Episode 5: Satellite Strategy, Smart Glasses and Data Trends What Is Hype and What Is Real?

Peter Jarich (00:00)
Hi everyone and welcome to Intelligence and Noise, the podcast from GSMA Intelligence, where we try to filter through the noise of what's going on in the telco and tech markets and try to provide a little bit of intelligence to help you make sense of it all. My name is Peter Jarich. I head up the GSMA Intelligence team, and as usual, I am here with Tim, Tim Hatt, who heads up our commercial content. Hey Tim.

Tim Hatt (00:25)
Hey Peter, how you doing?

Peter Jarich (00:27)
Good. I am super excited when this will drop. This will drop just a few days before July 4th. USA 250. Go, go, baby. So this is the point at which I go, it is really, really impossible to see through these glasses. And so I'll leave the scarf on because it is jaunty. And a reminder that I think we're all doing okay in the World Cup, right?

Tim Hatt (00:56)
Well Peter, times have never been better. We're through to the sixteens. You know, tidy win first time in the knockouts, first time in the sixteen, so looking good at this point. Perky.

Peter Jarich (01:07)
Not too bad. And yes, but we're not here to talk about football. We're here to talk about what is taking place in the market that we think is important, that we want folks to know about and to help provide a little bit of intelligence, as I just mentioned.

So for those of you who are new to the podcast, welcome. We try to do this through the lens of the weekly CEO newsletter that comes out from the GSMA and GSMA Intelligence team, helps to support that, looking at what got covered, what we thought was particularly interesting, and then what didn't get covered, what probably should have.

But I'm gonna start as I usually do with a bit of a story. So we'll talk a bit later about the travel that Tim and I have both been on — Shanghai for Tim, Copenhagen and DTW for me.

But you know, I was struck by a story from DTW and, due to some flight issues, I didn't get there in time for Monday afternoon. There was a great analyst day from Nokia. But I did manage to catch up on the next day with their Chief Technology and AI Officer, Pallavi Mahajan.

And if you don't know her, she's great. I think a lot of you might have seen the news coming out of MWC, all over obviously their strategies around AI RAN, AI Grid. They're obviously doing a lot with AI and the NVIDIA partnership and NVIDIA investment and everything.

So clearly she's taking a huge role there, coming from Intel, HPE, Juniper. But it was a great conversation. I think one of the interesting pieces, even before we got into the discussion of what was going on with the event, what was going on with the industry, is she looked and she asked — she saw this and said, hey, it is a Whoop wrap.

And we talked a bit about fitness trackers and data and then the AI of it all. And at some point it was interesting as we got to talking about how one of the colleagues of the other Nokia folks had actually, or someone they knew — this might be apocryphal, who knows — looked at a way to link their Whoop data.

And you could probably do this with Fitbit or Garmin or anything, with Outlook, with the whole goal being to show which meetings created the most stress. Right? Hey, you know, I have meetings with Tim, I'm super relaxed. I have meetings with Ralph and I'm not. And there's no one that I know of named Ralph at the GSMA, so I'm not spilling any tea there.

And it was an interesting concept, an interesting idea of what you can do with data. And to be honest, for me it's also an interesting data point of how easy this is. 'Cause like, yeah, I mean, it might take you a few hours to do. I went home, I just asked Copilot, write me a script to do it, and Copilot did. And, you know, maybe it might take two hours to implement.

But the interesting part was not just what you're gonna do with this, but with the follow-on, which was I was talking later that day to another analyst who was complaining about how data's being used and how different events use data.

No shade on the GSMA there, actually, surprisingly, because this analyst tends to throw shade our way. And, you know, I was saying, but Outlook, you're okay with? It's like, yeah, yeah, I trust Microsoft. I trust, I know, I trust they know what they're doing. LinkedIn's fine.

And I went back and said, yeah, but let's think about this, right? If I can link my fitness tracker with Outlook in a number of ways, how do I know how my data is being exposed? You trust Outlook, but if people are out there and you can connect these things and, you know, your data is in someone else's calendar because you had a meeting with them, then you really don't know sort of where this data is going.

