GRTiQ Podcast: 196 Anna Kazlauskas

The following podcast is for informational purposes only. The contents of this podcast do not constitute tax, legal, or investment advice. Take responsibility for your own decisions, consult with the proper professionals and do your own research.

Anna Kazlauskas (00:00:17):

They experience fear of like, “Hey, what if an AI takes my job? What happens to me from an economic perspective?” And my view is that I think with Vana, you have a system where users actually own the AI models that their data creates.

Nick (00:01:02):

Welcome to the GRTiQ Podcast. Today I’m speaking with Anna Kazlauskas, CEO and co-founder at Vana. A builder at heart who discovered programming through graphing calculators, Anna’s journey spans from mining cryptocurrencies in MIT’s dorm rooms to pioneering decentralized AI solutions. Her unique background combines computer science and economics, including work at the Federal Reserve, World Bank and Celo, where she focused on making blockchain accessible on mobile devices.

At Vana, Anna is revolutionizing how users own and monetize their data in the AI era, drawing inspiration from her Filipino heritage and passion for creating global opportunities in tech. A testament to her innovative spirit, she built her first startup after automating document processing while working at the World Bank. And now she leads Vana’s mission to establish a new paradigm for data ownership in the age of artificial intelligence.

I started the conversation with Anna by talking about her upbringing in Montreal, Sweden, and Minnesota, and how her early curiosity led her to take apart appliances and study instruction manuals, shaping her systematic approach to understanding technology.

Anna Kazlauskas (00:02:19):

I was born in Montreal, Canada, and then I moved to Sweden. My dad’s a professor of biochemistry, so he was on sabbatical and we moved over there, and then moved to Minnesota where mostly I grew up. Really curious kid, I liked to read instruction manuals. I would be so excited when we’d get a new appliance because that would mean that there would be a new instruction manual for me to read and then I could take apart the old one. And so yeah, just builder at heart. Got into programming through graphing calculators, just figuring out you could kind of code on those, and then dove into the world of modeling data through code.

(00:02:57):

The first real programming language I knew was actually Stata, which is this statistical programming language that mostly economists use. And I came across that because I was working at the Federal Reserve since I was very, very interested in just currency, picture of Janet Yellen in my high school bedroom. Just obsessed with the central banks and currency as the foundation of the economy. Those are the early days.

Nick (00:03:24):

I want to ask you this question about reading the manuals and taking things apart. So after almost 200 Episodes here, I’ve interviewed a lot of founders building in web3, and I find it pretty common for builders, when they were younger, this theme of taking things apart to understand how they work. So it seems like you’ve opened the door that this was something that was true for you, but do you relate to that? Was that something part of your childhood of trying to figure out how things work by taking them apart?

Anna Kazlauskas (00:03:52):

Totally, yeah. My dad has a story he’ll tell where in kindergarten they were teaching us how to make sugar cookies and they gave us some recipe and my dad was asking, “Oh, how much sugar do you put in?” I think maybe we were making them at home. And I was like, “Well, it depends on how you want the cookies to taste,” which is a very silly thing I think for a kindergartner to say. But I think trying to understand fundamentally what actually is going on? What is the cause and effect? When you poke one thing, what happens in the system has always been the way that I like to see the world. I think it’s a fun state of existence to just be curious, try to figure out this system that we exist in.

(00:04:33):

I think maybe in terms of how it relates to being a founder, it seems like it’s a trend you notice. I think it’s often when you have that systems-level understanding, you can take things apart, then you can also build new systems and you can put things together in a new way.

Nick (00:04:49):

You said you got started coding quite early, you were working on that calculator and it sparked your interest. Do you remember, and maybe that calculator is the wrong example, but was there a pivotal moment or experience where the light bulb went off of technology is something I’m super interested in, I want to learn more about how this works?

Anna Kazlauskas (00:05:09):

I think it started so early on, my dad was a huge Mac enthusiast and so I would always be excited for when he would get a new computer because that means I would get the hand-me-down and I would literally sleep with the laptop under my pillow. And just ever since I was a kid, loved the computer. Early on, I was playing RuneScape and playing games and early MySpace, figuring out how to edit the HTML.

(00:05:35):

There was a period where my parents were limiting how much time I could have on the computer, so I figured out if I hid the computer under my brother’s bed at night and then I could go over there and take the computer from over there because they wouldn’t find it in my room and get on it at night. But yeah, just kind of integrated. It’s oddly soothing for me just to be around computers, just grew up around them.

Nick (00:05:59):

In addition to that, you recently did an interview with BusinessWorld. You and your co-founder talked about your background and your Filipino heritage and talked a little bit about why that’s important to you and something you pay attention to as you shape your own career and the types of things you work on. Do you mind exploring a little bit about that heritage and why that’s also something that’s important to you?

Anna Kazlauskas (00:06:19):

Yeah. So my mom was born in the Philippines. My dad is Lithuanian, so have a whole blend of different cultures. I went back to the Philippines a few years ago and got to see where my mom was born and where she initially grew up. And it’s like a plot of land in the rural Philippines, which is kind of crazy if you think about it. Now, because my grandparents took the risk of taking us to the US and providing us with more opportunity, just immensely grateful of we got access to really good higher education. We got to have much stronger opportunities and different ambitions that are available to us but might not otherwise be if I grew up in the Philippines instead.

(00:07:04):

And I think too, talent is very, it’s evenly distributed across the world. It’s opportunity that’s not. And so a lot of what’s motivated me and my co-founder, who’s also from the Philippines, is how can you create global opportunity for people to get exposure to AI to this new economic wave that’s forming? That’s some of the foundations from back in the Philippines.

Nick (00:07:31):

This sounds like a trick question, but it’s not. It’s a question about how you perceive yourself given this background you have with your folks. And the question is, again, a theme that comes up all the time on the podcast is do you view yourself as a child of immigrants? And has that been a story about how you frame maybe obligations to opportunity and to your parents and sacrifices? I mean, how have you thought through that?

Anna Kazlauskas (00:07:55):

That’s a good question. Not immediately so. I would say that’s not the core identity, but I do think that anyone who comes from a mixed background or is the child of immigrants learns how to work really hard from an early age, especially on just that instinct, especially among Asian immigrants is work hard. And I think that’s a really good mindset to have. I’m grateful for that mindset and have gotten to pick that up. And my Filipino grandparents are super proud of me and I work hard for them.

(00:08:28):

So I left school and my grandparents really wanted me to finish school because they’re like, “We risked a lot coming over to the US, and now you drop out of MIT? Come on.” And so I actually have graduated on paper and figured out a way to actually get that degree because it was really important to them, and I think too, just worth honoring that.

