Rethinking the Internet with Bob Frankston - Episode 640 of the Community Broadband Bits Podcast

Bob Frankston (00:08):
The Internet is so mind bogglingly, simple as a concept that it's hard for people to grasp.

Christopher Mitchell (00:14):
Welcome to another episode of the Community Broadband Bits podcast. I'm Christopher Mitchell. I'm at the Institute for Local Self-Reliance and I'm in St. Paul, Minnesota. Today I'm speaking with Bob Frankston, who I'm going to guess is in [00:00:30] the Boston vicinity.

Bob Frankston (00:32):
Yes I am.

Christopher Mitchell (00:32):
And you can find Bob's writings @frankston.com. And Bob, you've been someone who's been working on this stuff, we were just talking about it, but for quite some time. One of the things you've done that I've always thought was the coolest thing is that you enabled home networking. We're going to start there and talk more about your views of Internet access networking, more importantly communications. [00:01:00] And I should also say that we spoke last in 2014, so it's been a minute, but you have been on the show before and I'm excited to get back into it with you.

Bob Frankston (01:09):
You mentioned brings back lots of memories and background because if theme at home networking, this was the mid nineties and broadband, everything we knew and remember these were the cable TV companies. They were used to charging with set up boxes and they wanted to charge for each PC because it was just like [00:01:30] as if they were giving you a motive, each IP address was treated as a dialup line. That was the mindset. Now we have to go back, let's say 60 years to the origins because I was not only interested in computers, it was a lot of areas like psycholinguistics, AI and things and all these concepts of them. Psychology come into play as we'll see. But I remember one class I took, this was in 1973, we studied ALOHAnet. [00:02:00] Now ALOHAnet was a hack, you can call it. It was really a network, really research, it was called when people do something at university called research.

(02:09):
I don't like fancy words like that. So the problem was in Hawaii, how do you interconnect all the computers at high speed without spending a lot of money on these lines and everything? Use radios. The problem is radios are not reliable. So what is possible in that by the 1970s [00:02:30] was departments had work group computers. So instead of having the phone company to take responsibility for reliability, they would just program around it. If the packet was lost, well, you write a sub routine that retries and it very simple algorithm. But what it suddenly meant, and we really didn't appreciate it back then is those radios did not form a telecom network. They formed a network. But the problem is we [00:03:00] talk about a telecommunication network, we're talking about a very physical wires and as I mentioned to you earlier, the word network is so nonspecific. We can both use the word network, I mean entirely different things.

Christopher Mitchell (03:13):
One of the things as you're describing that, what I'm trying to imagine, first of all, I have trouble imagining how one uses radio to communicate because in my mind I didn't grow up taking apart radios. But let's leave that aside for a second. What I think the audience should be familiar with [00:03:30] is that if you're designing a communication system, and I don't want to get lost in this, but you have two choices. One is you can have a system that is sort of very intelligent and in some entity or device is choosing how everything talks when they talk and making decisions like that. And the other is that every device on the network has a simple set of instructions that it can use to try to communicate. And the first system I think is quite brittle because you [00:04:00] have to have such a complex system that can control so many different possibilities. It's bound to fail. But you're talking about what you had was these devices that, and people that had a set of simple instructions and if they follow them, maybe it'll work, maybe it won't at a given time, but if they follow them over and over again, they will be able to communicate.

Bob Frankston (04:18):
It's actually much more subtle and deeper than that. The network is an application, not a physical thing. Number one,

Christopher Mitchell (04:26):
The network is an application, not a physical thing.

Bob Frankston (04:29):
Networking was [00:04:30] a technique. You did networking. There was no physical network. You did networking over anything available.

Christopher Mitchell (04:36):
There are physical things that are involved, but that is not the network.

Bob Frankston (04:39):
You can use facilities. You can send postcards, same thing. You didn't do radios, you can send postcards a little slower, but that'll work in London. In fact, in the early 19 hundreds you get a few mail deliveries a day, you can exchange postcards back and forth and have a conversation. We forget all this nomatic tubes. There are all [00:05:00] sorts of techniques. But the important thing is content of telecommunication network goes back to ities before when you had run is carrying messages. So the big invention was to carry a message reliably using a telegraph. And that was modeled after the post office. In other words, you gave responsibility for delivery the message to the courier. That's the way communication networks worked Until the 1970s. [00:05:30] You had the network take responsibility for delivery. Now the thing about what Aloha showed is you didn't need the network to take responsibility.

(05:44):
There's a fundamental conceptual difference between the ANet and the Internet. Unfortunately we use the term Internet because we really did such a good job. We didn't have to confront the difference in meaning that [00:06:00] the fundamental and paradigmatic difference between the Internet and the ANet is the atric responsibility. The Internet, nobody anywhere takes full responsibility for delivery. They don't even know what they're carrying. So basically anything you want gets encoded as bits with no trace of the meaning. One my essay, these are the keywords, is VoIP inflection. So voice [00:06:30] over IP was inflection point because Claude Shannon digitized analog communication. He emulated the analog communications system. What VOY does is saying took a fundamentally different approach. It turned everything into tokens. It's a fundamentally different concept because everything was encoded meaningless tokens and you didn't even depend on all those tokens. So I pointed out, if you said the word hello, you [00:07:00] might encode the phone names in tokens, but if you missed a few, no big deal, you could program around it.

