Podcast
Join Viktor, a proud nerd and seasoned entrepreneur, whose academic journey at Santa Clara University in Silicon Valley sparked a career marked by innovation and foresight. From his college days, Viktor embarked on an entrepreneurial path, beginning with YippieMove, a groundbreaking email migration service, and continuing with a series of bootstrapped ventures.
Links
Follow Me
Follow Me
Join Viktor, a proud nerd and seasoned entrepreneur, whose academic journey at Santa Clara University in Silicon Valley sparked a career marked by innovation and foresight. From his college days, Viktor embarked on an entrepreneurial path, beginning with YippieMove, a groundbreaking email migration service, and continuing with a series of bootstrapped ventures.
A Global Mission to Connect: Unveiling Giga's Journey with Chris Fabian
Play On
25 FEB • 2024
1 hour 10 mins
Share:
In this episode, I’m joined by Chris Fabian, co-founder of Giga, to explore the ambitious goal of connecting every school in the world to the internet. Chris’s journey from setting up ISPs in East Africa to driving innovation at UNICEF offers unique insights into the challenges of global connectivity.
We start with Giga’s innovative approach to mapping schools using open-source machine learning and satellite imagery. What particularly caught my attention was how this method has revealed significant gaps in government data, with Giga having mapped an impressive 2.1 million schools so far. Chris’s explanation of their data collection methods, combining government data with software probes in schools, shows the complexity of tracking real connectivity.
The conversation gets especially interesting when we explore the economic and educational impact of school connectivity. Chris shares compelling research showing how increased school connectivity boosts both years of schooling and GDP per capita. His vision of schools as nodes in a decentralized network, supporting both education and financial inclusion through blockchain technology, reveals the broader potential of this initiative.
I was particularly intrigued by our discussion of the technical challenges in connectivity. We explore the roles of different technologies - from fiber to radio and satellites - and the regulatory hurdles they face. Chris’s critique of initiatives like Facebook’s Internet.org raises important questions about the difference between true internet access and walled gardens.
If you’re interested in global development, education technology, or the future of connectivity, you’ll find plenty of practical insights here. Chris brings both technical expertise and social impact experience to the discussion, making complex connectivity challenges accessible while maintaining their technical and social depth.
Transcript
Show/Hide Transcript
Joining me today on Nerding out with Victor is my good friend Chris Fabian.
One of his friendship goes back to memorable encounter at challenges in Barcelona, or rather at a beachside bar afterwards.
Chris's remarkable career, particular at unicef, has been about technology for global good, focusing on open source solutions to complex problems.
Recently he has taken his vision further as a co founder of giga.
So without further ado, welcome to the show, Chris.
Hey, Victor.
I wish were on a beach in Barcelona right now.
That sounds a lot more compelling to me than a cold England or Switzerland.
Switzerland indeed.
Cool.
So maybe we should just start with what's giga?
Give listeners some kind of overview of what it is and why it matters.
Sure.
Well, so, I mean, I think as a lot of us, I started using the Internet at a very early age.
I love what it gave me.
I think recognizing that there's a lot of stuff that's wrong with the way the Internet's built right now, it still has been a very powerful tool.
And one of the things that I've always been interested in is how other people can have access to that same opportunity.
There are about 2 billion people who aren't connected to the Internet right now in any reasonable sense of the word.
And GIGA is a project that we've started about four years ago to try to make sure every school in the world has Internet connectivity.
And the basic idea is that if the school's connected, the community's connected.
But I mean, Victor, you wouldn't.
Oh, I didn't believe when we started it, nobody knows how many schools there are in the world, but for sure nobody knows how many schools are disconnected.
And we think it's more than half by a large margin, probably around 70% of schools in the world.
That's crazy.
So to take me back until about 2019 or so when it all started, you were back at the time you were at Innovation Fund, correct?
Maybe you start there a little bit and kind of use a stepping stone off.
I guess so.
I mean, even before that I, like in my earlier career, I set up a bunch of companies.
One of them was a web portal and sort of Internet service provider in East Africa.
And I saw when I was doing that on my own how powerful it was to provide Internet to people who Maybe the normal ISPs didn't think had money.
There wasn't revenue there.
And so we did a lot of interesting things that grew users in a very organic way.
Fast forward a bit.
I was at unicef, you know, for my sins, working for this Organization that is in 190 countries around the world.
I mean, UNICEF is huge.
It's the world's leading organization for children.
I didn't know very much about it when I started because I was coming from tech and kind of building open source products.
But it's got a really large footprint.
And one of the things that UNICEF does is has access in every country to the kind of parts of government that work with kids, ministries of education, ministries of health and so on.
And so in 2017, 1617, we had started at UNICEF, the first crypto fund in the public sector.
So it was entirely crypto denominated BTC and eth.
And right before that, about three years before that, we started the first venture capital like fund.
It was a grant capital fund.
So it did just do grants, it didn't do investments, but it mimicked a lot of the VC due diligence, the kind of term sheets and the way that we worked with our companies.
And so those two things, the crypto fund and the venture like fund, had really led us to see that there was an opportunity to find these exciting companies in emerging markets that were doing technology that could help people, whether these were drones in Malawi that were doing surveys of fields, or whether they were AI based chatbots in Brazil that we could then port into other languages and other places.
But all of these companies stopped growing where the Internet stopped, which makes total sense because they're digital companies.
So like no digital company.
And so we, you know, I started going back to the thing that I'd started with which is like, how do we get more Internet to more places?
And that's really where Giga began, was with the desire to help our portfolio companies grow.
And at the same time the crypto stuff was really interesting to me because were able to move money, like do our capital calls from investors and get that money to investments in like seven minutes.
So you could go from an LP for our fund out to a company in Mexico in seven and a half minutes.
But again, if that company wasn't well connected, if they didn't have good bandwidth, that digital transaction, which I think is so powerful, wasn't going anywhere.
Yeah, so that was like, that was, you know, that was the problem at the beginning of the journey basically.
Yeah.
And I remember I visited you in New York back then, see the Innovation fund and it really felt like an incubator in the sense that you would like instablank VC incubator, but like with the framing of doing good, I guess, in open source.
So that was super interesting.
Cool.
So you started giga.
We covered Giga and done some digging here.
And then essentially it comes around large, what you're mentioning around schools.
Right.
So to talk me through kind of the mapping process, because it's as correct as pointed out, like, never been something I even thought about, like the dark number really, in how many schools are unconnected.
And how did you go about even finding these schools that were not connected in any way?
And it's such a random data point to even think about.
So there were like three really important people who influenced the beginning of giga.
And so we had this intuition, right, that if we could extend Internet to places in emerging markets that it wasn't, we would be able to help companies grow faster.
And that's kind of like bubbling around in our heads.
There were three people who said things to me that were much smarter than anything I was thinking.
One of them was my mentor at the time he passed away this last year.
About a year ago.
Exactly.
Named Sharad Sapra.
Sharad was a brilliant man.
He's the reason that I joined unicef.
And at some point during this venture fund, in the crypto fund work, he said, chris, you know, why don't you use your tech skills and start doing something useful rather than all of this vc, blah, why don't you actually figure out like the answer to a real problem?
And I was like, okay, Sharad, you know, as my mentor who says things like that to me and inspires me, what's the real problem?
And he's like, you know, you idiots don't even know how many schools there are in the world.
And I was like, well, what does that have to do with anything?
And he's like, well, maybe it doesn't, but isn't that weird that you don't know that?
Very simple.
Like, wouldn't you think you should know that?
And so he sort of said that.
And I was, that was bubbling in my head.