Nick (00:08:54):

I love that. Well, I appreciate you sharing that. So something you said just a few minutes ago was that in high school you interned, I believe at the Federal Reserve, you had a photo of Janet Yellen up in your room. There’s something there. Talk to us about that experience. And I guess if we’re thinking in parallel, as a young person, you’ve got this interest in tech, but now you’ve clearly got this interest in economics. Talk to us about that.

Anna Kazlauskas (00:09:16):

I think that I always like to just model the world and understand how does this work as a system. And when I found economics, I found it, I think, through reading The Economist and then reading different economics blogs and understanding how economists see things. And I liked it because I was interested in pure math. But what was nice about economics is it gave you a really strong application to be able to also have impact in the world and see if you do really cool math, that can help people and that can come back and actually impact the economy in some way.

(00:09:52):

I think I had seen and learned about Janet Yellen from reading The Economist or something similar and I was like, “Whoa, being chair of the Fed, that is interesting.” And so yeah, just ripped the page out of the magazine, hung it up and it was like, “All right, how do I become chair of the Fed?” These days, I’ve discovered decentralized currency is in a whole different world, and so the end goal is no longer chair of the Federal Reserve. It’s actually like, “Hey, how do we build a new Federal Reserve for data and how do we build it in a decentralized way as well?”

(00:10:22):

But yeah, I think having that role model early on and then also getting to work at the Federal Reserve. There aren’t that many kids super excited about the Federal Reserve, so I’m really grateful to the economist there who took me on and let me do some research for him and learn about how the Fed works. They have in the basement an area where humans aren’t allowed to go, only robots can go there and manage all the money, move it around. It’s just interesting, right? I liked getting to see how the system works, just another way to poke at it and be like, “How does this thing work? What’s going on here?”

Nick (00:10:57):

Did you have any epiphanies? It was when you were younger, so maybe the way you think about the world now clearly would be different than somebody who’s in high school. But did that inside glimpse to how that all worked shape a little bit of your opinion or perspective on centralized banking and the Fed?

Anna Kazlauskas (00:11:16):

Yeah, absolutely. I was interested in the history of currency, how did currency come to be? What does a dollar even mean? I collected currencies from different countries. When my dad would travel, I would have him bring me coins and stuff, so I had a collection of, oh, actually each country has a different currency. As a kid, just really curious about that.

(00:11:36):

I think I also like to look at things where someone might just take it as a given and just assume, oh, that’s how it works, a dollar is a dollar. And I would always want to go deeper of what actually is this? What does it mean? How did it come to be? What might people assume that is actually not necessarily true.

(00:11:55):

I have a bag of shredded dollars somewhere where they get rid of dollars from circulation and shred them. Usually they’re just getting rid of the dollar bill and they replace it directly. But yeah, just hyper curious and interested in assumptions that people might make about currency that might be true or might be not true or people might just not be aware of them.

Nick (00:12:17):

So as you mentioned, you went on to MIT after high school to pursue degrees in computer science and economics. No surprise given everything we’ve talked about. Can you talk to us just a little bit about what your vision for your career was at that time by choosing those degrees. And if you’re willing, is there more that meets the eye here in terms of how economics and computer science might be related to one another?

Anna Kazlauskas (00:12:40):

So initially I went in just focused on economics. I had been interested in programming and coding, but for me it was always a tool and so economics was really the end goal of how can you effectively model the economy and then put different policies in place or other interventions. I was interested in monetary policy to make the economy more successful. Going into school, the default was, become a professor of economics, do research, work my way up to one day become chair of the Fed, and that was the default.

(00:13:16):

MIT has a really good quantitative economics program and such good economists, a lot of them I look up to. David Author, who I did research for, looks at how do you model the effective technology on income inequality and runs different experiments there. My advisor, Bengt Holmström, won the Nobel Prize for contract theory and is really a very deep thinker from an economics perspective. I think the idea that he can just come up with this new way of modeling the world that is super impactful today in business, I think is just so foundational.

(00:13:51):

But then studying economics, I kept on wanting more and more advanced ways to model data. So I wanted basically to push what was possible with statistical techniques, which is why I started to look at machine learning and AI of if we want to model this economy, can you actually use AI as a way to do that? And so then started really diving into AI. Got to learn from Regina Barzilay who’s at CSAIL, which is MIT’s AI lab, do research in the early days of natural language processing, which today is this whole field of large language models and what’s possible with AI models where we can actually generate text. At the time, you could really just classify text and use it to make simple decisions of is this related?

(00:14:38):

I was doing some research around, I had a bunch of transcripts from village meetings in India and looking at are the policies changing people’s sentiment around these different institutions? But yeah, just really cool ways to model data and always trying to push the frontier, and that’s when I also came across decentralized currencies, which has become a really important piece of how I think about the world.

Nick (00:16:05):

Let’s dive into that and not a very uncommon question, in fact, I think I try to ask it in every interview, but can you take us back to that moment when you first became aware of cryptocurrency and decentralized systems, blockchain, and what those first impressions were? Maybe they’re different than what you think now, but can you take us back and tell us about that?

Anna Kazlauskas (00:16:24):

Yeah, yeah. So I was a freshman at MIT living in a dorm, and actually one of the students who also lived in my dorm, who’s super, super privacy-oriented, so I won’t share his name, but he’s super hardcore into crypto. And just to give you a profile of him, this is a guy who lives on Soylent. If you go in his room, he has all these computers and what he’s doing is mining Ethereum. He never wears shoes. I think he doesn’t have a bank account, which is very impressive in 2015 because you could actually convert, I think, Bitcoin to whatever MIT’s cash equivalent was, and then you could just eat in the dorms and in the grocery stores and stuff. And so this dude, he has no bank account, he does not wear shoes, and I’m like, “What?” I’m like, “This is so cool. This is crazy. Who is this?”

(00:17:15):

And so he introduces me to cryptocurrency and he tells me about what decentralized currencies are and how they work. We have a whiteboard and he’s explaining things and I am just fascinated. I’m like, “This is insane. I didn’t even realize you could make something like this.” It’s basically a central bank, but he’s running part of it in his dorm room over there.

(00:17:41):

And so from there, started to really dive in and join the five-person MIT Bitcoin Club, which was this guy who doesn’t wear shoes and doesn’t have a bank account and loves cryptocurrency, and four other people who are very similar to that, super radical. And that’s totally my vibe, I’m like, “Yes, let’s push it to the extreme.” Then started going to the MIT Bitcoin Club, starting this super small group and then also started mining. And so I would go to the loading docks of MIT, which is just their dumpsters where they get rid of stuff and go collect GPUs the labs had gotten rid of and set them up to mine.