And it's getting more and more complicated. I think that was an interesting takeaway for me, was how these things are all interconnected as we're looking at AI making coding easier, but making data more accessible, how we need to think about data, how we need to think about sovereignty.

And that probably leads us into the first story. So with no further ado, let's dig into the stories. Tim, remind me, what were the two stories that made it into the CEO newsletter this week?

Tim Hatt (05:32)
So we had a story on agentic AI, which covered a number of announcements, largely coming out of DTW, Mavenir, NVIDIA, and others. And then a second story on India's sovereign satellite strategy, which I think gets into a number of implications there.

Peter Jarich (05:47)
And we'll talk about the DTW stuff towards the end, as I mentioned that. But yeah, I think let's dig in a little bit to the India sovereign satellite story.

So just for context, right? You've got Airtel and Jio who since last year have partnerships with satellite folks, Starlink. But then you've got Jio who then just recently are obviously in time for the newsletter, announced sort of a plan within two or three years to launch a constellation, 1,600, 1,650 satellites, low Earth orbit, 650 kilometres, to essentially provide their own capabilities there.

And that's kind of the crux. I have my own thoughts of why, and I think I see three different explanations for this, right? And the quote that I think was interesting was, you know, Jio connected India on the ground. Now we must connect India from the skies.

So it feels very inspirational there. And obviously it is a big country. I've called this out a number of times, although we often talk about India and China in the same breath, because they're huge countries. They're hugely different demographically, where you've got China about one-third rural, two-thirds urban.

And, you know, India being the opposite, right? About two-thirds rural, one-third urban. So a lot of rural folks to connect. So you can see that side of things. There is the sovereign side of things, which kind of goes back to my story.

But there's also the coincidence, maybe not so coincidence, that we've got a Jio IPO coming up. And obviously these sort of exciting plans help to drive value behind a company.

Which one is most important, Tim? Which is the main driving factor, do you think, or is it just it is all of them?

Tim Hatt (07:42)
It's all of them, but I think the link between the sovereign compliance and the commercial ownership of connectivity that covers the full stack, and I include satellite within that, is critical here.

If you look at the announcement, I mean, first thing to note, India is the world's biggest country now. It has more people than China, it's got a billion mobile subscribers, Jio and Airtel, its competitor, both have hundreds of millions of customers, which means that there is a very big scale story here.

I think as well, when you look at the sovereign side of it though, India is cottoning onto the fact, as have other major jurisdictions like Europe, like China and the US, that NTN connectivity is a nationally strategic asset that should be protected.

And the government is taking a firm stance on that. With a lot of the telco companies, again, not just in India, being important providers and deliverers of that sovereign connectivity.

So I think when you peel it back, it's both a regulatory story here and a commercial story. But ultimately, the timing, call it a coincidence or maybe not a coincidence with the SpaceX IPO, I think is important here because we're seeing a lot of quick reaction.

But yeah, what's your take on that?

Peter Jarich (09:04)
And I think you can, there's some definite indications to the sovereign side of the story, not just being sort of a nationalistic play here to say, you know, and every country right now has a heightened level of nationalism versus maybe eras gone by.

But I think there's a number of things here to say it is more than that, particularly if we're looking at the Jio case. I mean, so if you look at it, it was last year you saw some new requirements being put on the satellite players in India.

You saw, it was just this month, right? India froze some approvals to operate commercially, right? This was Starlink was about to operate, they pulled it back. This was due to concerns about were they operating in Iran, when they weren't supposed to be operating in Iran.

And I think you've seen those concerns around some of the satellite players from India. And so I think, you know, against that backdrop, it's not like all of a sudden they said, hey, we're gonna do something sovereign.

There is this sort of history there of them going, we need to put some regs on this. We need to look at these folks with a little bit more scrutiny and being able to have that sovereign capability.

And obviously they've talked up sovereign in a number of other ways in terms of rolling out data centres and building local tech talent. I think we called that out last time, how sovereign is more than just where data is processed. It's also those local communities.