And he said, you know, you think you're so smart and you're counting rivets on bridges with some of the drone technology that you've been investing in, and you can figure out whether a bridge needs to be repaired, but you can't find the schools.
And so I started looking into that a little bit and I was like, that's.
It's a fair point.
You know, government says they have a bunch of schools in their country, but do they really?
Do we?
Is it really true?
And has anybody aggregated that?
The answer is no.
So I was kind of there.
And then another friend named Greg Weiler.
So Greg is a satellite guy.
He founded OneWeb and O3B and has a passion for all things space and satellites and connectivity.
And Greg said, as were going through this, I was like, I'm gonna figure out where all the schools are, because that can't be very hard.
Really?
What does that guy Sharad know?
So then said to talk to Greg, and Greg's like, wow, wouldn't it be great if you had these if you knew where the schools are?
Which you should know, right, Chris?
Like, obviously, everybody knows where the schools are.
Wouldn't it be cool to have, like, a dashboard where you could see how connected the schools were in each country?
And I was like, interesting.
I guess you'd need to know where the schools are to do that, but that would be good.
And he said, you could have, like, a little green dot if it's connected and red dot if it's not.
And then the third person was a woman named Doreen Bogdan Martin, who's now the head of itu, the telecommunications agency for the union.
ITU is the International Telecommunications Union, and it's the, like, the oldest UN agency.
They were built to regulate the telegraph originally.
They're now doing, like, 6G policy, so go figure.
And Doreen said to me that, you know, wouldn't it be interesting if we could combine the work that UNICEF has on schools and kids with ITU's regulatory power?
And she wasn't the head of it at that point, but, you know, couldn't we bring something together?
So that was the sort of.
Those were the three points of real inspiration.
A basic question that sounds so easy to ask, like, where all this goes in the world?
That's super hard to answer, right?
A kind of an ability to tie that to technology that I care about, which is QoS, like monitoring and kind of backhaul and traffic shaping, monitoring, and then the ability to connect to a regulatory agency that could do things that are interesting that are way outside of my scope and knowledge?
So that's how it started.
That was like 2017, 18 and then.
And people were kind of like, why do you need to connect schools anyway?
And then Covid happened, and that question was no longer ever asked, but still were stuck with this question of, like, where are all the schools?
So you want to hear what we did?
I do.
One of the companies we had invested in put money into grants through this venture fund, quasi venture fund at unicef, was able to use satellite imagery to identify Various features and to do feature extraction of things like swamps or wetlands or even small puddles.
And were using that in some countries to look at where malarial mosquitoes might be putting, laying their eggs.
And so that used imagery from Planet and from Maxar, that sort of like 30 centimeter.
So pretty high resolution at the time, 2017, 95% cloudless.
So pretty like fine grained stuff to figure out stuff on the ground where it is.
And were also doing some work with drones and UAVs to have them do.
Not vertical imagery, but kind of diagonal imagery to rectify satellite imagery.
And so we thought, could we use that satellite imagery to find schools?
Because that should be pretty easy.
It turns out that schools actually look different in different parts of the world, but in contiguous parts of the world, they all look the same.
So schools in Central Asia, there are only seven types of schools.
They were all built on the kind of Soviet imprint and architecture.
And if you can identify schools in Kazakhstan, you can do it in Uzbekistan, which is a machine learning problem and a training problem.
So we built a bunch of ML open source ML with some partners like Development Seed and UCLA and some other kind of partners, academic partners.
And we had this pipeline of satellite imagery and we started to find schools and train basic open source ML defined schools.
And we had kind of 95% true positives, which is okay.
Now we're up to 99.5%.
And that was really neat and it showed us that were totally wrong.
When I said to Sharad, like, of course everybody knows where the schools are, we didn't.
So when we did the mapping in Columbia, we took this satellite imagery and we found out of 43,000 schools that we have now in the database.
And everybody can go.
If you want to go to ProjectConnect World, you can go and see this live.
So go to ProjectConnect World, you can click on Columbia and you'll see what I'm talking about.
We found out of 43,000 schools that 6,000 of them that we'd identified from satellite imagery didn't exist at the national registrar.
So like the government was working without a full deck of cards.
And it's not that nobody knew where that school was like somebody does in the district or the town, but it's not necessarily at the capital city.
And we keep finding this in every country where we do mapping that.
There's a disconnect between the government data about this very central building that is pretty important to society ostensibly, and what's actually There.
And so even that first piece of work has been incredible.
We've mapped about 2.1 million schools and we still don't know what the denominator is.
Right.
We don't know how, we don't know how many schools there are in the world.
And we've been doing this for four years, which is crazy.
Sorry.
So that's like the very beginning of the story.
That's like there's a lot more, but that's where it started.
That's super cool.
So, so where, how did you end up mapping these schools then?
Because if there are no common denominators, like you said, like you can't look at building types.
I mean some countries you could, but in general, like, oh, you look at like stream of people like over time or how are you actually mapping this?
Right.
I think the technical term is it's AI, magic blah.
But I'll give you like a slightly more technical view.
We do a bunch of things.
So one of the reasons that we're at UNICEF is because UNICEF has access to this massive labor force.
We have people in every country in the world which, if you know, machine learning, the biggest problem is always annotation and training data and training sets, right?
Like that's the tough thing to get.
So if you're going to map the school in Sudan, which we did before the war had started there, you need some pretty good ground truth data to train your agents on.
We are able to deploy people using UNICEF's network so we can subcontract and we can say to people, go on a motorbike, go and find the school and either go to find, give us the centroid of the school or give us the four corner points of the school or whatever, just go and do a GPS notation, send that to us.
So we pay for that.
And we would usually do, you know, 5 to 10% of what we government.
Government will give us a number.
They'll say we have 20,000 schools.
So we'll try to get 500 to 2,000 schools or something like that in different parts of the country.
So scattered all over because the ones near an urban area look different from the ones far away.
You get that feed into the ML, train it, and pretty quickly you're able to identify what schools look like.
Now there's other stuff, there's other signal that you can add on top of that to fortify it.
We don't really know, and this is why I said it's sort of a magic blah.
You don't really know what The ML is picking up on.
We believe this is our best guess from it.
We believe it's things like there's a football pitch.
Most schools have some playing ground, some area.
As you said, most schools have a queue of kids walking there in the morning, walking back in the afternoon.
So depending on what time of day your satellite imagery is, you know, from, you can see that there are certain building shapes, as I said, like in Central Asia, that just represent a school, often a school or a prison, which is a weird thing about the education system, but, you know.
Well, I mean, because you need some.
I mean, I guess the building structure has.
Does have some constraint, right, because you need to house a lot of people in.
You got to keep them inside and make sure they get out.
And, and so things like that are the.
All the tells that go into, feed into it.
We also then work with partners who do some manual annotation as well.
We've built some games for crowdsourcing.
So you may be familiar.
About 10 years ago, there was a malaria spot game where people could be waiting for their Starbucks coffee and they could flick left and right like swipe left on a malarial parasite.
You could train the user.
Yeah, but what are you doing when you're waiting for your coffee?
I work from home, Chris.
This is back in the old days.
They found that if you train people up on what a malarial parasite looks like, you could have, you know, I don't know, if you had 17 people look at the slide, it was as accurate as a doctor doing it so you could figure out malarial parasites.
So we did spill some games like that, like school or not kind of thing.
Swipe left or swipe right on your school.
And so we've kind of built up these sets of tools to just get the physical data about where schools are.
And if you look at that project connect that world map, you'll see a lot of countries have purple dots, which are just the physical locations.
But that still didn't get us to Greg Wyler's question about how to turn the dots different colors.
And that's tough too.
So tell me.
