AI0001: The Map Says Covered? Why Aren’t They Connecting? Also, Introducing TelePost.ai

Three pieces from this week, and all of them circle the same question: who actually gets left behind when the technology works fine. SpaceX has told the F-C-C that Starlink makes billions a year in rural subsidies pointless, and I think the terrestrial side is going to lose that fight on the technical merits. So I'd rather argue the part that's winnable, which is what a household can afford to pay.

Then there's a research finding about fiber and A-I usage that reads like good marketing and lands like a warning, plus the tool I built so that the world revolves around me instead of the other way round. One phone call a day, 500 news feeds, and a draft I still have to approve myself.

SpaceX tells the F-C-C that Starlink has made rural broadband subsidies pointless, and why local still wins. Fiber users ask 28.5 questions per A-I session. Cable users ask about fourteen. That's a digital divide forming in real time. I ran out of excuses not to blog, so I built a system that calls me every day and interviews me.

Everything in this episode, in full:

• The fiber-AI finding is good marketing and a warning about who gets left behind
• I Ran Out of Excuses Not to Blog, So I Built Something That Revolves Around Me
• Satellite Already Won the Scalability Argument. Affordability Is the Fight That Matters

Made with TelePost.

Transcript

SPEAKER_01 0:30

You can walk into some of the most economically challenged parts of South Texas, and the map will tell you it's covered. Multiple providers, wireline, wireless, sometimes fiber right down the street, and the household still isn't connected because the bill is $100 a month. Coverage isn't a signal. Coverage is whether people can afford to subscribe. This podcast is generated by telepost.ai using my words, my voice, and my ideas. Check out Telepost and make sure to tell a friend. I'm Drew Lentz, and this is the Waves Weekly Rundown, where I go

Coverage Versus Affordability

SPEAKER_01 1:00

through the week in connectivity and tell you what I actually think about it. Three things this week. SpaceX has told the FCC that Starlink has made rural broadband subsidies pointless, and I'll tell you why I think the terrestrial side loses that fight and should stop fighting it. Then a piece of research about fiber and AI that's great marketing and a genuinely worrying signal. And finally the tool I built that calls me on the phone every day and interviews me because I ran out of excuses not to publish. SpaceX has told the FCC that Starlink has made billions a year in rural broadband subsidies pointless. Rural telcos

Starlink Says Subsidies Are Obsolete

SPEAKER_01 1:34

and electric co-ops are pushing back on the technical merits. Light reading has covered the filing. I want to skip past the part everybody is arguing about. I think the terrestrial side is going to lose that argument, and I think we should stop spending energy on it. Here's why. Satellite is infinitely scalable, it seems, because there's no stopping the amount of birds these guys can put in space. Starlink V3 is pushing a terabyte per second, Amazon Leo is pushing a terabyte, and laser communications between the satellites themselves means the same number of birds in a constellation can just keep growing in capacity as the chips get better. The only time satellite actually failed in rural connectivity was latency. Low Earth orbit, fix that. So from a technology perspective, it's going to work, it's going to do everything they need it to do. That's a moot point now. Affordability is not. So look at who wins if the FCC buys the argument. Obviously, the satellite providers and not just the ones flying today. Amazon, Leo, Hughes, VISAT, and whoever pops up a constellation next. The losers are the incumbents, the wireless internet service providers, the local telephone companies, anyone who took a subsidy to build or maintain the network they already have, and municipal broadband. Think about a city like Far Texas, which built its own fiber network to connect one of the most broadband challenged communities in the United States. That's the kind of build that stops getting federal support, whether it came through USDA or the FCC. If big money wins this, big money gets to keep doing what it's doing, and it chokes out the smaller providers who are actually ingrained in the community, the ones sponsoring baseball teams and local events and giving back to charities. All of that gets washed out. And if USF dries up, a lot of the money smaller operators run on comes from there. They'll be chasing grants and economic stimulus instead. Meanwhile, the cost to start up a small internet service provider pales next to the cost of launching rockets. So the people with the money end up taking the money short game, sure, they deploy more broadband faster. Long run, small investments over time probably yield a better result for the communities. So if you're a co-op breaking ground next quarter, don't change your plans. Keep the build on schedule. When someone tells you your county is covered, go measure whether residents can afford to subscribe. Publish community pricing that a $50 or $100 satellite bill can't match. And start mapping the funding paths you'd use if USF goes away instead of waiting to find out. At the end of the day, you can rely on your neighbor in your community more than you can rely on a corporation to do what's right for your area.

Why Local Providers Still Matter

SPEAKER_01 3:56

28.5 questions per AI session. That's what Recon Analytics says people on fiber connections are doing. They survey 6,000 people a week about AI, and they describe fiber sessions as 50% more intensive than cable sessions. Do that math backwards, and someone on cable is asking about 14 questions per session. Now the research came from the Fiber Broadband Association. So the fiber providers obviously went from this, and it's already starting to show up in broadband funding arguments. I don't know exactly where the correlation comes from.

