and it depends on who you talk to, but a lot of times if you're talking about ISAC, I find I get sort of one of three responses from folks who know nothing about mobile networks and technology.

Like, "Wow, that's cool, you can do that." And then maybe it's scary. Great. Will the operator actually know where I am? Right, down to tracking individual people. And we've talked a bit about that, about some of the obstacles in my face.

Two, you know, people, folks who really love the innovation and how networks are going might be, "Yeah, that's cool. And I can see some opportunities."

But I think more often than not, I see a lot of folks go, "It's just going to be a niche, we're talking about it. I don't think it's going to go anywhere."

But there's a lot of naysayers, right? And I get it. Whenever there's a big buzz and there's a lot of buzz around ISAC as one of the really new 6G use cases, there's naysayers.

But if you do think it's an opportunity and someone like Ericsson obviously does, then you really need to show off. You need to show off what's possible, right? You need to show that it works, you need to show the use cases, you need to give an idea of really what's possible.

And I've seen that every time I have a conversation with my friend Ronnie, Ronnie Heraldsvig from Cohere, obviously another big ISAC proponent. He's always going towards, like, the distance of, "Hey, in the future you could do this, and cars could help. You know, cars may not know what's around the corner, but if the network could tell them because it's using ISAC, you could do this."

And for me it's a little bit getting ahead of ourselves. And so what I think I see here is trying a combination of that. Talking about what's happening in the future with 6G, but using 5G technologies, showing it, but perhaps most importantly, and yes, there might be those other mission-critical use cases, linking it to something that people care about and are willing to pay for right now, right?

And that is defence. It is security, because yes, we can see digital twins and environmental sensing and smart cities and traffic monitoring and people monitoring, and there's all sorts of things.

But if we want to get it going now, focusing on that critical infrastructure.

And what's interesting is it reminded me that, you know, two things, and we'll end this with this because we probably want to get out. We don't want to speak too long for this, right?

For one, these two angles here that I'm not seeing that I would like to see a little bit more is one: how do we see some of these moving beyond the US?

Right, a lot of those demos that I mentioned, a lot of that interest seems to be coming from the US. And maybe, I think as we've called out, there are less privacy concerns or things you can get around in the US. We're sensing this less sensitive, so to speak.

There might be some opportunities there. In completely unrelated news, German investment fund DTCP closed their first defence-related fund. So it's an investment fund, around $455 million in commitments. Deutsche Telekom was one of the companies to go in there. There's even a quote from Tim Höttges on the launch.

So there are defence interests outside the US. How those actually tie into this, I think, is there.

And then, you know, the other piece that I don't think that we see nearly enough is that this also goes back to the whole idea of network for AI and physical AI.

We sometimes don't think about these things as physical AI, and I think physical AI is another thing like ISAC, where people go, "We're all talking about physical AI and robots because they're sexy and cool, but how big's the opportunity?"

And sometimes people poo-poo it.

But this is another demonstration use case example of ultimately physical AI, using AI to deal with the physical world, provide an insight into the physical world, with network supporting it, telco networks.

And so I think we need to cast it in that way as well, which again, you can tell I think it was a particularly interesting story.

Cool. Let's close this. Yeah, shoot. Shoot.

Tim Hatt (19:39)
Just one other thing, Peter. Well, just one other thing before we go on to the next one.

And just I think you make a really important point on the networks for AI. The other thing, though, is if you strip this back, ultimately you've got to have a buyer for the technology. You acknowledge that. I acknowledge that we need that.

Interesting thing in the announcement here from the client at the Finnish state-owned critical comms operator is that they can integrate this technology to their command and control centre.

Why does that matter? You have a long-term anchor tenant for the service, and you then modularise the mobile network as part of the integrated service.

I think that's a real change from how networks have worked in the past. And if we can see that here, the chance of scale is much higher.

Peter Jarich (20:22)
Hundred percent, hundred percent, I think. I think if you do have some issues or concerns around ISAC, I would say the commercialisation, how people make money, and again, this is part of the conversations that I've had with folks: excited about the technology, cool about those implications, but who's going to make money?

And there's oftentimes, say, "Well, we've got to get excited to get this stuff out there." It's like, no, telcos need to understand if they're going to invest in this. How do they make the money? And if there is this opportunity that they're going to lose out?

Again, in a market like the US, where the administration is particularly strong, there's an interest in getting in with the Department of Defense, there's a lot of interest in showing that off. I get that. We need to see that elsewhere.

Tim Hatt (20:59)
Yep. Yep.

Peter Jarich (21:01)
Second story that made it into people's inboxes this week: the launch of a travel eSIM offer by Deutsche Telekom.

Tim, tell us a bit about it. What is it?