(07:05):
If you think about it, the one problem with I'm getting very far field, but it's all problem is everything is relevant. There's a lot of background. You would hear a click in traditional things because you had dumb end points. You had two team cans and a string, which is basically what telecommunications is. So microphone speaker and a long wire, all the telecom companies did was preserve [00:07:30] that wire over a distance. They stretch that wire electronically preserve the wave form. But if you lost it, you'd hear a big click. Well, once you have computers at both ends, they knew how to bridge the gap. They can fill in the blanks, they can buffer a little, just a few milliseconds and that's how VoIP works. But they can also, that was the early stage. You could just program around it a little bridging. What is profound is with VoIP is you can now reassemble the packets out of order handle jitter. And the interesting [00:08:00] thing about VoIP, it was not invented as much as discovered that the origin story from vocal tech is they were using this nice algorithm with a little buffering inside their office. They were surprised to see people using it on the open Internet.

(08:16):
And the reason it happened is important to understand it did not happen was anybody anyway did anything to enable it. But it had was a virtuous cycle which increased capacity so that [00:08:30] the odds of it working were high. And this is very fortunate because we actually did this experiment. We built two networks, one where there was no guarantees, VoIP couldn't work and it took 50 years for VoIP to start working. We did another one where we built VoIP bin where you could pay somebody to get first dibs to get priority to get a guaranteed path. It cost a trillion and it would cost you back with inflation $10 a minute to call [00:09:00] across the country, but you got it to work and which one survived? Right now the telecommunications in entire telecommunication industry running ambulation mode, it's, it's a fake. And that goes back to the idea that you needed to preserve the weight for, so lemme go back to that class in 1973 there was a kid sitting next to me in class doing a cute little project. I remember what he was now sitting class, I scooted [00:09:30] back to listen. So instead of using over the air where the FCC ruled, he used the coaxial cable as the ether.

Christopher Mitchell (09:39):
I'm thinking this is a guy named Bob also

Bob Frankston (09:41):
Bob Metcalf. Yes. So I just happened to be there when he first in-house his class project.

Christopher Mitchell (09:48):
That's pretty remarkable. So this is getting us back to home networking and sort of some of the thought that you had, the background [00:10:00] you had, why you thought what you're about to describe was important, but to give people a sense of what you've been talking about here is that we have seen empirical evidence, we've tried it, we've run the experiments to see whether a more centralized complicated system is able to handle certain tasks or a more simple system. And your point is that even my discussion in that way [00:10:30] is missing the point that you can communicate in any number of ways.

Bob Frankston (10:35):
What does it mean to communicate?

Christopher Mitchell (10:37):
Yeah, I mean I sort of feel like you're a professor now and I'm on the spot.

Bob Frankston (10:41):
Yeah, telecommunications. The word telecommunications because 200 years ago sending a letter was the only way to communicate,

Christopher Mitchell (10:49):
Which is about moving information from one point to another point in my mind,

Bob Frankston (10:53):
But not really because that's the freight model. It's really getting somebody to understand what you're saying that [00:11:00] has nothing to do with the wires. And this is where the cycling roof background is a hundred percent outside the facilities. In other words, lemme give you a message. How many bits does it take to go someone?

Christopher Mitchell (11:11):
I have no idea.

Bob Frankston (11:13):
No, that's not the right answer. The answer is that's a stupid question. Okay, and how many bits are in a pause versus a pregnant pause?

Christopher Mitchell (11:21):
Well, it would depend on the encoding scheme.

Bob Frankston (11:23):
No it doesn't. The answer again, that's a stupid question.

(11:28):
The reason why we can even [00:11:30] form that question is Claude Shannon, he defined an abstract model with clock a certain bit right on the wire. So you could say a one second pause. It takes so many bits because that's a clock time. That's nothing to do with the real world, the mathematical models that absolutely nothing to do with how we communicate. And that's the problem with all of these words. They embed a certain theory of what it means to communicate and then we build a system that based on that, [00:12:00] and it turns out according to central control creates scarcity by necessity.

Christopher Mitchell (12:06):
Okay, so if I'm following what you're, you're discussing in your corrections with me are that, don't get frustrated if I'm missing this.

Bob Frankston (12:15):
No, don't worry. I'm very well aware. I mean the point I'm trying to make is

Christopher Mitchell (12:19):
No, I was going to say is what I feel like your key point is that we have systems that were designed by people that don't understand how humans really communicate. [00:12:30] And now the way the system works has influenced the way we think we communicate.

