Balaji Srinivasan is the former CTO of Coinbase and General Partner at Andreessen Horowitz, and is an early investor in many successful tech companies and crypto protocols. He is also the author of the book The Network State: How To Start a New Country, set for release July 4, 2022.
This podcast was recorded on February 26, 2022.
Shane: Balaji man. I think the rest of the world is going to tell us never to talk again. The last time we talked, it was right before COVID and now we got the Ukraine/Russian thing going on.
Balaji: Well, I think that the world is going to be an eventful place over the years to come. So, probably every podcast will be a few months before something crazy. Correlation is not causation. This is not investment advice, various other incantations, perhaps with that we can begin.
Shane: What’s the most interesting or surprising thing you’ve learned or changed your mind about in the last three months?
Balaji: Last three months? I don’t know, but the last year I actually wrote an essay on this for Barry Weiss’s blog. And essentially what was surprising to me was that the reduction in state capacity was bipartisan, in the US. And that is to say under Bush or under Obama State capacity had been moderately high enough to do, even if you don’t agree with it, the invasion of Iraq is a logistically complicated thing. Even operations in Syria or there are domestic actions, foreign actions that were taken that required some degree of coordination. Under Trump there was a huge, visible decline in State capacity under Biden that seems to have continued. Why is that surprising to me? Last summer I had this little tweet, which is a little bit surprising, but I said, the Federal Government is losing control of affairs, both domestic and international. Inflation, Afghanistan, China and Taiwan, France on AUKUS, El Salvador and Bitcoin, 17 states pursuing the Feds on education, Texas on abortion, vaccines, many others and of course, obviously now the Ukraine issue.
What I thought might happen, and the thing that I was concerned about, was that new government would come in and with total control, they’d be able to implement all kinds of crazy surveillance or whatever tactics. And in fact, this is what folks had thought about Trump as well. And it actually, in both cases, in both Trump and Biden, it has so far turned out to be all of a Potemkin village, right? They say, some folks thought Trump would be this competent surveillance state. Some people thought Biden would be able to do that. But neither has been able to do that. And that’s really interesting because it makes one think, okay, let me step back and be like, why did I expect that? And why is the result actually completely different from where the words are and where the conversation is?
And so I wrote this mini essay. So this is the thing that I changed my mind on. So I’ll just read this little poem issue thing. For some time, I’d been increasingly concerned that instead of marching into a future of freedom, we were heading into a time of tyranny. I still think something like this may take place in China’s sphere of influence. Arguably it already has. But as the events of this year have unfolded, it appears we are not on track for a dictatorship of either the right wing or left wing variety, for something more like American anarchy. If you squint past today’s half ignored TSA like COVID regulations, you see a half ignored TSA like COVID regulator, namely, a failing state that people can half ignore, and arguably must half ignore because the USA itself is now the TSA and the TSA we know is safety theater.
In the territory governed by this inept bureaucracy you see power outages, supply chain shortages, rampant flooding, and uncontrolled fires. You see riots, arsons, shooting, stabbings, robberies, and murders. You see digital mobs that become physical mobs. You see a complete loss of trust in institutions from the state to the media. You see anti capitalism and anti-vaxism. You see states breaking away from the US Federal Government, at home and abroad. And you see the end of power, the revolt of the public, the defeat of the military, the inflation of the dollar and, looming ahead an American anarchy. What’s coming isn’t fascism or communism like a left wing and right wing pundits will have you believe, even though they don’t believe it themselves. What’s coming is the exact opposite of that. A world where the civilized concepts of freedom and equity are extrapolated to their de- civilizational limit, where you ain’t the boss of me and we are all equal.
Where all hierarchy is illegitimate, and with it all authority. Where no one is in charge and everything is in chaos. You can argue that this may be preferable to the status quo, if you’d argue that the chaotic rush of the 90s was unbalanced better than the authoritarian Soviet Union of an 80s. You can argue it may be inevitable. As the Chinese proverb goes, the empire long divided must unite, long united must divide. And you can argue that this transitional period of anarchy may be lamentable, but that it’s better than the other team being in charge and that we can build a better order on the other side. And maybe so. But prior to any rebundling, I think we’re on track for quite the unbundling. That is what I changed my mind about.