But I think that backdrop for me says, yeah, the sovereign side of things is more than just — it's not just lip service, right? I mean, I think you see this in the background, though it's not just them.

The EU's 2 GHz proposal, where they've allocated, or they plan to set aside, most of that spectrum for domestic providers. And obviously we know that Elon and the Starlink side of things basically said, hey, this is gonna hold you guys back. You can't do this, this isn't great, this is bad.

But again, that was a focus for keeping that capability, recognizing that those satellite services are important.

And then again, I think for me, the big piece goes back to, and I think this was a bit of the conversation I had a few times at DTW as well, is that we talk so much about sovereignty as where your data is being processed.

And I know I sort of led this off with, hey, it's a data side of things and how your data is exposed. But it's also about one, having control over critical infrastructure, right? Of which satellite is being seen as part of that, to the routing of where your traffic lives, right? And maybe even where that traffic is processed.

And obviously, as you know, Elon talks about space data centres. That's part of it.

So I do give this one, I think there's been a lot of sovereign washing. I think it's very hard to see anything that's infrastructure oriented nowadays that doesn't include the word sovereign in it.

Because it's an easy thing to throw in there and it is such a broad term, as we've discussed, that you can kind of attach it to anything. But this does feel like it's like there's more than a kernel of truth there.

Tim Hatt (12:21)
And I think there is, just leaning into the sovereign dimension, there's another implicit kind of insight from this, which is that, not to be a regulatory geek, but there's something called ex ante and ex post regulation.

Essentially it means before the fact or after the fact. And I think governments, including India in this story, are realizing that with satellite connectivity, if you're gonna have a sovereign constellation, you can't wait for other providers to come in outside your country and then put the regulation down.

You've got to do it ex ante. And so for various reasons of how you actually regulate constellations that by default are sitting above national land masses.

Peter Jarich (13:00)
Yeah, yeah.

Tim Hatt (13:01)
And so they're saying, okay, guys, we need to get this in place now. We're going to use a domestic provider to create a domestic ecosystem and let the situation evolve from there, knowing that there are competitive implications, security implications, geopolitical implications if you do it the other way.

Peter Jarich (13:20)
It does call, I think, into question the question of, you talk about being a scale game, right? Although China and India are the biggest countries, and we know China has a lot of satellite plans as well.

What I think is interesting is, I mean, the whole point, at least as I understood it, of the scale of satellite, is that it is global, right? It's that, hey, I'm not talking about a domestic network. I'm not talking about one network just for my market.

I'm talking about I'm going to scale my investment over eight billion plus people, there's huge scale there. And you see that with what Starlink has done in terms of the number of satellites they've gotten out there and the number that they plan to deploy.

It's interesting then to think about how do you position that then in terms of, okay, here's a constellation that's going to get to — does not even have aspirations to get to a fraction of what Starlink has. But that's okay because its primary goal is to focus on one country.

So I think it does bring in the question of: have satellite economics, the scale requirement, probably changed either because launches have gotten cheaper or satellites have gotten cheaper, or it's just required.

But I mean, it's a bit of a head-scratcher in terms of how we think about scale.

Tim Hatt (14:36)
I think the point on scale is that you have an advantage if you're vertically integrated. So that's basically SpaceX, it's Amazon, why they own the rocket launching facilities, they own the satellites, they own the ground routing capabilities, tick, tick, tick.

That massively improves your economics. You see it in China with their three native domestic constellations. Talk about that later when we get to Shanghai.

In India, it's less the vertical integration and more the scale of base at the end of it. And the fact that they can see this as serving — goes back to point of making a huge population over a very long term.

So it's not necessarily something you're looking at a one-to-two-year ROI. This is in this is for the long term.

Peter Jarich (15:21)
Yeah.

And by the time — I mean, on the scale topic, I mean by the time this drops, the whole Iridium acquisition by Rocket Lab will have so... I think I think we do continue to see the sort of acquisitions and mergers and consolidation because scale is important.