Bob Frankston (12:35):
Actually, I'll give Al Bell more credit than that. He knew how people communicate, but remember he was a teacher of deaf.

Christopher Mitchell (12:42):
Did you know him calling him Al?

Bob Frankston (12:46):
Well, he's a local guy, I thought.

Christopher Mitchell (12:50):
Yeah, he probably

Bob Frankston (12:51):
Is. I've tried to humanize it. Actually I won't get off any Elisha Gray story. That bell maybe stole the patent from Elisha Gray who retired about two miles away [00:13:00] from here too. It's all local. And I think actually there's a little sidebar here. One of the things about being in the middle of all this is that this is real. In other words, I always play with the people know the people, friends of mine who keep doing the Internet, all this and that. In other words, this is not an abstract thing you can't understand. It's all simple stuff. And the point is the Internet is so mind [00:13:30] bogglingly simple as a concept that it's hard for people to grasp.

Christopher Mitchell (13:34):
Yes, I agree. I

Bob Frankston (13:35):
Think there must be more mechanism to that. Which lemme one more story before I get to the whole networking. So Bob Metcalfe Metcalf, his advisors were telecommunication specialists. So we had to prove to them in his thesis that is an efficient use of a coaxial cable, which is utterly absurd. Who cares? And those [00:14:00] of us, I was excited, 2.994 megabits, which is here reminds me was the speed that was so far beyond mono speed. Even if it ran a 10% of that, we'd be thrilled. We were so used to thinking of communications in a certain way and telecommunication in particular that it never entered the head of who cares. And this is why getting a little ahead, I want to create the opportunity for discovery rather than making this building all of these things in just create opportunity. And what's separating [00:14:30] up the bits from the meaning means you now had this fungible connectivity, which is a resource with no promises and discovery table. Now do you know what else is based on this idea of separating meaning from the tokens?

Christopher Mitchell (14:45):
No

Bob Frankston (14:46):
LLMs.

Christopher Mitchell (14:47):
Oh right.

Bob Frankston (14:48):
All of AI is based on meaning not being. So that's another essay I can give you which goes into that abstract thing. But that scares people even worse than Copernicus did. [00:15:00] So I was exposed to ethernet as something you did yourself. You could just run wires around a computer and locally, which is work. Now getting back to your comment about how do they do network build networks? Well again, you broadcast networks, you build them by getting people to work together, cooperate shows any kind of things in network. So the way you scale into the ethernet is you bridge the networks is one way that is learning bridge and that doesn't [00:15:30] scale enough. So then you come up with an Interneting protocol, which basically says any facilities we want we can use to network. And ethernet is a facility we use. The packets on the ethernet are not IP packets, they're ethernet packets.

(15:50):
And that becomes a resource, not a dependency. So again, I can easily get far too much into the details, but jumping ahead [00:16:00] to the home networking, this was the mid nineties, broadband was new, actually A DSL was repurposed 1950s technology, but that's another story. So I realized that within a home or local network, you didn't need all this fancy stuff at that point. Call networking meant home control. People were very confused. You built purpose built system for each purpose. And what I knew from my experience is that all [00:16:30] we need to do was exchange packets and that would work. Remember, there are two things I say remember you might remember, but most of you listeners probably don't. There were a couple of things about that. One at the time is that the idea of people running their own networks was considered impossible. I mean networks, you need real network engineers, rocket scientists, something as if throwing a rocket at the moon was that big. A hard a problem

Christopher Mitchell (16:54):
At this point. Most homes have zero computers in them. Some homes have one [00:17:00] computer in them. And so

Bob Frankston (17:01):
Yes, and when I was at Microsoft trying to explain what I was doing, they didn't understand it. Nobody wants more than one computer at a home.

Christopher Mitchell (17:09):
And so I remember the only time my home had multiple computers was when I tell this story frequently was when friends of mine wanted to play doom on Doom two and you could do four person networking. And so we would take the computers from their parents and bring them over to my house and in the basement we would try to figure out, I learned networking by trying to learn how [00:17:30] to play video games on multiple computers and

Bob Frankston (17:33):
You learned something you could do yourself,

Christopher Mitchell (17:35):
Right? And that was exactly it. You follow instructions on the Internet and it doesn't work and you follow a different set of instructions and that doesn't work. But then you start learning between the two and figuring it out and eventually you find something that works and by you've either forgotten how you got it to work or you've lost interest. And either way you never document it for the next person. That's one of the key things I learned. Yes,

Bob Frankston (17:58):
By the way, it just is a little side [00:18:00] story. I remember I visited Paris from Microsoft back about 96. I remember we went down past versaci about everything, but so were going to show me something wonderful. It was basically people playing video games on the land of the house so much for exotic powers every side.

Christopher Mitchell (18:19):
But this is not a time when many people were imagining homes having multiple computers, let alone tens of devices.