(00:18:21):

I actually met one of my really good friends at the dumpsters. We were just hanging out. I went back a few months ago when I was at MIT giving a talk and ran into some students who were going and foraging for parts and stuff. They actually let me borrow their screwdriver and their wire snippers because I wanted to take a motherboard back. But yeah, such a throwback. I mean, it became the sport of it where, I mean, obviously there’s not going to be a GPU at the dumpster every day, so you need to be going by the loading docks a few times a week. You figure out when the labs actually get rid of stuff. You try to time your visit, I don’t know, Tuesday night or whatever it is, so a higher likelihood of getting a GPU and then setting them up to mine.

(00:19:04):

I mean, the labs would get rid of things because they were broken in some way. So typically you’d need to find, okay, the GPU works, but the power supply is broken and piece everything together. Had it set up in a milk crate. My friends thought I was really crazy in school. I had all these broken computer parts scattered throughout my room, and you’re like, “What is she …” They were kind of loud. I was like, “I’m helping to run a distributed central bank. It’s this educational project.”

(00:19:33):

So yeah, that was the beginning. I mean, it was just so much fun, I think, to get to directly participate in the network and really from the very beginning figure out what is this thing? What is possible? And I’m really grateful to the people who helped me learn about it really early on. I got to meet some amazing entrepreneurs. David Vorick, who he didn’t go to MIT, but had joined the MIT Bitcoin Club because the MIT Bitcoin Club, anyone can come by. He was the first entrepreneur I had met, and he had a decentralized storage project that was basically just a hackathon. And he went on to become Sia, but it was a hackathon project, and I was like, “Whoa, that’s crazy. You can start something from scratch and just build it up?” And so then I started to learn more, see that as an option.

(00:20:22):

I think MIT is really comfortable with people, you just do your thing. Everyone just has their own view of the world, a way of doing things, and I think that it’s really good to support that. I lived in a dorm which they closed down called Senior House, which I think was actually really influential to feeling comfortable in more of a counterculture type movement of it was just this crazy place where Anna’s mining ETH in her dorm room. Someone else is, I don’t know, writing an opera. Someone else is proving some math thing on the walls. So yeah, it was a ton of fun there.

Nick (00:21:00):

When was the moment, and maybe it’s happening earlier than I’m postulating here, but when was the moment where it turned away from being sport and experimenting on something new and had some novel interest maybe in it to something more grounded like, wow, this is new tech, this is a web3 type of a thing? Was there a moment where that happened for you?

Anna Kazlauskas (00:21:23):

I almost still see everything as just emerging tech to be explored. There was a moment where I think just from mining, you start to see like, “Whoa, this is really interesting. This is actually real currency that’s being created.” And so I think that is a moment where it shifts from just a project to this is connected across the world. I’m part of this network of miners. I think actually trying to maintain that spirit of almost sport of, oh, what can be built is really important in just the process of building and keeping it fun, staying curious, and not getting too bogged down in the today and instead being like, “Okay, what is the next frontier?”

Nick (00:22:10):

So let’s talk about what you did after university. So is it correct that the first stop out of university was interning at World Bank and having a role there? What can you tell us about that?

Anna Kazlauskas (00:22:22):

Yeah, yeah. So following my sophomore year of school, I decided to go work at the World Bank. I’d worked at a bunch of different banks, so I had been at Federal Reserve over a winter term. I had worked at the European Central Bank. And so the World Bank was interesting to me in that it’s not quite a central bank, but it’s another important banking institution. And they were doing some interesting research around labor markets, which is a lot of what I had focused on, understanding labor economics and those different dynamics.

(00:22:56):

So yeah, went to D.C. for the summer, worked at the World Bank, and one of the things that I came across is the interns around me were assigned this really boring task of sorting documents. I was like, “Oh, you can automate that.” I saw it and I was like, “That’s a machine learning problem. You can actually automate that.” And so I was able to build document sorting software, which did this categorizing of documents. So you look at a given document and you’re like, “Is this related to the environment? Is this related to labor markets?”

(00:23:30):

A lot of organizations in Washington D.C. were using this to basically do accounting and stuff like that of how much are we spending on different projects of different types? It worked very well, and I was able to actually save them a lot of time. I had never worked at a for-profit company or even learned about start … I had known a little bit about startups through my friend who started the decentralized storage company, but I didn’t really know how they work. And so I was like, “Okay, I guess I have this company now, and we’re selling document sorting software.”

(00:24:06):

And so then I was flying to D.C. all the time and honestly missing class a lot my junior fall. Then I got into YC and I was like, “All right, I guess I pack up and move to Silicon Valley.” When I moved out here, I was living in a hacker house with 12 other founders, a lot of … I mean, most of my friends are still in school because I was still midway through undergrad. But I got to live with, I call them almost my startup older brothers of this whole crew that I got to learn from.

(00:24:39):

I think they really looked out for me, showed me the ropes, helped me understand the world of Silicon Valley and are now some of our earliest investors and good friends of mine. That I’m super grateful for, I think. That hacker house experience is a really great way to get started when you first come to Silicon Valley.

Nick (00:24:58):

Anna, I want to ask you this question about the state of the art of AI, because you were working on it at university, you went to work at World Bank, came up with your first startup idea, which is related AI, its machine learning. When you fast-forward from that period of time to present day AI, are you saying, “Boy, people are making much ado about nothing, we were doing similar things earlier on and you’re just now finding out”? Or are you saying, “Oh wow, the advancement of the state of the art has taken us to this point? No wonder people are so excited”? How are you looking at that?

Anna Kazlauskas (00:25:39):

It is definitely the latter, and that’s a really good question. So actually back then, like 2017 is when the paper Attention is All you Need, which is the foundation of the AI models that power ChatGPT. That paper had just come out, and the question we were always asking is like, “Hey, today, you can use this idea of attention to do a better job analyzing text and categorize documents or whatever it is you want to do. But the big question is when can it generate text? When is generative AI going to be possible?” And I remember talking to my professor and she was like, “It’s like five years out at least.”

(00:26:19):

And so now we are, I don’t know, five to seven years out, and it works in an incredible way. And so I think that from a technical perspective, it is really incredible how far it’s come and the models are, I mean, it’s just like, little Anna was looking and was like, “Oh, I can’t wait for that to work.” And now it works and it’s honestly better than I think some have hoped and dreamed. So really impressive technical milestones that have been achieved within AI and the early signs of it were there in terms of the research.

Nick (00:26:53):

So returning to your story then, we’re in 2008 period here, you were working at World Bank, you applied some of your university learning with machine learning to a problem they were having reading documents. You started your first startup, Iambiq, if I’m getting the name right there. Do you mind just telling us about that experience, what you learned about startup land and some of the key lessons that this first experience taught you?