Tim Hatt (21:16)
Yeah, really interesting story from Deutsche Telekom. And I'll just frame the news first, and then we can get into the so-whats here.

So, just for our listeners, as Peter said, Deutsche Telekom has launched a new Travel eSIM product, T Travel. It is available in more than 200 countries, at least five languages across Europe, including, of all languages, Croatian, because T-Mobile has an opco there.

You can get data from between one and 30 days and choose your flexibility. And, it being eSIM, the difference here with a traditional travel SIM is that it activates seamlessly on your phone once you hit the country that you are travelling.

There are a number of roaming partners obviously established to support the service here, but it's a global product, flexible data, works through eSIM. Initial reactions to it?

Peter Jarich (22:15)
Yeah, yeah. I mean, obviously, to your point, global, 200-plus destinations, 600-plus operator partners. Classic.

I mean, I think it's a travel eSIM offer, right? I think at this point people are fairly familiar with the big names who are dedicated players: the Airalos, Holaflys, Nomads, Saily, some names. When I was just doing some research on the big names, even familiar with, like, Ubigi, MobiMatter.

So, yeah, it's just another. But I think it's important to recognise that, you know, they're not the first in terms of telco, right?

So this is a big, you know, as we've worked with operators and we've talked with some of the big ones, travel eSIM has been on the radar for a while as, "Is this going to help? Is this going to hurt us in terms of taking away some of our roaming revenue?"

How worried should we be?

As this goes more mainstream, as we see particularly the big names like Holafly and Airalo gain traction and become consumer brands, right?

I mean, I got to the airport the other day and there's a big ad for Airalo, right? Going to Paris. "Hey, Paris is better with Airalo."

You can't watch a piece of BT content. Yes, I'm one of the people that actually watches things on the screen, doesn't bring my own content. And so I have to endure the Holafly ad.

The ad isn't bad. It's the guy running the ad has gotten better over time. But they've spent I don't know how much money, they've spent a good amount of money to put ads for Holafly at the beginning of all the content.

So they're becoming consumer brands. Telcos worry. And I think one of the responses and pieces of advice we've had is, look, you have relationships. You have roaming relationships. You can do this if you wanted to.

And the reality is lots have, right? So, yes, Deutsche Telekom has, but Orange has, Telefónica has an offer, EE, Telstra, Singtel, SKT, Docomo, Orange, too. I mentioned Orange before. I mean, there's no shortage of operators who have already done this, right?

So, you know, the interesting question, I guess, is okay, so now you're just another. It's great you can compete. It's great you can now say, "Hey, you don't need to go use Holafly, you don't need to go use Airalo."

But those guys are doing a good job of advertising their services. What's the differentiation? Like, why, as a consumer, or if you're giving advice to Tim, like, how do they differentiate? What is the way to get folks to go to them?

Tim Hatt (25:07)
Well, it'll ultimately come down to the simplicity of how you activate the service, the global coverage and price. So three factors: simplicity, coverage, price.

The challenge they've got, just to play kind of devil's advocate here, you're right. A lot of operators have launched travel eSIM. DT is far from the first, and there is a long line of non-native telco competitors.

The challenge they've got is eSIM is still a small share of the base.

Right now, globally, we estimate it at around 5% of smartphone users. Okay. As in 5% of people who own a smartphone have an eSIM smartphone.

That will rise by the end of the decade. We expect it to get to around 55%, largely driven by the iPhone base, which went eSIM-only in 2022 in the US. But that's going to take time.

The other thing is that it takes a long time for people to shift their buying attitudes.

In our global consumer survey, and you can check this out on our website, only around 10% of people who took a trip last year used eSIM. Right. So even though the availability through the providers you talked about is high, people still default to what they know.

And so I think the significance is you've got to differentiate it in a different way through the rates and the ease of activation.

Peter Jarich (26:28)
And you can take a step back on that, because I know that 5% number will sound really low to people, but to your point, it was just a few years ago that Apple went eSIM-only, but just in the US, so then it took some time to roll it out elsewhere.

So you're talking about newer Apple devices, you're also talking about global, so those are global numbers, right? Obviously, in a market like the US, much more developed markets like Europe, you're going to have a higher penetration.

And it's interesting, Scott, go back to that point you made in terms of people sticking to what they know.

Doing a little playing around, sort of research, trying to play with some tools, throwing out some prompts to say, "How would I evaluate the value or comparison of different travel eSIM providers?"

After playing around, you've got a number of different sort of criteria, right? So to your point, coverage, geographic coverage, simplicity, which is kind of distribution breadth, product breadth, right? Does it have what you need?

You talked about price. As long as they're similar, I'm not too worried there.

But one thing that really came up that was interesting is sort of ease of use of app. How is the app? How popular is the app?