So now we have some data set of schools, right?
And we can correlate that with if they have connectivity, I presume, based by fiber lines where they are drawn, because they're probably better data on that.
Well, actually, that's the next question, I guess.
How do you take it from.
Okay, we have a school.
Do we even know if it has connectivity?
And if it does have connectivity, what does it look like in terms of QoS and all that.
So there are two steps between finding the school and figuring out actual traffic data.
The first is exactly as you said, to figure out what's the rough umbrella of signal that it's under.
And so if you take GSM maps, so you can go to gsma, which is the kind of umbrella agency for all the mobile service providers, and you can request GSM data from them.
So 2G, 3G, 4G, 5G coverage.
So every cell tower is registered in that database.
Yeah, they won't give you the cell tower locations, but you can see like what's underneath a tower umbrella, basically.
Right.
And signal strength.
So you can kind of get range.
Roughly, I presume, plus, minus.
And you know, telephone operators are kind of cagey still about like what radio is on what mast, which they don't really need to be anymore, but it's sort of they consider protected information.
Ten years ago, when we started doing the venture fund stuff, we couldn't even get tower locations.
It was like that was a proprietary thing to be like, that's.
Nobody can know.
And it's like, okay, why?
Because we're putting our tower there.
Somebody else is going to put it one hill over and then they're going to take all our users away from us because they're super concerned about churn.
Right, right.
That's a little less.
Now everybody does tower sharing, so that's not really going to share anymore.
And we actually get a lot of tower placement data from like partners like IHS Tower, which has a ton of masts in Africa, which they rent, you know, radio space out on to competitors.
They give us their tower location data.
We're able to process that and use it.
But you can go to a service provider and you can be like, if I give you 5, 5 lat long points, can you tell me, you know, what umbrella they're under?
Is it 2G, 3G, 4G?
And they'll sort of do that for you.
And there's some published maps that do that.
And then you can also go to big companies like Meta and if you have the right agreement with them, they'll also give you some coverage data.
Now, none of this is live and none of it is really linked to the location itself.
It's just where an RF wave would hit.
So you're sort of like, if I'm on this point on the map, should I be able to get some frequency of something undocumented?
Okay, yeah, but it doesn't feel like it.
Exactly.
Right.
And so a mountain being in the way is not an ideal condition.
And it's not necessarily captured on their coverage maps because those coverage maps are pretty much like the world is flat and there's no trees.
You know, you're under our network, so but you can you get that map.
And a lot of the countries that you'll see on Project Connect World have what we call, you know, proximal data or potential connectivity data that is not real time, that's not actually linking into the school.
That's just where the network should be.
And it's usually pretty good.
It's like pretty generous.
There's a lot of coverage in the world, it's just there's not a lot of access.
So still, As I said, 2 billion people don't have access, but many of them live under a cell signal of some sort.
There's another question about fiber nodes, which I'll talk about in a little bit.
But, but generally you can get like a pretty good cell plus signal map that's not super hard to generate, but that generates a lot of schools that look green because they are under a cell signal, but they're not connected.
So the third thing we do, the most difficult thing is figuring out in real time whether a school is connected or not.
In order to do that, we do a bunch of things.
So one thing we do is connect into governments network operations centers, if they have them, or depending on the kind of complexity of the government contracting, sometimes they'll have an ISP that's the government contractor that will have their own API that basically represents some of the data from their own network operation center.
And if you're not familiar with network data that basically is ping, you get roughly ping data from a client and from the client details it'll spit back something so we can get some of that.
Whether or not you trust that type of data is another question.
So we also have our own software probes that we can install that are chrome extensions that we can put on site that's behind the router.
We'd love to do firmware like at the router, but at least like we can get it on some appliance in the school, a computer, laptop, whatever, and that pings back.
So we've got about 70 or 80,000 schools providing that real time data.
So the first data point you get for the ISP is basically like ISP to the edge gateway essentially.
Then then obviously you have packet loss and whatnot from that to the school.
So like you want to have the end device experience.
Yeah, I mean, I'm happy with either of them.
It's a.
We have a huge argument on our team of like, is ping good enough for bandwidth?
Like for bandwidth monitoring?
I believe it is for most of the stuff that we need to do.
But this is like a religious argument on the team.
That's how.
Right.
So like, no, because it doesn't really give you throughput, like at the level that you need for real bandwidth monitoring.
It doesn't give you any real traffic details.
Ping is pretty brutal.
But at least lets you.
Yeah, yeah, it at least lets you know that there's something there.
Right.
And so for a lot of the stuff that we're working on, like, I just need to know that there's a heartbeat.
I don't need to know like the size of the patient body really.
Like, it'd be great if I could.
Now the other problem is the way that an ISP does their like, allocation of IP addresses may not be statically assigning them to schools.
Like they might just do a block.
So they may be giving you pretty arbitrary stuff and who knows really what a third party provider puts in the data that they give to you.
So I like to have something that links a geolocation to, you know, to some ID in our database that can give me a signal from that.
Yeah, I guess our appliance.
There's a massive incentive for the ISP to overrepresent their performance against their bodies, right?
Well, for sure.
Especially if it's a public service contract where there usually are some uptime guarantees and things like that.
Right.
So if you are legally liable for stuff, you're gonna be probably pretty generous about what your uptime is.
For example.
Not trying to point fingers, but you know.
Well, yeah, it's good to cross check it, absolutely.
But then that means sometimes you gotta get something installed at the school, right?
So that's the bummer of it, right?
You gotta go to a school that probably didn't even have electricity a few years ago, that there may be only one computer on site.
Cause maybe they're running just a, like a projector off of the thin Internet that you're giving them and you got to be like, listen, we got it.
We've got a chrome extension that we want to put on here and hopefully that gets done.
But we've done that successfully.
That's in 80,000 schools.
Now.
Every school in Mongolia is reporting data.
We've got something like 60,000 in Brazil that are doing it.
And they do it every four hours during the School day.
And it's cool.
You can actually see like the MBPS coming in.
And it's kind of neat, right?
I mean, that's because that's a trade off as well, right?
Doing a band, if you have a very slim connection, you wouldn't want to eat up that connection.
Doing benchmarks off the connection.
Not too frequently, right?
Yeah.
And I mean, you know, I think we need to be realistic also about what this project is for.
It's, we're trying to get a community that has never had Internet and probably doesn't have electricity in their school to have the school be the node for the community.
WI fi.
I mean, the real trick of this is that you connect the school and through that you bring a lot of other services to the community.
What I mean by that is if you can get the public contract from the government to put the tower on top of the school, for example, say it's a cell tower or say it's like a radio receiver, you can propagate that signal through a WI FI antenna to the community for a few hundred meters for relatively low cost.
You can do that after school hours.
You can then create a WI FI hotspot that people can use for a lot of other things.
And the school then becomes the hook that drags this connectivity out.
But you create a public access point which gets more people to use the Internet, even people who aren't in school, which then gets me to my goal of having more people connected to positive, you know, software and technology and data that can help them with their lives.
And so for us, like, do I really care exactly what the throughput is for the school?
No.
Do I care that there's a signal in a community which maybe didn't have electricity a year ago?
Yes.
So, see, three colors of dots in Project Connect in red, yellow and green.
Green is like you're getting some juice.
Red is there's nothing.
Yellow is like there's a trickle.
It may be like 2G speeds, you're talking Kbps, but honestly, that's better than it's.
There's, it's.
That's so many orders of magnitude better than zero that you know, I'm happy even with yellow.
Yeah.
So you found, so you looked at antennas, you figure out some kind of fiber links in the city or in the region.