What's interesting, it seems like it's that scale and consolidation on the legacy side of things, right? Where it's these folks that have been around, they're out there for a while, sort of getting rolled up.

Tim Hatt (15:51)
Yeah.

Peter Jarich (15:52)
To your point, also vertical integration, right? The Iridium side of things is also just taking their partnerships, their business model along with the manufacturing capabilities and everything else of Rocket Lab.

Tim Hatt (16:05)
And it was very much the last chip on the table with spectrum available. And that's the main purchase driver there. As we've seen in a number of the consolidation — the Link, the Omnispace, the Eutelsat JV, SES Intelsat — they all boil down to having spectrum and infrastructure scale to be able to compete.

Peter Jarich (16:24)
There...

I mean, there's still spectrum, there's still spectrum auctions out if I saw correctly. SpaceX picked up two licences in the most recent US auctions. One covering, as our colleagues at Mobile World Live call it, the Gulf of America, USA 250, but...

And one covering, I think, some other market in the Midwest. I'm really wanting to know what they plan to do with those two licences for a small amount of money.

That remains to be seen, and we have other topics to talk about.

So the two stories that didn't get picked. And Tim and I had had a bit of an arm wrestle about which one we should talk about. The Canada moving to the round of 16 tilted in his favour of getting to win of which one.

The one that we're not gonna talk about, but probably because it's gonna be in the news, is sort of the new silicon moves that you're seeing out of Qualcomm.

That you're seeing out of OpenAI. You saw OpenAI's silicon — so much for no more side quests on that front.

You saw Qualcomm coming out of their Capital Markets Day talking about the acquisition of Modular to help them on the software side of things to roll out solutions along with plans to roll out a new data centre. It's the Dragonwing data centre AI chipset, with expectations that by 2029 they plan to have one-third of revenues each for automotive, IoT, handsets and data centre.

As a reminder, it's now 70% handsets, 30% roughly IoT and automotive. So there's a long way to go, but with some big hopes there.

But, and just last thing there, I mean, I will give them credit, Qualcomm, which is always great on the marketing side of things, talking about scaling from milliwatts to kilowatts. I mean, they always have rhymed, but that's a good one.

Tim Hatt (18:30)
Yeah.

Peter Jarich (18:30)
The story we're gonna pick, Tim — glasses. And not just the horrible kind that I had on earlier, but smart glasses. Not dumb glasses like the ones I was wearing.

Tim Hatt (18:42)
Smart glasses. We don't often talk about smart glasses, Peter. Meta, two stories actually. Meta announced new partnerships with the iconic glassmakers, Ray-Ban and Oakley, both to support their augmented reality capabilities, so-called AR glasses.

Separately to that, Snap, the social media provider, announced a much more high-end, deeper technically integrated solution, more targeted at developers, but again, in the same AR glasses point.

And I think the reason I think this is interesting, so just with the Meta point, is that for a long time, people have hyped up AR and VR.

And this has been going pre-COVID. VR, AR, they're gonna be the next thing. It's a way to monetize 5G. People will have immersive entertainment experiences. Largely hasn't happened.

There's been peaks and troughs, but if you look at the data from our consumer survey, VR headset ownership is only at five percent of households, even in advanced economies, and AR glasses is about two percent ownership.

Doesn't really change if you look at Japan, the US, UK, Germany, it's all around that 2% mark. So it's a strong minority.

But so you have this dichotomy where people think there's promise and yet nobody owns them yet. So there's a big gap in that.

And I think the interesting thing for me from the stories are that with AI compute advancing so rapidly, can you get to a point where the use cases and the cost critically come down to mass market levels?

And I'll get your perspective in a minute, Peter, but I think for me the big question marks are with Meta, they're trying a fashion statement here. I want to embed AR into stylish glasses.

Snap is much more trying to get at the developer audience to develop new applications for it. And the price point for the Meta glasses has come down. It's about $300 for those retailers, which is in range for maybe not everybody, but I would say the mid-market consumer who values glasses.