Anna Kazlauskas (00:27:22):

Yeah. So super early on, I had never worked at a for-profit company, incorporated a for-profit company. And so there was all sorts of stuff where I was like, “How does this even work?” And so I think that what was really valuable about going through YC is getting to be alongside a bunch of other entrepreneurs. I was 20 years old at the time, just learning and growing and understanding. I remember, I think they had some events where they were at a bar and I wasn’t 21 yet, and so I couldn’t go.

(00:27:53):

And so I was really just getting started in there, but I got to meet a bunch of other founders. And I think one of the most valuable things too is getting to see that play out over time. Because at the time, during the YC batch, there were companies that were the hottest in the batch. There were companies that were perceived as not doing as well during the batch. And then you play it out over five years and it literally doesn’t matter how hot you are as a YC company in the batch, right? It’s actually, I think, often the underdogs that do really well.

(00:28:22):

I think in our batch, the big winners were Vanta, which is the data security SOC 2 company, Replit, Substack was in the batch. And at the time, I mean all of these ideas were pretty early. With Substack, people were like, “Paid newsletters, what the heck?” And now we look back and we’re like, “Oh, they created this whole new category.” Generally for any founder looking at doing YC, it’s just a super valuable experience in getting a bunch of training data and understanding.

(00:28:53):

And then for building out the document sorting company I was working with, honestly, I had chosen the wrong person to work with. I was 20 at the time, and it was like, we’re going to need a business person. And so, one of the main criteria I was looking for in that co-founder was someone who had worked at a company before, which now if you look, you need much more than that to be the right co-founder. But early on when you don’t know anything, I was like, “Well, yeah, she’s worked at one. She probably knows how to incorporate one too.” And so I think honestly, it was just super early, but super grateful for that experience.

Nick (00:29:34):

So returning to that question, but asking it slightly differently because it’ll be fun, your brain, as you’ve described it thus far, likes to take things apart. You like to read manuals as a young person. You like to figure out how things work. Do you apply that lens to startups or entrepreneurship? What’s the one observation that you could give to listeners as a form of advice of I’ve done this, I’ve taken it apart. Don’t start, or before you get started, make sure you do this thing. Is there one thing?

Anna Kazlauskas (00:30:11):

At the time I didn’t realize how long-term of a commitment starting a company is, where if you really want to build something that is generational and has a very large impact on the world, it’s probably going to take 5 to 10 years. It’s probably closer to 10 realistically. And so I think that I had almost stumbled into this document sorting thing, and I think as I started to understand the level of commitment involved in building an organization, it is massive.

(00:30:41):

And so you really want whatever you’re building to be your life’s mission, your life’s work, and really be willing to dedicate a huge chunk of your life. I’ve probably spent 15 to 20% of my existence thinking about how do you get people to own their data? So you want a problem space that you care that deeply about.

(00:31:02):

So yeah, I guess just finding a problem that you’re extremely motivated to work on, almost to an irrational point because I mean you’re willing something into existence. It doesn’t exist. It’s going to be very hard. And so it really needs to be very worthwhile and motivating for you in order to actually achieve what you’re going after.

Nick (00:31:22):

Well, after Iambiq, you join Celo as an early engineer focused on making blockchain accessible on mobile. Longtime listeners of the podcast know I’ve featured an interview with Marek before, another MIT alum, on this podcast, and really a thrill to have met Marek and go through that story. What can you tell us about the backstory here, how you ended up at Celo and why you made that move in that direction?

Anna Kazlauskas (00:31:46):

So I was at this phase where I was figuring out, I was like, “Okay, I don’t want to commit my life to sorting documents.” And so I was like, “Okay, this is not …” But I wasn’t sure, okay, do I start something new? Do I join a project? Do I go back to school? I was back in Boston and one of my friends was actually doing some consulting work for Celo super early on, and I was known as a crypto person. And so he kept asking me these crypto economic questions and I was like, “What are you working on? What is this thing?”

(00:32:18):

And so then I met Rene and Marek and just chatted. I was like, “Oh, this is kind of interesting.” Because I had actually spent some time in the Ivory Coast explaining crypto to regulators from the MIT Bitcoin days and trying to have people understand crypto and how it can be an alternative to centralized currency. And as part of that, I had met someone from Zimbabwe who showed me a picture of literally Zimbabwe dollars in a wheelbarrow. I was like, “That is crazy. You guys definitely need cryptocurrency.”

(00:32:49):

And so I think that I saw a really strong need for crypto and stable payments, especially in emerging markets. So chatting with Rene and Marek, got really interested and just started working with them for a month or so in the early office, which was actually in the Tenderloin. It was tiny, we work in the Tenderloin. A few of us stepped into a room and just loved it. Felt like I was learning a lot, felt like it was a really strong, really ambitious problem space to go after of can you create a stable store of value that works on anyone’s phone so that anyone can have a bank account even with a low end Android device?

(00:33:32):

And so yeah, decided to join the team, thought really highly of Rene and Marek who are almost like my entrepreneurial uncles at this point, have shaped the way. I think Celo is also very mission-driven and has attracted a really great group of people, so gotten to learn a lot from them too.

Nick (00:33:49):

That’s definitely something I remember from that interview with Marek, the mission-driven nature of what he was working on at Celo and the vision that was driving him personally. I want to ask you this question about cryptocurrency. You talk about meeting with people in Zimbabwe. You talk about getting mission aligned with Celo and the problem they were working on.

(00:34:09):

In the United States, a lot of the talk around crypto is really related to speculation and asset class, if you will. Outside of the United States, it’s something far different. And I’ve interviewed a lot of people that give evidence to the fact that it’s a currency, it’s a hedge against inflation, it’s a way out of poverty, it’s a way to fund a business, it’s a way to sell goods and services.

(00:34:33):

So the question is, do you see that in the work that you’ve done to this point where the adoption and utility of cryptocurrency outside of the US tells a different story than maybe we’re hearing inside the US?

Anna Kazlauskas (00:34:46):

Yeah, I think that crypto offers a whole new paradigm of building technology. And so you can build stable currencies that don’t require trust in any single government. You can also build just broadly decentralized systems where for Vana, it’s really about giving users ownership of their data and the noncustodial nature of crypto where what it means for you to own something is to have it in your noncustodial wallet. You have the key that permissions it out.

(00:35:18):

So I think that it’s still so early in terms of where crypto as an industry is headed. There are lots of early applications that are still finding their footing. And it’s crazy because if you look five years back, we’ve come so far. But also it’s still super early, so it’s important to remember that. And I think the core primitives that crypto introduces really allows for just new kinds of applications. With Vana, it’s about how do you have these almost labor unions around data, these data DAOs that can negotiate, they can choose to train AI models.

