Ask Me Anything! With Abhi Vyas and Don Walpole

Jase Wilson: Hi everybody. I think we just lost a guy. So, does anybody in the community know any good jokes? There's Don.

Abhi Vyas: There he is.

Jase: Hey Don, Don? At least it's the letter D [laughter]

Abhi: He should be joining us any second.

Jase: And so that's why it's important that you have really great connectivity for broadband and all the great applications that you can do with it. So.

Abhi: Yeah.

Jase: Awesome. I'm gonna give Don another couple of seconds, but, Drew, are we good to go apart from Don? 

Drew Clark: Yeah, we're live. We're live now.

Jase: Okay. Alright. We're live. Maybe Don and I will go into the webpage...

Abhi: He's... He's coming on.

Jase: That's sweet. Okay. That is awesome.

Abhi: Yeah, we could get it started.

Jase: Okay. Yeah. Broadband Grants Community, thank you for making time. Happy Friday. It's always great to have really smart folks on the call. But today's call is extra smart and we get a double dose of some next generation visionary folks. We have Abhi and Don from Mem Protocol and... We have Abhi and maybe we'll have Don in a minute, but it is...

Abhi: Yeah. He's joining us. He had technical difficulties on his end, so he's just going to join on my end here.

Jase: Oh yeah. Sweet. I'll see you guys on like in the same space. That's awesome.

Abhi: Yeah.

Jase: This is an example of the importance of broadband and bandwidth and making sure that it's adequately sufficient. Hey, you all, it's cool. Okay. This is even better. So Abhi, Don it's great to be with you all and, we're excited to have you at the Broadband Money Community for Broadband Grants folks and we are really excited. There's a lot of folks in the audience that are, really interested in what you all are building and how, what you're building will impact the broadband space and then an important point that I think we're gonna make a lot in this call together is that, broadband is... As special as it is to the folks in our community and as big of a topic as it is, and foundational to the functioning of society in 2022 and beyond, it's a means to an end.

Jase: In fact, it's a means to several ends and it's really an honor Abhi and Don that we get to hang out with you guys. We know you're very busy. We know that things are moving very fast for you and that what you're building is insanely large and important and complex. So, welcome first and foremost to the Broadband Grants Community. And, thank you for sharing your time with us, I'd love it, you know...

Abhi: Thank you, Jase for having us. I know like, the work that you've been doing at Ready and the broadband community, is... It's much needed. And, we like the 75 billion dollars that is going to get allocated over the next decade to localized Ps and, giving more connectivities to rural America is so important. So yeah.

Jase: Thank you, Abhi, and also, you know, cities too, right? 

Abhi: Yeah.

Jase: There's a lot of places in the United States in the land where the internet began, where even in like a dense area that you would assume has like great coverage, there are parts of town and not all parts, but certain parts of town that have, total crap if they have anything. So, thank you. Abhi and Don if you would like, we met through Nathan Black, our mutual friend.

Abhi: Yes.

Jase: And it is super exciting what y'all are building, but I'd love that if you guys could give the community a quick intro to yourselves.

Abhi: Yeah.

Jase: And then talk a little bit about what Mem... What you're accomplishing with Mem.

Abhi: Yep. I'll kick it off. So, I'm Abhi, co-founder here at Mem previously I worked at early stage companies, as generalists. I did everything from like building a basic web app to marketing ans sales growth, product design, you name it, I did it. Prior to that, I was in college, and, left a little early [laughter], to go work at startups full time. One of the problems that I've been obsessed with for the longest time is information asymmetry, this idea that there's someone out there that knows something that I don't, and I don't even know the question to ask to identify that thing. So about a year ago, I called my friend here, Don, who is a mathematician. And I told him, I've been exploring all these different blockchain primitives of like bounties, data interoperability. And I wanted to build a better version of Quora.

Abhi: So you ask a question, what is blockchain? You post a bounty a hundred dollars. Someone answers it. And, they earn that bounty, but not only that, but they earn reputation that they can take with them, off platform, right? So into the real world on LinkedIn, wherever they go. We started building this and what we really quickly realized was, hey, one, if we... Like, why would you start your reputation from zero you have a college degree, you have, reputation that you built on Reddit, Quora, Stack Overflow already. Why don't you bring that with you? And two, if we allow you to take your reputation app off platform, we have to allow you to take your data and identity with you too. And so in essence, about six months ago after we tested that and came across that insight, we realized that what we really needed to do was build an open source protocol for trust and identity on the web. And that's what we've been building here at Mem Protocol. And I can... I'll go into the explanation, but before that, I'll let Don give a brief intro first.

Don Walpole: Yeah, I'll keep it short. So Don, the other co-founder here at Mem. So before joining Abhi I was over at Comcast working as a machine learning engineer, working on commercial operations, trying to do things for quality of service improvements.

Jase: Where? 

Don: Over at Comcast [laughter]

Jase: Yeah. Oh, I've heard of those guys.

Don: Yeah. Yeah. They're fairly well known. I think. So.

Jase: They're the ISP, right? That's the ISP.

Don: Yeah.

Jase: Okay, great. So you were doing machine learning at Comcast. That's interesting Okay. Okay. Cool.

Don: Yeah. So like parsing, server data, trying to improve service for our commercial clients, before that it was at grad school in Rutgers. That's where Abhi and I met, I ended up leaving there with a master degree in statistics. Did some machine learning stuff over there as well. Did minor grad at Berkeley in applied math and statistics.

Jase: Very cool. So when you say in your profile on Twitter, I do the math. You're not joking? 

Jase: So, this is awesome. Abhi when you and I met, I had this sense that like, I didn't know Don yet.

Abhi: Yeah.

Jase: I'm super thankful that, we're all connected and friends now, but like when I met you, I was like, holy crap. What have I been doing with my life? A, this dude is like, he's already got like a whole career, but, I was also really struck by what Nathan shared with me first, which was your website, which is, I think it's abhi.nyc.