Bob Frankston (18:25):
You might want to do this a multi episode because there's so much to go into here. [00:18:30] Okay, the key thing I'd learned from that class is that if you throw packets on ethernet, eventually it'll get through and that's all they had to do. But the other part of it is my career started, it was on Wall Street actually, but aside from learning about the underbelly of Wall Street, again, I was online in 66, I was creating software for people. So my career is creating tools for others, not giving them solutions, giving them tools and power. [00:19:00] So part of my thing was how do you make it? So one thing about the spreadsheets when Dan and I were doing it was I did not think of it as a program. The program was just the means. We were creating an experience, an illusion, and if you forgot you were using the spreadsheets computer, that was a zen experience. You were getting into the task, the challenge. So one of my ironclad rules for home networking was no installer. [00:19:30] I mean some people would eventually get it. So I try to look at everything that made it hard. Now, by the way, if you want the background story to why I did home networking was my wife's fault.

Christopher Mitchell (19:44):
Okay?

Bob Frankston (19:45):
You had Microsoft people wanting me to move to Seattle. You had my wife who didn't want to move, guess who won?

Christopher Mitchell (19:52):
Well, you're not in Seattle right now. Yes.

Bob Frankston (19:55):
So that was actually combined with A DHD. Fortunate, I probably would've been kicked out of Microsoft [00:20:00] money sooner. So I had multiple computers at home of course. So I networked all my computers at home. But what I discovered was a technique called an Nat network address translation. It allowed me to share single connection to the world so I can make my home part of the Internet, not just dial into it, but become part of the Internet. And that's the real philosophical shift because back in that point you would dial in periodically, get online, [00:20:30] then get off. We wouldn't just have messages pop up saying there is mail and all. This is why AOL's message is such a big deal back then. So I went into his attitude of I want to interconnect people with the Internet. It was a time when broadband was becoming available. Now watch broadband, there's a whole history of the word, but basically it means fat pipe,

Christopher Mitchell (20:54):
Right? Always on fat pipe,

Bob Frankston (20:55):
Always on fat pipe. But don't worry pretty little ahead, but don't worry [00:21:00] about it. You get a phone call, the phone company will use it to give you services or the cable company you want know the phone line, they'll key in a message and now hit a phone line. This is VoIP except you were paying 30 bucks a month instead of zero bucks a month for each VoIP line. You want to basically do home shopping. They would do the home shopping for you. So built, so broadband was basically a way for the phone company to deliver all the services. And that's why at t spent an exorbitant amount for media. One [00:21:30] was they thought they'd monetize all the services by putting the gnat there. Suddenly that pipe was turned around. All they saw meaningless packets that they could not monetize

Christopher Mitchell (21:44):
Because they didn't know what service it was. They had

Bob Frankston (21:47):
The service didn't necessarily exist. Yes, in the days when I still multiple phone lines for other purposes, they were opposed. There was a dial tone on the line. It turns out it was coming from inside my house as a build goes because [00:22:00] I always had to voice a device.

Christopher Mitchell (22:03):
But they were the mindset of you think fax machines and other things where you're paying for a line per device and so they can,

Bob Frankston (22:12):
But it's full price for each line, for each device.

Christopher Mitchell (22:15):
And so you're not allowed the homeowner to then start putting different devices in and didn't need permission from the network owner didn't have to pay an extra fee to do it. And then they were [00:22:30] actually in charge of their own network. And this is something that we talk about when I am doing the tribal broadband bootcamps and other things is that we feel like networking is super hard and yet almost everyone has built their network in their house themselves.

Bob Frankston (22:45):
In 1995 when I was trying to say networking is easy, even a child can do it. Or in those days only a child can do it, but Seattle. But yeah, that was my body symbol because all we had was a packet. Same for home control. You [00:23:00] didn't need to build a special network for home control though that took a while. So the way it sold this idea at Microsoft was sharing the web that two people at home share the web. Now as I said, it was a hard sell because why would you have two computers at home those days? That's what marketing was telling me. Luckily because of iCal, I was able to get away with doing it Anyway. So that was basically the value proposition. You can share the web connection. [00:23:30] And if you think about it, when I take credit for home networking, say it was so simple all I didn't really do much. I just took an off the shelf net put into Windows. I designed an order configuration. But once the HCP came out, it just went with the HCP, which wasn't as good as what it was doing.

Christopher Mitchell (23:47):
And let's pause for a second because I do think an alternate history in which people have to pay per device and these companies are trying to figure out how to maximize their revenues, they're probably [00:24:00] thinking that they're going to be charging a lot per device. And then that would lead to a situation where people are less likely to have home devices.

Bob Frankston (24:09):
Lemme make it more extreme than that

(24:12):
In those days, the idea of ip, remember IP was considered complicated. People said you couldn't put IP on a small device. Of course, absurd. That's absurd because there's no protocol simple in ip. So that was part of the mindset that you do everything for each purpose. So we had all these, but the other thing is [00:24:30] we wanted to build smarts into the protocols. But the key thing, again, we go with main directions, but I want to focus on one aspect. So there are two reasons I take credit for home networking. One is imagine if Microsoft didn't build into windows. Yes, you can have hackers do it, some people would do it, but none of that succeeds. So being in the position to put it into windows and took a while, I had to sneak it in your hotspot in Windows.