The Snap is above $2,000, but again, that's a developer product.

So the question now is can we get to a point where people start to say, hey, I want to have the ability to look at weather or alerts or a sports score on my glasses, or is that just something that is a nice-to-have but not a must-have?

Peter Jarich (21:22)
And that's the — I mean, I think what I think is interesting here is, to your point of how do we drive this out. It is one part technology, one part branding.

Because you're right. So Meta and obviously Meta has been partnering with EssilorLuxottica, the owners of Ray-Ban and Oakley. I think they own more eyeglass, more sunglass companies than probably people people realise.

But what's interesting is they're not actually branded, right? They're not Ray-Bans, they're not Oakleys. They're sort of just — you can see the branding with EssilorLuxottica like on the glass, but apparently they didn't want to dilute that brand to some extent, of the sort of more premium Ray-Ban and more premium Oakley and any other premium brands when you lower that price point.

But I think it's interesting, right? If I go to buy a new pair of Ray-Bans, summer holidays coming up, time to replace my glasses. Let's say I want to buy some of the new A$AP Rocky aviator puffers because I'm a huge A$AP Rocky fan. You can probably tell that by — you can just probably guess.

I mean those are gonna run me what, like close to $200, an extra hundred for something that maybe isn't, you know, maybe is a different style but has capabilities in it. Yeah, it makes it a much easier, a much easier purchase.

I do get the impression that the experience is just completely different, where one will be maybe I can get turn-by-turn directions. I'm sort of speaking into it for AI. I am just taking some pictures, doing some image recognition.

That feels a lot, I think it's a lot also an easier entry point for folks versus the big question mark of what Snap is trying to do.

And Snap is really, you know, I've seen some folks out there saying if you're still investing in Snap, you're investing in the future of AR and this vision that they have.

You know, they supposedly invested around three and a half billion in bringing this to life. They picked up sort of the Illumics to support their AR vision before.

And I mean, this is gonna be more of, you know, a real AR sort of a real AR experience. And I don't know what that use case looks like.

You know, one of the ones they're showing off, right, is this idea of gaming where I look at someone else in the eyes and we do some sort of interactive gaming.

That's the hard part. And I think I'm not saying this is another Apple moment, but one of those moments where you go, people don't get this, but I do.

And people won't understand it intuitively. But if you're building it correctly and it is something that they find interesting, because we have no idea — will people take to this or not? — well, they go, okay, this AR stuff is actually really cool when done right.

And done in a way, you know, you talk about where Apple has gone with the Vision Pro, right? Obviously, a much different experience. They're trying to deliver AR, right? That thing you've got a puck. They're huge. What do they weigh? I've got down here, 26 plus ounces,

Tim Hatt (24:28)
It is.

Peter Jarich (24:29)
compared to around four, right? With four hours of battery life. They're not — I don't care what people say.

They don't look like regular glasses, slightly larger. They are definitely a statement that either says you're going to the Met Gala and you're trying to look weird on purpose, or you just... yeah.

So they don't quite look normal, but I mean they've got two Snapdragon chipsets in them. They've got batteries.

I mean, so I think that's the thing of, to your point about where we are. I mean these are different experiences. But I think we know what you're getting with the Metas.

We don't necessarily, I think people have the experience with what Snap's trying to do and we don't know if people take to them. And that's the big question, like will that — will people go, wow, this is great and drive some sort of recognition that goes beyond the two percent.

Tim Hatt (25:29)
Yeah. And I think just one other point on that, which is the reason it has stayed at two percent and five percent for VR headsets is that there has been no killer use case and largely the hardware and software just hasn't got there yet.

You know, nothing has coalesced to make this, you know, to bring out an iPhone moment, to use that phrase.

Peter Jarich (25:52)
Yeah. No, the killer use case — at least as I understand it, talking to people and looking out there — is music, right?

I've got, and not for AR VR, right? I think those aren't gonna be in the numbers that we're talking about, right? But the killer use case is you buy some Metas and maybe I'm not taking pictures with it, right? 'Cause I don't want to be too much of a creeper.