Abhi: Yep. Yep.

Jase: It's a geographically misrepresented now, right? You're not actually in NYC anymore? 

Abhi: Yes, yes, yes, yes, exactly. Yeah. It's geographically misrepresented. [laughter]

Jase: Yeah. So abhi.nyc.

Abhi: Yeah.

Jase: The reading list alone is worth like a bookmark. Right? 'Cause it's ebbed in the thoughts that you've... And the stuff that you've curated for inspiration is like, it just... It's a small window into some of the things that led to the current you. Right.

Abhi: Yeah. Absolutely.

Jase: And that's awesome. Okay, cool. So Abhi and Don, like, if we could, you talked a little bit about it, it's open source protocol for identity, but can you, help us understand? Like especially me, you know me, like you gotta talk to me like I'm eight. And let's do...

Abhi: Let's start by like three to five years from now. Let's, start, by talking about experiences. Right? 

Jase: Please.

Abhi: Like what do the... Like the infrastructure and technology we build, what do we want it to enable? So Jase three years from now, when you're booking a restaurant reservation.

Jase: When I'm 11.

Jase: I'm with you.

Abhi: When you're booking...

Jase: I'm doing a the math too, Don that's your influence rubbing off buddy.

Don: Got some allowance money saved up. [laughter]

Abhi: When you're booking a restaurant reservation at... And whatever the booking engine is let's say it's OpenTable or Resy, whatever it is, we'll ping your vault which has your preferences. Yes. Which has all of your preferences and everywhere you've ever eaten before. And so it will ping that and essentially give you a curated menu on arrival.

Jase: Wow.

Abhi: And so in essence, what we're saying here is that Mem protocol is essentially building the best safe on earth for your social event, history, your contact book, your preferences, your location, history, all of the data that you generate when you're using any application online.

Jase: Where does that data go today? 

Abhi: Today, it lives in the servers of the applications that you use, whether it's Facebook, whether it's Google, whether it's Amazon, any, or even Apple, any one of those applications or services essentially has that data on their blockchain.

Jase: The internet today, if I'm understanding correctly, is filled with platforms and services that every time I log into one of them is getting like a huge amount of info about me, about like my situation, my activity, it's probably somewhere there's like satellites flying around, there's like beeps beeping and, dot matrix printer, like printing out stuff. That's like, based on X, we think, Jase's XYZ, blah, blah, blah. And there's, they're probably making some assumptions, right? Like, I mean, you doing the thing.

Abhi: You can see it for yourself if you, right? Like you can pull your... If you have a Facebook account, and if you've had it for a while, you can request a dump of the data there, right? You can get it in like a JSON format or an HTML format. And then one of the files that you'll get will show the different labels associated with you exactly like that. Right? So the... I mean, we all know about like surveillance capitalism, targeted advertising, the whole industry that's sprung up along that. So, and it kind of comes from the end to end principle that was used to design some of the lower level infrastructure. The basic idea is that today, third parties control your data and the data that you generate online does not serve you.

Abhi: Right? So like, what does... This is a bit of a jarring concept to wrap your mind around of actually owning your data, because we've never done that before. We've never had ownership of our data. So, I gave you one example of the curated in menu arrival at a restaurant. Another example I love sharing with people is that one of my goals right now is to lose weight. However, DoorDash and Yelp don't know this. If DoorDash and Yelp knew this, they would never recommend the pizza shop four blocks away. They would always recommend salad or a healthier option instead. Right? 

Jase: Whoa.

Abhi: So if I like... So that's step one, right? Step one is like me telling the system my goal is to lose weight. And then the applications I use reflecting that in the experiences I have. Two, is actually the system inferring based on all of the connected devices that I have. So I have an Oura ring, I have an Apple watch where I track all my workouts. I have a weight scale that's connected to the internet. If I'm losing weight week over week, a smart agent that lives on top of my vault would actually tell me, "Hey, Abhi is one of your goals to lose weight right now? If so, I'll recommend healthier options the next time you go to a restaurant."

Jase: Damn dude.

Don: It's really about the experience that you're gonna have online. If you take the consequences of this kind of control form of ownership over the read and write access to the data. Then you... What we're really talking about is moving away from this kind of attention oriented economy to this more goal oriented economy once each connected agent on the network is in control of its own representation and presentation and how it interacts online. So as Abhi said, the smart agent now can serve your goals as opposed to the economic problem of this principle agent problem that we have now, where because these platforms are the third parties that are mediating all these interactions and they have potentially conflicting objectives, your goals aren't being served directly. So it's kind of reorienting the way these interactions happen and potentially changing the way these economic interactions happen.

Jase: That's okay. That's extremely exciting for somebody in the broadband and connectivity space that you... What's you're aiming for is in the end, it's like an internet that works more for the person.

Abhi: Yes. Yes.

Jase: Like it does more for the person while also not necessarily needing everything, every last piece of information about the person.

Abhi: Exactly. Exactly.

Don: So you can have your goal oriented economies but you can still have the business logic, all the information you need for your business logic for the interaction on the other end of the line.

Abhi: So our thesis is basically that in the future, rather than users signing into applications, applications will just request permissions from users who have control and custody of their data.

Jase: Rather than I don't log into the application, the application requests specific permissions from me. So this is analogous, so if I sign into something by Google in the future, there's gonna be a Mem. I'll say just talk to my Mem.

Abhi: Yeah, exactly. Yeah.

Don: In fact it should become a more seamless experience. So a lot of that stuff should become automated and happen under the hood, long term.

Jase: It's beautiful. Why... So can I step back for a second guys? 

Abhi: Yeah.

Jase: Like we dived right in there, but before we get into... We got a lot of great questions here.

Abhi: Yep.

Jase: We got some more questions that have been coming in.

Abhi: Yep.