Christopher Mitchell (25:00):
[00:25:00] Yes.

Bob Frankston (25:00):
Now that's one of the side byproducts of putting an that into windows. So a bay is a very simple thing. The other thing I tried to do was no new wire networking. I was overkill because the idea was you could use phone wire to do networking, messy phone wires and not interfere with the dial tone that might've caught on. But fortunately, Wi-Fi came in in time though if you remember, it took years for Wi-Fi. But important point is if I try to explain to people the future of all these IP devices [00:25:30] that wouldn't work, they wouldn't have understand. And if anything, I think that's far greater value than simply the web access. And the web by the way, is a great story of the virtuous cycle playing out. Tim Burners league didn't create a single solution. He created the ability for people to create this solution. Just like people think spreadsheets or financial tools, no, they're just scraps of paper and packets are just fungible packets. So because broadband was there, I was able to repurpose [00:26:00] the broadband facilities. Unfortunately, broadband has turned toxic in the following sense. The story I sometimes tell is think of a bicycle trail. How do you build a bicycle trail? Well, you find an abandoned railroad track and you repurpose it,

Christopher Mitchell (26:18):
Right? Because you want something that's relatively flat that goes a long way without interruptions and turns. And it's hard to do that in most areas because people own the land and you need something that has right of way.

Bob Frankston (26:30):
[00:26:30] But what if people take away in order to build more trails, you need to have more tracks. So they start laying down more tracks just to repurpose them.

Christopher Mitchell (26:40):
This is something that we think of as a path, which I think is a fancy term for what you're describing, right?

Bob Frankston (26:45):
Exactly. So people think they need broadband. The problem is broadband is the anti Internet. We were able to repurpose the pipe, just like the Internet repurpose. The entire telecommunication network is just one segment, one hop. [00:27:00] Fortunately I saw it before we had dial mean. The whole dial up Internet is a story that happened because universities shared among themselves for free. But then we had bulletin board systems that stuck backends into the universities to get on the larger Internet. It was not an infrastructure, so you had to find some way to get to it, which is double down on the old model. So the reason we're talking about accessing the intent is that's the way it dialed up. You dial up modem into a [00:27:30] OL and then stay on all day and then backend into the Internet. Now, by the way, little sidebar, the reason that the US led here is because we didn't trust the phone companies. You have to trust the phone companies. So there was no unmeasured service. We had unmeasured service. The US was just cheaper than detailed billing that led us to talk about access. And basically the Internet was treated as just another long, oh, you listen, won't remember the days when you were up for long distance carrier and a local carrier separately.

Christopher Mitchell (28:00):
[00:28:00] I do remember that as well. And then the abuses that also came from all kinds of new names for things where people would switch their long distance carrier and to one that charged them more. And

Bob Frankston (28:13):
Well that's like how many people call you every day to sell you new energy plans?

Christopher Mitchell (28:18):
I don't live in one of those states fortunately. So it's not my issue. But some of the listeners do. I'm sure,

Bob Frankston (28:25):
Yeah, there's scammers everywhere. But when you sign up at DSL, [00:28:30] it was DSLs from phone company. You would sign up for a separate Internet company and a long distance company when you signed up for your DSL line.

Christopher Mitchell (28:41):
But before that, I think what I would say is kind of the glory days, there was multiple companies in every metro that had modem banks and we could use the system to make that phone call connect to who we wanted to. It was as DSL became more important as we wanted faster connections [00:29:00] that a series of government decisions allowed the telephone company sort of the monopoly of the area effectively to kind of take everything back over. And that's I think what you're heading toward is that we made a series of decisions and we believe now that our only options for communicating over the Internet is to go through broadband from one of these large companies that built a network for other purposes decades ago.

Bob Frankston (29:29):
Yes. [00:29:30] So there are lots of different directions. But let me follow this one in particular. What is the proper model for broadband now? So you know the origin story of Comcast?

Christopher Mitchell (29:42):
No, I know there's a small company that bought a ton of other companies,

Bob Frankston (29:46):
But you know why a group of Haber dashes bought Comcast?

Christopher Mitchell (29:50):
No, I do not know.

Bob Frankston (29:52):
This is their own story. I'm not making this up. It's direct from had they were Haass Bel pants came out, [00:30:00] they thought,

Christopher Mitchell (30:01):
I barely know what a habashi is. I don't know what those pants are.

