WEBVTT 00:00:00.000 --> 00:00:03.299 Welcome to episode 246 of the 6G podcast. It's 00:00:03.299 --> 00:00:06.179 the latest inside scoop on everything 6G. And 00:00:06.179 --> 00:00:09.560 we cover six topics in 30 minutes, and it's brought 00:00:09.560 --> 00:00:11.740 to you by More Inception Strategy. I'm Anshul 00:00:11.740 --> 00:00:13.560 Saag, and joining me again this week is file 00:00:13.560 --> 00:00:16.940 analyst Mike Gano from Ookla. Mike, how has your 00:00:16.940 --> 00:00:19.980 weekend been so far? Yeah, no, it was a good 00:00:19.980 --> 00:00:22.460 weekend. I celebrated Easter, so that was really 00:00:22.460 --> 00:00:26.039 good. I had a nice relaxing time. I am. So I've 00:00:26.039 --> 00:00:27.760 got a rock climb that I've been working on, but 00:00:27.760 --> 00:00:29.679 I didn't get it this weekend. That was my goal 00:00:29.679 --> 00:00:32.100 for this weekend. Was not able to do it. Good 00:00:32.100 --> 00:00:35.579 and bad. How was yours? Lots of driving. I took 00:00:35.579 --> 00:00:38.920 advantage of Easter to drive back home from the 00:00:38.920 --> 00:00:42.719 Bay Area where I attended a media tech analyst 00:00:42.719 --> 00:00:47.799 summit and also spent Passover with family. Also 00:00:47.799 --> 00:00:50.780 a very busy weekend and slash week. I'm just 00:00:50.780 --> 00:00:53.880 happy to be able and happy. I actually drove. 00:00:54.590 --> 00:00:58.369 back to Sonoma to watch the Artemis 2 launch. 00:00:58.530 --> 00:01:01.530 I'd love to roll that into my first topic, which 00:01:01.530 --> 00:01:03.329 I know I'm sure you're excited to talk about 00:01:03.329 --> 00:01:05.549 too. I know. I've been watching your posts on 00:01:05.549 --> 00:01:08.670 social media about it too. I'm as excited as 00:01:08.670 --> 00:01:11.329 you are, I think. We've both been watching it 00:01:11.329 --> 00:01:13.769 really carefully. Yes. And I've been waiting 00:01:13.769 --> 00:01:16.569 for the images to roll through. I'm very glad 00:01:16.569 --> 00:01:18.849 NASA still uses Flickr so I can get the images 00:01:18.849 --> 00:01:21.569 at their absolute best quality. And yeah, speaking 00:01:21.569 --> 00:01:24.459 of the images. How are we getting them? It's 00:01:24.459 --> 00:01:30.180 a new laser link called O2O, which is just short 00:01:30.180 --> 00:01:35.500 for Orion 2 optical link. And it's a smaller 00:01:35.500 --> 00:01:37.700 version of something that we had actually heard 00:01:37.700 --> 00:01:40.920 about in 2013, where they were able to do the 00:01:40.920 --> 00:01:43.500 gigabit link. But basically what it is, they 00:01:43.500 --> 00:01:49.579 are limited by the receiver on the capsule. And 00:01:49.579 --> 00:01:52.969 it's really interesting because They, this is 00:01:52.969 --> 00:01:56.569 the size of the house cat, which is a very interesting 00:01:56.569 --> 00:02:01.109 scale that they chose, but they're able to do 00:02:01.109 --> 00:02:06.810 about 260 megabits per second. So they're uplinking 00:02:06.810 --> 00:02:10.389 at 260 megabits per second and downlinking at 00:02:10.389 --> 00:02:14.229 20. They're actually streaming at 4k from the 00:02:14.229 --> 00:02:16.629 capsule, which is very cool being able to see 00:02:16.629 --> 00:02:19.590 that. What's interesting is they're able to. 00:02:20.349 --> 00:02:22.830 There's so much data that they're sharing constantly, 00:02:23.150 --> 00:02:27.509 whether it's video, images, sensor data, all 00:02:27.509 --> 00:02:29.710 kinds of things. They have to be very careful 00:02:29.710 --> 00:02:32.349 with what they're sharing and when they share 00:02:32.349 --> 00:02:34.629 it. And that's why we didn't get some of those 00:02:34.629 --> 00:02:36.909 really awesome images from the dark side of the 00:02:36.909 --> 00:02:41.030 moon until a few hours or almost a day later, 00:02:41.150 --> 00:02:43.270 because there's a bunch of other stuff to upload 00:02:43.270 --> 00:02:45.509 that's more important, like how they're getting 00:02:45.509 --> 00:02:48.530 home. Which, by the way, they're splashing down 00:02:48.530 --> 00:02:51.680 in Fandina by the clock. I'm going to be on the 00:02:51.680 --> 00:02:54.620 coast looking for them. So I'm going to take 00:02:54.620 --> 00:02:56.780 my kids with me and we're going to watch a spaceship 00:02:56.780 --> 00:03:01.159 land in the water. Oh my God. Fantastic. How 00:03:01.159 --> 00:03:03.280 do you beat that? I'm going to turn my little 00:03:03.280 --> 00:03:06.560 daughter into a space nerd. What a nerd. Yeah. 00:03:06.560 --> 00:03:09.300 So in addition to that, the one thing a lot of 00:03:09.300 --> 00:03:11.419 people didn't realize is they have iPhones with 00:03:11.419 --> 00:03:14.099 them up there in space. And while many of the 00:03:14.099 --> 00:03:18.490 pictures. are being taken on Nikon V9 and V5 00:03:18.490 --> 00:03:21.169 professional cameras, which if I were still in 00:03:21.169 --> 00:03:23.310 the Nikon world, I would probably have. They 00:03:23.310 --> 00:03:27.330 are also using iPhones to take selfies. And what 00:03:27.330 --> 00:03:28.789 a lot of people didn't realize was some of these 00:03:28.789 --> 00:03:31.150 really cool pictures of them with the planet 00:03:31.150 --> 00:03:34.969 illuminating their face in a selfie were literally 00:03:34.969 --> 00:03:38.789 the iPhone 17 center stage camera, which is an 00:03:38.789 --> 00:03:41.710 18 megapixel camera that when it's in center 00:03:41.710 --> 00:03:44.639 stage mode takes a 7 megapixel image. And then 00:03:44.639 --> 00:03:46.159 when you need the bigger version, it gives you 00:03:46.159 --> 00:03:49.039 a full 18 megapixel sensor. What's cool about 00:03:49.039 --> 00:03:50.539 it, and I'm going to nerd out a little bit over 00:03:50.539 --> 00:03:53.180 this, it's a square sensor. So whenever you take 00:03:53.180 --> 00:03:56.039 a picture, you can take, you can hold it in portrait 00:03:56.039 --> 00:03:59.479 and get both in portrait automatically with having 00:03:59.479 --> 00:04:01.759 to turn it on. Because the sensor is perfectly 00:04:01.759 --> 00:04:06.439 square. So you get both anytime you want. And 00:04:06.439 --> 00:04:08.300 I think they added explosives on that technology. 00:04:08.539 --> 00:04:10.180 So I don't think anybody's going to be able to 00:04:10.180 --> 00:04:12.020 copy that for quite some time. But it's very 00:04:12.020 --> 00:04:14.360 cool. And it's cool because you can like, I was 00:04:14.360 --> 00:04:17.519 using Flickr to look through the stuff, and I 00:04:17.519 --> 00:04:20.680 remember first thinking, yeah, it's the selfie 00:04:20.680 --> 00:04:24.199 camera. This wasn't a IM camera. This is a selfie, 00:04:24.279 --> 00:04:26.439 and it looks really good. And then if you actually 00:04:26.439 --> 00:04:28.139 look at some of the other selfies they've taken 00:04:28.139 --> 00:04:31.540 on it later on through the process, you'll know 00:04:31.540 --> 00:04:34.180 that there's already some dead pixels in the 00:04:34.180 --> 00:04:36.759 images because the radiation is destroying the 00:04:36.759 --> 00:04:40.480 image sensor. So even with the amount of protection 00:04:40.480 --> 00:04:42.060 that they have in the capsule, they're still 00:04:42.060 --> 00:04:47.259 being irradiated. by cosmic rays and yeah that's 00:04:47.259 --> 00:04:48.759 all i want to say i don't want to take too long 00:04:48.759 --> 00:04:51.360 that's incredible that's so interesting i don't 00:04:51.360 --> 00:04:54.160 know it really has been such a joy to be able 00:04:54.160 --> 00:04:57.279 to live through this week and watch all the adventures 00:04:57.279 --> 00:04:59.839 of the artemis crew and really cheer them on 00:04:59.839 --> 00:05:02.300 and we wish them a happy and safe landing and 00:05:02.300 --> 00:05:04.279 i just can't wait for that yeah to them for them 00:05:04.279 --> 00:05:06.379 to splash down and to get all the stories out 00:05:06.379 --> 00:05:08.000 of that but yeah i think it's what i think it's 00:05:08.000 --> 00:05:09.899 interesting too is just as apart from all the 00:05:09.899 --> 00:05:12.079 technological stuff it's like it's neat to see 00:05:12.829 --> 00:05:16.430 what brands get associated with stuff like this. 00:05:16.490 --> 00:05:18.810 I know, I think you actually tweeted about it 00:05:18.810 --> 00:05:21.670 too, is what watches the crew are wearing. I 00:05:21.670 --> 00:05:23.350 think they've got a version of the Speedmaster, 00:05:23.629 --> 00:05:26.490 which is obviously the Omega. And they're mass 00:05:26.490 --> 00:05:30.589 issued. Oh, there you go. Oh man, so cool. It's 00:05:30.589 --> 00:05:34.870 cool. It's so cool. It's neat to see. They arrived 00:05:34.870 --> 00:05:38.160 on fighter jets, which is awesome. They didn't 00:05:38.160 --> 00:05:39.980 just come on a helicopter or plane. They were 00:05:39.980 --> 00:05:42.019 like training jets, but they arrived on jets 00:05:42.019 --> 00:05:45.139 because a lot of these guys are ex -fighter pilots. 