(00:35:56):

And what’s important about these data DAOs is that your data on its own, it’s not really valuable unless you combine it with other people’s. You need to have enough data that you can actually train an AI model. And so I think that crypto as a primitive offers many different interesting tools that allow for a new user-owned internet, user-owned economy.

Nick (00:36:22):

Okay, so let’s then turn our attention to Vana. And if then your first startup Iambiq came out of an observation or an experience you had at the World Bank, in 2021, you got inspired to launch Vana and went to work on it full time. What was then that catalyst observation or experience you had at that time in your life that fed this idea of maybe I’ll try and start up again?

Anna Kazlauskas (00:36:45):

So what I had seen from working with AI models really early on is that ultimately data is the main differentiator. Any AI model, it’s just data. That’s what it just comes down to. There’s some training code. I mean, there’s a whole bunch of stuff that goes into the process, but in terms of what at its core is this thing, it is something that has captured the patterns that were available in its training data.

(00:37:14):

And so then the question becomes, okay, in a world where AI is really valuable, which is the trajectory that I think I had started to see, where are you going to get all that data to train AI models? And how will you be able to use a decentralized system so that everyone controls that all-powerful AI model that gets created? And so that’s the early thinking around Vana.

(00:37:38):

Super early on, we were doing on-chain data labeling where we’re having people label data, earn for their contributions, and really finding ways to just bring many people into having exposure to AI. And also allowing AI researchers and folks who are training AI models to access data sets that they really need to train. So yeah, that was the early days.

(00:38:02):

I would say honestly, we were a little too early to market where people were like, “What the heck are you talking about?” I think that ChatGPT hadn’t happened. There wasn’t a broad awareness of even the connection between AI and data. Today, when you talk about data and AI, people are like, “Oh, AI learns from data.” That didn’t exist three and a half … And people were like, “What is this thing?”

(00:38:24):

We were quite early, but I think that we really had conviction on the space of at some point there’s going to be a time when data is almost like the new currency or this new asset class in an AI-native world. And so how do we build an economy around that and the system where you can actually interact with data more like a currency than how it works today, where typically data just sits inside of the walled gardens of the different big tech platforms that people use.

Nick (00:38:57):

So if we spring forward to today, you had that original thesis. You were probably too early, but you got to work on an important problem early. So that’s an advantage to you and the team. But as we spring forward to today, can you just reset the vision for Vana, the problem set it’s working on, and how it works presently?

Anna Kazlauskas (00:39:20):

So pretty much exactly the same vision and now we have just built it, which is the big difference. It exists and there are real people using it and real builders building on it. What Vana is going after is how do you create a world where users own their data and the value that it creates. That’s a little bit abstract, so to become even more precise, it’s really about almost turning data into an asset class where there’s more liquidity. So you can have different people combine their data together to power new kinds of products, specifically AI models. That’s what most people think of today. So that’s a really important component of Vana.

(00:40:00):

The other piece I’d call out that you get with data ownership is data portability. So when you log into an application today with your crypto wallet, all of your funds are there with you. And with Vana when you log in, all of your data is there with you too. So rather than being in a world where each company holds a copy of your data for you, instead you hold that copy and you bring that with you throughout the internet.

(00:40:27):

There are a few projects that really pioneered this, Solid project, which is based out of MIT started by Tim Berners-Lee, who’s the original creator of the internet, and Urbit were the first to introduce this model. But broadly, it has failed to get adoption. And so I think what we’ve really cracked at Vana is both having the pure federated system where users host their own data, but also being able to tie incentives to it. With the Reddit Data DAO, they saw 140,000 users who connected and contributed their data.

(00:41:00):

And you’re actually seeing real adoption because Vana allows you to build a new kind of business. It’s sort of like a labor union for data is what I would call a data DAO. And that is basically driving adoption of bringing data onto the platform, and then anyone can access it if they get the permission. So you can train different AI models. We can also just build an application that uses a user’s data when they come in.

(00:41:27):

From a technical perspective, Vana is a layer 1 blockchain that is designed to work with private data, but for most people, you don’t have to think of it that way. So for a builder, it’s a way to build a new kind of application with user-owned data. So a new way of interacting with data. For users, often actually a user will interact with a product that a builder has built. So they’re like, “Oh, Vana is just what the 23andMe DAO is built on top of.” And then for someone who’s consuming the data, so like a machine learning researcher or a business, Vana is a way to get data that would typically be stuck inside of a walled garden.

Nick (00:42:06):

Everything you’re describing about how Vana works and obviously a really important problem as we try to mature AI tech to something that’s more useful, everything you’re talking about there makes me feel like the time is now, like this technology and this problem you’re working on must be happening now because you’ve got this marriage of cryptocurrency and incentives, and then you’ve got the emergence of AI. So is it serendipitous that these things happen side by side, or do you think these are two tools that smart people brought together to solve important problems for the world? How do you think through that?

Anna Kazlauskas (00:42:41):

It’s a very philosophical question of almost how did the world come to be? I think that the two core technologies are in some ways very ideologically different. AI is actually naturally very centralizing because it requires amassing a huge amount of resources to train models. They can cost upwards of $100 million to train today. I think that’s probably going to go into a few billion dollars, which is just insane from the perspective of compute data and research.

(00:43:12):

So AI, deeply centralizing, and then crypto is the polar opposite, which is like, hey, this is naturally a distributed network, naturally a decentralized network. And I think that they work really well together because they actually balance each other out. Where one has a tendency to be more centralized and potentially concentrate power too much, the other has a tendency to almost be too distributed. If you think of a DAO where it’s like maybe the stereotype is nothing gets done. It is actually quite good that they balance each other out.

(00:43:47):

And so yeah, I think that the timing, it’s just right. And in terms of what the core driver of that timing is, it’s hard to say. It’s like would this have happened in a thousand different iterations of the world? Deep philosophical questions, but in the current state of the world, absolutely, they fit together in this really complementary way. And I think they offer a path towards having really powerful super intelligence where you have an AI model that can produce real economic value.

(00:44:22):

I mean, today, OpenAI does $3.4 billion in revenue. That’s growing, but I think that over time, that’s just going to grow even more in terms of revenue that comes from AI models. And so the amazing thing that crypto and decentralization offers is rather than having one all-powerful AI model that rules us all and I think a lot of people are scared of, we can have an AI model that’s controlled by everyone and a healthy ecosystem where different AI models as they’re used, the users are actually getting paid out. It’s even really beautiful from an artistic perspective of what’s been possible with the two technologies

Nick (00:45:04):

I had the opportunity to interview during Episode 171 on this podcast, Dr. James Hendler, maybe a name you may or may not know, but he worked with Tim Berners-Lee on publishing that very influential paper on the semantic web. And my question is, are some of the things that you’re working on and some of the things that we can do in web3 decentralized systems, is it bringing that semantic web vision to life where web2 probably couldn’t do it or failed to do it?