Bob Frankston (30:04):
Oh, they sell ties and bell suspenders. Okay, okay. Well Truman was a habashi for a while. You have to know these words like Arian and things. Of course my puzzles help. So they were spooked. So they bought this company that basically did remote antenna and then years later they discovered they could do other things, which they had a whole broadband

(30:28):
All this capacity. [00:30:30] They were not limited by over the years. So it took a while to discover that they eventually got T-B-S-C-N-N Tech turner, all these evolutions along that path, but they were still building on. They own the facilities model and the facilities are there for this one purpose now world forward. And now the fungible packets, we've totally split those services no longer in the wire. So if you think of it, there was days when [00:31:00] trolleys would offer you a ride, but once you discovered you can own a car and do it yourself, the business model of roads is very different than the business model of railroads. They're fundamentally different concepts. Now, unfortunately, financial I say they're all transportation. No they're not. Roads in McAdam pavement that you can tear up trade and sell you rides. And there's a good book by Robin White called Railroaded wonderful. He's a great writer, wonderful history of railroads, a [00:31:30] creation of the ICC. So both railroads and the phone companies provided a service. One was rides, the other was carrying your messages freight. So they're very eerily similar. And we're talking about the long distance railroads in the west, which offered those facilities. But the problem they had was first the facilities were cost of doing business.

Christopher Mitchell (31:55):
Right? I was just going to say, this is where I read that book and a few others [00:32:00] about them. And my takeaway, which I wish I had known more when I picked it up, because when you're new to something, it's harder to figure out what the important bits are. But my takeaway was that the railroads were built by people who were seeking to use financial tricks to get rich and they happened to have to build railroads as a part of their more or less,

Bob Frankston (32:22):
I'm not quite that skeptical. Even if they wanted to do well, there was no way that when you have [00:32:30] high capital costs and no differentiation, the business work.

Christopher Mitchell (32:33):
Yes.

Bob Frankston (32:34):
So the ICC was a solution to that.

Christopher Mitchell (32:36):
Yes. Yeah, the Interstate Commerce Commission

Bob Frankston (32:38):
Right now, what other business had high capital costs and no differentiation from

Christopher Mitchell (32:44):
Companies

Bob Frankston (32:46):
And you know where the FCC came from?

Christopher Mitchell (32:47):
From the ICC, but long after we already, the government struck the deal with veil in order to create

Bob Frankston (32:54):
The yes. But the point is that both [00:33:00] markets were broken and at the best idea we had was regulation to fix it. And that's people demonize regulations. They forget governments the way we work together, corporation creation, government. So these are all attempts to make things work. Unlike we won't discuss the brain dead people now. So the ICC went away, but railroads offered a better solution to competition. No longer had a choke point. They had department of transportation. Now people, game transportation, all that. [00:33:30] But that's another topic. So the best thing we can do in 1934 was to have a managed marketplace to assure communication. This is where at and t became like the post office Wasi governmental company largely funded for research because they were guaranteed a return on their investment and they were also legitimized by the idea that you needed to preserve the waveform in order just to carry speech. So when you didn't [00:34:00] have a wire, we created a virtual wire out of whole cloth the spectrum allocation, which is one of the worst algorithms ever

Christopher Mitchell (34:07):
Because we're not using the spectrum very efficiently.

Bob Frankston (34:09):
Well, it's worse than that. It creates dead zones. I mean efficiency is a measure. Let's say we took the abundant commons and split it up as if we put a to separate narrow lanes. I mean it turns out now we'll get to a way to work around that solution. So spectrum allocation basically that there are only so many frequencies [00:34:30] it it's just like when Newton said there are only seven colors, which meant you have only seven people in a room because they eat. How else would you distinguish between people except use the search color. So it was the same kind of thing. So the solution to this broken marketplace was basically licensing. So we created all the phone companies and all this. What happened, the reason why low high net was it showed us if we turn all into packets first, it shows a technique.

(34:57):
Again, when you're in the middle of this, you don't see the full implications [00:35:00] of it. It took a while for the virtuous cycle to play out. The virtuous cycle of the web means each thing we did create a value, therefore we created more capacity of the web. But then we discovered what else we could do with the capacity. I think that's much more important than Moore's Law. It's actually a variation on the same thing of sort of a market driven abundance. If we take the entire tele network and convert it into packets, we get abundance. You can use all the capacity it becomes pull it all together, it suddenly all these wires, you take a look at the, remember the days phone calls had these big bundle. [00:35:30] Do you know the effect of speed of a copper wire for phone call?

Christopher Mitchell (35:34):
No, I do not.

Bob Frankston (35:35):
It's about one bit a second.