Jase: We got some questions from smart asses, we've got some questions from smart people, community if you had any questions for Abhi and Don at this juncture go ahead and weigh in. I think we're gonna ultimately end up not having time for all these, but I do have a sort of question based on something that I heard from you Abhi early on, and that's like you said something to me that really struck me, and it was about how the internet impacts your life. Can you share a little bit about that? Cause I think it's important people in the audience understand you in a lot of ways you had a lot of education and training that came from the internet. So...

Abhi: Yeah. I mean just to share more context here. So I grew up in a little town called Avenel, New Jersey. I don't necessarily come from any kind of wealth. My dad was a truck driver, my mom worked at the bakery at Dunking Donuts and growing up the only way that I could learn was... Outside of school was through the internet. Like I didn't know what startups were, like I used to... I remember watching when I first got my laptop I would read voraciously. I would read like everything I possibly could, I would watch these keynotes by Steve Jobs like the Apple events. And that was like a... That was a moment in time. It was like... You'd sit in front of your computer for an hour before it started and watch the screen just like [laughter] alert. It was a big thing for me as a kid, and I would just dream about what technology would enable from all the things that I was reading and like all the interests that I had.

Abhi: And I would keep asking questions and when you come from an immigrant household where your parents don't speak English, you're pretty much the person that's managing like your parents 401ks at like 14, and you don't necessarily know how to manage your parent's 401ks. So like you go Google, and read everything you can. And that's how you discover David Swensen and portfolio management. And I probably went deeper than I probably needed to at that time, and most of it didn't make sense, but hey my parents 401k is still doing pretty well. [laughter] so...

Jase: That's fantastic. Okay. So the internet helped raise you, like the internet helped to teach you, right? 

Abhi: Yeah, yeah absolutely.

Jase: And for folks in the room that are wondering like if you're facing the oncoming tsunami, there's $75 billion incoming. State broadband directors in the audience, like this can be overwhelming, but it's super appreciated what you're doing to make sure that you're not just getting the money into places that already have service. You're getting it to everywhere that doesn't have service. And this is a living breathing example of what happens when there is at least connectivity. Right? 

Abhi: Oh, absolutely.

Jase: You, as a curious young mind, you were able to impart, educate yourself using the internet. And now you're not doing so bad. Like, can you guys share a little bit about like... I'm not the only one excited about Mem, right. Like...

Abhi: No, no, no. Yeah just...

Jase: Maybe you're maybe new to the Broadband Grants Community, but you're definitely no strangers in Silicon valley. Like you've...

Abhi: No. No. No.

Jase: You share anything about like what is involved with it.

Abhi: We haven't shared the latest update, but last year we did... Last August we did close around funding led by Andreessen Horowitz, and some other great investors.

Jase: Awesome.

Abhi: That are helping us bring this visions to life.

Jase: So Andreessen Horowitz, top tier absolutely astoundingly, valuable folks. Mark was a big part of like getting worldwide web to folks with a browser that didn't completely suck. And they've always been obsessed with infrastructure, they've backed some of the great infrastructure firms.

Abhi: Yeah.

Jase: Martin Casado's a partner there, inventor of software defined networking, which is an AR world and broadband world and connectivity world that's somebody that brought about an order of magnitude cost, structure collapse that allows far more internet for far less money. Right? 

Abhi: Yep.

Jase: As one example. So that's a stacked group of folks, and they've got Katherine Boyle, David Ulevitch for American Dynamism talking. If folks are on the call and they're wondering who the hell we're talking about, like go to a16z.com or.co or whatever it is. And then look at also the American Dynamism stuff, because it's super important. I think that folks understand that these two on the call Abhi and Don are backed by some of the best. And there's a belief that firm like in rebooting America essentially from getting back into the business of being innovators and our position to them is like, well that first starts with really strong information infrastructure everywhere, but yes. So you have the blessing of one of the greats, you have a lot of other amazing folks involved I'm sure.

Abhi: Yes, absolutely.

Jase: So you're not alone, right? We're talking to the two of you, but you actually stand for a pretty interesting community in a very real sense.

Abhi: Yeah. And incredible team as well. We couldn't be doing this, if it wasn't for all of the people on our team Dylan, Cameron, Boicheck, Russ, like all the incredible people that we have. Yeah Nick.

Jase: Okay. Awesome. Also, FAQs having some questions. We've got several that have come in. We've got some, some don't like, some of them look like they're kind of smart ass, but they actually have interesting components of first up is like a question from Greg from Twitter. And it's what do you think is gonna happen with web 1 and web 2 now that you're here, like what's... And is web 3, like replacing web 1 and web 3 what's the story? 

Abhi: Yeah. And he has a follow up question too, which I thought was also really good. He was asking you basically about like, what's gonna happen with web 4, 5, 6, 7, 8, 9 [laughter] which kind of points out a interesting point. I mean, obviously he is framing it as a joke, but these terms, web 1 web 2 web 3, they're useful terms for people trying to orient themselves. Especially when they're talking about these different technologies and the capabilities associated with them, but really it's a solution to a more under fundamental underlying problem, which is we're talking about this evolution of capabilities and kind of control of your experience online. So the issue there is like, yeah, web 3 is useful for orienting yourself, but you have to be careful because web 3 means different things depending on who's saying it and what they're talking about.

Abhi: So, yeah. I mean, like we can, we can slice and dice web 1, web 2 and web 3, like on many different axis, whether it's functional, whether it's temporal time based any one of these different axis. I like to use them as like markers of like what was possible in terms of experiences at different points in time. So like in terms of web 1 specifically the idea was that you could read, you could read any content, that someone was posting. And again, you have to keep in mind the time, as well, like there were very few people on the internet at the time. There were very few people who could actually, post things because most people didn't have computers and web.

Abhi: That also means that read was kind of just there. It was again... Kind of from the design, it was like, just if it was there, anyone could read it. And in terms of web 2, it was actually read plus write. You could actually like now add content. It was easier for people to add content to the internet the average person. Like all these blogging websites, social media websites started popping up and going... Becoming more mainstream. So this coincides with like development of CDNs and things like that. Right? 