Shane: Is there anything you’d change in that now?
Balaji: The big thing is, so I think your reference might be Canada, right? And the fact that Canada has implemented this financial surveillance and what have you. And will you actually see that top down dictatorship? And the thing about it is, that’s actually kind of a piece with this where, yes, Canada did implement these unprecedented financial controls. Yes, it did affect these truckers. But because they were unable to successfully censor social media, Canada suffered probably the gravest PR disaster that I’ve ever seen in my life. I don’t know about you, but in the sense of abroad, I’ve never heard people talk about Canada that much. And certainly not as negatively. This crackdown demonstrated insecurity. When you need to do something like that, it indicates that you feel you can’t just go and put out a megaphone and be like, “Hey people, I’m listening to you. We’ll figure it out.”
And there’s a mutual trust, right? You feel you need to do, not quite shooting at a mob, kind of thing, but certainly not going out and trying to meet people halfway, that the government felt extremely insecure. Second is, it was much easier to argue against in some ways, because it wasn’t, what’s happened in the US for a while is a bunch of corporations just happened to all do the same thing at the same time. And then people can argue, “Oh, that’s just the total coordination of the most powerful companies in the world. It’s just private decision making. What, are you against the free market, huh?” And, now in practice that is quasi coordinated or extremely coordinated behind the scenes by an incoming administration or by people who are aligned or even just by social beliefs that they all share. But in this case, the fact the government had to push the companies to do it was useful, because it showed a centralized hand behind the whole thing and it gave people something clear to argue against which is state power, something people are skeptical of.
So first, they had to use a lot of power. Second, they had to use it highly visibly. Third, because they couldn’t censor social media, they took a major PRL. And fourth, they just aroused a lot of resistance and what did they do with this? They harmed some poor truckers. And, the thing about this by the way is, I’m pro-vax, right? I’m one of the earliest advocates of vaccines and so on, on Twitter. Very, very, very, very beginning of this I’m like, we need to do vaccine development. We need to defray the cost. I thought there was going to be a huge controversy over vaccine cost. There was to some extent, it’s more after the fact. The one thing I hadn’t forecasted was that in the presence of a potentially deadly virus, that people would be against vaccines.
Now I get it. The thing about coronavirus and the thing that we’re actually really fortunate about, it’s not the Spanish Flu, so it’s not 100,000,000 dead. The way I think about it is it’s like the Richter scale, right? It’s not 1,000,000,000 people dead. It’s not a 9. It’s not 100 people dead. It is about 10,000,000 people dead worldwide. It’s about 10,000,000 people. It’s going to be, it’s like 6,000,000 people are dead worldwide, right now, I think from COVID, right? If I look at Google, what the number is. It’s about 6. Okay? That’s reported, of course. So, there’s some countries where the stats might be lower or higher, but ballpark, as the thing keeps burning its way through the global population on the order of 10, certainly way more than one. Right?
So it’s just bad enough to have been certainly like an important event, but fortunately not so bad that it actually broke down civilization or something like that. At the beginning of it, though, if you look at COVID deaths and you look at the curve, all deaths all time, the first several months going into basically mid April 2020, it just looks like a straight exponential. And what you don’t know is where that exponential is going to top out. We’re very fortunate that it tops out at something like 7,500 deaths a day or something like that. And then it sits at that set point-ish for basically the next 2 years. Okay? And that’s the sum of all kinds of curves around the world, but who knows that could have just gone, voosh, like this, and it could have topped out at 75,000 deaths a day.
That’s just like an experiment, which we’re very fortunate that it turned out to be just serious enough that people could argue about it. Just bad enough to be really bad, but not so bad that it was like a complete undeniable to everybody that it was bad. One thing I find is that people, if you’ve got a vaccine, then, forget about masks and everything else, just go back to quote, normal. Right? And then you can ask, “Okay, do you actually need that many booster shots or is that going to give you more resistance?” And I think that’s an empirical question, and there are new tests that are fast tests for immunity, right? Determine the presence of antibodies. And, that should allow you to determine and evaluate both herd immunity and the effectiveness of immunization programs. And you can sort of get a sense of the number of NABS that are present individuals, will give you some measure of whether you possess protective immunity to the virus.