But I can always have music, right? I can listen to it and they're always glasses, right? If nothing else.

I think that's been the killer. It's low barrier to entry. And adding this, what they say, 26 styles, it's not really — I mean it's a couple models in different colours.

You can get the Kylie Jenner ones, because yeah, you know, along with A$AP Rocky, huge Kylie Jenner fan. But that's — you make it easy, right? Make it easy for people.

Tim Hatt (26:37)
Yeah. And Meta has the luxury, less so now given they're investing so much into AI infrastructure, but has the luxury of using this as a sort of test bet.

It's not meant to drive financial results at this point. This is looking at, and Snap to a certain extent as well, trialling the latest AR on a set of glasses to see how it works, what the take-up is, iterate, move on, etcetera.

What I think is interesting — the interesting questions for me here is when you look at the tech side, it is a lot more advanced than it was even two or three years ago.

We've benchmarked silicon companies across the world on their capabilities of their flagship chipsets to deliver on AI use cases.

And actually, Peter, what we found was all bar one, so this is like Qualcomm, MediaTek, others, all bar one could service AR glasses right now with their flagship, but they don't rate it as a priority.

Why? Because they don't see the demand.

So there's some lead-lag effects here where if we start to see people taking up these Ray-Bans, if we start to see Snap going off, that balance might start to shift.

So I'm not writing it off. I think we're at an interesting point where AR is probably more possible than it ever has been, but we've got to see if people respond to the utility of the device.

Peter Jarich (27:57)
And I would just, and before we talk about sort of the what we've been up to lately side of things, it is also interesting to note because we think about AR VR versus sort of the AI glasses, and that's what Meta is calling them.

And I think there's a direct quote from the release that glasses are the most exciting hardware category of the AI era.

Tell the server folks that, because I think they think servers are the most exciting. I mean, maybe it's not exciting, but I mean, what I think is interesting is what you don't necessarily have with AR/VR to the same extent, because the use cases are different, is some of these questions of, am I filming someone?

Right. And I think you're seeing that sort of becoming a broader societal issue.

And it's interesting that at the same time, what was it that we recently saw, CAICT out of China releasing its voluntary, voluntary code of conduct for AI glasses manufacturers, right?

And that includes things like provide notifications to users if you're filming them, make sure people get consent, watch data collection, make sure that the processing takes place on device if you can.

What's interesting is, you know, is that coincidence or not? I don't know. Obviously, you've got a lot of AI glass manufacturers in China.

But it is a reminder that those are the sorts of things that could hold back the market if we don't get it right.

And I think that's sort of the question of maybe there are less of those issues with AR/VR glasses. If I'm in my own little world on VR, VR glasses, people are less concerned about what I'm filming and how I'm using that data.

Of course, AR is about interacting with the real world, so you've got that, but it just seemed a little bit coincidental that that voluntary code of conduct came out the same time as Meta.

Anyways, speaking of China, now is the time for our favourite section of the show. On our minds and on the road.

And there wasn't much on our minds because we were both on the road last week.

So speaking of China, Tim, MWC Shanghai, I heard that it was much more international than years past. And a good show by all accounts.

Tim Hatt (30:19)
Absolutely. So MWC Shanghai is a sister event for our listeners to MWC Barcelona and the other MWCs.

Historically, it has been a China-focused event. However, in recent years, there's been much more of a diversification of the attendance space to other countries from Asia and indeed Europe and further afield.

I think to your point now, the top-line takeaways for me were international representation was up over 30 percent.

We see a lot from Southeast Asia, we see from Australia, the Gulf states very present. And that's interesting because that speaks to China's gravitational pull as a tech centre within the Asia region and an influencer there.

So that was clear.

I think on the topic side, Peter, what was interesting for me this year was, you know, obviously AI was front and centre as expected, but it was much more advanced along the lines of agentic and physical AI.

Peter Jarich (31:21)
Sure.