So it really comes down to marketing. So I think as I think about, to some extent, what telcos need to do in terms of getting this out there, it is just marketing. It's like, "Hey, making it clear. Look, if you're doing this, stick with us."

You don't have to, hey, you know, you could be a non-DT customer and use our Travel eSIM, but, one, getting that news out to their embedded set of customers, but I think there's a bunch of other things they could do.

And we've been talking a bit about this, right? It's, you know, yeah, you can compete on price, number of countries, but competing on quality, speed, network selection. Hey, we've got more partners, we can do better network selection, right?

Playing to some of the things that they can do.

Most of the travel eSIM is really, you're looking at those data packages, which is what most people care about. If you can figure out how you tie together that identity in a continuous way, particularly with voice, right?

Where you go, great, travel eSIM provider is wonderful, but if you turn off your home SIM and you have no connection to your home SIM and you get some SMSs or voicemails or you need to call.

Does it happen that often? No. Do you want that peace of mind? Maybe.

Do you want to make sure you keep your existing phone number, all your voice services, all your security that's going through the home-routed network? Right. We do a lot of stuff when we route stuff through your home network.

When you're just doing regular old travel eSIM, who knows what network you're on? I'm not saying to play up the FUD side of things.

And then, of course, there's the enterprise side, right? So on the enterprise side of things, you think about APIs, you think about slicing, you target enterprise travel.

I think there's a lot of ways where, if they're serious about this, the trust and the roaming and the connection to home services are something that telcos can play to.

Doesn't mean they will. It could just become a tick: we're going to compete with those guys, we have our travel eSIM offer on our job.

But I think there's a bigger opportunity. And then likewise, bigger opportunity for the dedicated travel eSIM providers to look at that and go, okay, maybe we need to be thinking about those same things.

Tim Hatt (29:50)
Yeah, and I think it's maybe a last point: it's interesting that in this launch, DT has made this service available to their competitors' subscribers as well.

So you don't have to be a DT customer to use it. You can be a Vodafone customer. You could be an Orange customer.

They still obviously need to market to those people, but the fact that you're opening this up to the wider market shows they're at least serious about scale. We'll see where it goes.

Peter Jarich (30:16)
Hundred percent. Hundred percent.

Now we get to our favourite section, the unofficial Willie Nelson section. What's on our minds, on the road and on our minds.

We know where I am now. I'm on the road. Tim and I'll be soon together in India with a bunch of the team for the board meetings. So the on-the-road part is out of the way.

On our minds, I'm going to kick off. Meta Connect put forth a whole lot of different things, including a bunch of glasses.

They really pushed sort of, they've got glasses without cameras, better ones with cameras, and there's a whole lot of discussion of, hey, does this mean a non-camera set of glasses? Is that they're hearing people's privacy concerns and they're taking—no, no, that's just silly, right?

I mean, if they only started offering them without cameras, rather, glasses without cameras, if they only were doing audio-only and had no cameras, then maybe. But they're still offering the ones with cameras.

They recognise that heck, a lot of people just use this for the audio anyways. And heck, if you only do audio, you can do smaller, thinner devices, better battery life, cooler form factors.

But the big thing seemed to be, more than anything, is tying this all to Muse.

Muse being their personal AI agent.

And what's interesting there is, if you're not in the US, you probably haven't seen this. There are, I've seen multiple TV ads just in the time that I've been here. They're doing a big push on getting it out there.

And early, early reviews say it works. I started playing with it.

You can't use it if you're not in the US. I don't know if this is just the early nature of it or if they just don't care that much, but you can, if you're not in the US, use a VPN.

I was like, this is not going to work. I'll just attach the VPN to the US and it'll probably say, "Nah, nah, nah, Peter, we still nope, nope." Like, sure. Hey, so a tip: if you're not in the US, feel free to play around with it. It works fairly well.

And probably not too surprising. They are a good consumer product company. They know how to roll out consumer products.

It works really well. And I think the interesting part will be to see then, okay, well, what does this mean next, right?

Because the glasses are sexy, right? The glasses are cool, but they're going to try to connect this to this.

They also rolled out the Muse Charm. You've probably seen it. It's probably memeified by now. It's a tiny little thing. Looks like probably the size of a little bigger of an Apple Watch.

And all it is is it's a device that's meant to be your agent. All right.

And just to take a step back here for folks that don't know, I mean, we've seen a lot of folk focus on agentic AI. We've talked a lot about how OpenClaw has taken off, particularly in some markets, people going, particularly in China, having it loaded on their laptops, loaded on their computers.

But that's more of like, if you know what you're doing, you can then use this. This is the consumer side of things. We can just ask it questions.

It's like it's like agentic AI meets vibe coding. It's much more user-friendly.

And then you're going to have this charm thing. I guess my question is that thing will obviously need to attach to the network. It's got 5G in it.