Jase: What is CDN for.

Abhi: Content Delivery Network. Right. So where people host their...

Jase: Like CloudFlare.

Abhi: Yeah. Stuff like that.

Jase: Groups... Okay. Got you.

Abhi: Now we're talking to 2013... 2008 Bitcoin white paper, 2013, Ethereum, where now you have the ability to actually, write smart contracts, on a blockchain, by trying to complete a blockchain. So we're getting towards the concept of verifiable computation. Which is kind of one of the fundamental problems with cloud computing, which is all right. I need this function to be executed on my data, not on my local device, but I need it to be verified that the actual correct computation was performed and none of the data was tampered with right. All the information security stuff about, well, confidentiality is an issue with some of these blockchain implementations, but the integrity and authenticity of the data. Right? And the availability of the data and the new capability that was added for users was ownership for the first time, you can actually own, digital things online.

Jase: Read, write, own.

Abhi: Yes. Right.

Jase: Web 1 read, web 2 read, write.

Abhi: Yes.

Jase: Web 3 read, write, own.

Abhi: Yes, exactly. And again, this isn't necessarily a framework that I came up with. Like Christensen talks a lot about this.

Don: Even own is a bit nuanced. Right. This is why you have to be careful with kind of the context of when people are using these terms. Because the thing that we're talking about with making our experiences more intuitive. So that the web is more like we make the machines basically match what humans are doing as opposed to the other way around, which is kind of what we've been doing. With the web so far, we've been trying to get our, our behavior to match up with the way we've built the machines. If you really own your data, we're talking about a digital thing, not a physical asset. So you really have to think about the access control over it. Which is like who is even allowed to see the data itself. Because it's a digital medium being able to read it...

Jase: Like CHM.

Don: You can copy it. Right? So.

Jase: A little Linux humor there.

Don: Sorry. I missed that. Jase, what was that? 

Jase: CHM like little...

Don: Yeah, exactly.

Jase: Yeah so this is fantastic. So thank you for breaking it down for us. Like that's a good segue into next question. It's from, Ben Kahn, who, shout out to Ben recently joined the team. He was a very skilled talented reporter at Broadband Breakfast and was mentored by Drew Clark, one of the greats. And he's now helping us like engage more folks and wrap everybody's minds around awesome stuff. Like what y'all are doing here Ready and Broadband Money. But he asked, like he wants specifics, like what type of interactions are you most excited about for web 3 like what now that we can, like, now that we have the own permissions activated for people and people are able to do that, like what does that mean? Like, how's that exciting? 

Abhi: I walked you through the two examples earlier. The one that the cured menu on arrival at a restaurant. The second one is like applications actually knowing about my goals. They both speak to the higher level category of performance based applications that serve our goals as user agents. And what I... When I say a performance based application, what I mean is that if you decouple the user table from any specific application, what this opens up is that in the future, like... I imagine there'll be an app store where some developer is going to build a social network just for your family, Jase. So there will just be a social network for the Wilsons where each individual in that network has their own data that they're bringing into this context. So imagine like a mini Facebook just for your family. That is something that is now possible.

Jase: Dig it.

Abhi: Yeah, the kind of overarching concept is trying to make a more seamless and intuitive set of experiences for people in digital... These digital spaces that we're building that is more in line with our physical experience, right? 

Jase: Yeah.

Abhi: Like that's the idea here. So just like you decide who to talk to when you're walking around in the world, you should have that level of control and consent in your interactions online.

Don: A good example of this is when you go into the login page for Goodreads, there are five signing options. It says, sign in, log in with email, log in with Apple, log in with Amazon, log in with Facebook. Now the issue here is that Amazon, even though they own Goodreads, Amazon and Audible, all they know about you is what you've read on Audible or the books you purchased on Amazon or the books that you've reviewed on Goodreads. However, they don't actually know all the content that you read on the internet every single day and so the recommendation algorithm on Goodreads for the next book that you read is actually incomplete. It's not going to be as good as it can be if they knew everything you read. And when I say know everything you read, I mean, like in a privacy-preserving way using techniques like homomorphic encryption, zero-knowledge proofs to make sure that your data is secure at rest, correct? 

Don: Yeah, so to get a little bit into the technical aspects of this. There are some cryptographic techniques that have been developed in the last... Within the last five years, or sometimes even more recently, but this... The stuff that Abhi was talking about. So being able to generate presentations of data that is held private, that still is... Satisfies the requirements for the receiver in terms of the business logic and is also verifiable. So one of the examples we like to use is with purchasing wine, right? If you go to any kinda alcohol vendor in the US, then you do an age check to authorize you to purchase, are you at least 21 years old? 

Don: So if you think about how this workflow typically works in today when you go to a store, they will ask you for ID, you will present, typically like your driver's license or some state-issued ID, and then they will inspect that, maybe they'll scan it through a machine and then that'll check your birthdate and verify your age, that you're in the valid range that of now minus 21 years. If you think about all the information you're revealing through that interaction, it is way more than the vendor needs to learn about you in order to perform the compliance check. Really the vendor just needs to know does your age fall within the valid range? They just need a boolean true or false. So some of the techniques Abhi alluded to like zero-knowledge proofs, homomorphic encryption, functional encryption again, depending on the specific use case straight off analysis, whatever, what you can do is you can have your PI... So for instance, say a digital version of your driver's license that is issued by the state, and verified cryptographically that is held privately in the vault, Abhi already described it before.

Don: And then dynamically a presentation is generated of your... Using as zero-knowledge proof, for example, that just has the data true, you are at least 21 years old, and there's a cryptographic verification proof alongside it.

Jase: Okay.

Don: So your birthdate and your home address, your eye color, your height, all of that remains private to you, is never even set over the wire, but the vendor receives a message and response that has the required assertion that they can use for their compliance check. So all the overhead for maintaining PIA and everything goes away for them if they're using the protocol.