What other tools do you have at your disposal to kind of figure out or connect to the dots here?
Yeah, well, so fibernotes are interesting.
The itu.
So the third part of my story.
So we got Sharad's thing done.
We got Greg Weiler's thing done, and then we needed to get Doreen Bogdan Martin's thing done.
ITU has access to all of the fiber data and kind of infrastructure data for broadband.
They run a thing called the Broadband Commission.
They work on behalf of all the regulators of the world.
And so one of the things that's been really interesting over the last year is that we're able to start taking their data about fiber node location and fiber itself and starting to overlay that on the information that we have about school connectivity.
Then you can see some pretty interesting stuff.
Like, well, there's actually a lot of fiber that's only in Kenya.
Like, something 95% of it is only 2 kilometers away from a school or less.
So it's like right there.
But you don't really know that if you don't have that fiber map.
So then you can get a provider like Liquid Telecom, which is a big East African isp, mno, mostly an isp, to kind of pop up off of that, off that fiber, pop up a hotspot directed to the school.
And if you can have that, you can create efficiencies for planning.
So the tools that you do, this map, the map is pretty important.
But after the map, we have some infrastructure planning tools and also some financing tools that we bring in to help a government get their schools connected.
I'd say that the map, though, is itself a really great starting point.
So what we found in Kyrgyzstan was that just by doing the map, we showed that there was a whole bunch of schools in the middle of the country that the government thought had been connected.
So they're like, we just did a big public procurement.
All the schools are connected.
And we did the map and we saw these red dots right in the middle.
It says, like, green to the end of the top bottom third, and green to the bottom of the top third, this swath in the middle.
And like, why are those ones red?
Turns out that the providers, they kind of contracted two providers and they both sort of stopped in what they thought was the middle, but it wasn't actually the middle.
It was just the end of the third.
So we're like, wow, great.
There's a whole bunch of the country that's not connected.
Showed it to the prime minister.
He saw it and he's like, what's going on here?
And I was like, I don't know, dude.
Like, that's a you problem, not a me problem.
But, you know, this is just a map.
And they immediately redid the bid.
So they did a new bid and they got a new company to come in to actually cover that intermediate intermediary space.
And it lowered the price of gigabytes by 55% for schools.
Wow.
So just the map itself, like, created some market dynamics which brought pricing down because it brought in a new isp.
So the map is important.
That being said, the things that happen after the map are pretty neat as well.
And we've helped governments now, like, arrange loans from some of these big banks, like the World bank or Islamic development banks, and financing, you know, players that can give them cheap loans for infrastructure.
And we've also helped to talk about, like, regulatory structures in governments that could open up money for public infrastructure.
So governments often have these universal service funds or universal access funds that are taxed revenue from network operators.
It goes into a pool that's supposed to connect poor people.
And spoiler alert, it never does.
It gets used for other infrastructure projects.
So we also help work with the regulatory side to open those up.
So, I mean, I guess.
Sorry, that was like a lot.
That's like giga in, you know, four minutes.
There's a.
There's a whole lot in there.
There's a lot to unpack.
Right.
And I think that the interesting part is, one of the interesting things that we ned out last time I saw you in person was about how you can kind of reduce the cost of all this because you have kind of a blueprint of how it should look like for both what it should cost.
Because I would imagine in a lot of these countries that are, I guess it goes without saying that a lot of them are very corrupt and a lot of people trying to reap the benefit of all these money, like, diverged elsewhere, I guess.
And so talk to me a bit about, have you kind of solved that problem with both fighting corruption and reducing the actual cost for the school?
At the end of the day.
I believe that every country is corrupt in its own way.
Every bureaucracy is corrupt, everything, you know, these things, all corrupt.
And I mean that in a very technical sense, that they are corrupted, by which I mean that they're inefficient or they're rusted or they're broken or they're old.
And I very firmly believe that data is the solution to that open data fixes that.
It's like the way you clean the rust off of pipes.
When we look at the policy of broadband in the United States, for example, it's terrible.
If you look at indigenous lands, if you look at, like, Navajo land, For example, you'll see almost no connectivity.
The ability to set up like local open networks is zero.
In the US, a few big telcos have the lock in on it, and that's because they spent an enormous amount of money on policy and advocacy, lobbying.
Yeah, this stuff goes throughout the world.
And I really don't like it.
And I don't like it because it's unfair and it's not balanced.
And I just.
That kind of stuff bugs me.
If you have data about it, as you're saying, you suddenly create a picture that you're not arguing like against a company, you're just showing the truth.
So if you look at the Project Connect map of Brazil, I'm not.
I don't have like a political view in that map.
I can just show you that there's like a whole bunch of the country that's red.
And that's not really fair.
So I'm not like advocating for or against something.
It's just data.
And that's pretty nice.
It's very apolitical.
I mean, in a sense, the thing that you're on the side of poor people, which I'm.
Okay, I think that's like a very reasonable place to be positioned from like a philosophical perspective.
But you're also using data to show the potential opportunity.
So when you show that red map, you know, red and green map, an isp, a smart isp, a micro ISP or a wisp or wireless Internet service provider who operates below the level of the three or four big government telcos can actually see a space for their business to be made.
They could see where they could position their towers or their assets, you know, or do their own little local, like a small local ecosystem that they can plug in.
And that's very exciting.
And that's.
I think so I think you solve the problem in both of those ways.
You make the data publicly available so that everybody can see it and they know that you're not like trying to mess up this person's business or that you're just trying to show something and you make this accessible to the players who have the energy and the drive to try something new.
whether that's by trying a new technology solution like using LTE in new ways or using kind of point to point for what, getting in and getting Internet, like over the next hump or over, up to the top of the mountain, stuff like that.
Or that they just are providing service to areas that they might not have considered because you're like Suddenly there's like a whole bunch of people here who you didn't see and those are customers.
And if you get in there, forget the new technological innovations, like you can just scoop up 20,000 customers really fast.
So I think at the ground level, those two are really interesting drivers of change.
Yeah.
And I think, I mean, I can resonate what you're saying even in multiple countries.
I've lived Both in the UK and in the US and in Sweden as well.
The best ISPs that I've ever had to deal with are small independent ISPs.
So that kind of holds because they're hungry, right?
Yeah.
Like, and they have the best bandwidth, best service.
Exactly.
Like by far, Google Fi springs to buy in the US and they got kind of shut out.
Sorry, not Google Fi, Google.
What's Google Fiber?
Yeah, Google Fiber.
Yeah, yeah.
They got shot at multiple cities by Comcast and the big ones.
Right.
Because they were, sorry, you're too good, you are too cheap.
But Google Fi went the other direction and went and worked with the mnos in a very interesting way.
Like, I love the Google Fi model for many reasons, but they sort of, they went exactly the opposite way and said from a policy point of view, so many partnerships with these different MNOs that they could just resell service and kind of did the back end of it.
But yeah, I mean, I think that there's a tremendous amount of innovation happening in these smaller providers and that if you could, you know, often they're lacking capital to go a little bit bigger, so they can do the 20,000 person build out, but you know, that's like what, three towers or whatever.
They can't necessarily do 200,000 people, 500,000.
So there is a kind of capital gap.
Yeah.
I was looking at one of the reports.
I think it was the report with the Economist that you guys work with on.
It was in the intelligence unit of the Economist and it was a quote in there that I thought was really interesting, which is, schools connectivity significantly enhances educational outcomes and economic prosperity.
A 10% increase in school connectivity can raise the effective years of schooling by 0.6% and GDP per capita by 1.1%.
That is pretty profound.
That's pretty powerful.
It's, it's nuts.