Fiber, AI, And A New Divide

SPEAKER_01 4:26

I do know who paid for it, and there might be some advantage in why they're mentioning it, rather than someone mentioning it at a telco or a cable expo. But here's the thing: even if it's good marketing, it's a big red flag for me. Because this is another digital divide forming in front of us, we already know there's a divide between people who can afford high-speed connectivity and people who can't. Now it isn't only that they don't get the same services on a level playing field, it's that with advanced applications like AI, not being on that kind of network puts them at a real disadvantage. Think about South Texas, rural markets, tribal lands, places that don't traditionally have fiber. It puts the students at a disadvantage. And even if fiber were available, plenty of those households couldn't afford it. Does it change the math for a fixed wireless operator? It has to. You have to start considering what your customers are actually doing on those connections. Luckily, there are products like Tirana delivering gigabit over wireless and Starlink pushing higher speeds. That could level the playing field. We just don't know what it looks like yet because I doubt anybody researched the fixed wireless side. I genuinely love to see those session numbers. And then there's the part everybody forgets where does a gigabit connection actually fall apart inside the house? My son just moved into a new apartment and called me because he needed to figure out how to get Ethernet cabling into the place. He games, he streams, it's his main source of income. Wi-Fi is everywhere in homes now, and there's still that demand for copper or fiber inside the walls. If the connectivity in the house is poor, it doesn't matter what your pipe to the internet is, which brings me to latency. For years people

Home Networks And Latency Reality

SPEAKER_01 5:55

cared about throughput. How big is the pipe? How fast is the speed test? Most AI sessions are a few K of text going back and forth. That's a latency conversation, not a capacity one. But then someone says, make me a picture of a cat playing a piano and they want it instantly. Now you need the low latency for the exchange and a big enough pipe for the image. It's a combination of the two. Have you ever seen speed tests look great while the applications failed? That's just about any enterprise network. Which is why when I design one, I look at the devices, the people using them, the services that make them work, and the applications running on them. Miss any leg of that, and you've left yourself open to a poor quality network. It started with a fairly selfish idea. I wanted the world to revolve around me. I wanted the tasks I'd normally sit down and make time for to come to me at the moment they needed doing, not a reminder, the actual task knocked out while it's in front of me. The thing that kept coming up was blogging and podcasting. I used to do a lot of it. I got into it because Tom Hollingsworth told me he thought people would value my opinion, and I'd honestly never thought about that before. The old process was ours. Read the news, research what's going on, go through all the social channels,

A Tool That Forces Publishing

SPEAKER_01 7:01

see what my friends were doing, form my own take, write it out or think it out, or scribble notes, then sit in front of a microphone or a camera and go through all of it, and then it stopped because things like that get easier to look past. It's too early in the morning, it's too late at night, nobody's online right now, so nobody will read it. I don't have time to edit. I didn't do my hair, I'm not wearing the right shirt. There's always an excuse. So I built the thing that's producing these words right now. Every day I get a phone call at a set time about a handful of topics. Behind that call is a list of about 500 news feeds running through RSS, a pile of trade magazines, and eight different social channels. The system synthesizes all of it into a few top line stories, and I pick one, then it interviews me, it asks questions, it responds, it behaves like a real interviewer, it records my voice exactly like it's doing right now, then it transcribes it and turns that transcript into an article written in my voice. And here's the line I care about most. When the draft is done, I go read it and I approve it. I approve that it sounds like me, that the words are actually mine from the interview. It doesn't invent sentences for me, it doesn't invent storylines. It might fix a grammatical error here and there, but it uses my voice and my words to build content that's genuinely mine. Why open it up? Because I made it easier for me. So I I can definitely make it easier for you. If you've never blogged, this is a way to start without having to come up with everything from a blank page. The system prompts you, it asks you. It's based on who you are and what you know, and it builds the questions that make the most sense for you. It brings out your voice, not just something that sounds like you, but the best version of you. It's called telepost at telepost.ai, and I'm opening it to a small group by invitation to start. Email me if you'd like to be considered, and if you're not going to, do this instead. Write down the excuse you actually use for not publishing, then work out which part of it could be automated away. That's the whole trick. Three different stories this week, and the same thread runs through all of them. The technology keeps arriving, the question of who can actually reach it keeps getting harder, and it's the question almost nobody's arguing about it. If you take one thing from this episode, go read the satellite piece in full, because the list at the end of it is the part I'd actually act on Monday morning. I'm Drew Lentz. Take care of your neighbors, and I'll talk to you next week.