Jase: This is fascinating stuff guys. So can I share briefly what my most... What most excites me is digital services, guys. It's the future of digital services. So you've made it clear that you're not replacing web 1, you're not replacing web 2. You're enhancing and extending it and you're unlocking a key piece of what the internet not is today, but what it can become.

Abhi: Yes.

Jase: To me, it's all about those digital services and back centuries and for a second, they wrote a great piece of 2018, 2019, the future of services, like service marketplace concept, right? 

Abhi: Yeah.

Jase: That was one of the inspirations for how we went about building Ready. Along the way, we got to know about digital services at the edge, hence there's... Once the information infrastructure is in place when you can transmit, which is what everybody obsesses about in broadband, right, is like getting information to and from in a good way, but also compute and store very near the application the point of context with the application. So if it's like a sort of a consumer and a service provider model, it's like there's a doctor, right? So like Justin Fulcher at RingMD, right, one of the things that he's obsessed with is making sure that they get exactly the data that they need to do telehealth connections into the hospital, right? 

Abhi: Yeah.

Jase: And Abhi, I know that you trained as an EMT. Huge respect for you to do that, right? Like, can you imagine a future in which a meme is part of the equation once these broadband directors deploy the money into really strong information infrastructure, where we might see a future where it's like people can live wherever the hell they want, and they can connect with amazing services, even if those services are being created a thousand miles away and they don't have to necessarily transmit their entire life story every time they go to connect to one of those services? That's what excites me.

Abhi: That, incredibly exciting and I think one of the questions that I had read on the page was what would we tell broadband directors who are deploying this capital. I read a report that came out in 2021 about the FTC looking at what data ISPs are pulling about users, and most people don't realize this, but I know in popular media, they talk a lot about how the big tech is using user data for targeted advertising and misusing user data for targeted advertising. And when they say big tech, they're talking about like FAANG companies like the usual...

Jase: What's FAANG for folks outside...

Abhi: Meta, Amazon, Apple, Netflix and Google, right or Alphabet. [laughter]

Jase: Okay.

Abhi: So I was reading this report and they said that it was actually...

Jase: Man, it's man, patriarchy. I knew it. [laughter] Okay. Sorry.

Abhi: 98% of the mobile internet market is actually taken by six core service providers. T-Mobile, Google, Comcast, Charter communications operating and cell code partnership. And AT&T.

Jase: Say that again.

Abhi: AT&T so there are like six ISVs that make up 98% of the mobile internet market AT&T cell code partnership. So that's Verizon charter communications operating Comcast cable T-Mobile and Google fiber, right? Those are the six major players there and most people don't realize this, but like they're tracking like all these ISPs have an unencrypted stream of your data, of what you're watching when you're watching it. Your IoT devices, your ring, your security cameras, your voice assistants, all of these different like devices that are transmitting data are going through these networks and in unencrypted data streams and the other thing that most people don't realize is that many of these larger companies actually have third party data brokers in house that they own. And so if you have now, like when you combine that data where they can actually verify your identity because you're paying, someone in the household is paying for whatever service this, and you combine that with all of this data that you're generating they can build profiles of you and sell that to other third parties without you necessarily knowing, right? And so it's quite shocking what these ISPs do when they amass these large pools of sensitive consumer data.

Abhi: And I mean this applies that infrastructure as well, right? If you look at kind of the way certificate authorities work with TLS. You have similar sort of market share by the top performers there. Right? So like the, I think top five certificate authorities are responsible for like 90, over 90%, right, of the certificates, the extended verification certificates, which if you're a business you want, because that not only is a certificate, which confirms the control over the domain name, but also has associated identity information with the business. Right? Like one of the certificates authorities handles 60% of that as of last year. Right? So again, you have kind of this capture of kind of the experience online, right? 

Abhi: Yeah. So in terms of the advice that we would give to a broadband director, who's allocating this capital, I would say two things one is authenticated data streams, so that whenever well one we need more ISVs two... [laughter]

Jase: Right.

Abhi: With those ISVs we need to make sure that there are standards for the data that they're transmitting on behalf of users. Yeah. So it's not just that the standards that need to be there, they need to be adopted right. Throughout the network. So getting that adoption and buy-in from every participant in the network is gonna be crucial and it will lead to lower OpEx and just more efficient and ask capabilities right. Of experiences online.

Jase: This is fascinating stuff. We could do like seven hours. I know we don't have that much time though. We're actually getting pretty short on time, so yeah. We're gonna start to move in to...

Abhi: 'Cause Jase, your question was kind of like multi-part, right? You kind of talked about like the edge and kind of like bringing compute to the data stuff. Right? So I can go into a little bit of the details there. So I mentioned some of the different graphic techniques, right. So functional encryption is one approach that may be useful depending on the use case again. Right? So the idea with functional encryption is that essentially, you encrypt the data, but there's also some function that you wanna evaluate over the plane check. So let's say you have some sort of JSON record and you wanna run some average over some fields or something. Right? Or again, it could be like the zero knowledge proof example I gave earlier, right? Where you've got JSON with all your verified driver's license information, and you wanna generate a verifiable boolean that you are at least 21 years old. Right? So now you don't wanna reveal any of the information or structure, so of the data, right? So that over time they can't correlate anything, right? You wanna have semantic security as well as just like the text sorry, for text security.

Abhi: So what you can do is you basically, in through the encryption method, the encryption method also encodes the function you want to evaluate, right? So let's say, I wanna evaluate this average over this array that's inside of the JSON right. But I don't wanna reveal any structure about the underlying JSON so the encryption I perform also encodes this average value, and there's a key associated with the evaluation of the average, the only person who can then learn the average is anyone who has this key. So rather than having to send the compute to the data, I can pre-compute it and just transmit the key to whoever needs to evaluate the function. Right? So that's one of the ways you can do it as opposed to something like a homomorphic encryption approach or zero knowledge approach where a lot of the compute has to happen on the client side. Right? So there's trade offs and different approaches here, right? There are techniques, depending on the different use cases for where we need the compute to happen.