You can look at your system and see if you’re topped up in terms of immunity and whether you might really need a new booster shot. That’s a very crude measure. And then you have to get into the specifics of are you immune to this variant and so on and so forth, or partially immune because it’s multivalent and all that stuff. But, you should be able to assess whether you need a top up versus just mindlessly getting it or whatever, if you wanted actually some feedback to see is this thing working versus just taking injections over and over? This gives something in between, “Oh, just take a shot versus, oh, don’t take a shot.” It’s like, okay, there’s a metric that you can look at that should determine whether you have residual immunity, whether there might be a cost benefit to taking it.
And whether you can evaluate that and you know, there’s probably going to be some trade off at where that’s worth it. And, and you can just look at that sober. And this is a broader thing, which is one of the most tragic aspects of all of this is that probabilistic cost benefit decisions are turned into morality plays where there’s an obvious 0-1 one for everybody of a 1,000,000 people on Twitter. And that is just not what reality is. Do you know the term pharmacogenomics?
Shane: No, never heard of it.
Balaji: So pharmacogenomics are just refers to the fact that different people have different drug responses. Some people, for example, are fast metabolizers and others are slow metabolizers of caffeine or of alcohol or of drugs like warfarin and so on. And so yes, height and weight and gender, and so those are all inputs into a regression algorithm, but so are your genotypes, CYP2A9 and VQRC1.
And so there’s actually genetic impact on your drug dosage, as we kind of know, right? But that also affects immunity. You have immunogenomics. And again, we kind of know this, you probably heard this story about how the trait that gives sickle cell trait also gives potentially partial immunity to malaria and infectious disease. Right? So that’s immunogenomics, right? It’s a case study. The cystic fibrosis variant gives some potentially partial immunity in the heterozygous condition to Yersinia. I believe that’s the case. Let me just actually check that.
Shane: So is the implication of this giving people information to make the choice, or do you feel like they should be forced to take the vaccine? I’m just trying to tie this back to the Canadian and truckers.
Balaji: So this is a complicated one. We know that their sickle cell trait can give some resistance to malaria. We know there’s mutations that have been hypothesized to give resistance to everything from HIV to the black death. And so that is probably also the case for COVID 19. There’s probably, almost certainly, and I haven’t looked at the literature recently, but almost certainly at this point 2 years in, we know some genetic variants that might give more or less susceptibility to COVID 19, with millions and millions of cases and deaths and so on. We’ve probably done these studies. I haven’t looked, but if I just dug into it. And so what that means is that one size fits all may not actually be the case.
The maximum that we’ve been able to communicate so far is, oh, older people are more likely to get COVID, overweight people are more likely to get COVID. Those have strong signals. And so therefore if you’re in those risk categories, then you should go get it. But ideally we would have everybody having a personal genome and you having a regression function where you could look at your phone and you could see an estimate based on all of your characteristics, and then if you wanted to you could drill down into it and you could see the underlying raw data and where those coefficients came from. And even if you didn’t believe it, you could have your software engineer or academic or whatever friend, diligence that equation, right? I’m not saying it’s just a software engineer, of course, because it’s not just code, there’s also medical knowledge there. So, whatever quasi priest you’d want to call on, you could have them, your most quantitative friend, you could have them go and look at that calculation and look at the raw data and look at the citations there.
And then you could feel like, okay, I understand why they’re reporting this risk, where it comes from, where the data is going into it. And I’m going to make the decision on that basis. It’s like it an explainable score for your credit score or something like that. Why am I getting this score, right? And, that’s possible. What I just described is actually not impossible with today’s technology. If everybody, if we had not held back personal genomics in the 2010s, if we had not so heavily regulated it, we would have a billion people with personal GMs, just like we have a billion people with, or hundreds of millions, at least, with fitness trackers. Right? And so the thing is rather than saying, oh, it should be mandatory. Oh it should be voluntary. Those are just like essentially the ramifications of a complicated decision into the government should do X or Y. It’s like this 20th century state where the only outcomes are, everybody has exactly the same rule. There’s no gradation of things. And this is the difference from Seeing Like A State to learning like a machine. You know the book Seeing Like A State?