Tim Hatt (31:22)
We see a lot of demos, probably more than you do in Barcelona, of specific enterprise verticals coming to the show to show how, for example, the agentic AI is being deployed in factory robots or in a hospital setting to test that out.

So a lot of...

Peter Jarich (31:36)
I mean, you take manufacturing, advanced manufacturing capabilities, you take the building of gadgets, you take AI, and to be clear, right? I think everyone talks about China being somewhere, you know, six months behind or so the frontier side of things in the US.

So they clearly want to show what they can do, right? And so I think you've got the perfect storm to show off physical AI in all those forms, which is why I think — and I'll be honest, right? I'm always happy for a self-own.

You know, several years ago when we had customers who were planning MWC Shanghai events, like here's the Latin America Summit at Shanghai, or here's the Sub-Saharan Africa Summit, I was like, well, I mean, there's one, obviously, the interest of Chinese companies in expanding into those markets, but it was always like, wow, you could — and I think that to your point, like it was sort of a, well, duh one.

The Chinese companies want to expand those markets too. Those markets want to learn from everything that is crazy and cool going on in China.

And so it definitely makes sense.

Tim Hatt (32:43)
6G was on the agenda a lot more than years past, and the transition to that, I think we're closer.

I think China will be one of the leaders there, but we see it, you know, from the Koreans, we see it from other Asia countries.

And then satellite was again a huge topic. We ran our satellite summit, we had an analyst wrap-up panel at the end, various roundtables. It comes up everywhere.

I think the interesting contrast there is how the Chinese domestic ecosystem is shaping up to deliver their sovereign capabilities versus what's happening in the West, 'cause it is still largely kind of a closed shop in China.

So interesting questions about how international they can go.

Peter Jarich (33:20)
And why we love MWC Shanghai? Because satellite summit in Barcelona, standing room only, and so you kick people out because health and safety regulations say if you have too many people you get in trouble.

Shanghai they just sit on the floor or hang from the rafters.

Tim Hatt (33:35)
Shanghai, we let in. They can sit on the floor, they can go on the outside.

Peter, you were in closer to home though. Tell us about DTW.

Peter Jarich (33:43)
Yeah, so a place where 6G and token economics was not so much on the agenda.

Yes. So DTW Ignite is the TM Forum's big event in Copenhagen. And, you know, I think it was around four — I mean, well, last year was around 4,000 people. It felt like it was about the same size this year.

And, you know, for those folks who don't know the work of the TM Forum, traditionally they've been the ones that, if you're thinking of OSS, BSS-type stuff, they're the place to go to.

And in recent years, automation has driven a lot of their work, as well as open APIs and API exposure.

So I've been there the past few years in terms of supporting some of the Open Gateway work that we've done at the GSMA, but obviously it's become a place to talk AI.

And I think you see a bit of them trying to pivot into taking automation into AI. 'Cause obviously there's a link there, right?

But we've been doing automation for years. And of course, when you've been doing something for years and you're trying to keep it fresh, what do you do? Well, hey, we add agentic AI.

And we say if we really want to extend automation to its full promise, we want to get to this, this level-four self-driving networks, level-four autonomy and beyond. We're gonna need agentic.

And so as Tim mentioned earlier, agentic was part of the, you know, what made it to the news, driven by a lot of the stuff from DTW.

Wasn't too surprising though. I mean, I think we probably all expected that was going to be the case.

What I thought was interesting were a couple things. One, you know, we've been talking a bit about how if you're looking at AI and applying AI to different parts of the network, the RAN gets a lot of attention, but the core might be an easier sale just because it's a natural evolution way of doing automation.

And so you saw some of that messaging from Huawei.

You actually saw the messaging, I think, out of Sean and I from ZTE with their AirCore launch. And then what was particularly interesting was Ericsson talked about extending their EAIP, their automation platform, into the core with what they're calling C-Apps.

So there's R-Apps and xApps, which are all about how you link — you know, how you have sort of automation applications and different applications you link into the radio access network.