Jase: Oh this is beautiful. So you guys are describing like what Mem is up to. To me, you're creating like a far larger possibility space for digital services, applications, like stuff that goes on the internet. And to me, it's like, you're onto something that could, I don't know make the experience better. Like you said, it's like, it starts to like... Our relationships with the machine. It starts to become a little more ambient.

Abhi: Yeah.

Jase: You can solve through some of these silly ass things that you gotta do. Like the friction of like, yeah. Tell me about the caption. He's like, you know I'm a human, because I'm like, I'm signing some stuff over here and you don't need to know everything about me, but you can you ascertain that I'm not a robot, at least mostly not robot. I always wonder about caption and cyborg discrimination. Like that's... [laughter]

Jase: There's...

Jase: Let's explore that in a the follow up.

Abhi: Yeah.

Jase: I do wanna ask you guys though one question real fast is, we work a lot with tribal communications and folks that are building, you know, really great tribal broadband infrastructure and getting folks involved with... In travel tribal communities to sort of get their networks, right? Like, you know, 'cause you know, they need connectivity everywhere. And even if they're like very remote from other places, but one of the things that the leaders at tribal communications kind of like, you know, help me understand that I never understood before was like, you know, there we occupy like, you know, similar parts of the geography of what people consider to be America, but it's not necessarily the case that everywhere in what it looks like America is America. Right? 

Jase: And these programs like there's billions of dollars that are being set aside for tribes. Right. And it's that's awesome. Right? But there's a challenge with the data, right? It's like, and this is where I didn't really understand this fully, but it's like, you know, and they wouldn't necessarily want, you know, all their data being transmitted to, you know, from a sovereign nation to another. Right. And that's really fascinating. Can Mem tech, is there like a near term use case of Mem where it's like, you know, like for example, there's the affordable connectivity program, right, and if you're a household of need, you get $30 a month credit on your internet bill, your ISP needs to be able to facilitate this process. And most ISPs, the way they're doing it is like really total trash, like just asking everything under the sun and then storing, like you said that the set of records needed to verify your eligibility.

Jase: And that's part of the reason why we built turn key ACP, which is like the fast way to discover and win, you know, and get ACP like set up for your household. But in tribal context, it's $75 a month. Right? And that's really good news, right. There are a lot of folks that are gonna be able to benefit from that, but they wouldn't necessarily wanna transmit all of this like information about themselves. Is this something that might meant, maybe we could talk about this, like in a follow up, but I don't know, I feel like there's like some potential use case there where it's like, you can tell us what we need to know. We can tell USAC what they need to know. We don't necessarily have to go into the details beyond that. So.

Abhi: Yeah. I mean, some of the stuff I was just talking about, right. Again, we need to know some more details about the requirements for the data flows and the confidentiality requirements and everything. But yeah, some of the techniques I was talking about, right? Functional encryption, zero knowledge proofs can definitely do that. Right? So like we were talking about never revealing the PII for purchasing a bottle of wine, right. That is a... That is through these mathematical techniques that are, fully general. So whatever you wanna store in your vault, right? If you want that vault to literally be on a set of digital devices, like your phone and your backup, like cold storage and that's it, you have that option. Right. And then through the protocol, right, you can have fine gain control over what gets transmitted and what the privacy requirements are. Right? So you can generate presentations that are useful for the business logic, but again, don't reveal any of the underlying PII. Right. So just like the driver's license example with buying a bottle of wine.

Jase: Beautiful. Speaking of wine, I don't know if you caught it, but we were able to catch up with Vint Cerf.

Abhi: Yes.

Jase: Is like your godfather of the internet protocols. And on that AMA, you know, we talked a little bit about Mem and, you know, what's what y'all are trying to build with identity and he had some interesting thoughts. I don't know if you caught that, but he is also a huge wine enthusiast, so when you guys, when you guys sit down and talk about how to make this really awesome, make sure to mention your wine bottle use case.

Jase: But two Tim Wessels community member asks, like, do y'all have any contact with Tim Berners-Lee, Sir Tim Berners-Lee. Right? The father is at www, with Tim Berners-Lee, if you're listening, my baby boy's initials are www. So.

Abhi: Definitely thanks. Definitely, no, Tim Berners-Lee has done some incredible work. We have not been in contact with him, but his paper on the, he, oddly enough most people don't know this, but in '08 he coined the term web free when he was talking about the semantic web and the giant global graph before the Bitcoin paperwork, yeah, I believe it was before the Bitcoin paper. You can go read the giant global graph blog post that he posted back then.

Jase: That's amazing. Okay. I used to walk by his office and I never gathered the courage to knock on the door seriously. How old do you want, you know, I invented the world wide web, but I'm busy, but.

Jase: Tim Wessels you asked the question, if you're connected to Tim through some kind of Tim to Tim protocol, like you wanna introduce the Mem guys, like definitely you make that happen 'cause I think what you guys are doing and definitely gets into a lot of what was so inspiring to me back in the day of studying, you know, what they were doing, especially W3 folks. And that'd be really cool to sort of, to see all like work y'all work with them already though in some cases, right? 

Abhi: Yeah. We're looking at standards like decentralized identifiers, verifiable credentials out of the W3C. We're also looking at authenticating change data containers at the IETF, we know, Tim Berners-Lee was working on solid protocol, right, to handle some of the data storage issues, model there. We're really looking at, the access control and, the dynamic and complex system requirements there.

Jase: Okay. Wonderful. This is fantastic guys. So we've got more questions than we have minutes, so...

Abhi: Yeah.