And if, you know, if you think about that, if you connect a school and a kid goes to that connected school for 12 years, you've given them an extra year of schooling, basically.
Like, that's crazy and that's really special.
That means they can either leave earlier or just do more.
But we also saw that if you connect all the schools in Sierra Leone, which is a lot of school, look at Project Connect, you'll see there are a lot of disconnected schools.
You raise GDP by like 19%.
I mean that's insane.
Fairly significant.
So we just got a five million dollar loan from the Islamic Development bank for Sierra Leone to connect more schools.
The president is very committed to it.
And it's both because of the education side but also because of the economic, you know, economic drivers that it has.
And that brings me to the other really important part of giga, which is that it's a blockchain based project.
Like we are blockchain native.
The whole stack is Deepin, which is decentralized physical infrastructure and networks.
The everything that we do is about creating these nodes on a network.
So for us the dots represent schools, but they are also nodes in a network.
They are decentralized nodes for provisioning services like Internet services, but also financial services.
So you connect to school and then Victor's out there using the WI fi after school hours to do his business or Sarah's using it to do her business and they're transacting.
How do they transact?
Do they transact in fiat currency?
Probably not.
Is it maybe mobile money?
Sure.
Is it possibly crypto or digital currency or central bank digital currency?
I believe so.
And so a lot of what we're trying to do is to also create a series of blockchain layers that go all the way from the most fundamental and low level all the way up to accounting and payments that I'm tremendously proud of.
And they're all open source, they're being built out of our tech center in Barcelona.
And to me, the blockchain part of it is possibly the most exciting because it's where I think we have a real edge and we can really change stuff at a global level.
So you essentially are building a connected city operating system in a sense, I guess.
Yeah, you could call it a big global isp.
I mean if one was being ambitious, if one wanted to call it gigafi, not Google Fi.
But yeah.
So if you think about it, what do you need to allow for a school to be that hub for economic activity?
That school's data all needs to be captured on chain in some kind of data object that has some permanence but also is able to be updated.
There happens to be a really good model for that.
It is non fungible type of tokens that exist on chain, not that retail NFT stuff.
That's Kind of gross.
But if you actually go back to what distributed databases can be, that NFT model for data storage is pretty powerful.
It means that also schools could have their ownership of their own data.
And you could update it, as we do with our current data on Project Connect, which is all in centralized database, which becomes really powerful if you think about then the next level up, which is giving the school a wallet.
Cool.
So school has data about it, like, where is it located?
How tall are the walls?
Are there trees?
Around it are mountains, including the certain degree, certain angle of radio frequency.
Second, what's the wallet for that school?
Can you pay a teacher through that wallet?
Can the school pay for its own connectivity through it?
Can you offer incentives for test scores or performance or pay village workers for providing health results or things like that?
That's kind of neat.
So you're like, so that's the.
So first layer of the stack is just physical objects instantiated on chain.
That's kind of important.
Second is kind of a payment, some kind of payment architecture.
And we're now doing payments in crypto in a few countries.
We're testing this out in sandboxes.
We're using staking rewards from Ethereum nodes to pay for school connectivity.
And then the third layer is like, if schools are objects and they can then pay things, and we can monitor contracts, all these complicated contracts that governments do, we can monitor those on chain.
Can you write smart contracts to pay automatically?
And the answer is like, yes, the dot turns green, you make the payment.
Now that's frictionless.
It's real.
The school is actually an oracle.
That's super cool.
And then the last layer on top of it all, after you've done all that, is can you create a marketplace around this where the connectivity itself becomes sort of tokenized, where you could exchange gigabytes, where Victor or Sarah in After Hours, maybe they're not using it for free, maybe they're paying a little bit for it, but the school gets that money back.
And they get it back because you're somehow locking that set of, you know, that cluster packets of that gigabyte for Victor into a token, issuing it to him.
He gets to access it.
And there are mechanisms for doing all of this in the blockchain world.
There are protocols that are kind of natively close to this type of stuff.
And we're able to play with those, to use those and to look at like rolling them out really globally, which is very exciting to me.
That's super cool.
So how does it.
So this space, you know, a Lot more than I do about mobile banking in particular in Africa.
Right.
Where they leapfrogged a lot of credit cards and bank accounts.
Where it's mobile banking is, it is phone banking essentially.
Right.
How, how do you see these two things kind of connected together?
Are they competing or synergetic or how.
Do you see that it's totally market differentiated like it.
And it depends on the regulations from the government about both CBDCs, their central bank digital currencies, like how do they see that and their fintech stack as it exists now.
So like a country like Safaricom, a country like Kenya which has Safaricom has a huge amount of their like national liquidity in the Safaricom bank account.
That's their mobile money account and people use M Pesa they may be less willing to experiment and in fact they're, I think their crypto stuff was around $8 billion last year but is down from 16 billion the year before.
So there's a big drop but still like a lot of that's remittances.
But then you look at a country like Nigeria where you've got tens of billions of dollars of crypto activity going on and there's a sandbox, kind of regulated sandbox that's being built up by the government and there's not this like one company that owns the whole fintech stack.
So it really, it's really varying.
And then you have smaller countries that don't have the capacity to do the regulatory stuff at all and they're kind of looking and watching and looking for help and we can provide some of that.
So we support governments in there.
And something that I love doing, talking to central banks about kind of what the CBDC landscape looks like versus the open source kind of more permissionless blockchain which I'm a proponent of ecosystem and so offering some pathways to grow their own economies, ecosystems just by like learning about really educational, educating their own governments.
Right.
Their own national actors.
But so I think the answer to your question is a lot of it's going to be picked up by network operators and this is going to be just native part of a network operator stack in the same way they have their billing, they built their billing stuff around.
I mean the way that they monetize mobile money basically runs on the same engineering stack that SMS runs on.
It's just like a little bit more sophisticated.
Right.
It's just like billing on top of that.
So I think they'll do something similar with crypto but it offers a type of Interoperability across network operators and then to other industries that they don't have with mobile money, it's like a lot more fluid.
And so I think you'll see a lot of really interesting growth happen in emerging markets as the definition of what a network operator is changes.
And they become service providers for a lot of other software packages like education, like health and so on.
And they can actually pay for that content and they can track and monetize it because blockchain gives you the ability to do that content monetization.
That's like the long term.
I think that's like the long term M and O play for a smart operator.
Right.
So going back to the mapping process and like connecting schools in general, what is the most meaningful deployment to you that you've seen so far on this journey?
There are, I mean, so I love the big national scale stuff because I think that's really important.
We have seen countries like Kazakhstan reach almost full, you know, full connectivity.
Kyrgyzstan has now 99 connectivity.
But I also love some of the stories of very remote communities that just never had access before.
So there are like 30 schools in Kyrgyzstan which even after we did that mapping stuff there like this, 30 schools were.
They're so far up in the mountains that they don't even have like an FM signal.
They don't have, there's no radio.
These schools, these villages are so disconnected and because of the mapping and because of a guy named Talent Sultanav in Kyrgyzstan and he works with isoc, which is the Internet society of Kyrgyzstan, he's connected these 30 communities, these 30 villages, and for the first time ever, they have the ability to be part of like a global discussion.
And that's super meaningful too.
And then, you know, every time I talk to teachers in schools, that teachers tell us that the connectivity has changed their way of working and the quality of their experience, which I think you can translate into anybody having the Internet.
So really interestingly, like we don't do any of the ed tech stuff, right?
We don't do like what kind of software you put in the school or do you do tablets or laptops?
That's just not our thing.
We just kind of get the traffic to the router basically.
We don't do internal like land shaping, we don't do any of that building up.
But we had a teacher in Kenya, this teacher named Paris.