00:05:45.759 --> 00:05:48.339 Which is also so cool. So cool. Everything about 00:05:48.339 --> 00:05:50.259 the Artemis project has been just wonderful to 00:05:50.259 --> 00:05:52.540 watch. And so, yeah, it's exciting to see that 00:05:52.540 --> 00:05:55.779 continue into Artemis 3 and 4 throughout the 00:05:55.779 --> 00:05:57.560 next couple of years. But yeah, it's neat to 00:05:57.560 --> 00:06:01.699 see the Omega watches, the iPhone brand that 00:06:01.699 --> 00:06:03.949 they're using. And you've got a phone there that 00:06:03.949 --> 00:06:06.089 you're showing. My screen. Your screen is so 00:06:06.089 --> 00:06:08.709 cool. I do not. I know. It's just great. Yeah. 00:06:08.790 --> 00:06:11.230 Yeah. Very excited for them to splash down and 00:06:11.230 --> 00:06:14.050 yeah, see what else happens to them. So my first 00:06:14.050 --> 00:06:18.089 topic is not nearly as cool as Artemis, that's 00:06:18.089 --> 00:06:21.930 for sure. But it is relevant to our podcast in 00:06:21.930 --> 00:06:26.029 that it is a little bit more detail from the 00:06:26.029 --> 00:06:30.540 recent 3GPP meeting that happened in Japan. from 00:06:30.540 --> 00:06:33.920 space will pivot into Japan. And so the 3GPP 00:06:33.920 --> 00:06:36.959 just finished a meeting there. And those big 00:06:36.959 --> 00:06:40.620 3GPP meetings are really worth watching. They 00:06:40.620 --> 00:06:42.279 happen every couple months, but this was a pretty 00:06:42.279 --> 00:06:45.279 pivotal one in that they are making some concrete 00:06:45.279 --> 00:06:49.120 decisions about the 6G standard. And again, the 00:06:49.120 --> 00:06:52.500 3GPP is the association that sets cellular standards. 00:06:52.680 --> 00:06:55.220 They're really the ones to watch in terms of 00:06:55.220 --> 00:06:58.120 what technical standards are going to be included 00:06:58.120 --> 00:07:03.899 in. The 3GPP releases batches of new technologies 00:07:03.899 --> 00:07:07.680 every year or so, and they're currently discussing 00:07:07.680 --> 00:07:11.839 what should be included in the release of technologies 00:07:11.839 --> 00:07:13.879 that will be the official 6G specifications. 00:07:14.339 --> 00:07:15.920 And we talked a little bit about it last time, 00:07:15.959 --> 00:07:18.300 but we got a little bit more. information about 00:07:18.300 --> 00:07:20.800 the meeting that happened in japan recently and 00:07:20.800 --> 00:07:22.339 actually it's i thought it was interesting it's 00:07:22.339 --> 00:07:25.160 the post was written by guy daniels who is one 00:07:25.160 --> 00:07:27.360 of the guys over at telecom tv he works with 00:07:27.360 --> 00:07:30.120 uh ray la maestra who i know quite well and has 00:07:30.120 --> 00:07:32.439 is just a solid dude and so it's interesting 00:07:32.439 --> 00:07:34.879 to see it's interesting to see guy daniels covering 00:07:34.879 --> 00:07:37.139 it's interesting to that he's writing about this 00:07:37.139 --> 00:07:40.019 on substack and what he wrote is actually quite 00:07:40.019 --> 00:07:42.529 interesting also so A couple of the takeaways 00:07:42.529 --> 00:07:45.009 from his post that he wrote. First, he talked 00:07:45.009 --> 00:07:48.230 about the timeline. So it'll be 3GPP release 00:07:48.230 --> 00:07:51.129 21. That'll be the initial 6G specifications. 00:07:51.189 --> 00:07:55.810 And the 3GPP has agreed to freeze work on this 00:07:55.810 --> 00:08:00.730 standard in March 2027. So almost exactly a year 00:08:00.730 --> 00:08:04.129 from now. And that's an important milestone because 00:08:04.129 --> 00:08:06.329 as soon as they stop work on the standard, it 00:08:06.329 --> 00:08:08.470 means they can't change it anymore. And it means 00:08:08.470 --> 00:08:11.060 that... equipment vendors and handset makers 00:08:11.060 --> 00:08:13.860 and other companies can start making products 00:08:13.860 --> 00:08:15.980 that adhere to those standards. It doesn't mean 00:08:15.980 --> 00:08:19.000 that it's finished, but it means that the initial 00:08:19.000 --> 00:08:21.779 batch of standards are completed and that work 00:08:21.779 --> 00:08:23.600 can start in the hardware and then they'll update 00:08:23.600 --> 00:08:26.800 it with software as time goes on. But that stage, 00:08:26.800 --> 00:08:28.660 they call it a stage one freeze. That stage one 00:08:28.660 --> 00:08:31.519 freeze is a really big sort of milestone in the 00:08:31.519 --> 00:08:34.799 development of the 6G standard and previous standards 00:08:34.799 --> 00:08:37.250 from the 3GPP. So that's the first one. A couple 00:08:37.250 --> 00:08:39.269 of other takeaways from this post that I thought 00:08:39.269 --> 00:08:42.250 were really interesting was that it sounds like, 00:08:42.289 --> 00:08:45.549 based on this recent 3GPP meeting, it sounds 00:08:45.549 --> 00:08:49.350 like they have officially settled on the using 00:08:49.350 --> 00:08:54.250 5G standalone, the SA, 5G SA standalone architecture. 00:08:54.789 --> 00:08:59.809 They want to use that as the base for 6G, which 00:08:59.809 --> 00:09:02.509 is pretty interesting. It's not a surprise. 5G 00:09:02.509 --> 00:09:05.210 standalone is far more capable. than the initial 00:09:05.210 --> 00:09:07.909 version of 5g which is the non -standalone version 00:09:07.909 --> 00:09:11.470 of 5g the nsa version of 5g but the fact that 00:09:11.470 --> 00:09:14.049 they've officially standardized on the standalone 00:09:14.049 --> 00:09:18.309 version of 5g as a base for 6g is important moves 00:09:18.309 --> 00:09:20.809 everything closer to sort of reality and it also 00:09:20.809 --> 00:09:23.750 creates a little bit of motivation i assume it 00:09:23.750 --> 00:09:26.470 creates a little bit of motivation for the 5g 00:09:26.470 --> 00:09:29.450 operators out there who have not yet begun upgrading 00:09:29.450 --> 00:09:32.649 from the nsa version to the sa version of 5g 00:09:32.649 --> 00:09:35.700 i assume that this will give them another reason 00:09:35.700 --> 00:09:38.460 to upgrade their systems to the standalone version 00:09:38.460 --> 00:09:42.120 of 5G. And just a side note here is that my colleague 00:09:42.120 --> 00:09:44.460 Luke recently put out a report on standalone 00:09:44.460 --> 00:09:46.679 5G, and you really do see some really impressive 00:09:46.679 --> 00:09:49.860 performance improvements when operators upgrade 00:09:49.860 --> 00:09:52.820 to the standalone version of 5G. Not a surprise 00:09:52.820 --> 00:09:54.980 that's a version of 5G that everybody's been 00:09:54.980 --> 00:09:57.620 watching for, and it's also not a surprise that 00:09:57.620 --> 00:09:59.440 it will be the sort of baseline architecture 00:09:59.440 --> 00:10:02.960 for 6G. the underlying baseline architecture 00:10:02.960 --> 00:10:05.700 for it. But still, we're making progress. We're 00:10:05.700 --> 00:10:07.940 making progress towards 6G. And then final takeaway 00:10:07.940 --> 00:10:10.980 that I'll briefly mention from Guy Daniels' post 00:10:10.980 --> 00:10:13.100 is he talked a little bit about non -terrestrial 00:10:13.100 --> 00:10:15.220 networks, the satellite networks that are out 00:10:15.220 --> 00:10:18.500 there. His basic takeaway was that they're going 00:10:18.500 --> 00:10:22.379 to talk more about that. So not a surprise. Again, 00:10:22.580 --> 00:10:25.340 the idea of smartphones connecting to satellites 00:10:25.340 --> 00:10:28.929 is still a relatively new one. There are... technologies 00:10:28.929 --> 00:10:31.090 out there that already do this. There's a number 00:10:31.090 --> 00:10:33.549 of standards that they're using now and will 00:10:33.549 --> 00:10:36.429 into the future. But it really is a sort of collision 00:10:36.429 --> 00:10:38.789 between two industries. It's the cellular industry 00:10:38.789 --> 00:10:42.250 and the satellite industry. And they're really 00:10:42.250 --> 00:10:45.929 in the process right now of a slow moving collision. 00:10:46.350 --> 00:10:49.909 And the work at 3GPP is part of that. They're 00:10:49.909 --> 00:10:53.330 essentially integrating cellular standards with 00:10:53.330 --> 00:10:56.190 satellite standards as part of their work on 00:10:56.190 --> 00:10:59.279 6G. And so it sounds like work will continue 00:10:59.279 --> 00:11:01.600 on that and we'll probably get more details as 00:11:01.600 --> 00:11:03.980 time goes on. It certainly doesn't, it's not 00:11:03.980 --> 00:11:05.519 going to stop any of the work that has happened 00:11:05.519 --> 00:11:07.379 so far. It's not going to affect any of the commercial 00:11:07.379 --> 00:11:10.580 deployments that have happened so far. But it 00:11:10.580 --> 00:11:14.440 is an indication that satellite is increasingly 00:11:14.440 --> 00:11:18.039 important in discussions about 6G. I think that's 00:11:18.039 --> 00:11:20.240 the real takeaway is that when you talk about 00:11:20.240 --> 00:11:22.570 6G, you talk about a couple of things. And one 00:11:22.570 --> 00:11:24.230 of the things that you'll talk about in 6G will 00:11:24.230 --> 00:11:26.629 be satellites, which I think is great. And it 00:11:26.629 --> 00:11:28.370 goes back to our conversation about space, right? 