Balaji: Seeing Like A State, there’s this is book about how governments, in order to see things, in order to make the population able to be governed, they make them legible. So everybody has to have a last name and a first name. You need to fit yourself into these boxes on a form. You need to fit yourself into these categories governments created, for example, Asian American. People of south Asian and east Asian descent didn’t really think of themselves as part of the same category until that artificial creation of Asian American, like by the US, which then groups folks in a way that they weren’t grouped before. It’s like saying, okay, now you’re a basketball team and you’re a basketball team and you’re a basketball team. And then people start looking like, okay, I guess we’re on the same team now.
Balaji: So it makes these kind of classifications so that the state can see. But here’s an interesting aspect of that. Most of those classifications are almost necessarily binary thresholds. It says, did you make more than X or less than X this year? Binary threshold. Are you part of this category or not? Binary threshold. And so the way the state sees is in terms of these sort of binary decisions. It’s mandatory or it’s forbidden. And one of the things that was a discouraging result in machine learning a long time ago was that neural networks that were just based on perceptrons, so called just zero-ones, couldn’t actually get that far. Right?
Balaji: Like there’s a, 1969, a famous book, Perceptrons by Minsky and Papert shows impossible for these class of networks to learn an XOR function. Okay? They’re only capable of learning linearly separable functions, right? Basically, perceptrons in neural networks, it’s like this sort of zero-one decision criterion. Okay? And for single layer perceptrons they can only learn simple kinds of functions. They can’t learn XOR and there’s a book that came out that proved this. It actually turns out that multilayer perceptrons can learn more complicated functions. But the point for our purposes is it’s a little bit imprecise, but these bureaucratic zero-one boundaries that the state comes up with cannot actually model properly the complexity of the human population underneath, but machine learning can. And so if you have that data in a database, you can potentially draw much more smooth curves rather than these industrial age lines. You don’t ask somebody, are you inside or outside the set, you have N variables that you can actually give a continuous decision boundary, right?
For example, to bring it back to earth, you have their genome and they have millions and millions of snips, these variations. And then you can infer what their risk is based on that. And you can say you have a 97% chance of coming down with a bad case if you don’t get the vaccine. Or you have only a 5% chance. Maybe this is just the scientist or whatever in me talking. But I do believe that force is really not, I mean, it’s such a last resort. You really should try to persuade much more than one has. And I don’t feel that the persuasion and the evidence presentation has been that good on this. Because just changes all the time and people say, oh, masks don’t work and now they do. The science says this, follow the science, the science changes every day.
And the issue is, fundamentally, that the people who are coming up with this stuff, they’re not distinguishing between science in the sense of Maxwell’s equations and science in the sense of a paper that came out last week. Maxwell’s equations, that’s capital S science. That’s real, that has trillions of independent replications. Okay? Every single time you pick up the phone and you message me or somebody else, you’re doing an independent replication of Maxwell’s equations. If our theory of how the electoral and magnetic field worked, were not correct, that would not work. Right? So that’s like science, okay? That’s solid. That deserves our respect, right? A paper that came out last week that does not have any independent replications is not science in the same way Maxwell’s equations is. And that quantitative access is very important. Science is not about prestigious citations. It’s about independent replications.
Shane: Well we’re confusing the term science, right? We’re just using science for everything.
Balaji: We’re using science for the form and not the substance.
Balaji: The substance of science is independent replication. If I cannot independently replicate it, it’s not science. If your data is hidden, if your methods are hidden, okay. The whole point of scientific publication was to allow others to replicate it. Right?
Balaji: Now what that’s evolved to in the modern age is something where it became prestigious to come up with these discoveries. And then it became just a pure prestige game where people often hid the data or the methods behind their discoveries, or wrote them in a purple prose kind of way to keep scooping their competitors. And, you could have whole fields where it was just this prestige game and nobody replicated each other’s work because they were all reviewing each other’s work. And that’s like the reproducibility crisis in psychology and other areas.