What does it have in it? So, two-inch touchscreen, front and rear cameras, 5G, fingerprint sensor. In theory, it will be available in December.

It feels like the classic stuff's running in the cloud. The processing, it's a fairly dumb device, the processing in the cloud, which means you need a connection there.

What does this mean? I think for me it's interesting because we see where things are going with AI, but it does come back to this idea of the network being particularly important.

Tim Hatt (34:32)
Yeah.

Peter Jarich (34:32)
It does.

Tim Hatt (34:44)
Well, I think it's going to take time. But short story, yes, in phases.

So if we step back right now and we track AI traffic with a number of mobile operators across the world, and, you know, right now it's kind of two stories.

It is a low share. It's just about 0.12% of total cellular data traffic at the moment, but it's growing very fast.

And we look at the last six months, it's gone up by 3x compared to just slightly rising traffic overall. So it's outstripping it, and that's because more people are using AI. They're using Claude, they're using ChatGPT.

But right now we're still seeing broadly low-bandwidth use cases.

As you look ahead, you start to get into more video, you get into image analytics, higher-processing applications like AR glasses, like VR.

That load potentially goes up.

And the question then is, no matter what it gets to as a share of overall traffic—and just for reference, video is about 65–70% of mobile traffic right now—it's nowhere near that.

Whatever it gets to, you've got to have intelligent ways to process that traffic depending on where it is in the network. You know, at the device edge, the network, in the cloud.

And right now I think the implication, Peter, is: do you get into intelligent offload mechanisms to allow that to happen between the device and where it can seamlessly switch to the network or the cloud depending on the use case?

Peter Jarich (36:15)
Yeah, I think it'll be interesting.

You know, let's go back to the glasses, let's go back to Charm. I can't see this driving tons of traffic.

The question, though, is: is it important traffic? How do you deal with the new sessions? How do you deal with it being always on? How do you think about where the traffic is placed?

Yeah, I'm, it's, it's, we're getting to the point where, you know, our numbers are showing that.

I think you said this from the work we're doing with the AI Observatory. We've seen the numbers from Ericsson, which showed that uplink is growing much faster than downlink on the traffic side of things.

This is obviously going to push that. And whatever you're saying in terms of thinking about AI traffic, I think it's thoughts about how we need to architect the network and rethink the networks.

Tim Hatt (37:08)
Exactly right. Yeah. And when we look at—we've actually done some interesting benchmarking—the silicon flagship chipsets can largely handle AR glasses right now, but when you get down to the rest of the base, not so much in the chipset.

So there is some—

Peter Jarich (37:21)
Yeah, and then the last thing—

Tim Hatt (37:22)
architecting across the use cases.

Peter Jarich (37:25)
And last thing before we jump off this, I would just say is, you know, I think it's going to be a lot of folks that see this and go, "Great, we've tried this already with, now, we tried this already with the Humane AI Pin, we tried this with the Rabbit R1."

I think the difference here is those were sort of hardware solutions that the software needed to get fixed.

This is really coming from a software position. So I do, for those folks who are also probably naysayers on this, maybe give it a little hope.

Tim, what's your last story that we're going to go out on?

Tim Hatt (37:56)
Airtel Money IPO. Significant move.

It's not new, but we're seeing a move to London. I think what's interesting here is that they've disclosed, as much as it was announced earlier in the year, what the IPO is looking to value at: about $8–9 billion.

Two, that the IFC, International Finance Corporation, will be a cornerstone investor, converting its existing capital to the IPO.

And three, I think it's interesting, Peter, just because it highlights kind of London as a centre for investment towards emerging markets.

We see so much AI floats and driven IPOs in New York, to a certain extent Singapore and Hong Kong. London's kind of missed out on some of that.

So now we see it here happening as a fintech mobile money play.

I think it's interesting to show the depth of institutional capital that can fund these services. And it's a big growth area for mobile operators, particularly in Africa and India.

Peter Jarich (38:57)
Hundred percent, hundred percent.

We've been talking about this for a while. It's good to see it sort of move forward. So I guess we'll watch it.

With that, we are coming to the end of our time. We've come to the end of the news.

I would say, folks, I would ask you to check out what's taking place after Nova, but it's Chatham House rules, so maybe you'll see some news. We'll talk about it a little bit more, but with limited details there.

But I will say, check back with us next week. Tim and I will be doing this from Delhi, which should be good. We've got some great news coming up.

Again, if you're enjoying this, leave us a review, follow us. If you're not, let us know why. Tell us what you'd like to hear.

In the meantime, I would like to say, as you're looking for the intelligence, feel free. Remember, you need to embrace some noise.

And with that, thank you, Tim, and thanks to everyone else.

Tim Hatt (39:55)
Thanks, guys. See you soon.