00:11:28.429 --> 00:11:29.789 How are these satellites going to get up there? 00:11:29.990 --> 00:11:32.049 They're going to use big rockets just like Artemis 00:11:32.049 --> 00:11:34.529 did. Probably not that size of rocket, but similar 00:11:34.529 --> 00:11:37.149 rockets. Interesting takeaways from Guy Daniels. 00:11:37.350 --> 00:11:39.669 Yeah, they're most likely something like a Starship. 00:11:39.730 --> 00:11:43.509 But what I would say on this topic is I'm very 00:11:43.509 --> 00:11:45.649 glad that the conversations we've been having 00:11:45.649 --> 00:11:49.110 about 6G that have led up to this point have 00:11:49.110 --> 00:11:52.500 culminated in the agreement. that NSA was a mistake 00:11:52.500 --> 00:11:59.019 and that 6G will have to be SA only. And obviously 00:11:59.019 --> 00:12:01.860 there will be a 5G network. There will be tons 00:12:01.860 --> 00:12:06.639 of 5G in coexistence with 6G. But if you want 00:12:06.639 --> 00:12:09.419 to leverage 6G, you have to do SA, which is great 00:12:09.419 --> 00:12:12.259 because my whole argument for the longest time 00:12:12.259 --> 00:12:17.179 was that carriers were lazy and cheats and did 00:12:17.179 --> 00:12:21.789 NSA and didn't really like... make the investment 00:12:21.789 --> 00:12:25.429 to go SA. And we were just talking, Turkey didn't 00:12:25.429 --> 00:12:28.830 even roll out 6G until last week. 5G. 5G until 00:12:28.830 --> 00:12:31.950 last week. 5G until last week. It's very frustrating 00:12:31.950 --> 00:12:35.169 because people are upset about what they've been 00:12:35.169 --> 00:12:38.230 promised with 5G and what they've done. People's 00:12:38.230 --> 00:12:40.409 immediate reaction is, why do I need 6G if 5G 00:12:40.409 --> 00:12:42.710 didn't deliver what I wanted? I think NSA was 00:12:42.710 --> 00:12:45.529 the main reason why that happened, which is really 00:12:45.529 --> 00:12:48.679 unfortunate because... Now 6G will theoretically 00:12:48.679 --> 00:12:51.580 deliver what 5G could have done because it will 00:12:51.580 --> 00:12:54.259 force NSA, right? Yeah, I'm very excited about 00:12:54.259 --> 00:12:56.960 this. Everything seems to be looking pretty good 00:12:56.960 --> 00:12:59.559 in terms of things we've been talking about in 00:12:59.559 --> 00:13:01.700 the last year or so about what might be at 6G. 00:13:01.799 --> 00:13:03.799 Yeah, I'm looking forward to that first freeze 00:13:03.799 --> 00:13:07.200 as well as the culminating second and third freeze 00:13:07.200 --> 00:13:10.639 that kind of solidify things. Having that happen 00:13:10.639 --> 00:13:14.720 in 27 does set the opportunity for some pretty 00:13:14.720 --> 00:13:18.539 standard. potential compliant deployments in 00:13:18.539 --> 00:13:22.539 28 and 29. I remember the timeline. I was looking 00:13:22.539 --> 00:13:24.879 at the article you were talking about. That timeline 00:13:24.879 --> 00:13:27.539 looks like everything will be pretty buttoned 00:13:27.539 --> 00:13:30.919 up and ready to go by March of 2029. That does 00:13:30.919 --> 00:13:34.879 set us for a potential 2029 full commercial launch 00:13:34.879 --> 00:13:38.779 and easily a 2030. I'm getting the sense that 00:13:38.779 --> 00:13:41.879 2029 is when people, when infrastructure vendors 00:13:41.879 --> 00:13:44.960 want things to happen. And maybe carriers are 00:13:44.960 --> 00:13:47.279 pushing for 2030. I think we might see maybe 00:13:47.279 --> 00:13:50.460 a T -Mobile push for 2029, but I think everybody 00:13:50.460 --> 00:13:54.039 else is more likely to be 2030 timeline if they 00:13:54.039 --> 00:13:57.860 can nail that. But yeah, I'm excited and I appreciate 00:13:57.860 --> 00:14:00.899 you looking into the latest RAN meeting from 00:14:00.899 --> 00:14:05.480 3GPP. That does lead me to my next topic, which 00:14:05.480 --> 00:14:10.740 is AT &T launched a new plan called OneConnect. 00:14:12.269 --> 00:14:15.090 In the surface, it seems pretty cool because 00:14:15.090 --> 00:14:18.990 it's a one subscription, one monthly price, and 00:14:18.990 --> 00:14:23.409 you get both fiber and wireless, which is pretty 00:14:23.409 --> 00:14:28.230 good. You get a gig of fiber, and it's basically 00:14:28.230 --> 00:14:32.350 $90 a month for one person, which is pretty good 00:14:32.350 --> 00:14:34.889 because I pay $90 a month for my fiber right 00:14:34.889 --> 00:14:38.629 now. So I would basically be getting a line for 00:14:38.629 --> 00:14:43.309 free on my AT &T service. But it's $120 for two 00:14:43.309 --> 00:14:47.570 people and $225 for unlimited members up to 10 00:14:47.570 --> 00:14:50.429 lines. So if you've got a big family, this is 00:14:50.429 --> 00:14:53.450 a deal. Because it comes out to $22 a line, your 00:14:53.450 --> 00:14:57.190 fiber is basically free. But I think the big 00:14:57.190 --> 00:15:02.429 thing here is it's only for people who have access 00:15:02.429 --> 00:15:05.409 to AT &T Fiber, which is a limited footprint 00:15:05.409 --> 00:15:08.909 on the national scale. It is growing, but it's 00:15:08.909 --> 00:15:11.980 still limited. Also, you have to bring your own 00:15:11.980 --> 00:15:16.179 eSIM compatible unlock smartphone. So this is 00:15:16.179 --> 00:15:19.460 not for somebody on a contract or someone who 00:15:19.460 --> 00:15:24.039 want this subsidized contract. It's a BYO situation. 00:15:24.659 --> 00:15:29.419 There's also, it's not prioritized traffic for 00:15:29.419 --> 00:15:32.179 the wireless. So if there's any kind of congestion, 00:15:32.580 --> 00:15:35.460 you will get throttled. So this is like, how 00:15:35.460 --> 00:15:37.740 do I put this? This is a cheapo plan. There's 00:15:37.740 --> 00:15:42.190 like a prepaid plan almost. And it does exclude 00:15:42.190 --> 00:15:46.470 satellite. So that's also interesting. But my 00:15:46.470 --> 00:15:49.889 colleague from another firm, GreenGuard, did 00:15:49.889 --> 00:15:52.789 catch some of these terms and conditions. And 00:15:52.789 --> 00:15:55.110 it was really interesting because there's almost 00:15:55.110 --> 00:15:58.389 more terms and conditions than there are actual 00:15:58.389 --> 00:16:02.429 features to the plan. So I think it's really 00:16:02.429 --> 00:16:05.669 important to understand that it's a family plan, 00:16:05.870 --> 00:16:08.860 but it's also a bundle. And it's interesting 00:16:08.860 --> 00:16:11.840 because it like really locks you in, right? Because 00:16:11.840 --> 00:16:15.860 it's so cheap. But I think that the fact that 00:16:15.860 --> 00:16:20.039 it's bring your own device and that laptops are 00:16:20.039 --> 00:16:23.299 in support and it's not priority wireless. But 00:16:23.299 --> 00:16:25.100 it is interesting, right? Because it leverages 00:16:25.100 --> 00:16:28.779 fiber. It's a strength AT &T has. And I think 00:16:28.779 --> 00:16:31.539 their wireless business is doing okay now. It's 00:16:31.539 --> 00:16:35.340 not hurting. It was before. And it's innovative, 00:16:35.480 --> 00:16:38.360 right? I could see Verizon doing this too right 00:16:38.360 --> 00:16:40.039 now that they have a much bigger fiber footprint. 00:16:40.340 --> 00:16:43.620 It's like the return of bundling. Yeah. But a 00:16:43.620 --> 00:16:45.940 little bit more creative. Yeah, I definitely 00:16:45.940 --> 00:16:49.360 see it as I would take this. The launch of this 00:16:49.360 --> 00:16:53.799 plan is TNT and others now are beginning to react 00:16:53.799 --> 00:16:57.799 to the cable bundle. So I think what we saw last 00:16:57.799 --> 00:17:01.299 year through to the end of 2025 was that cable 00:17:01.299 --> 00:17:04.490 companies like Comcast and Charter. were really 00:17:04.490 --> 00:17:08.130 working on offering very competitive prices for 00:17:08.130 --> 00:17:12.450 their bundle of fixed internet service and MVNO 00:17:12.450 --> 00:17:15.190 services. And so they were really offering some 00:17:15.190 --> 00:17:18.049 very competitive pricing plans to the point where 00:17:18.049 --> 00:17:21.789 mobile was almost free. And so I would expect 00:17:21.789 --> 00:17:24.490 that this plan from AT &T, I would expect other 00:17:24.490 --> 00:17:27.089 plans to come out that are similar to this in 00:17:27.089 --> 00:17:29.250 terms of, it's essentially the telco response 00:17:29.250 --> 00:17:32.150 to the cable threat or to the cable pricing strategy. 