Jase: I'm gonna start running through and there's another community member Dave, that he works with the Rubin infrastructure technical advisory group and they're focused a lot on what he thinks of as of buffer bloat. Right? Like the idea of like latencies that are stacking up and the difference between what, you know, is speed test considers to be latency versus when you layer in all the applications and all the different stops along the way and add all that up, and it turns out to not be that great. Like he's asking a question, like, what do you think about that? And how does that play into your vision of like, you know, these future web apps and services. So.

Abhi: It depends which aspect you're talking about there. Right? So like for a speed test, for example, right? Like if you wanna have verifiability of the speed test and be able to tease apart the, different aspects of your instrumentation that you were just talking about there, Jase, right? Like, is it really testing what you want it to test? Right? When you use things like authenticated data streams, you actually know which process is emitting which events, right, because of the cryptographic verifiability of the streams. So now you just have in your, receipt of your streams, your data streams, where each event is associated with which process authored that. So you can then feed that into whatever your speed test is. And then that output can be, cryptographically verified and then encoded in one of these verifiable credentials. Right? So now you have like a certificate that is portable and presentable, right, of the performance of the speed test at that point in time. Right? So now you can do verifiable comparison or performance if you want.

Jase: Whoa, that's really interesting. That's getting into an interesting groundwork that you're laying for enforceable SLAs, and...

Abhi: Exactly.

Jase: Oh my God.

Abhi: And automating that sort of stuff at the infrastructure level. Right? That's how you get to these seamless experiences that we're talking about.

Jase: I see why, Andresen and lots of great folks were really excited about y'all like the even the insurance industry, even the insurance verticals like the idea of like your ability to sort of play a role in assertions that will happen in the future of these digital services like in telehealth it's gonna happen. Like there's gonna be a failure mode somewhere and it's gonna be somebody in, oh my God. Like, y'all being able to like actually help. Right? Sort of evaluate track risk, understand that stuff at every level at every step that's gonna be, man. So let's transition for a second. Like then ask another question. Like how do you come up with the name Mem can you tell us more about, [laughter] why Mem? 

Abhi: Oh man, yeah so, Mem is the Latin based fruit and root for mind and memory.

Jase: Mind and memory.

Don: So short answer is we are nerds [laughter]

Abhi: So if you think about, like memento memorabilia memory like they all have Mem in it and we really liked that notion of mind and memory. Right and so Mem protocol, initially, as I mentioned, we started with the hypothesis of building a better version of Quora, helping you get answers to questions. And it's just fortuitous that the name still works. And in terms of the protocol that we're building now.

Don: And if you think about the analogy we're drawing about making these experiences in these digital spaces that we're building more intuitive and like our physical experiences. So we can actually use our human behavior right? The way we are used to then it's basically like having this ownership over data is kind of like not having your head open. Right? It's like actually having like control over who can see your thoughts. Right. So the Mem thing kind of works there too.

Jase: And you got a sick logo. Can you tell, since we are on the topic of the name, like what's going on there and how did you do that? 

Abhi: So we had a designer who worked with us, for a bit and, he helped us come up with this logo and it's essentially, M to the exponent. So like, it's supposed to be like an exponent M, if you look at it closely, it's like essentially a little, perched there.

Jase: I read it as like a me and I think of the interesting points that, event and lots of great folks we are working on where there's, a sort of there's a need for a permanence to some of the aspects of what we put in the internet and legacies and stuff like that. And it's just, it's really interesting to think about, that right. If you're, able to actually pull this off and, get our identity to be portable and stuff like that you could also maybe be a sort of repository of like what the person wanted people to know. Right? And in age's sense, like me to the power of something else, you know? 

Abhi: It would really be like whatever the user wants though. Right. Like that's the key component here. It's all about the agent and the individual involved or the entity.

Jase: I would do fave cat gifs [laughter], the correct pronunciation of gif [laughter] the recipes, good hiking routes. [laughter] and let's see. We got 10 minutes left actually we got nine minutes left. That's funny how time flies when we're having fun. This is really fun guys. Let's go through the questions like been asked another one just using here, but it's super interesting to think of all the new ways ISPs could serve their customers through web three, improving energy efficiency, speed, general latency, customer service. This will now be part of an equation once providers can better understand and predict with the security. That's not a question. It is not a question. Doesn't have a question mark. It started off like amazing. Okay. With the security offered by the tools to, okay. I'm sorry. What about the legal and economic incentive structures that need to be in place to enable your vision? Right? These are control data, power of different apps. That's gonna be important, these broadband directors and folks. So? 

Abhi: So I mean you hit on some of it with the stuff about SLAs, right? Jase. So like obviously, right? That's kind of where this is going to, right. Once we get some of the foundational stuff in place there. But I mean...

Don: So like, GDPR and CCPA, are our privacy policies, like regulators have done a great job of like releasing those policies and essentially making it so that users can actually pull data from any of the applications that they use. They can create data exports from Facebook. You can, and anyone can do this right now, go try it yourself. You can go to Facebook, Google, Amazon, any one of these services and pull up data export of all of your data. The issue is that even though you can pull that data, you don't really have alternate places to use it, so it's not like you have options in the market of where to go with that data. And then the other issue is that, yes, those regulations are in place, but a lot of these platforms have such outsized power that they can basically comply with the letter of the law, but not with the spirit of it. So like when you get the file dumped, it'll have stuff that is only relevant internally to their systems, so you have to kind of sort through that and get rid of it. So, one of the things that we can get to with the protocol is automated enforcement and encoding of this.

Don: So just like you can have verifiable performance analysis, you can actually have verifiable computation and assertions about whether people are complying with regulation, whether the data flows are actually compliant with the requirements of the law.

Abhi: So I guess the thing that I would urge regulators to do is actually make it a continuous data stream so that these applications are actually... It makes easier for users to actually take this data and make use of it outside of the context of the application that they're exporting that data from. Yeah.

Jase: Okay.