Christopher Mitchell (35:36):
Okay,

Bob Frankston (35:36):
It was 99% of the time it wasn't in use, but you had a dedicated wire to each purpose. So on average it was like one bit per second,

Christopher Mitchell (35:44):
Right? Because so you have periods where you're getting a lot of data across it, but then you have whole hours where nothing goes across

Bob Frankston (35:50):
It. We weren't even getting a lot by today's standards, but you're getting some. You had a voice call 64, 56 kilobits, reversal, the koac, you get abundance by saying we're going to treat this. [00:36:00] All the bits the packers can use any available thing. You bundled the water, you get a match. In fact, I proposed basically turning line cards and central office to be DSL line cards, all DSL, which would've been the same cost, but by the 1990s the electronics was so much you didn't need somebody going out tuning everything. You can just use it. It's not about speed. It would've been made one bag of it, which was high speed then. But the point is it would've been created opportunity that would've rapidly driven the virtuous [00:36:30] cycle for people would then more to invest and own the facilities. So by normalizing all the spectrum to packets as we did, the wires starting to get abundance.

(36:39):
So here's the problem. So go back to broadband. That broadband copper wire is no longer a service pipe. It's no longer a cost of doing business because the service is a hundred percent decoupled from the use. So in that case, the better model is sidewalks and roads not [00:37:00] a billable service. So the economics of broadband is fundamentally broken. Broadband creates scarcity because the business model of selling service no longer makes sense. Those are apps why you're paying a service fee to use that wire. The only sane model for that is you and your neighbors own the copper. Just like you own your driveway and the community owns the sidewalks. And that's very hard for people to get to wrap the minds around, which is why [00:37:30] I want to understand it. But I want to find people who see the business opportunity and shifting the model. And right now I've got three broadbands past my house plus three cellular carriers all whom can offer. I've got, I'm experimenting with the so-called home Internet wireless as well as wired. So basically I've got six broadbands around me. How many do you need? How many electric grids do you have? One. And the importance of having one [00:38:00] electric grid is that's what enables competition in energy sources. Otherwise you have to barrier entry. You need to build your own own facilities.

Christopher Mitchell (38:07):
But let me, as we're about to wrap up a note of skepticism. If we had been doing for the past 20 years as you described, don't you think that we could have ended up in an area where we were more brittle because there wasn't a business model to build out duplicative

Bob Frankston (38:25):
Infrastructure. It's a business model,

Christopher Mitchell (38:27):
But is there a business model to have both fiber [00:38:30] and wireless, I mean not fiber, but to have both a wire and a wireless service

Bob Frankston (38:34):
Available? Well first, who cares whether it's why or wireless? If wireless works use that. If we need fiber to use that,

Christopher Mitchell (38:41):
Lemme ask. So the question is though, is there is a business model. If we say, well, we're going to be using all the infrastructures here, why would we go to the trouble of building out a new, who would be going to the trouble of building out a new, I guess

Bob Frankston (39:00):
[00:39:00] I've got an answer. No, that's trivially easy to answer. What's the business model of a road?

Christopher Mitchell (39:06):
I mean, well roads typically built by local governments is to enable others to use them to travel on, to move goods

Bob Frankston (39:14):
On and how much does the road cost,

Christopher Mitchell (39:16):
Right? But no, here's the thing. The thing is, is that what I like about our system, imperfect as it is, is that when there's a problem with the wire line system, there's a wire list system I can jump on. And [00:39:30] the thing is, nobody has a second driveway because they're like they have the one driveway.

Bob Frankston (39:34):
No, you're missing the point first. You're making the case against broadband. Why do I confined to a pipe at all? Why don't they just mix wide wireless without, why aren't they all just one fungible thing? One resilience. The current business model doesn't do that because it requires preventing free riders. It's as if you put a turnstile between your house and the sidewalk.

Christopher Mitchell (39:58):
Yes. No, [00:40:00] I get that and I think it's really sharp. The question is, your system recognizes that a free rider is actually adding value and we have a lot of benefits when everyone can use it.

Bob Frankston (40:10):
It's a mindset issue. It's just like the days of medicine was bloodletting. It was very hard to get rid of. So lemme answer, I'll give you the answer to, okay, so we put a million dollars a mile into a road

Christopher Mitchell (40:22):
More than that, I would assume much more than that.

Bob Frankston (40:26):
It's a thousand dollars a mile or $10,000 a mile for [00:40:30] fiber. Okay, $0 a mile for wireless. So what idiot is now working very hard to prevent us from using the wired and wireless facilities even though the cost benefit is far larger than that road. Well you can argue, but the point is the business model is core infrastructure and economics. The same business model is road. Same business model sidewalks. Except [00:41:00] I argue that if you went to an infrastructure model, your cost would go down by a hundred percent. Now let's say we have, I've got say six. Broadband simply could sell a common infrastructure should reduce the cost by youth margin. But we also pay for separate infrastructure, police, separate infras sector. If we combine all these infrastructures, basically the cost of the broadband part that you see, it becomes a trivia part of the mix.