00:17:33.230 --> 00:17:35.289 I would also say, I forgot about this because 00:17:35.289 --> 00:17:37.029 this happened this morning and I was half awake 00:17:37.029 --> 00:17:39.750 when I saw it. T -Mobile also launched a bundle 00:17:39.750 --> 00:17:42.730 plan, which I forgot to mention, which is $45 00:17:42.730 --> 00:17:46.930 a month, which is half the price. And it's called 00:17:46.930 --> 00:17:51.809 5G Home Minternet because it's using Mint, an 00:17:51.809 --> 00:17:54.910 unlimited wireless plan with 5G Home Internet 00:17:54.910 --> 00:17:58.910 for $45, which is absurd. But again, it's not 00:17:58.910 --> 00:18:01.670 going to be very fast, right? Because it's FWA. 00:18:03.000 --> 00:18:05.539 It's mint, so it's going to be a deprioritized 00:18:05.539 --> 00:18:08.359 mobile connection. But if you don't care about 00:18:08.359 --> 00:18:10.559 the price or if you don't care about the speed 00:18:10.559 --> 00:18:13.400 and you do care about the price, I got to say 00:18:13.400 --> 00:18:17.000 45 bucks a month for home and for your cellular 00:18:17.000 --> 00:18:19.900 line is pretty impressive. It's a pretty good 00:18:19.900 --> 00:18:21.980 deal. And it really does. I think what it does 00:18:21.980 --> 00:18:25.220 is it reflects the this is the new competitive 00:18:25.220 --> 00:18:27.660 landscape in the U .S. telecom industry today. 00:18:27.720 --> 00:18:31.099 It's really industry where. these competitive 00:18:31.099 --> 00:18:34.299 bundles of fixed services and mobile services? 00:18:34.680 --> 00:18:37.299 And what are the terms and conditions? How can 00:18:37.299 --> 00:18:40.039 they reduce the price of one in order to sell 00:18:40.039 --> 00:18:42.559 the other? I think that is the game that we're 00:18:42.559 --> 00:18:44.240 going to see this year. I think it really is 00:18:44.240 --> 00:18:48.599 a game of convergence, bundled services, lower 00:18:48.599 --> 00:18:50.759 prices in some cases, but they're still, like 00:18:50.759 --> 00:18:52.480 you said, it's the terms and conditions are really 00:18:52.480 --> 00:18:54.680 where this is. How can they eke out a little 00:18:54.680 --> 00:18:57.440 bit of profit by selling extra services to these 00:18:57.440 --> 00:18:58.880 customers? I think that's where you're really 00:18:58.880 --> 00:19:02.609 going to start to see In general, the U .S. telecom 00:19:02.609 --> 00:19:05.230 industry is moving to this new level of competition, 00:19:05.390 --> 00:19:09.170 which used to be free phones. It used to be additional 00:19:09.170 --> 00:19:11.950 mobile services. It used to be you get Netflix 00:19:11.950 --> 00:19:14.829 for free. And now the game has changed a little 00:19:14.829 --> 00:19:17.150 bit in that it's a bundled service offering. 00:19:17.210 --> 00:19:19.470 You get fixed and mobile together. I think that's 00:19:19.470 --> 00:19:23.789 where things are headed right now. Yep. Yeah, 00:19:23.869 --> 00:19:27.089 fun, fun stuff. So I've got another topic. I'm 00:19:27.089 --> 00:19:29.700 going to pivot to my second one. which is a very 00:19:29.700 --> 00:19:31.319 different topic than what we've been talking 00:19:31.319 --> 00:19:33.420 about. And frankly, I think it's a lot more complicated 00:19:33.420 --> 00:19:36.099 topic, but I think it's worth bringing up here. 00:19:36.180 --> 00:19:38.759 It's about the sort of the future of the wireless 00:19:38.759 --> 00:19:42.140 network, what it's going to look like. And I 00:19:42.140 --> 00:19:44.759 bring it up because I've really been thinking 00:19:44.759 --> 00:19:48.500 about what, as operators upgrade their networks, 00:19:48.759 --> 00:19:51.099 their mobile networks, what are we going to see? 00:19:51.180 --> 00:19:52.539 What is it going to look like? What are they 00:19:52.539 --> 00:19:55.410 going to do in order to remain competitive? And 00:19:55.410 --> 00:19:57.309 a big topic that they've been talking about, 00:19:57.369 --> 00:19:59.490 as you and I have discussed in the past, is this 00:19:59.490 --> 00:20:04.910 idea of installing GPUs inside the network, inside 00:20:04.910 --> 00:20:08.150 base stations, in order to conduct AI inferencing 00:20:08.150 --> 00:20:12.930 business. This is a big part of the AI RAN story 00:20:12.930 --> 00:20:15.369 that a lot of companies have been talking about 00:20:15.369 --> 00:20:19.490 in relation to 6G. And we keep getting additional 00:20:19.490 --> 00:20:22.769 insights into what this battle is really about. 00:20:23.369 --> 00:20:26.349 I remain unconvinced as to the value of putting 00:20:26.349 --> 00:20:28.829 GPUs inside of base stations on a widespread 00:20:28.829 --> 00:20:31.950 basis. I think that it may make sense in certain 00:20:31.950 --> 00:20:34.789 circumstances. Certainly T -Mobile and SoftBank 00:20:34.789 --> 00:20:37.309 have been very vocal about that. But the reason 00:20:37.309 --> 00:20:39.930 I bring it up today is because we get a little 00:20:39.930 --> 00:20:42.890 bit more sort of commentary on this topic. And 00:20:42.890 --> 00:20:45.750 so I'm going to mention two, I think, two very 00:20:45.750 --> 00:20:48.549 telling quotes about this that I think casts 00:20:48.549 --> 00:20:51.930 a little bit of cold water on the idea of... 00:20:52.299 --> 00:20:55.519 putting a set of GPUs inside of a base station 00:20:55.519 --> 00:20:59.500 in order to run AI inferencing, in order to sell 00:20:59.500 --> 00:21:02.359 inferencing as a service and for operators to 00:21:02.359 --> 00:21:04.680 make more money that way. The first quote comes 00:21:04.680 --> 00:21:08.599 from Egal Elba at AT &T. I'm sure you've talked 00:21:08.599 --> 00:21:10.500 to him in the past. He's a real outspoken guy. 00:21:10.619 --> 00:21:13.259 He's a real smart guy. He really has a real good 00:21:13.259 --> 00:21:15.960 outlook on the industry. And he speaks in a way 00:21:15.960 --> 00:21:17.579 that I think is understandable and relevant. 00:21:17.799 --> 00:21:20.200 So one of the quotes that he just gave is that 00:21:20.200 --> 00:21:24.319 he talked about... the proliferation of computing 00:21:24.319 --> 00:21:27.920 across the nation in all metro areas, he said, 00:21:28.000 --> 00:21:31.059 I'm not sure that there's much value in extending 00:21:31.059 --> 00:21:35.119 that compute all the way to the far edge to save 00:21:35.119 --> 00:21:37.680 another millisecond or two milliseconds of latency. 00:21:38.059 --> 00:21:40.539 And so he was discussing that in the context 00:21:40.539 --> 00:21:44.099 of whether it makes sense to install GPUs inside 00:21:44.099 --> 00:21:46.940 of AT &T's network in order to sell AI inferencing. 00:21:47.140 --> 00:21:50.140 And essentially, he's having trouble seeing the 00:21:50.140 --> 00:21:52.480 value in that idea. And I think that's interesting. 00:21:52.799 --> 00:21:55.240 Obviously, operators have very different approaches 00:21:55.240 --> 00:21:57.539 to strategy. Some operators go real heavy on 00:21:57.539 --> 00:22:00.680 5G, some don't. So it does make sense to have 00:22:00.680 --> 00:22:03.240 a different view on this particular topic from 00:22:03.240 --> 00:22:06.140 AT &T compared to T -Mobile or others. But I 00:22:06.140 --> 00:22:07.579 think it is interesting. And then the last quote 00:22:07.579 --> 00:22:09.519 I've got here that I think is also interesting 00:22:09.519 --> 00:22:12.440 and talks to this. topic specifically is from 00:22:12.440 --> 00:22:14.680 i guess it's from the wall street journal but 00:22:14.680 --> 00:22:16.660 i'm reading it through axios but they talked 00:22:16.660 --> 00:22:20.640 about um the recent you talked about it the nvidia's 00:22:20.640 --> 00:22:23.119 new products that nvidia released at the at their 00:22:23.119 --> 00:22:29.359 big conference the gtc conference yeah ai grid 00:22:29.359 --> 00:22:32.160 stuff and the quote here though is just talking 00:22:32.160 --> 00:22:36.519 about it's from a from a guy named paul kadrosky 00:22:36.519 --> 00:22:40.900 a venture investor anyway he told it OK, yeah, 00:22:40.980 --> 00:22:42.339 I've never heard of him, but I thought this quote 00:22:42.339 --> 00:22:45.400 was interesting in that he told Wall Street Journal 00:22:45.400 --> 00:22:48.380 that the inferencing stuff is incredibly important 00:22:48.380 --> 00:22:52.440 to NVIDIA because it's all efficiency driven. 00:22:52.759 --> 00:22:56.680 And that basically his point is that NVIDIA is, 00:22:56.720 --> 00:22:59.720 quote, desperately trying to find a way to extend 00:22:59.720 --> 00:23:04.460 the franchise into inference. And I thought that 00:23:04.460 --> 00:23:08.279 was interesting. Essentially, GPU or sorry. NVIDIA 00:23:08.279 --> 00:23:11.799 has built their business on selling the chips 00:23:11.799 --> 00:23:14.619 that make AI training possible. And obviously 00:23:14.619 --> 00:23:17.480 AI training is extremely compute intensive. You 00:23:17.480 --> 00:23:19.539 really need tons of GPUs in order to get the 00:23:19.539 --> 00:23:23.940 AI training working. But when you talk about 00:23:23.940 --> 00:23:26.720 AI inferencing, which is after you train an AI 00:23:26.720 --> 00:23:29.079 model, then you use AI inferencing in order to 00:23:29.079 --> 00:23:33.420 deliver the results from that AI model, not necessarily 00:23:33.420 --> 00:23:37.349 as computing intensive. And so NVIDIA is in this 00:23:37.349 --> 00:23:39.970 position where they need to stay relevant as 00:23:39.970 --> 00:23:43.509 AI inferencing becomes more important. And so 00:23:43.509 --> 00:23:47.190 that's a big focus for NVIDIA now, which that 00:23:47.190 --> 00:23:51.190 just ties into the idea of putting GPUs inside 00:23:51.190 --> 00:23:52.990 of the base station. Obviously, NVIDIA wants 00:23:52.990 --> 00:23:56.730 to expand their market dominance and put their 00:23:56.730 --> 00:23:59.569 chips into more and more places. So they've been 00:23:59.569 --> 00:24:02.109 talking about making their chips relevant to 00:24:02.109 --> 00:24:04.269 telecom operators, putting their chips inside 00:24:04.269 --> 00:24:06.880 of base stations. I think it remains to be seen 00:24:06.880 --> 00:24:09.980 whether big equipment vendors like Ericsson and 00:24:09.980 --> 00:24:14.380 operators like AT &T will adopt that strategic 00:24:14.380 --> 00:24:17.880 effort by NVIDIA and others. Obviously, we've 00:24:17.880 --> 00:24:20.539 seen Nokia. Nokia has a huge investment from 00:24:20.539 --> 00:24:24.500 NVIDIA. So obviously, Nokia is a proponent of 00:24:24.500 --> 00:24:26.819 this idea of putting GPUs inside of base stations. 