And that just became something where we used these proxies for the underlying variable. And then we focused too much on those proxies, right? And that’s a, that’s a fundamental problem in Western society, now. You have prestige, which was a proxy for replication. Then you lost sight of that underlying thing of replication. Okay? Or you have elections, which is a proxy for the consent of the governed, but lost sight of that underlying thing and start doing militarized police and all this type stuff, just thinking, oh, I get 51%. That means I can Ram it down somebody’s throat as opposed to realizing actually it’s a continuous thing. It’s not just get 51% and you have the power of 99%. 51%, you need to govern very humbly and modestly. Right?
Balaji: This also happened on Google, right? Google had this great thing, where it was able to use backlinks as a imperfect measure of search quality.
And now they’ve lost sight of that. And search quality is actually quite low for many queries, as you’ve probably seen, right? So this is a common thing where there’s a cheap proxy for something. And then people game that proxy, there’s actually a term for this. It’s Good Hearts Law, when a measure becomes a target, it ceases to be a good measure. We have gotten away from what science is supposed to be, which is independent replication. It’s become prestigious citation and insistent repetition. Somebody who can’t replicate the study, just like, “It’s science, it’s science. Whoa. I can’t believe you’re against science. Oh my God.” And there’s a thousand retweets, you know what? I don’t care about a single one of those retweets. What I care about is did you download the code at least? And the data set and rebuild all the figures.
Right? And that’s like V one, by the way, that’s like the minimum threshold of replication. Just to dig into this for a bit, because I gave a talk on this. It’s actually a very important topic and I’ll come up to the vaccines thing. Because the reason is, I don’t want to just say, oh, here’s my opinion on vaccines or here’s my opinion on this or that what I want to do is basically just say, okay, here’s how I’m thinking about it. Here’s my thought process. Here’s my algorithm for thinking about this stuff. And then you can attack that or you can listen to that and then you can go and do it on your own. Right? Not in a naive, do your own research way, which says basically just look at Google, but if Google is corrupted, then that doesn’t even work that well either.
Right? But okay. More deeply. Here’s here’s just a few concepts. First. I talk about it as the transition from Fiat information to Crypto information. The way that our system is based is academia study something, they write up a paper, the media reports on it and then the government regulates on that basis. That is the sense in which our society is based on quote “science”. And if you think about it, almost every decision can be tracked back to that. That is the information supply chain. Okay? From production to dissemination to regulation. Right? Okay. So, once you think about this, okay, academia is the source, it’s upstream of all of this stuff. If there’s a dispute at the government level, what did the Harvard scientists say? What do the Stanford scientists say? Let’s go back to them, right? Let’s have a study. And there’s some good to this in the sense of trying to study things, dispassionately and do a proper writeup and so on. But the problem is a few fold. First is, there’s this concept, have you heard this concept called reproducible research?
Shane: No, but I intuitively think I understand that.
Balaji: Yes. So there’s an intuitive term for it or there’s intuitive definition, but then it also refers to a technical thing. And the technical definition of reproducible research comes from about 20 years ago, this guy, Jon Claerbout at Stanford, a Geophysicist and another guy, David Donoho in the Stats Department, started realizing, oh, training new grad students is a huge pain because you have to communicate all of this context and so on. How do you onboard them? And eventually they came with something very clever that at first you might not think was sufficient, but actually turned out to be really good, which is every paper in stats and many other fields of, let’s say applied math is essentially a database plus code. And the code reads that data and generates a paper with a bunch of figures. Okay?
Anybody who’s listening to this who’s written code that takes from Postgres and templates, a webpage and makes some graphics. It’s basically just like that. That’s what an academic paper is. Okay? It just, output is PDF rather than HTML, to first order. Okay? You’re making some graphs and some tables and some charts. So you distribute a package where you hit enter and it takes the data, which you include and the code, which you include, generates a PDF. The reason that is sufficient to train a new grad student is if you wonder how they got this number of 0.73 in this paragraph, in the paper, what do you do? You look at the source and you’re like, okay, it was this function call and that called this data set and I can look at the whole thing and I can hit enter and I can rebuild it. Okay?
In R there’s a package called Sweave that does this. In Python, there’s Python notebooks that do this, okay? There’s a website called distill.pub, which is all based on this. But this is a concept that’s been around for sometime, it’s called reproducible research. Okay?