This is extending that with the same interface into the core. It's not necessarily an industry thing yet, more of an Ericsson thing yet, but it does show the focus on the core.

Data was a big thing I called out last year. You saw Ericsson and Nokia, and you had messaging from everyone else talking about getting your data platforms in place.

This idea that unless you have clean data, accessible, not necessarily all in one place, but accessible and in some way that you can actually access it — I guess that's what makes it accessible — that AI is not going to work.

And you saw platforms last year. This year you saw a bunch of great stuff from Nokia talking about partners, which says, hey, it maybe it's not going to live in some place. I'm going to have to partner with Microsoft or I'm going to have to partner with Databricks or whoever.

So it was mostly extensions of what we saw in the past. Nothing sort of breakthrough. But we probably shouldn't expect that anyways, right?

Tim Hatt (37:02)
Well, one question I had on that, and it's kind of linked 'cause we heard a lot on token economics in Shanghai and this whole notion of how — is tokenization a way to make money from AI? Is it a way to manage data traffic? Is manageable?

And I think there was part talk, but still a lot of questions there.

From your sense of automation at DTW, do you get the sense like level-four autonomy — like is that gonna be a reality? And if so, do we get towards this zero touch, or is that just the...

Peter Jarich (37:29)
It will at some point, right? I mean, I think at some point we're moving in that direction, whether or not the bar moves. I think the reason you have these levels is so that there is a goalpost.

I think I have heard some complaints that it becomes an objective in and of itself as opposed to what are we trying to achieve, right?

You don't want just a badge that says I've got level four, and you can also say, well, I'm level four in this part of my network versus that part.

So I don't want to say the system's being gamed because it's not, but you could.

So I did hear some pushbacks.

But what is interesting is, to your point though, is I didn't hear so much on the token economy side of things. And that's good, I think, because it's still new.

I did actually have some conversation with other colleagues who were at Shanghai, and even they said yes, although the Chinese operators have been the ones to really push forward on that.

Even they've said it's early, right? They're still in trials, they're still trying to figure it out, which is a good indication that — or it's a good sign then — that it really didn't make it into being buzzworthy at DTW yet.

Tim Hatt (38:37)
Yeah, it will be interesting to see where it goes. I think just on the tokenization, it is a China and India story right now.

But I think as questions start to get answered and around what the store of value looks like, how you measure it, etcetera, etcetera, we probably will see more of that.

That's gonna be something to watch over the next, in the next year.

Peter Jarich (38:59)
Yeah, it's not — it's fungible, right? I think it's a... Even if someone does understand that if I watch an SD movie for this long, it's gonna take up this much data, it's not the same thing as saying if I use this model versus that model.

And I think, as we all called out, and these are the times when we can put in some quick plugs before we leave, as we called out in our analysis on token economy, it's also based on the fact that the Chinese operators have deployed a lot of AI capacity.

So they need to find some use case for it. They have the capabilities. Check that out if you haven't.

Check out our recap of DTW Ignite is up on the site. There will be one soon around MWC Shanghai.

Quick plug if you haven't seen it yet. Christina from the team put together a blog post on the Qualcomm Capital Markets Day.

So covering a lot of the insights about what's going on in that side of things. So a lot of good content.

Tim Hatt (39:56)
Good read.

Peter Jarich (39:57)
It amazes me all the time when I talk about all the stuff that we have that actually fits with what's going on in the market because it's what we're here for.

So anything, any last thoughts, Tim, before we go?

Tim Hatt (40:11)
Just be there Saturday night, Canada, Morocco, six o'clock UK time. Be there, be square.

Peter Jarich (40:15)
Canada-Morocco.

Yeah, I mean, I guess I'll leave with this and say, one, as I said in the past, don't leave a review yet.

Two, I think we've gotta have some sign-off, some cool — so send in your ideas for what would be a good sign-off language.

In the meantime, USA 250, and we'll see you next week.

Thanks, Tim. Thanks everyone.

Tim Hatt (40:42)
See you there.

See you guys.