00:24:27.019 --> 00:24:31.220 But I think it remains to be seen whether that 00:24:31.220 --> 00:24:34.799 will be adopted by other... players in the industry, 00:24:34.940 --> 00:24:38.779 such as AT &T and Ericsson and people like that. 00:24:38.960 --> 00:24:42.519 So I think this is, I think the real battlefield 00:24:42.519 --> 00:24:45.460 to watch is as 6G standards are finalized, it's 00:24:45.460 --> 00:24:48.700 like how many, how many 6G players are going 00:24:48.700 --> 00:24:52.039 to want to adopt NVIDIA, Silicon and their base 00:24:52.039 --> 00:24:54.140 stations? I think that is an open question. And 00:24:54.140 --> 00:24:56.240 it is one where basically billions of dollars 00:24:56.240 --> 00:24:57.980 hang in the balance as to whether they decide 00:24:57.980 --> 00:25:00.640 to chase AI inferencing and this new business 00:25:00.640 --> 00:25:03.519 model. I think it remains to be seen. But I know 00:25:03.519 --> 00:25:05.359 you have a slightly different take on that topic. 00:25:05.420 --> 00:25:07.680 I don't know. It's a big one. A lot going on 00:25:07.680 --> 00:25:10.039 there. But I don't know. I'm collecting these 00:25:10.039 --> 00:25:12.740 quotes to understand these different perspectives 00:25:12.740 --> 00:25:14.460 on this industry to understand what's happening. 00:25:14.559 --> 00:25:16.000 So at least I can understand what the battle 00:25:16.000 --> 00:25:19.420 is. Don't know what to say. I do think NVIDIA 00:25:19.420 --> 00:25:22.299 is absolutely the key places to use inference. 00:25:22.799 --> 00:25:25.779 I would say their architecture is not, wasn't 00:25:25.779 --> 00:25:28.619 originally designed to be as efficient for inference 00:25:28.619 --> 00:25:31.089 as it was for training. But they've worked through 00:25:31.089 --> 00:25:33.470 a lot and they've re -architected their GPU delivery 00:25:33.470 --> 00:25:35.730 generation to be better at inference. There's 00:25:35.730 --> 00:25:37.509 other, they have more competition in inference 00:25:37.509 --> 00:25:39.589 than they do in training. So I actually think 00:25:39.589 --> 00:25:42.650 there's more compute available to the industry 00:25:42.650 --> 00:25:46.609 for AI inference that's very efficient. So I 00:25:46.609 --> 00:25:48.349 think that's something to consider as well. It's 00:25:48.349 --> 00:25:51.309 not just an NVIDIA story to be told in inference. 00:25:52.029 --> 00:25:55.430 They're just the loudest voice for all AI, whether 00:25:55.430 --> 00:25:58.799 it's inference or training. And I think. Paul 00:25:58.799 --> 00:26:01.519 Kudrowski is right in that sense. I will say 00:26:01.519 --> 00:26:05.259 Egolyba is definitely one of the better voices 00:26:05.259 --> 00:26:07.700 in the industry. And I do believe a lot of what 00:26:07.700 --> 00:26:10.799 he has to say. But I also believe that AT &T 00:26:10.799 --> 00:26:16.500 is behind T -Mobile in deploying standalone and 00:26:16.500 --> 00:26:21.559 probably 6G as well. While I think AT &T is on 00:26:21.559 --> 00:26:23.619 the right path, I'm not sure that they're necessarily 00:26:23.619 --> 00:26:27.319 the leader in the space to be. making leadership 00:26:27.319 --> 00:26:31.160 statements. And I want to see them be a leader. 00:26:31.259 --> 00:26:34.319 I want competition. I want Verizon to also be 00:26:34.319 --> 00:26:36.779 more competitive. Right now, T -Mobile is like 00:26:36.779 --> 00:26:38.720 the one carrier that's really talking about AI 00:26:38.720 --> 00:26:41.400 rant, right? Out of the big three. And I think 00:26:41.400 --> 00:26:44.440 part of that is just because they've basically 00:26:44.440 --> 00:26:47.420 been running on standalone for way longer than 00:26:47.420 --> 00:26:49.400 their competitors have. Whether you're talking 00:26:49.400 --> 00:26:53.980 about their 600 megahertz or the 2 .4, that. 00:26:54.220 --> 00:26:57.019 So I think... They're mentally in a different 00:26:57.019 --> 00:27:02.299 place, I think, in D &D and Verizon. Also, their 00:27:02.299 --> 00:27:04.319 capex cycle is totally different too, right? 00:27:04.380 --> 00:27:06.420 Like they front -end loaded so much more of their 00:27:06.420 --> 00:27:09.519 capex. And I think they're going to be doing 00:27:09.519 --> 00:27:12.579 a refresh like they said earlier too. So I think 00:27:12.579 --> 00:27:15.640 the reality is going to be somewhere in the middle, 00:27:15.700 --> 00:27:17.599 right? It's not going to be zero, but it's also 00:27:17.599 --> 00:27:21.380 not going to be 100%. I believe it's going to 00:27:21.380 --> 00:27:25.349 be a mixture of stuff at the RAM and stuff. sitting 00:27:25.349 --> 00:27:27.289 somewhere in the near edge, but I don't think 00:27:27.289 --> 00:27:30.289 everybody's going to need it. And I think the 00:27:30.289 --> 00:27:32.109 CSPs are probably going to be the ones who actually 00:27:32.109 --> 00:27:35.349 drive it in terms of demand. I do think T -Mobile 00:27:35.349 --> 00:27:38.049 is showing that there are some ways to do AI 00:27:38.049 --> 00:27:41.150 in the core, like with the live translation feature, 00:27:41.329 --> 00:27:43.710 which is just one of many. But I also have a 00:27:43.710 --> 00:27:45.309 feeling there will be some applications that 00:27:45.309 --> 00:27:48.250 they want to deploy that will require like RAN 00:27:48.250 --> 00:27:52.170 -based compute for really low latency applications. 00:27:52.930 --> 00:27:55.299 And to Egal's point. He's right. There probably 00:27:55.299 --> 00:27:58.259 aren't any applications today that require kind 00:27:58.259 --> 00:28:01.420 of the latency. But if we're going to do stuff 00:28:01.420 --> 00:28:04.119 like remote drone operation, you're going to 00:28:04.119 --> 00:28:06.240 want as low latency as you can get. Then again, 00:28:06.279 --> 00:28:08.619 that might be fully autonomous, right? But again, 00:28:08.759 --> 00:28:10.900 what's going to be driving that? Maybe you need 00:28:10.900 --> 00:28:14.200 to run an agent in the RAN to control the drone 00:28:14.200 --> 00:28:18.339 efficiently and faithfully. So I think we're 00:28:18.339 --> 00:28:20.579 so far out in terms of actually figuring out 00:28:20.579 --> 00:28:25.210 what the topology is going to look like. In my 00:28:25.210 --> 00:28:29.329 mind, I would be looking to the GOs and the T 00:28:29.329 --> 00:28:31.450 -Mobiles of the world on how this is going to 00:28:31.450 --> 00:28:34.369 work, less so the ones that are staff followers 00:28:34.369 --> 00:28:37.789 or even people that are trailing. I wish AT &T 00:28:37.789 --> 00:28:40.990 wasn't behind, but they very clearly are. I want 00:28:40.990 --> 00:28:44.589 to see what the leaders who are leading are doing. 00:28:44.849 --> 00:28:48.529 And maybe we'll see some competitive hopscotch, 00:28:48.589 --> 00:28:51.450 right? We can see someone leapfrogging another. 00:28:52.009 --> 00:28:54.829 Which we saw with 5G, right? T -Mobile very clearly 00:28:54.829 --> 00:28:58.609 was a distant fourth or third, depending on how 00:28:58.609 --> 00:29:00.769 you measure it. And they now are number one. 00:29:00.890 --> 00:29:04.470 There are ways that things can change. And I'm 00:29:04.470 --> 00:29:06.809 optimistic. I love competition. I don't want 00:29:06.809 --> 00:29:09.329 to see T -Mobile be the one and only leader in 00:29:09.329 --> 00:29:11.589 the space. I want them to be challenged more. 00:29:11.869 --> 00:29:14.029 Like I really thought Jio was actually going 00:29:14.029 --> 00:29:16.089 to be more aggressive with how they deployed 00:29:16.089 --> 00:29:19.470 5G. And we're going to see more unique 5G use 00:29:19.470 --> 00:29:22.500 cases. And they did deploy some slices, but it 00:29:22.500 --> 00:29:24.740 didn't really seem like it took. Yeah. I'm really 00:29:24.740 --> 00:29:27.019 curious to see how this pans out. I think your 00:29:27.019 --> 00:29:30.220 skepticism is fully warranted. I just think that 00:29:30.220 --> 00:29:32.660 will probably land somewhere in the middle. Yeah. 00:29:32.859 --> 00:29:35.220 It'll be interesting to see who makes what decision 00:29:35.220 --> 00:29:37.000 and when. Now is the time when they're really 00:29:37.000 --> 00:29:39.420 charting out where they're going to invest, how 00:29:39.420 --> 00:29:41.339 they're going to do it. And I think it'll be 00:29:41.339 --> 00:29:45.750 interesting to see how. different companies arrive 00:29:45.750 --> 00:29:48.150 at different decisions given the same inputs. 