Don: And again, some of these digital services that we're all gonna be working on are things that basically we'll have to work, like kind of like a data type Oracle, so like you've generated this credential or certificate, it has certain data types in it, what are the meaningful operations that you can perform on this? Because now that this data's portable and we're in an open system with the internet, with the web, you don't know where that data's gonna be presented, so you kinda like, "Okay, whatever applications uses in the future here are the kind of global operations." This is the type of operations you can perform on this struct.

Jase: Okay. Dig it. So real quick, our digital community director, Sarah Sterling asked an important question like, "What's the gap here?" What needs to happen besides these important regulatory enhancements that you're describing? What else needs to happen to enable this world? Like, what do you need? 

Don: Once the protocol is in place, if an application conforms with the protocol, then the data-flows themselves will be in this more mutual setup. So it's not like you have to twist a platform or application developer's arm to get them to push data back, because the application itself will also have its own smart agent interacting through the protocol and mediating and negotiating the data flows there.

Abhi: So a couple of events, building on what Don just said, one is, we need to create applications that push data back to user's vaults. These have to be novel experiences that people want, and we're working with a few of those. We haven't talked about them publicly, but we will have a few releases this fall that we're really excited about. So once we launch those, then we will essentially go to, the Web Creek community, who's also building applications that are within this ethos of pushing data back to users' vaults. Once we show the interoperability of that data across multiple places, when you see multiple... Your data across two or three applications, you'll get that magical experience, and ideally, as more applications start pushing this data back, users will start to expect applications to push this data back. And that's what we need to get to.

Don: And from the developer side, it actually gives a competitive advantage, because of the network effects. So if you are a developer, who's trying to provide a service to your customers, if you have no longer to worry about the operational overhead of managing PII, you just get the required data for your business logic that's compliant, and you never have to see what you don't need to see. You get to focus on the actual functional aspects of your service, and that allows you to out-compete whoever is there, if they aren't also participating in protocol. So if you are building a more personalized and intuitive and overall better social network, you now have a chance at competition with some of the existing bigger players.

Abhi: So over time as the value of the data to user's vault increases, other applications like the incumbents Spotify, Facebook, Google, the bigger companies, will actually be incentivized to look at the data in a user's vault, because the user is actually the one person who sees the picture from all the different places that they exist. Facebook never sees Google's data, but the user sees both Facebook and Google's data.

Jase: Yep.

Abhi: And so being able to pull from a vault that has both of those data streams is...

Jase: It's really fascinating.

Don: You're the common factor in all of your experiences.

Jase: That's a richer... That's such a... That's a far richer experience, and there's less hassle, there's less friction, and I can turn to both of those platforms, that example, with a more seamless and fluid interaction between the two, that's a beautiful future, man.

Don: Yeah.

Jase: That's awesome.

Don: Yeah. The other thing to keep in mind is that data doesn't exist in and of itself, people generate data in order to do things, right? 

Jase: Yeah.

Don: And the thing about that is, data has kind of like a freshness to it, like produce. [chuckle] So as more people move on to applications that are compliant or conforming to the protocol, participating in protocol, then whatever data is still on these legacy platforms becomes staler and staler. So that also is kind of like another economic nudge over here as we get more adoption.

Jase: All right. Beautiful. Guys you have a couple minutes left, wanna give a quick shout out to some upcoming Ask Me Any Thing's that we have. August 26th, Nathan Stooke from Wisper, he's the founder of Wisper. He's a visionary WISP builder, kinda the godfather of WISPs in a lot of ways and big part of Wisper, they're industry organization. And he's been working his tail off for quite a few years now, connecting folks in rural America to better broadband. And then after that, we have John Chambers from Conexon, he was former FCC, FCC's coming up in September, make sure y'all tune in and subscribe to Broadband Grants Community events stream, 'cause there's some really heavy hitters coming up.

Jase: And then we have Christopher Mitchell, he's a big pioneer in helping community networks understand the space and helping communities get broadband access. And so there's a lot of really exciting what's coming up, we've got another five that we're posting up next week, announcing some really awesome folks, so stay tuned. One last question is... We're gonna actually go to turn to Drew Clark's last question, he asked a really important question, like "How does Mem protocol compare to Richard Whitt's GLIA foundation and Jared Lane's concept of mediators of individual data. Did you guys work with those guys or? 

Don: Yeah. So I looked a little bit into those projects, so kind of the vision and the recognition of the problem, we share a lot of the same opinions there. One of the differences between us and the GLIA Foundation opinion is we think that the issue is not just kinda at the application layer of the protocol stack, it actually is kind of more fundamental. It's a little bit nuanced, where the issues are there because it's not... It depends on how the thing is actually implemented, like when you look at DNS, yes, in principle, DNS is decentralized, but the registry of names in DNS is by the central authority, IANA. So there's different aspect... I already talk about TLS, there's kinda centralization aspects to that as well.

Don: But yeah, the idea of having sovereignty over and control over data in order to have more agentic experiences online is definitely in line with everything that we were talking about for the GLIA Foundation. In terms of the mids, again, I need to look more into the details of that, but from my little bit of reading through some summaries of what they're trying to do there, it kinda sounds like they're essentially trying to do a data Dow that will functionally be union for influencers. So that obviously could have applications, but what we're talking about is a much more general approach where you just have an automated smart agent negotiating on your behalf, and we're moving towards a more goal-oriented economy for everyone and every agent connected on in these digital spaces, not just influencers and endorsements.

Jase: Okay. Abhi and Don, you guys are a couple of beautiful minds doing something pretty large and important. We're very excited about it. I think personally that it's gonna be a major influencer in the broadband space over the years. If nothing else, it's a reminder that broadband is a means to several ends, and what you're doing with... As you make progress and succeed, you're removing a lot of these frictions, you're promoting security, you're enhancing the experiences you put it. And you're building towards unleashing the full potential of the internet, is how I kinda think about it. And I really appreciate you both making time. I know you're super busy, we're over and we wanna thank you from the Broadband Grants Community, everybody at Broadband Money for all you're doing here. So, thank you.

Abhi: Yes. Thank you.