(41:27):
So a stable capitalistic business [00:41:30] model says we've decoupled from the wire. The only business model of core infrastructure is core infrastructure and that's all the people who put the copper in support. All those people have businesses and get hired by the city. Just like the city hires people and people are, the city can't maintain it. Well it's so trivially easy that even a child can do it compared with water work, sewers and roads, those are hard. So it's too trivially easy and when you have resilience, so why it goes down [00:42:00] when a Comcast connection goes down, they roll a truck to my house because they don't have the protocols they should call me to tell me when Verizon, my Verizon connection, I experimented with number of them had a problem first I couldn't convince them because they pinged the most people was ping, they pinged them. I think it got through, there was something wrong internal, they couldn't detect it. But they did offer me a previous premium plan in which they would come in and fuck up my TV because for them the TV is a value. They fundamentally. Now the big advantage Comcast has here is [00:42:30] Comcast is out of the business because Comcast, the Haass were always in the make money business. They had no problem becoming NBC. Verizon is proud of the networking expertise. So they got themselves into hole. When at t Verizon tried to get out of it by buying media companies, it did not end well. Let's put it that way.

(42:52):
So there's a very simple model, I call it public packet infrastructure, which you start locally. And the key thing though [00:43:00] is I realize I can't explain that to all the city municipal people. So what I'm trying to do is find entrepreneurs who want to start a business because it's so easy to disrupt telecom at this point. All it takes is to go to cities, offer them a deal here, we'll cover this for you, but remember telecommunications business model is a loan you can never pay off.

Christopher Mitchell (43:22):
We did cover that in 2014 as well. I was just reacquainting myself with, I'm

Bob Frankston (43:26):
Glad you remember. The problem is, I keep repeating some is that my story evolves [00:43:30] and I get better at it. But yes, if I look back at some of the things I was trying to explain way back then, but the problem is people don't understand finance.

Christopher Mitchell (43:39):
No, I agree. And it's what drives me nuts. I mean we're in a situation in which we pay far more than most people who actually, and they haven't even embraced your system, Bob, they just have a better system than we do in terms of their,

Bob Frankston (43:54):
Yeah, they're the wholesale retail residents. But in a sense it's also the stuck there was, I think it's so perfect [00:44:00] when really they're using a railroad business model for roads,

Christopher Mitchell (44:04):
Right? But we're stuck in this place and I am frustrated. I feel like it is extremely hard to get people to understand.

Bob Frankston (44:11):
In a sense I've given up and I'm trying to reach out to entrepreneurs and people interested in doing it as a business because I think trying to explain to people as I've discovered over the last three years is funeral and this is where to form a business. You were, except with empathy to basically [00:44:30] go to cities and now other companies are put in dark fiber now, but they're still stuck in the old model. We'll put in dark fiber and then other people rent it. No, that's the wrong model. It's like go to city. Oh, we're going to put the streets in here, but we're going to put a toll booth in front of every house and create the concept of unaffordable stroll.

(44:47):
What you need is something more like the companies that do sewers or the people collect garbage for you that go to the city say we can do a better deal than you can now. So in a sense, just here has the simplest model. [00:45:00] Say you still got Comcast or franchise, you just go to single payer model. A single payer model would mean you can then assume it's infrastructure. You do connect to healthcare. And I think the important part reason, earlier I mentioned web was a driver, but the real value of the device took a long time for people to understand how to build IP devices. Matter of fact, very recently that I can start assuming I like companies like home assistant stuff that help with that process. So initially people will just see cheaper Internet, it's free broadband, [00:45:30] it's wireless about the city carry my devices. Now I can take, let's say a glucose monitor and it'll just work everywhere. Instead of having I having to phone a big bucket brigade. I mean you try to explain to somebody how to connect the pill bottle to the Bluetooth and everything. So instead, if we could assume open connectivity throughout the city, then you can start to with connected healthcare. But it's going to take a while for people to realize the value and see it because it'll land of the blind. The one-eyed man is insane. He sees things out there.

Christopher Mitchell (45:59):
Alright, [00:46:00] we, there's many other threads that I would like to pull and I know that you're begging to pull as well, but we're going to wrap it there. Thank you Bob for sharing these thoughts. Now, I hope that there are people out there listening who are interested in trying to figure out how to make this model something that they can touch and feel.

Bob Frankston (46:22):
Yes. I hope I didn't scare too many people. Let scare them. It's okay.

Christopher Mitchell (46:28):
Yeah.

Ry Marcattilio (46:29):
As long as they come back.

Bob Frankston (46:30):
[00:46:30] Yes,

Ry Marcattilio (46:31):
We have transcripts for this and other podcasts available at communitynets.org/broadbandbits. Email [email protected] with your ideas for the show, follow Chris on Blue Sky. His handle is at Sport Shot. Chris. Follow community nets.org stories on Blue Sky. The handle is at Community Nets. Subscribe to this and other podcasts from ILSR, including Building Local Power, local Energy Rules, [00:47:00] and the Composting for Community Podcast. You can access them anywhere you get your podcasts. You can catch the latest important research from all of our initiatives if you subscribe to our monthly [email protected]. While you're there, please take a moment to donate your support in any amount. Keeps us going. Thank you to Arnie Sby for the song Warm Duck Shuffle. Licensed through Creative Commons.