Anna Kazlauskas (00:45:35):

Yeah. So I’m most familiar with Solid project, and I think that it actually exactly is where early on it said, “What could it look like if users own their data? What could it look like if a user could …” I mean, early on at the MIT Media Lab, there’s a hilarious video I’d actually recommend watching, where they basically thought every user would carry their data around with them physically. So you would have, straight up, a hard drive and they style it, and you have all these nerdy engineers. It’s like a hard drive strapped to your belt, and that’s your data. You bring it with you.

(00:46:09):

And I mean, now the world looks totally different. In fact, the default is that your data is stored in the cloud on a server that is the company’s, not yours. I think that a lot of it is just economic forces at work. So there are some incentive to have driven that centralization, and generally that’s what things tend towards. They tend towards centralization unless you take active steps or use technology to break things up.

(00:46:36):

So with the aspect of bringing crypto in, I think that crypto has introduced these really powerful incentive systems where you’re able to actually build a new kind of economy that might not be possible in a centralized paradigm. And so that is what has really unlocked decentralized AI, the ability for users to bring their data with them and give them a reason to actually aggregate their data together through different data DAOs.

Nick (00:47:06):

Anna, as, again, someone that’s non-technical, I understand that Vana is doing some exceptional things in terms of getting user data available, train models, and the way it creates value. My question is about the structuring of that data. As a non-technical person, I’m envisioning just this endless ocean of data. I see builders trying to build on it, but if it lacks structure, can it work? And I guess I’m hearkening back to that vision for the semantic web a little bit with those standards about the way data should work. How do you think about that?

Anna Kazlauskas (00:47:43):

Yeah, that’s a really good question. The ability to work with data depends on knowing the structure that you’re getting. So actually each data DAO on Vana defines the structure of their data and is validating when a user comes in, what is the format in which they’re contributing their data? So like comments, messages, photos, audio files. That’s all defined at the data DAO level. And what that allows for is this known structure that then a machine learning engineer or application builder knows when the data is coming in.

(00:48:18):

That said, I think that as AI models become more and more advanced and as AI agents are doing a lot of the interactions on behalf of the user, you can get more flexible on structure. It’s actually okay if some of the data is messy. And so I think that we’ll see that shift as well.

Nick (00:48:35):

Clearly, there’s going to be listeners listening to this interview who want to get more involved with Vana, figure out what’s going on and how can they participate and how can they learn more. Given where you’re at in the cycle here of building this, how can listeners get involved? What’s a way to participate?

Anna Kazlauskas (00:48:51):

So if you want to dive in really deep and you’re a builder, come and build a data DAO. Vana is a permissionless network, so you can get started. If you go into our docs, there is a three-step guide to starting a data liquidity pool, which is the more technical term for a data DAO. Essentially what it’s referring to is being able to pool different data. Data DAOs are a subset of data liquidity pools because not every data liquidity pool actually operates as a DAO. Some of them just use a stable coin or something else. So if you’re a builder, dive into the docs on vana.org.

(00:49:27):

If you’re a user, a great way to get involved is to try out some of the different data DAOs that are building on the network. And so if you go to our website, vana.org, there are a bunch of different links to the applications that you can try out. And then we’ll also be releasing a data hub in the next couple of weeks actually, where you can easily explore a whole bunch of the different data DAOs, interact with the network.

(00:49:52):

And then if you’re a machine learning engineer and you want to train an AI model, I think getting in touch with specific data DAOs. There’s also a company called Tensor Source, which basically if you’re coming from a web2 background and you want to access the data, but maybe not think about any of the web3 aspects, they’ll make it really easy for you to be able to see that data. Following us on Twitter @withvana and checking things out. Those are the different touch points where you can get involved.

Nick (00:50:20):

Anna, as you may or may not know, a lot of my listeners are enthusiastic about web3 data and The Graph. Because of the origins of this podcast within The Graph ecosystem, a lot of people have been following the growth of that project and meeting builders that use and understand the importance of decentralized indexing and querying layer, like The Graph. As a non-technical person, and again, a lot of my listeners probably being the same as me, can you help us understand how indexing all that data and I guess ensuring that it comes from decentralized sources to a project like Vana, how are you thinking through all of that?

Anna Kazlauskas (00:50:57):

That indexing piece is really important. How do you allow for data to be discoverable on the network and actually have some structure to it as well? I believe there is, I’m not sure exactly what the status is in terms of The Graph and Vana, but I know that there is some kind of discussion, brainstorming happening over there to figure out how best to leverage The Graph in order to make data more query-able and more easily available. So excited to see where that goes.

Nick (00:51:29):

It’s weird to me, and maybe you just won’t even take the argument, but it’s weird to me that in web3, there still is a debate about how important decentralization is. And I understand the argument from the perspective of the market. Do devs really care? Do users really care? But it seems to me odd that builders are still arguing this. Where do you come in on that. I mean, do you see it as well that there’s a debate about this, and do you have really strong feelings about no, that’s a non-negotiable. This thing, the stack and the tech has to be decentralized?

Anna Kazlauskas (00:52:01):

Yeah, I mean, I think if it doesn’t have to be decentralized, then you can use a web2 database. So I think that I fall pretty strongly on the decentralization perspective. I think that often what works well for crypto is you can build in that resiliency. Also, there are aspects of regulatory arbitrage that come from decentralization. In the case of Vana, personal data is actually highly regulated. So a lot of people don’t realize you actually legally own your data, in the same way that when you park your car in a parking lot, the parking lot doesn’t own your car. When you put your data in a platform, it’s still legally yours. So because of that, you can always go to a platform and take your data back.

(00:52:41):

But if Vana were a centralized company, then that wouldn’t work. So you actually need that decentralization in order to be able to leverage individual data rights to build a system like this and make it possible. That’s in Vana specifically, but more broadly, yeah, I guess just a deep belief in decentralization. I guess the properties of decentralization that are important, I think you have sovereignty, true self-sovereignty and censorship resistance. So I orient towards more of those properties that you want to achieve, and then decentralization is one way to get there. For Vana, it is the main way to get there.

(00:53:20):

But I think that there are also technologies that are self-sovereign and censorship resistant that might not require necessarily a blockchain, but it could be more of a distributed system like the local large language model communities who are running AI models on their computer. They’re not using a blockchain, but they have self-sovereignty, they have censorship resistance. And so I always orient towards what are the core properties you want, and then how do you use these different tools to get there?