She was the head teacher of a school called Norkopan in Kenya.
It's like 35 kilometers outside of Nairobi.
She said she's like, this was amazing.
So Having Internet gave us like kids exam scores went up.
Scores went up 8%, 9% just in a semester of having this.
That's good.
Kids attendance went up by 10% because, like, it's interesting, it's more interesting for them to be a classroom where there's this thing.
But she's like.
But most importantly for me as a teacher, like, that's all very good.
But I also would have to travel every day to this, you know, every week to the town, which is like a little while away, and put in these scores and attendance and come back.
And that took me an hour every week for just no reason.
And I'd have to pay the Internet cafe a little bit of money and I don't have to do that anymore because I can do my admin work from school.
So it saves me an hour every week plus a little bit of my own money.
Imagine that.
Like, imagine having back, I don't know, 5% of your time, 10% of your time because you're admin.
Like, that's.
The Internet does that for us so often we don't even think about it.
So I love, I mean, I love those types of stories and then I like the successes.
So we helped open up in Brazil $600 million worth of financing for connectivity, and that's going to go connect 20,000 schools this year in the next 18 months.
So there's nothing big stuff like that.
But these really, the personal stories make me very happy because I know that we're doing something kind of meaningful.
Yeah, I mean, yeah, I could definitely see that.
What I'm really curious about also is, I mean, this is something we chatted about last time, but it's about the early days of the Internet that you and I experienced.
Like the BBS era and like the early days of the Internet coming online from zero today versus the experience of like the 90s is just so radically different.
It's almost like how do you even start?
I guess that's, that needs to be part of the curriculum.
Almost like how.
I mean, it's such a vast universe to navigate that it's almost hard to wrap your head around in so many ways.
Gavin, When I came online, like my entry point was like a man page for Telnet, right.
You know, it's like there weren't a lot of options.
It's like you could get, you could use Gopher and get to a few things and then, right.
You could do some stuff on Telnet.
And that was pretty much it.
You had to write A TDT commands to connect your modem to the.
Yeah, so it's really funny.
Like I actually came in right after the scripts.
So, like by the time I came on, scripts were like pretty much there.
But it was that like set of six months, right?
And I was like, I'm so glad I never had to do that.
Like that thing that was really old.
Now I can watch like ASCII pictures, you know, come on really slowly as the BBS sets up.
Like, I can watch like the name of the BBS come up in ascii, right?
And you know that.
So I think a few things like the.
It was so much harder for there to be information on there, that the information was generally higher quality.
There's for sure no AI, but there's a better, you know, kind of.
The stuff was a little bit more meaningful, but also it was a lot safer, the Internet.
Like there were still like creepy people on the Internet.
And I remember that from being 10 years old and being like that, somebody talking to me like a really weird way.
But it wasn't.
You couldn't really do that much with pictures.
You couldn't really do that.
You didn't have a phone, so they weren't in your space and it wasn't as dense.
One of the big problems with the Internet now is that it's super dangerous.
So if the Internet.
I think that if the Internet was like a physical place, was a building, I would certainly never let kids in there.
It's full of a lot of issues.
And when we think about bringing young people online for the first time, there are at least two things that we're trying to do as GIGA that maybe can be a little bit helpful.
But this is a much bigger issue than we're prepared to handle ourselves.
One of them is that we build a set of contracting safeguards in for the government.
So they contract 20,000 schools.
In the contract, it says if there is material that's dangerous to children stored on the ISP servers, you can cut the contract, you can take them to court.
Stuff like that.
Like that just didn't exist before.
The other thing that we're trying to do is to build up a cohort of what we're calling community information workers.
So these would be somebody like.
I'm also calling them human infrastructure experts, like people who are near the school who can be responsible for helping with some of the training and who are like government, slightly like government employees who aren't teachers, and they can provide some of that extra skills training.
Like what's dangerous online?
And so on.
And we're testing that out in a few countries, but it's really tough.
The flip side is that were, I was just talking to a young woman from Tuvalu in the South Pacific where they had no Internet for two years during COVID So like somebody hadn't paid the service provider bill or whatever and the Internet went down.
All the online banking was down on the island.
It's about, I don't know, 11,000 people on the island.
And she said she's like, all of us lost a year of our education, life.
It's very much to your point about the Economist report, like what is it, your connectivity.
But if you talk to a 17 year old, I think she's 17, 18 years old, about how devastating it is to be totally cut off while all of your friends on other islands and in Australia and New Zealand are like, they're all doing stuff.
And maybe it's like dumb tick tock stuff, but it's still stuff.
And to not have that is devastating for somebody who wants to be part of the digital economy.
And so I think we have to really balance these safety concerns, but also the economic potential.
And there aren't a lot of people who are really talking about that space.
I think in a way that's solution oriented.
Yeah.
Because it's very, I mean Internet censorship is, I mean something we talk about, spoken about, linked up before is extremely challenging problem in so many ways.
Not only from a technical but also from philosophical problems.
And so when you deploy these, when you help deploy these devices on routers, is there any sense of like filtering on that level in terms of what should be accessed in general?
Whitelisted, blacklisted or any level of parental control?
I guess in a sense, yeah.
So now you're getting into our 20 year business plan.
But I'll say right now, no.
So right now we don't actually have any router level control of anything.
We are exploring some collaborations with router manufacturers to say like could we have a branded router for schools, for example, that could come with some pre built firmware that might include, you know, some level of kind of packet inspection and understanding what's going on.
But then it gets a lot more expensive when you have to do a DPI and drives up the price.
Right?
Yeah.
And it also puts like a lot of processing in different places and you know, figure things out.
Flip side is like, could you also put other things on that router that could do Edge, you know, like, you could do tiny ML stuff, you could do edge compute.
Like, there's an interesting side there, but we're not quite in that yet.
What we do is have a software appliance that's literally a chrome extension that gets installed.
And you could think that thing could actually become a little bit of a sniffer for certain types of activities.
But we don't want to get into that right yet because it puts us in a different, you know, relationship to policy and everything than we want to do.
We want to be able to just say either the bits came in or didn't like.
We sort of stopped there for now.
Yeah.
One thing I wanted to cover, which is, I think, kind of interesting around this is a project that did have a lot of potential a long time ago.
At least the one laptop per child project that I haven't heard about in quite some time.
I mean, the objective was create a $100 laptop that could be essentially given to kids around the world.
What is your view on hardware in general?
And I realize it's not something you do right now, but seeing kind of an eye on the future with regards to these things.
Hardware is really hard, as you know.
I think so.
Nick Negroponte, you know, he left that, he started that project.
I talked to him.
He's pretty familiar with Giga.
He really changed the.
The way that the world thought about a lot of the various components of hardware.
So if you like, I have my one laptop or child somewhere.
If you look at it like it had screens that became Kindles, it had WI fi that became like a lot of our mesh kind of protocols that we use today came out of that.
So, like, a lot of things came out of it.
Crank for it.
Yeah, the crank one never.
Yeah, well, you know, pieces were there.
The swivel screen became like that stuff is like that hardware.
This hardware components were great.
The problem was that those machines, you know, one of the problems is they weren't like manufactured in or tested in the places where they were going to be used.
And the other problem was it's really hard to get governments to agree to buy things and actually get the procurement to be followed up on.
So I think there were a bunch of logistics issues and manufacturing technology issues, but it was so far ahead of its time.
I think that I have a very risky perspective on this.
I think that hardware mostly solves itself for the most part, or at least I'm not a big enough player in the world to solve hardware problems.
I think the market drives hardware Solutions and not an individual's desire or will.