00:29:48.369 --> 00:29:50.069 They all essentially have the same view of the 00:29:50.069 --> 00:29:52.029 market, but I expect that they're like, just 00:29:52.029 --> 00:29:53.450 like in 5G, they're going to make very different 00:29:53.450 --> 00:29:57.109 decisions about how to go forward. And like, 00:29:57.130 --> 00:29:59.890 it reminds me of the whole Verizon edge compute 00:29:59.890 --> 00:30:02.190 story, right? Like they deployed all these edge 00:30:02.190 --> 00:30:07.829 compute nodes all over the country and nothing. 00:30:08.690 --> 00:30:11.789 Yeah. Yeah. There is a penalty to moving too 00:30:11.789 --> 00:30:14.759 early. But there is a lot of rewards if you move 00:30:14.759 --> 00:30:17.279 at the right time. And I think that's the question. 00:30:17.599 --> 00:30:20.279 I assume eventually, yeah, AI inferencing will 00:30:20.279 --> 00:30:23.099 be everywhere and deploying it at the very far 00:30:23.099 --> 00:30:26.839 edge would make sense. But is that a 2040 situation 00:30:26.839 --> 00:30:31.799 or is it a 2030 situation or is it a 2028 opportunity? 00:30:32.500 --> 00:30:37.759 Really hard to know at this point. Yeah. All 00:30:37.759 --> 00:30:40.400 right. I'll move on to my final topic and we're 00:30:40.400 --> 00:30:43.440 back to talking space. and SpaceX specifically, 00:30:43.960 --> 00:30:47.640 which is now the parent company of XAI, and a 00:30:47.640 --> 00:30:52.900 joint venture with Tesla on TerraFab. This happened 00:30:52.900 --> 00:30:56.599 a few weeks ago, and I had some press reach out 00:30:56.599 --> 00:30:59.799 to me about it. And I was brutally honest that 00:30:59.799 --> 00:31:02.940 I don't think it's going to happen in the way 00:31:02.940 --> 00:31:05.799 it was presented, which was Tesla's going to 00:31:05.799 --> 00:31:10.160 build a fab with SpaceX to make one terawatt 00:31:10.160 --> 00:31:13.839 of... capacity available to those companies for 00:31:13.839 --> 00:31:17.359 compute. And guess what? Today they announced 00:31:17.359 --> 00:31:20.539 that they were partnering with Intel. And that 00:31:20.539 --> 00:31:23.880 makes way more sense. Intel's CEO was pictured 00:31:23.880 --> 00:31:27.640 and did announcements with Elon and the Terrapad, 00:31:27.900 --> 00:31:32.599 which will make processors for Tesla and for 00:31:32.599 --> 00:31:36.619 SpaceX. And that will mostly be for these data 00:31:36.619 --> 00:31:39.299 centers that they're building for AI, but also. 00:31:39.789 --> 00:31:42.369 For all the chips that SpaceX needs for its rockets, 00:31:42.490 --> 00:31:45.349 for Tesla and all its cars and all its robots. 00:31:45.970 --> 00:31:49.329 But I will say that this is not a near -term 00:31:49.329 --> 00:31:53.349 story. They haven't purchased any land as far 00:31:53.349 --> 00:31:55.450 as we know. They haven't built any facilities. 00:31:55.569 --> 00:31:58.289 They haven't ordered any tools. Tools alone, 00:31:58.569 --> 00:32:01.589 best case scenario, 18 months, probably more 00:32:01.589 --> 00:32:04.269 like 24. So minimum, if they had even placed 00:32:04.269 --> 00:32:07.559 any orders, would be two years from now. Realistically, 00:32:07.799 --> 00:32:09.599 I think we're looking three to four years from 00:32:09.599 --> 00:32:13.279 now. So that makes sense why Tesla's inked a 00:32:13.279 --> 00:32:18.460 deal with VIMS last year to basically continue 00:32:18.460 --> 00:32:21.880 its supply chain for automotives using the latest 00:32:21.880 --> 00:32:24.720 possible nodes. And it'll be interesting to see 00:32:24.720 --> 00:32:28.339 what'll happen in terms of how these fabs would, 00:32:28.420 --> 00:32:31.240 what these fabs will look like. They say it's 00:32:31.240 --> 00:32:34.660 two chip factories. So I have a feeling what 00:32:34.660 --> 00:32:37.740 will most likely happen. is that one will be 00:32:37.740 --> 00:32:42.000 a logic factory which is chips for and processors 00:32:42.000 --> 00:32:44.720 and it will be like a memory factory which will 00:32:44.720 --> 00:32:49.680 be probably samsung enabled application because 00:32:49.680 --> 00:32:51.019 they're already partnered with samsung today 00:32:51.019 --> 00:32:54.880 in texas like the way i look at it is that's 00:32:54.880 --> 00:32:56.819 not doing any of this on their own to begin with 00:32:56.819 --> 00:33:00.279 and chip design is much easier than chip fabrication 00:33:00.279 --> 00:33:02.920 especially if you don't have the know -how there 00:33:02.920 --> 00:33:06.130 are actively like five companies in the world 00:33:06.130 --> 00:33:09.670 that are producing nearly 100 % on chips on earth. 00:33:09.869 --> 00:33:13.890 If you take TSMC, Samsung, and Intel, you've 00:33:13.890 --> 00:33:17.450 already 70, 80 % of the market. Then you add 00:33:17.450 --> 00:33:21.690 SK Hynix and Micron, you've got 100%. It's just, 00:33:21.750 --> 00:33:24.049 there's not that many companies doing it. It 00:33:24.049 --> 00:33:26.890 costs tens of billions of dollars. I think the 00:33:26.890 --> 00:33:29.910 quote was 25 billion. Elon said, that's pretty 00:33:29.910 --> 00:33:33.049 realistic. I think if we're talking two fabs. 00:33:33.420 --> 00:33:35.799 But that's with help. If they did this on their 00:33:35.799 --> 00:33:38.440 own, that was a low number. It probably would 00:33:38.440 --> 00:33:41.000 have cost them double. And is this, they're going 00:33:41.000 --> 00:33:44.559 to build the manufacturing facility that will 00:33:44.559 --> 00:33:46.579 make the chips, right? That's what they're going 00:33:46.579 --> 00:33:48.680 to do? Are they going to do it in Texas or somewhere 00:33:48.680 --> 00:33:51.740 in the U .S.? Yes, a cluster of factories that 00:33:51.740 --> 00:33:54.599 they've built. My former colleague, Will, lives 00:33:54.599 --> 00:33:56.839 out in Bastrop, and that's where their center 00:33:56.839 --> 00:33:59.920 of orbit is. So I believe they'll probably be 00:33:59.920 --> 00:34:01.640 doing this out in Bastrop because they own the 00:34:01.640 --> 00:34:04.859 land already. Realistically, like they do with 00:34:04.859 --> 00:34:08.000 battery electric vehicles, the battery fabs are 00:34:08.000 --> 00:34:11.860 near the vehicle fab. I expect these chip fabs 00:34:11.860 --> 00:34:14.239 will probably be close to the data center and 00:34:14.239 --> 00:34:16.119 also the car manufacturing facilities because 00:34:16.119 --> 00:34:19.119 then you can make very quick turns. And if you 00:34:19.119 --> 00:34:21.639 look at like how the Chinese have been able to 00:34:21.639 --> 00:34:23.699 build these cars so cheaply and so effectively, 00:34:24.039 --> 00:34:26.219 it's because the supply chains are right there. 00:34:26.300 --> 00:34:28.360 So when they need to troubleshoot things and 00:34:28.360 --> 00:34:31.570 make adjustments and change things quickly. Everything, 00:34:31.710 --> 00:34:33.889 the entire ecosystem is near them and around 00:34:33.889 --> 00:34:36.670 them. And that's really how they're able to make 00:34:36.670 --> 00:34:39.570 things happen fast. So I understand the need 00:34:39.570 --> 00:34:41.670 and desire. Also, it gives them more control 00:34:41.670 --> 00:34:45.929 of the supply chain, costs. They can custom build 00:34:45.929 --> 00:34:49.090 things as they see fit. But I have a strong feeling 00:34:49.090 --> 00:34:51.829 that this is going to be massively accretive 00:34:51.829 --> 00:34:54.829 to Intel for their founder business. They're 00:34:54.829 --> 00:34:56.710 going to be able to claim quite a bit in terms 00:34:56.710 --> 00:34:59.389 of revenue and profit. I sort of think this is 00:34:59.389 --> 00:35:00.949 going to be a near -term story for them, but 00:35:00.949 --> 00:35:03.429 it does look good for their stock. Last I checked, 00:35:03.630 --> 00:35:10.030 it was up a good 5%. That's 4 % now. But in the 00:35:10.030 --> 00:35:14.050 last week, their stock is up 26%. So the good 00:35:14.050 --> 00:35:16.690 news is rolling for Intel. And this also means 00:35:16.690 --> 00:35:19.010 that an American company will be producing chips 00:35:19.010 --> 00:35:21.590 or helping to produce chips for another American 00:35:21.590 --> 00:35:24.590 company, which is very much part of what the 00:35:24.590 --> 00:35:26.849 tracking administration wants. And truthfully... 00:35:27.070 --> 00:35:29.429 I think every administration has wanted for a 00:35:29.429 --> 00:35:31.750 long time. Yeah. Yeah. It makes a lot of sense. 