Nick (00:53:48):

Anna, I want to ask you this question about your approach to learning. So you were successful at university, you’re still going to complete that work. You’re obviously successful in career, you’re working on some very knowledge base, you got to learn a lot to succeed. Are there any mental models or lenses you use when you approach a subject, something that you’re trying to learn that might be helpful for listeners or someone like myself to want to have the same appetite in learning that you’ve accomplished?

Anna Kazlauskas (00:54:17):

Yeah, I think very much by analogy. Actually, often people come to me and ask, can you figure out what the analogy is for the business they’re building of like, “Oh, it’s, I don’t know, Infura for mobile or whatever the analogy is.” So I think that being willing to lean on analogy, even if it’s not perfect, just because it’s helpful and then you can think through how is it similar, how is it different, how does it fit? That’s the number one thing I’d say.

Nick (00:54:48):

Yeah, and I’ve already seen you use that in some of your answers when you were talking about the labor union or you talked about the parking lot. So you’ve already shown how you use that. I want to ask you this question about motivation then. What motivates you as just a person? I know you’re certainly motivated about your startup. You’ve got this great idea, but is there a drive, a vision you have, something inside you that drives you?

Anna Kazlauskas (00:55:14):

Yeah, I think that there is this very strong conviction around the need for user-owned internet, which is what gets me off in terms of okay, how do we make this happen? How do we actually push this forward? I’m generally drawn towards hard technical problems that have very large scale impact if you can make them work. So some of it is curiosity. Some of it is just, I don’t know, it’s innate drive, really just hard work. I just have constantly this energy to put towards something and always want to be pushing something forward.

(00:55:51):

I think it comes from this place of really seeing what could be possible and being able to see that future and then ensuring that you guide the present towards that. And that’s the activity of building. It’s like this is the future we want to get to, and then how do we guide the present there?

Nick (00:56:11):

Are you afraid of failure?

Anna Kazlauskas (00:56:13):

No. I mean, I think that it depends. You can be tied to a specific outcome of this is what it should look like. But it’s interesting where for me, it’s like you see the state of the world and you almost know we’re going there and it’s just a question of how you get there as fast as possible. I think there are times where it’s like, oh, there’s this specific thing and we want it to go this way. But yeah, I mean, you can only control what’s in your direct control.

Nick (00:56:42):

Well, Anna, I only have a few more questions for you before I ask you the GRTiQ 10. These are 10 questions I ask each guest every week. They’re a lot of fun. Gives us a chance to get to know you a little bit better and hopefully gives us some ideas of things that we should check out or get more involved with. So first question I want to ask you is about AI. I mean, you’re tracking it from university to present. If you project out a little bit, what’s the next big thing you’re watching for as maybe outside of even the work you’re doing as just an enthusiast about this tech and what it could mean for the world? Is there a milestone or a mile marker that you’re watching for?

Anna Kazlauskas (00:57:18):

Yeah, I think that just model performance for LLMs, it just keeps getting better and better and really starting to see it be good enough that you can have agent-like behavior where agents are able to operate on their own and guide some of their decision-making. I think that that’s going to be a really big milestone. I think audio and video models are getting better and better. They’re still kind of clunky, still very expensive to work with, but I think that over the course of a year that there are going to be massive improvements there. I mean, it’s just honestly so great. It’s an amazing time to be alive and get to experience AI progress. Week over week, there’s something that comes out and it’s like, “Wow, this is so impressive.”

(00:58:02):

Yeah. Another thing I track pretty closely is model architectures. As AI models get better and better, are they going to follow a transformer model or are they going to be an SSM, these different model architectures. That’s more of a niche technical point I just find interesting. Another thing I track that I think is a big open question in the industry is whether most models are going to be run locally on people’s devices. Is it going to be what Apple is pushing towards where you have an AI model on your device that can work with your data and a lot of it stays local, or is it going to be something that is in the cloud? And I think that has a lot of economic implications. Basically, how much GPU demand is going to come from AI inference? And how much of that then is going to be in the cloud or is it going to be local?

Nick (00:59:00):

Anna, I want to ask you this question, and I think I’m going to fumble through it, so I’ll do my best to ask it, but it’s this question about the AI market and the types of problems and the types of things Vana is working on and others in that competitive set. And the question is, we know for sure there’s a market for AI and then we’ve got this decentralized flavor of AI, the types of things that you’re working on.

(00:59:26):

And so my question is, is that existing market resistant to this flavor of decentralized AI that’s emerging? Is there a segment of the market that’s highly attracted to that? Are you hitting up against that wall? I mean, help us understand, I guess, the market for decentralized AI and this broader discussion of this amazing rocket ship we’re watching with people using and wanting AI.

Anna Kazlauskas (00:59:52):

Yeah, I think that what’s really interesting about decentralized AI is many of the builders in the space are actually new to crypto. Some of them are AI people, some of them are engineers, but not crypto people. And I think that if you talk to even your parents or your grandparents, the idea of everyone wanting to control AI and collectively owned AI and decentralized AI, that actually really resonates, I think, with a broader user base and demographic than some of the more DeFi oriented products.

(01:00:30):

So what’s been really cool to see, and it’s still quite early, I think the general view of the open source AI community is still skeptical of crypto, but over time, you’re seeing more and more people come over. And one thing that’s really facing the AI community today is this data wall where we’re actually running out of data to train AI models on. And so because of that, I think people are more open to different decentralized approaches, including Vana, which allows for data to come out of these walled gardens where usually it would be stuck and then accelerate AI progress forward.

(01:01:06):

And that’s one area where the incentives of decentralized AI and user-owned technologies are actually very aligned with just AI progress. If you want to build better AI, you need more data. Good way to get data is to have users own the AI models that they contribute to. That’s this really cool piece where you’re not relying just on ideology of telling someone, “Hey, AI should be decentralized, so use this product because it’s more decentralized.” But you’re actually putting something net new in front of a user that says, “Hey, here is an AI model that’s only possible because you’ve contributed your data to it and because you’ve made it better.” And so I think that’s one of the really great tailwinds that’s helping decentralized AI right now.

Nick (01:01:50):

In another interview, you’ve said that AI might take your job, but you’ll own that AI model. Can you help us understand that? That’s a little bit of a twist on this mainstream fear that AI is going to take people’s jobs, but you have a little bit of a optimistic twist there, if I understand it correctly?

Anna Kazlauskas (01:02:06):

Yeah, yeah, absolutely. I mean, I think that as AI gets better and better, I mean, you can imagine I can’t wait for AI Anna. I want 10 AI Annas I can just deploy to go and operate as needed. But what’s important is actually having that ownership of that model. I think a lot of the time people think of AI and they experience fear of like, “Hey, what if an AI takes my job? What happens to me from an economic perspective?”