And even with like, I don't know, the ipod and the iPhone, like that was a market demand that Steve just like tapped into.
So I think what you're going to see is market demand for a certain set of devices that are much smarter than the devices that we have right now.
So for example, this device, bless it, is a thief, right?
Every time I turn this thing on, it sucks so much data from me, I don't notice it because like I live in a rich connectivity place.
I live in a green place.
I don't pay for my gigabytes.
But if I was on a bandwidth plan, imagine you turn your phone on and like immediately you lose a quarter of your gigabytes, right?
That's not fair.
So I think you're going to see like market demand for things that don't draw so much, that don't provide so much data back to all of the apps through SDKs that like limit that stuff.
And that market demand will cause hardware manufacturers to make those things.
I think it'll be that.
I think there may be some people who have keys into that, you know, kind of they go to key and they open the door and they go, this is the way it should go.
But, but I'm not sure that there's like a pure hardware play that fixes this stuff.
I also think the same for IOT again.
I think that's really a space that you know so well.
But IOT is going to be a huge driver of access because you can have like low bandwidth, like lower latency stuff that just works and that drives a whole market of low earth orbit satellites that don't really exist right now because there wasn't a reason to provide it.
But you can have stuff that just like pings every once in a while and gives you like a few bits that still is valuable for IOT that will, I think those things will come up.
I just don't think we need to invent them necessarily.
That's fair enough.
Do you see the end user devices being tablets and phones or do you think it will be PCs in the way we grew up in the sense of building PCs and all that stuff.
But I am so bullish on the phone.
Like I, I just believe that form factor is it.
And I know it's like hard to type an essay on the phone, but you don't need to because you dictated into ChatGPT.
Like I know it's hard to look up multiple things on the Phone, but it's not.
If you're 14 years old, you can do it.
Like, I, I just, I think that this form factor is so much superior to everything else.
I think that you'll see things where like, oh, you can connect a phone to a projector.
And I would still love like a low cost, ultra high luminosity projector that you can, you know, chunk around and it doesn't matter, it doesn't break.
But like, as a teaching tool that'll be great.
But I think this idea of having like power banks and charging multiple laptops and keeping them in cages just doesn't make sense to me.
And I'll do a quick story on that.
Like, I was in Brazil, in Amazon rainforest where there's no GSM at all, zero signal.
And I was in schools that have satellite connectivity for a few hours a day.
This is a few years ago.
And so for a few hours a day, all of Manaus, this capital city, connects all of the cities around all these little villages in the Amazon.
And there's like a huge wide area network for 60,000 kids.
I was in a place four hours away from Manaus on the river.
The kids had smartphones, they didn't have SIM cards.
I'm like, why do they have smartphones?
They have them for that three hours of wanna.
And during that time they're all like on their own big WI fi network, sending poop emojis to each other.
There's no SIM card in it because they've never used a GSM network.
So I just think like, it's going to be phones.
But I could be super wrong about that.
I just think probably right.
And I think the phone is a defined factor that I've realized that I am getting old because I can't spend more than a few minutes, type messages and I'm like, yeah, I'm just gonna put my laptop.
Yeah, I find it much easier to.
But I'm like, I'm mega old.
I've got multiple screens here.
Right?
Yeah, yeah, myself too.
I'm like, how could I do an actual piece of work on my phone?
But then the other day I was traveling somewhere and I'm like, I need to write something.
And I described most of it sent into ChatGPT.
It laid it out for me.
I did like two more revs of that and I had a perfectly workable thing that I was actually using, you know, chatgpt as a command line for transcription and that was great.
And so in like a year it'll be doing that for PowerPoints.
And what do I do anymore?
I just make damn PowerPoints.
You know, like, that's like the limit of my technical prowess these days.
But also you can code a game in it pretty easily too.
So I don't really need that.
I think ChatGPT has a long way to go for coding.
Actually the fun side project for that.
Well, that's a sad story.
But I did, I've done some coding with ChatGPT and it's pretty terrible today.
It gets you started pretty fast and it can do like, it can code across modules in Python.
So I was like playing with it the other day and like it can maintain state a little bit.
Yeah.
Which was surprising to me.
But it breaks really quickly.
But yeah, I mean you put it.
You can see the first 15% and then you're like, cool.
Now I can do the rest of it.
But I think your point still holds though.
Like, I think it's generational thing.
PCs or laptops is kind of a thing of the past.
We are the old people that are still drawn to the keyboard.
So I think your thesis career is probably going to hold water.
But also look at the, look at like the resources that a laptop takes both to manufacture and also to keep up and then to be repaired.
And it's like, I don't know, you're going to have.
If you have another 2 billion people coming online, they're not going to be using a thing that costs you $3,000.
No, no, you're absolutely right.
Okay, so I got a few things I want to cover more, which is one of them is open 5G and SDR is a big part in what's going on in the telco world right now.
Project like Magma or the OpenSD and some other project.
How do you see that's fitting into the whole connectivity world at large from your vantage point?
So I have to say, like, it's a regulatory game and I'm not a regulator.
Our partner.
So GIGA is a partnership between UNICEF and itu is the regulatory agency that governs that.
So I can give you my very limited perspective on it, which is that I think that the whole regulatory landscape about Spectrum, around spectrum licensing needs to be rethought.
I think that there is big potential for open 5G in the same way there is potential for TV white space.
But it's a licensing and regulatory question.
So if you can't get a government to give you dynamic licensing for spectrum allocation, then you're pretty limited on what you can do with spectrum.
Like that's kind of a defined thing, right?
Of course.
So like Magma was great because Magma provides software driven backend to do a lot of this stuff.
But if, yeah, if from the national level the regulator isn't super behind it.
It's really tough for me to see how you break monopolistic behavior on the part of MNOs.
In a nutshell, like yeah, very high potential, very high friction.
Yeah.
I mean what I find fascinating about is the commoditization of the telco stack which has been owned by a handful of companies for the last 20 or plus years.
And I think that's super interesting because that can potentially drive down deployment costs by order of magnitude.
I agree.
And some of the stuff that I've described with our Depin stack is an attempt to provoke activity in layers of that telco stack without being too aggressive and trying to take over.
You know, there are a few providers of billing backends and billing software.
There are a few providers of these different pieces and like you don't want to antagonize anybody but you can provide alternatives that they that are open source and that are licensed in a way that they can't, you know, commercialize and proprietize.
They can commercialize them but.
And I think that there's a lot there.
I think interoperability is key and I think once you have more interoperability you have more flexibility on for example spectrum allocation because people feel more comfortable with it and you can do that dynamic stuff.
So during peak hours you can give a block to somebody and then take it back.
But you can't do any of that if things are locked in three different providers, silos and backends.
And you can't cross Bill, for example.
Yeah, CBRS is probably only spectrum that's still, that's properly open around the world.
That's the only example I've seen, at least at a larger scale.
Yeah, like the ham.
Ham and CB.
Those nerds are good nerds.
Yeah, 100%.
That's where it all started.
Right.
They predate a lot of the nerdery really.
So you know Doreen Bogdan Martin from itu, the Secretary General is a huge ham radio fan.
Oh really?
That's funny.
Cool.
The last thing which I think this topic cannot be completed without is to mention Starlink and how that kind of fits into or satellite Internet and obviously everybody's a big fanboy of Starlink.
How does that fit into the whole GIGA playbook?
Does it actually solve A problem.
Is it way too expensive?
I, I, it's been in the news for a lot of things around the Ukraine war, for instance.
And I'm just curious more your.
View is on that it actually solves a problem and it's not very expensive and the technology is beautiful.
Yeah.
So full disclosure.
Elon put some money into Giga at the beginning.