00:35:31.869 --> 00:35:34.409 What I think is interesting, here's another example 00:35:34.409 --> 00:35:38.530 of the expansion of Elon Inc. Elon Incorporated, 00:35:38.590 --> 00:35:41.050 now he's going to be building chips too for all 00:35:41.050 --> 00:35:42.730 of his various businesses. I know he's talked 00:35:42.730 --> 00:35:46.690 about how important it is to have control over 00:35:46.690 --> 00:35:49.510 the supply of chips for everything from Tesla 00:35:49.510 --> 00:35:53.989 to SpaceX to XAI. I would not say it's a surprise. 00:35:54.190 --> 00:35:55.969 I think it's very interesting. I think I will 00:35:55.969 --> 00:35:59.230 assume it will accelerate the work that Elon 00:35:59.230 --> 00:36:02.829 Incorporated is doing. And will this bring us 00:36:02.829 --> 00:36:05.050 a million data centers in space? I know that's 00:36:05.050 --> 00:36:07.989 part of the SpaceX IPO. That's the plan is to 00:36:07.989 --> 00:36:10.210 launch data centers into space and have them 00:36:10.210 --> 00:36:13.170 run AI out in space because it's better and cheaper 00:36:13.170 --> 00:36:15.469 there. I don't know. We'll see. But presumably 00:36:15.469 --> 00:36:18.769 this is part of the plan. I am not a believer 00:36:18.769 --> 00:36:21.869 in the space data center thing. I do think there 00:36:21.869 --> 00:36:24.469 will be some space data centers, but not at the 00:36:24.469 --> 00:36:26.869 scale that Elon's trying. A million seems like 00:36:26.869 --> 00:36:30.349 a lot. Yeah. And the truth is that, like, I actually, 00:36:30.409 --> 00:36:33.730 when SpaceX, long time, when SpaceX first started 00:36:33.730 --> 00:36:36.489 launching reusable rockets, I was like, I'm going 00:36:36.489 --> 00:36:38.289 to invest in the SpaceX IPO when it happens. 00:36:38.489 --> 00:36:41.730 But now there's, like, XAI, which is losing tons 00:36:41.730 --> 00:36:44.070 of money, and Twitter, which is also losing money, 00:36:44.190 --> 00:36:46.829 or not as profitable as it should be. Like, I 00:36:46.829 --> 00:36:48.750 don't want to touch its stock now. Like, I don't 00:36:48.750 --> 00:36:50.739 want anything to do with it. I was such a big 00:36:50.739 --> 00:36:52.500 fan of that. I'm like, I still believe space 00:36:52.500 --> 00:36:54.159 is the future. You want to invest in these space 00:36:54.159 --> 00:36:56.079 companies before they become trillion dollar 00:36:56.079 --> 00:36:58.780 companies. And by the way, the IPO might be the 00:36:58.780 --> 00:37:00.980 first trillion dollar IPO. We'll see what happens. 00:37:01.079 --> 00:37:03.239 I don't want to touch it. Because let's remember 00:37:03.239 --> 00:37:06.519 how many of these satellite constellations have 00:37:06.519 --> 00:37:09.119 come in, right? We had some discussions online 00:37:09.119 --> 00:37:11.260 with some people on Twitter recently, yesterday 00:37:11.260 --> 00:37:13.500 or today. It's going to cost them billions of 00:37:13.500 --> 00:37:16.400 dollars to maintain these thousands of satellites 00:37:16.400 --> 00:37:19.630 as they... blow up, but it had to blow up recently 00:37:19.630 --> 00:37:22.769 as a decommission within the five -year span. 00:37:23.170 --> 00:37:25.789 Like having to launch more of them, it costs 00:37:25.789 --> 00:37:27.969 a million for the satellite and a million to 00:37:27.969 --> 00:37:30.670 launch it. It's $2 million a satellite. That's 00:37:30.670 --> 00:37:33.269 an expensive endeavor. So you've got to have 00:37:33.269 --> 00:37:36.590 that momentum constantly going and that's a very 00:37:36.590 --> 00:37:39.809 high risk, especially when you add a fab, which 00:37:39.809 --> 00:37:42.789 was a conversation point that I had. One of Intel's 00:37:42.789 --> 00:37:46.510 biggest costs is just depreciation of the foundries. 00:37:47.019 --> 00:37:50.179 And that's why there's so many, you know, fabulous 00:37:50.179 --> 00:37:52.019 semiconductor companies because nobody wants 00:37:52.019 --> 00:37:54.639 to eat that depreciation. It's depreciation is 00:37:54.639 --> 00:37:56.079 a blessing and a curse when you look at it on 00:37:56.079 --> 00:37:57.599 accounting, right? Because it's a write -off, 00:37:57.639 --> 00:38:00.000 but it's also like a bleed on your financial 00:38:00.000 --> 00:38:03.360 viability. I think this is good for Tesla. Maybe 00:38:03.360 --> 00:38:05.980 they'll take on the financial risk. I don't know. 00:38:06.300 --> 00:38:08.920 Those details were not disclosed. And I don't 00:38:08.920 --> 00:38:10.639 know if we'll know those details for quite some 00:38:10.639 --> 00:38:13.639 time. But maybe we'll get some news from Intel 00:38:13.639 --> 00:38:15.929 on earnings and see that. I feel like they wanted 00:38:15.929 --> 00:38:17.869 to wait till earnings, but they couldn't for 00:38:17.869 --> 00:38:20.730 some reason. Maybe it was going to be. But yeah, 00:38:20.849 --> 00:38:22.530 I think this is great for everybody. I can't 00:38:22.530 --> 00:38:25.130 really see a downside. No, it does make a lot 00:38:25.130 --> 00:38:27.550 of sense. Those two companies partnering to do 00:38:27.550 --> 00:38:29.929 the manufacturer. Yeah. Yeah. And so it does 00:38:29.929 --> 00:38:32.949 dovetail in a little bit with my last topic, 00:38:33.030 --> 00:38:36.750 which is really about direct to device satellite 00:38:36.750 --> 00:38:39.550 stuff. It is within the realm of satellite operations 00:38:39.550 --> 00:38:43.030 a little bit. So the news item that came out 00:38:43.030 --> 00:38:46.760 about a week ago. was that, and this is, again, 00:38:46.860 --> 00:38:49.320 it's a little bit related to the topic of satellites. 00:38:49.579 --> 00:38:52.260 So there's a company called Grain Management. 00:38:52.400 --> 00:38:56.039 It's an investment company. And they, about a 00:38:56.039 --> 00:39:01.019 year ago, they agreed to purchase some 800 megahertz 00:39:01.019 --> 00:39:04.559 spectrum from T -Mobile. Not a huge surprise. 00:39:04.960 --> 00:39:07.880 Grain has previously dabbled in spectrum ownership. 00:39:08.219 --> 00:39:11.360 They're one of several investment companies that 00:39:11.360 --> 00:39:15.420 has done so. owning Spectrum in the US. But anyway, 00:39:15.539 --> 00:39:19.119 so they have already agreed to buy this 800 megahertz 00:39:19.119 --> 00:39:22.980 from T -Mobile. The new thing is that Grain mentioned 00:39:22.980 --> 00:39:26.119 in one of their FCC filings that they might be 00:39:26.119 --> 00:39:30.380 interested in leasing that Spectrum to direct 00:39:30.380 --> 00:39:33.940 -to -device companies like potentially Starlink. 00:39:34.019 --> 00:39:36.460 They didn't mention Starlink in the filing, but 00:39:36.460 --> 00:39:39.960 they said they may lease their 800 megahertz 00:39:39.960 --> 00:39:43.699 Spectrum to... D2D companies, and they didn't 00:39:43.699 --> 00:39:45.739 name any of them, but there's only a few of them 00:39:45.739 --> 00:39:49.539 out there. But I think what it reflects is that 00:39:49.539 --> 00:39:53.960 this D2D market is hot right now. This is where 00:39:53.960 --> 00:39:57.760 a lot of the activity is happening. We see Starlink, 00:39:57.920 --> 00:40:01.239 SpaceX launching Starlink, getting into the D2D 00:40:01.239 --> 00:40:03.539 market, direct -to -device, where a phone connects 00:40:03.539 --> 00:40:06.059 directly to a satellite. We've seen a lot of 00:40:06.059 --> 00:40:09.329 interest in AST Space Mobile. Same thing. They 00:40:09.329 --> 00:40:11.869 sell D2D connections. Tons of interest in that 00:40:11.869 --> 00:40:15.090 company. Link is another one. They merged with 00:40:15.090 --> 00:40:17.530 a company called OmniSpace with backing from 00:40:17.530 --> 00:40:20.690 a satellite operator, SES. They also were chasing 00:40:20.690 --> 00:40:23.889 this D2D market. I guess it shouldn't come as 00:40:23.889 --> 00:40:25.809 a surprise that a lot of these companies are 00:40:25.809 --> 00:40:28.349 clamoring or more companies are coming into this 00:40:28.349 --> 00:40:31.230 space. Grain management is talking about leasing 00:40:31.230 --> 00:40:34.570 its spectrum for D2D operations. But you're right. 00:40:34.630 --> 00:40:38.500 You talked about, is this... Is there too much 00:40:38.500 --> 00:40:40.460 investment going into satellites right now? Is 00:40:40.460 --> 00:40:42.659 there enough interest in satellites in order 00:40:42.659 --> 00:40:44.780 to maintain this level of investment? I think 00:40:44.780 --> 00:40:47.440 that certainly remains to be seen. Grain Management 00:40:47.440 --> 00:40:49.739 is one of the many companies that's coming into 00:40:49.739 --> 00:40:52.099 this space, hoping to take advantage of all the 00:40:52.099 --> 00:40:54.639 interest that's going on right now in there. 