(01:02:34):

And my view is that I think with Vana, you have a system where users actually own the AI models that their data creates. So you align incentives much better, right? Then rather than saying, “Hey, I want to strike against AI because I see it as competition.” You see it as more complementary to you as a way to scale yourself.

Nick (01:02:57):

And then finally, Anna, as an expert on AI, somebody working in the space building really innovative things, maybe a bad question, but I still want to ask it. Which one do you use? I mean, I imagine with all the different things out there in the marketplace, there’s some competitive advantage to, I guess I’m familiar with Claude, I’m familiar with ChatGPT, obviously there’s Gemini. Is there one that you use? Maybe you’ve got insight that we could all learn from?

Anna Kazlauskas (01:03:23):

I use Claude by default. I can guide it in the way that I like, and then I have some local models that I run as well. I have little setups and experiments that I’m running, but my default is Claude.

Nick (01:03:37):

Well, Anna. Now we’ve reached a point where I’m going to ask you the GRTiQ 10. These are 10 questions I ask every guest, and I do it because first of all, we want to learn more about you, but I always hope the listeners will learn something new, try something different, or achieve more in their own lives. So Anna, are you ready for the GRTiQ 10?

Anna Kazlauskas (01:03:51):

Yeah.

Audio (01:03:54):

The GRTiQ 10.

(01:03:55):

This is the way.

(01:03:55):

10 questions for astronauts floating in space.

(01:03:55):

GRTiQ.

(01:03:55):

This is the way.

(01:03:55):

Roger that.

(01:03:55):

This is the way.

Nick (01:04:05):

What book or article has had the most impact on your life?

Anna Kazlauskas (01:04:09):

I think just reading appliance manuals as a way to understand the world and understand step-by-step, how different things work and how people choose to explain things. I realized that is not that interesting of content to dive into, but for me as a kid, just reading some appliance manuals, pretty interesting.

Nick (01:04:29):

Is there a movie or a TV show that you would recommend everybody should watch?

Anna Kazlauskas (01:04:33):

One of my favorite things to watch is I’ll look for really old YouTube videos of entrepreneurs and founders giving talks. You want things that have sub-5,000 views, and then you can watch them evolve as a founder over time. So it’s my favorite content to consume via video.

Nick (01:04:53):

If you’d only listen to one music album for the rest of your life, which one would you choose?

Anna Kazlauskas (01:04:57):

I would probably choose something relatively neutral to be safe. I mean, maybe I’m being too strategic about the game of you only get one, but I’d probably choose some kind of meditation soundtrack. If everything else is going to be taken away, I’m like, “That’s probably the safest bet I would want to say.”

Nick (01:05:17):

What’s the best advice someone’s ever given to you?

Anna Kazlauskas (01:05:20):

Okay, this is kind of counterintuitive, but it’s to not listen to other people’s advice. And I think that it took me a long time to understand where that comes from. And I think a lot of it is actually just that most people don’t have the full context. And so advice taken out of context can sometimes be more harmful than if you as a founder had just figured things out from first principles. But yeah, I guess just being very mindful about how you take advice, not as, oh, this is what I should do, but as a data point that you should take in as an input to your end decision.

Nick (01:05:53):

What’s one thing you’ve learned in your life that you don’t think most other people have learned or know quite yet?

Anna Kazlauskas (01:05:57):

I think that realizing that literally everything is just a skill that you can learn. I think that even learning how to give talks and how to talk to people or how to be a good friend or a good partner or a good manager. I think there are some things where people think of them as innate traits, and there are other things where it’s like, oh, it’s a technical skill, you can just learn it. Like learning to code, no one says, “Oh, I was born bad at coding.” Whereas sometimes people are like, “I’m just not good with people.”

(01:06:27):

But I think that the thing that I’ve learned is literally any skill you can learn, you just have to sit down and figure it out and be willing to try and mess it up and then just get better over time. So I think just seeing basically every skill as something that is learnable.

Nick (01:06:45):

What’s the best life hack you’ve discovered for yourself?

Anna Kazlauskas (01:06:49):

Sleep is really important. I think just really prioritizing sleep is so critical. And it’s hard as a founder. I think that there are times when you’ve got to find ways to squeeze out an extra hour, and sometimes that comes out of sleep when absolutely needed. But I think very, very aggressively prioritizing sleep has been super important to just making sure that you’ve got the right perspective and you’re thinking clearly.

Nick (01:07:16):

Anna, based on your own life experience and observations, what’s the one habit or characteristic that you think best explains how or why people find success in life?

Anna Kazlauskas (01:07:26):

I think it’s just persistence. I think generally it’s basically just how badly do you want it? And I think it just comes out like okay, if you really, really want something, if you really want to build towards this world that you want to create and you’re going to be relentless about it, like insanely persistent, then probably it’s going to work at some point.

(01:07:46):

So I would say just, you should do things that you really, really want to do, almost to an irrational extent. You should just believe that it is absolutely critical that you do it. If it doesn’t happen, it’s going to be a huge failure, you really need to make it work. So yeah, just desire and persistence is generally very correlated with success.

Nick (01:08:08):

And then the final three questions are complete the sentence type questions. The first one is, the thing that most excites me about the future of web3 is?

Anna Kazlauskas (01:08:16):

I was trying to think of a single word, but I think it’s broadly as we bring more and more people on chain, imagine we’re able to bring a billion people on chain and then train better and better AI models with everyone. And starting to have people think of web3, not just as a technology, but more as a new way of having true ownership and a new way of being able to build economic primitives or different kinds of business models where you’re actually an owner and a stakeholder in the products that you use.

Nick (01:08:50):

And the next one, if you’re on Twitter or X, whatever people decide to call it, you should be following?

Anna Kazlauskas (01:08:55):

I really like this blog, Marginal REVOLUTION, which Tyler Cowen writes. It’s an economics blog. They tweet their posts when they go out. So that’s a good one to follow.

Nick (01:09:03):

And then, Anna, the final question. I’m happiest when?

Anna Kazlauskas (01:09:08):

I’m building, just building something. And building takes many different forms. It’s writing code, it’s shipping, it’s helping other people ship. But yeah, just happiest when I’m building and creating.

Nick (01:09:30):

Anna, what a thrill to interview and welcome you to the GRTiQ Podcast. Your story’s not only inspiring, but incredibly educational, and it makes me optimistic about the future and what we might see in terms of data and AI. If listeners want to stay in touch with you, follow your work and the things you’re working on, what’s the best way for them to stay in touch?

Anna Kazlauskas (01:09:49):

So on Twitter, you can follow @anna_ kazlauskas, just K-A-Z, Kaz, that will find what you need. And then you can find Vana’s Twitter @withvana. That’s the open source protocol. You can also find Open Data Labs, which is the company that’s contributing to Vana, and so those are the different Twitter handles.