It was money that's not for procurement or anything.
Like helped, that money helped build the open source map that we use.
And it's not linked to like any, you know, we don't have to use any certain technology for anything.
The what we see in most countries is that you need a mix of technologies in order to connect everybody.
And LEOs themselves don't do it all.
Fiber does a lot of it and fiber does a lot of it.
Well, high bandwidth radio does a lot of it pretty well too.
Like 5G is pretty good for a lot of use cases and satellites are good for when you can't get the other two.
Yeah, they're also much cheaper than, and in some cases the only option rather than laying fiber or kind of putting up a mast.
So in those villages in Kyrgyzstan that I described, the roads are frozen for eight months out of the year.
You're not going to ever drill anything into the roads of these mountains.
You need to have satellites.
Same in a desert area where it's far away.
So satellites solve a lot of problems.
The question is also the regulatory piece, which goes back to why we're partnered up with ITU in Giga.
Governments have a lot of regulations about who provides Internet access.
And it's, you can't just bring Starlink in or any other leo.
You have to have an allocation from the government.
The work that's and I believe that allocation should be done to all LEOs.
It should be made open.
It shouldn't be given to one specific one.
I'll speak very broadly about the LEO industry, not about anyone in specific.
So this is not about Starlink, but I think that governments need to have a policy for how they bring in a new contender against the existing mnos and how they regulate that fairly.
So any LEO should be offered a smooth regulatory pathway to access a certain amount of spectrum and have landing rights in a country.
Their hardware should be allowed to be brought in without like excessive tax and excise and they should be allowed to play in the market, but there should be guarantees for the provision of public connectivity that protect pricing.
So my biggest worry with LEOs is they come in with extraordinarily low pricing on certain assets and they increase them arbitrarily over time very quickly.
And there's nothing you can do contractually if the contracts aren't carefully written.
We believe that we as Giga can help to make sure that's a fair transaction, that you have a, a price ceiling, for example on connectivity.
You can also have a price floor.
So you can protect everybody.
But that for public contracts that are written with public money you have some long lasting capacity to not be totally upended if something goes wrong or is change on the part of the provider.
Secondly, the hardware needs to be open access like terminals can't be linked only to one provider, I believe because if you have that, you can't reuse it.
If that company goes out of business and everybody's going to say we're never going to go out of business.
You need right of reuse.
You need to be able to repurpose things.
And so it's very dangerous for me to think about spending a huge amount of capital for public money on a specific device that can only be used for a specific provider.
That to me creates a type of lock in that we've seen is very dangerous in other parts of the telecom industry.
And I am highly nervous about that.
So those are the pluses and also the minuses of any of these companies.
Interesting.
I mean and that brings me actually it's something I should have brought up earlier which curious about it's restriction of access.
Right.
Remember Facebook tried a project in.
Was it India?
I believe they were.
They tried to connect people to the Facebook platform purely or meta I guess purely as their Internet at a heavily discounted rate.
But there was obviously not a proper Internet connection.
How do you see or how meta.
How does Giga see that worldview and how does that.
Is that a real thing that only limited that or have you seen that at a bigger scale?
Yeah.
So that was Facebook, that was called Internet zero or Internet.org or Facebook zero.
There are different names for it.
It was there.
They basically whitelisted all of the Facebook sites and then 20 partner sites.
It's really dangerous.
So like then you create your own walled garden.
There's no anything.
There still are countries where that exists that just think the Internet is Facebook.
Like that's just their user experience.
I'm obviously like I'm a huge proponent of open software, open protocols, open hardware.
Like I really don't like that from a genetic point of view.
It's also just dangerous.
So I don't love that.
I think that the other thing that I'm kind of always nervous about is like locally cached content, which is sort of similar, like, oh, let's just cache it all here and then we're going to control the cache.
I know it lowers kind of consumption costs, but it also creates these bottlenecks, these choke points for control.
So I think that these are things that need to be like, the more open you are on the network, the better you are able to have discussions about this stuff.
The more you create these pockets of closure, the less that conversation even has to happen.
So one of the people that I respect super highly in this space is a guy named Ste.
Steve Song S O N G.
He's like an Internet philosopher and thinker.
But also it does very good analysis on both the questions of open Internet and also of like, infrastructure that can be used for other things.
So there are like these big fiber rings that run between universities called nrens, which are national research and education networks that are like unused fiber basically in a lot of cases that can be used to propagate good stuff without getting that kind of proprietary lock in.
And Steve talks a lot about how these types of networks can be used, and I think he's a great person.
He also has been talking a lot about satellite LEO pricing for people who want to explore that a little bit further.
Yeah, I mean, because I think that's something that I would make.
Guess would make a lot of states nervous is when you have a fiber line into the country, essentially you can and put constraints on that.
However, if that goes up in space, you have zero control over what ingress or egress.
Yeah.
Unless you've got an agreement with the service provider that satellite is just providing backhaul for local MNOs.
Right.
In which case you're fine because you're still trunking it like locally.
And so in that case you're just providing like a dispersion vector for a local pipe.
So you.
That's what I mean.
I think that's what you're gonna see governments doing is that they kind of just limit.
Everybody has their sovereignty and the.
Like a network operator sovereignty technically ends 10 kilometers from the border, essentially.
So like, you don't see a lot of masts in most countries closer than 5, 10 km from a border.
It's a, it's a gentle person's agreement.
So it's not written in, you know, in blood anywhere.
But that's just kind of how it's been.
So that's.
You don't get interference from two different MNOs.
And so like that's.
Governments try to protect their sovereignty, but the world has just changed and like the Internet states are bigger than a lot of physical states and we're in a post sovereign era.
Yeah.
So like, people are gonna have to get used to that.
I think one of the, you know, most difficult things to watch has been over the last few months with the Israeli bombardment of Gaza, watching the Internet change there.
I mean, you've got Internet totally cut off and you can't get Internet access in Gaza now, which means that, you know, so there you have a case of a government like actually cutting off another people's access to this fundamental thing in addition to bombing, you know, a lot of bombing and deaths.
That makes it very hard to under, you know, to imagine what the future world looks like when everybody's going to be trying to cut off other people's access to information.
Again, I have some biases towards how decentralized networks fix some of that, but that's probably a longer conversation.
Yeah, no, I was just kind of curious if you've seen any of that in your deployments at Giga.
If they've been like, yeah, we're happy to provide Internet.
However, these are the rules of engagement.
You cannot grant access to X, Y and Z.
We haven't yet, because we have.
We only provide the raw bandwidth.
We did see as soon as the bombing of Gaza started, we connected 70 schools in Palestine and they had the live connectivity thing.
And we did see them go from green to black as the bombing started.
So we could watch that kind of stuff.
But we haven't had a government tell us not to provision something because at the end of the day they hold the contract with the service provider.
So it's really up to the government to dictate what goes in that service package.
Got it.
This has been super secret.
I think we could go on for another hour or two at least, but I think that we have to put a pin in there and keep on in a future episode perhaps and keep up with the developments at Giga and the telco industry at large.
So thank you so much to come on the show.
Real pleasure.
It's been a real lot of fun nerding out with you, so thanks so much, Chris.
Thanks, Victor.
It's been really fun and please, you know, we'll keep in touch and I'd love you to chat with some of the other members of our team who are doing the tech build and watch what's happening in our Barcelona center over the next year.
It's really been.
It's been fun to chat.
I'll be in Barcelona soon, so let's do that.
We can redo the beach.
The beach discussion that we had a few years ago.
Can't wait.
Cheers, buddy.
Have a good day.
Talk to you later, bye.
Found an error or typo? File PR against this file or the transcript.