00:40:54.900 --> 00:40:57.760 They didn't say that anyone had leased their 00:40:57.760 --> 00:40:59.500 spectrum yet. They just said it's essentially 00:40:59.500 --> 00:41:03.119 available for lease. They're asking the FCC to 00:41:03.119 --> 00:41:05.550 make sure that... They can lease that spectrum 00:41:05.550 --> 00:41:09.869 out for a D2D application is essentially where 00:41:09.869 --> 00:41:12.130 they are now. They have not yet completed the 00:41:12.130 --> 00:41:14.469 purchase from T -Mobile. So presumably when they 00:41:14.469 --> 00:41:16.489 do, if they are able to close their purchase 00:41:16.489 --> 00:41:19.130 of that spectrum from T -Mobile and also receive 00:41:19.130 --> 00:41:23.510 an FCC waiver for D2D operations in that spectrum 00:41:23.510 --> 00:41:26.190 band, then they could go about leasing that spectrum 00:41:26.190 --> 00:41:29.670 to companies like potentially AST SpaceMobile, 00:41:29.849 --> 00:41:34.150 Starlink, Link Global. We don't know yet. But 00:41:34.150 --> 00:41:38.949 this is the market jostling that goes along with 00:41:38.949 --> 00:41:41.389 a market that is unclear how big it's going to 00:41:41.389 --> 00:41:44.409 be. Is there still opportunity here? It's really 00:41:44.409 --> 00:41:47.090 in its early stages of developing. And so I think 00:41:47.090 --> 00:41:48.769 it's interesting. I think it's something worth 00:41:48.769 --> 00:41:51.349 watching. It's just an indication that, yes, 00:41:51.449 --> 00:41:53.110 there's still excitement going on right here 00:41:53.110 --> 00:41:56.349 in the D2D market and the wider satellite industry. 00:41:57.099 --> 00:42:00.440 It's a sector of the market that is really fun 00:42:00.440 --> 00:42:02.639 to watch right now because practically anything 00:42:02.639 --> 00:42:05.179 can happen. And anything can happen any week. 00:42:05.880 --> 00:42:08.639 We don't know if maybe Amazon will buy GlobalStar 00:42:08.639 --> 00:42:10.619 next week. That was the rumor that came up about 00:42:10.619 --> 00:42:12.599 a week ago. We'll see. I think a lot of this 00:42:12.599 --> 00:42:14.579 stuff is just rumor and speculation, but we'll 00:42:14.579 --> 00:42:16.719 see. There could be something big that happens 00:42:16.719 --> 00:42:19.639 relatively soon. We don't know. I was going to 00:42:19.639 --> 00:42:23.059 say, this is an extension. So Grain is getting 00:42:23.059 --> 00:42:25.869 this 800 megabytes from T -Mobile. which was 00:42:25.869 --> 00:42:28.809 originally supposed to sell it to Dish. But Dish 00:42:28.809 --> 00:42:30.989 didn't have the money to buy it, and they had 00:42:30.989 --> 00:42:34.590 first refusal. But what's funny is, what's not 00:42:34.590 --> 00:42:38.389 talked about is, in Grain getting this 800 MHz 00:42:38.389 --> 00:42:44.110 spectrum, and also paying some cash to T -Mobile, 00:42:44.289 --> 00:42:47.469 T -Mobile's also getting all of Grain's 600 MHz 00:42:47.469 --> 00:42:51.389 licenses. So T -Mobile will strengthen its lowest 00:42:51.389 --> 00:42:54.800 band and coverage. with more frequency, which 00:42:54.800 --> 00:42:56.860 they obviously need in the 600 megahertz band. 00:42:57.199 --> 00:43:01.119 So T -Mobile wins, Grain gets their money, and 00:43:01.119 --> 00:43:03.960 SpaceX gets more spectrum, which they are massively 00:43:03.960 --> 00:43:07.039 starved for because they're going to be powering 00:43:07.039 --> 00:43:10.739 all of United's fleet. And we found out that 00:43:10.739 --> 00:43:13.980 Delta is going with Amazon Leo. So I think there's 00:43:13.980 --> 00:43:16.599 going to be way more airline using Starlink or 00:43:16.599 --> 00:43:19.000 some other Leo service. So there's going to be 00:43:19.000 --> 00:43:21.550 lots of demand for that. Plus on top of all the 00:43:21.550 --> 00:43:23.070 other things that are happening, like I wanted 00:43:23.070 --> 00:43:26.369 a cruise late last year. It was Starlink powered. 00:43:26.449 --> 00:43:28.289 They had the six or seven Starlink terminals 00:43:28.289 --> 00:43:30.630 on the ship. And it was the best internet connection 00:43:30.630 --> 00:43:33.130 I've ever done on a cruise ever. Obviously there's 00:43:33.130 --> 00:43:35.510 tons of room for improvement, but yeah, I gotta 00:43:35.510 --> 00:43:37.269 be honest with you. I don't want to drag that 00:43:37.269 --> 00:43:40.429 any longer. I just don't think D2D as a market 00:43:40.429 --> 00:43:43.050 is as big as people seem to be acting like it 00:43:43.050 --> 00:43:45.570 is. I think if there is a market for it, it's 00:43:45.570 --> 00:43:48.210 in the commercial space and like for commercial 00:43:48.210 --> 00:43:52.800 logistics. And offering, I think Amazon's approach 00:43:52.800 --> 00:43:55.539 is a lot more what I think where the market is 00:43:55.539 --> 00:43:57.960 going, which is servicing businesses. Because 00:43:57.960 --> 00:44:00.360 I think consumers would love to have satellite. 00:44:00.579 --> 00:44:03.340 I think it's a great service to have. I don't 00:44:03.340 --> 00:44:05.219 think it's the majority of the population. I 00:44:05.219 --> 00:44:08.280 think it's great for airlines. As a user, I don't 00:44:08.280 --> 00:44:09.820 want to pay for that additionally because I'm 00:44:09.820 --> 00:44:12.820 not going to need it 99 % of the time. Now, you're 00:44:12.820 --> 00:44:15.320 going rock climbing. I'm sure you do not have 00:44:15.320 --> 00:44:17.550 service where you're rock climbing. I do appreciate 00:44:17.550 --> 00:44:20.969 it for that situation. Yeah. So I think there 00:44:20.969 --> 00:44:24.050 are people who absolutely see the value and their 00:44:24.050 --> 00:44:27.190 lifestyle matches the need. But let's be honest, 00:44:27.309 --> 00:44:29.730 that's probably 10 % of the population. It might 00:44:29.730 --> 00:44:32.590 be like 1 % of my month, for example, that I'll 00:44:32.590 --> 00:44:35.590 need it. Exactly. So I think I like the idea 00:44:35.590 --> 00:44:37.730 of having the service and it being available. 00:44:38.030 --> 00:44:41.130 I think it's going to be huge for emergency services 00:44:41.130 --> 00:44:45.110 and other places where people need to have coverage 00:44:45.110 --> 00:44:47.159 where there just isn't any. but I just don't 00:44:47.159 --> 00:44:49.199 think it's going to replace. The way people are 00:44:49.199 --> 00:44:50.739 behaving is it's going to replace smartphones 00:44:50.739 --> 00:44:53.360 and terrestrial services. And I'm like, no, it's 00:44:53.360 --> 00:44:56.500 complimentary, but it's necessary. I think at 00:44:56.500 --> 00:44:58.139 your point, it will be part of the 6G standard, 00:44:58.300 --> 00:45:01.960 but I just don't think, I think it's like a component 00:45:01.960 --> 00:45:05.559 carrier. It's not an anchor. So I think that's 00:45:05.559 --> 00:45:08.300 important to understand. And it's expensive to 00:45:08.300 --> 00:45:10.500 send satellites up. Like it's really expensive 00:45:10.500 --> 00:45:13.059 and to maintain them and to keep them fed with 00:45:13.059 --> 00:45:16.949 data. And that's why Starlink. is so successful 00:45:16.949 --> 00:45:19.550 because they have that criticalness. When you 00:45:19.550 --> 00:45:23.329 buy a Starlink terminal, there's a 99 % chance 00:45:23.329 --> 00:45:25.869 you're going to have coverage, right? Like it's 00:45:25.869 --> 00:45:29.010 very high. So that's why I think it's going to 00:45:29.010 --> 00:45:31.730 be a challenge for smaller services to just keep 00:45:31.730 --> 00:45:33.909 up. And I think we're going to see some consolidation. 00:45:34.250 --> 00:45:36.829 Somebody might end up buying GlobalStar. It would 00:45:36.829 --> 00:45:38.650 also end up being Apple. They already own a big 00:45:38.650 --> 00:45:40.610 chunk of it. So it wouldn't be a big lift for 00:45:40.610 --> 00:45:43.329 them. But yeah, we'll see what happens. I don't 00:45:43.329 --> 00:45:45.860 think we're anywhere near knowing. Where the 00:45:45.860 --> 00:45:48.139 end state will be in, most likely we'll know 00:45:48.139 --> 00:45:51.079 until 2030. Yeah, this is going to be a pretty 00:45:51.079 --> 00:45:53.400 fun ride. I think for the next few years, while 00:45:53.400 --> 00:45:55.719 everybody figures out the opportunity here and 00:45:55.719 --> 00:45:57.519 how much revenue there really is going to be 00:45:57.519 --> 00:46:00.179 in there. But yeah, for the next few years, it's 00:46:00.179 --> 00:46:03.880 going to be practically anything goes. And yeah, 00:46:03.980 --> 00:46:06.659 we'll be interested to watch it. Yeah, I think 00:46:06.659 --> 00:46:08.800 that kind of wraps up our conversation for this 00:46:08.800 --> 00:46:12.000 week. Thanks again, Mike, for joining me. I also 00:46:12.000 --> 00:46:13.599 want to say thanks to our viewers and listeners 00:46:13.599 --> 00:46:16.989 for following on. for this week's topics. And 00:46:16.989 --> 00:46:19.030 if anyone has any topics for a future podcast, 00:46:19.309 --> 00:46:22.510 please reach out to us on social media. I'm at 00:46:22.510 --> 00:46:25.489 Anshul Sog and I'm at Mikey Dana. We both meet 00:46:25.489 --> 00:46:28.590 on Twitter and on LinkedIn. And please don't 00:46:28.590 --> 00:46:30.389 forget to rate us and subscribe. And we hope 00:46:30.389 --> 00:46:31.510 you have a great rest of your week.