There’s a second concept that’s been around for sometime. It’s called open access. And this refers to the concept that when you publish something, one of the weird things about academia, people think that if you have a prestigious publication, you’ll get paid for it. Because that’s how it works in other spaces. That’s not how it works in academia. Once you get accepted to nature or science or something, you have to pay the journal. It’s this silly system because the government pays for your grant. You have enough money to do it. And the journals, they feel like they’re not charging you. They feel they’re charging the government and you have to pay them something for the paper to be online and you have to pay them even more for it to be open access so that it’s visible to everybody. And there’s different kinds of open access like green and gold. Is it accessible to people today or in 30 days or forever, blah, blah, blah, right?
And open access is this insane thing where essentially, you have to actually fight these for-profit journals to open source the research that the public funds paid for. And so it’s been this huge fight for the last 20 years on this. It’s gradually been moving in that direction, right? So reproducible research is you can rebuild a PDF from the code and the data. Open access is the public can actually get access to that PDF and that HML without having to go through some sign up and subscription process for an expensive journal. And you put those together and you combine a third thing, which is actually, what I love, the blockchain, okay? And you get the concept of truly reproducible research where you’re taking that code and you’re taking that data and you’re putting it on chain. So it’s permanent. It’s archival. Okay? It can’t be mucked with. It’s also truly open access because it’s globally accessible. Anybody can download the blockchain, right? Let’s say it’s at the arm chain or it’s a chain dedicated for this purpose. Okay. And if you hit enter, you can rebuild a PDF.
Now a lot of things become really interesting when you do something like this. First, citations become function calls. To cite a previous paper, you’re actually pointing to it. The whole composibility thing in defi becomes citability in Desi. Okay? Decentralized science. Okay. That’s huge. You can just imagine that you could actually, in theory back trace all the way back to Principia Mathematica or Maxwell’s publications, right? You can actually see how your manuscript relates back to these originals, because you’re doing function calls, which are more precise in citations, right?
The second thing is every data set that on route to your conclusion can be diligenced and people don’t understand how bad sometimes the citation path is. Especially in biomedicine, there’s factoids that people think they know, but when you root cause them, there’s a famous thing with like Popeye, spinach. Do you about that?
Shane: The root cause of Popeye’s spinach?
Balaji: Yeah. It’s like is spinach a good source of iron? Like the myth?
Shane: Well, you’d be able to trace everything back, right? So if something was unproven or didn’t replicate, then you would also know all the implications of it through all the different, how it’s weaved its way through science, the term used loosely there, not reproducible science, but how we’ve used that information to make decisions or how it’s influenced future studies or implications.
Balaji: Yeah, exactly. So here’s this, there’s an interesting study, which is basically about trying to back trace and find out whether, this story is complicated and I don’t want to paraphrase it. Academic, urban legends. Okay? Many of the messages presented in respectable scientific publications are, in fact, based on various forms of rumors. Some of these rumors appear so frequently that we can think of them as academic urban legends. The explanation for this phenomenon is usually that authors have lazily, sloppily or fraudulently employed sources and peer reviews and editors have not discovered these weaknesses in the manuscripts during evaluation. To illustrate this phenomenon a drop on a remarkable case in which a decimal error appears to have misled millions into believing that spinach is a good source of nutrition. Through this example, I demonstrate how an academic urban legend can be conceived and born and continue to grow and reproduce within academia and beyond.
And so, he just goes into this story and the submyths and myths and so on. Point is, something you might think was a solid problem, like is tracking back citations, is not a solid problem. In fact it’s actually something where the evidentiary base is actually not as firm as you might think, especially in biomedicine. And so what you actually want to do is click, click, click and see the original table in 1963 and the specific rows and columns that are giving this parameter that is being used on millions of people, right? It is basically just like in any complicated software engineering project today, you have the set of dependencies and dependency management is a whole thing. If you’re coding node or something like that, you have like 30 libraries that your program depends on and there’s specific version numbers, and this is a whole thing.
So now you bring that kind of rigor into actually the management of dependencies for scientific paper. And those are essentially the input premises that are then expanded upon by the paper that leads to the conclusion of that paper. And it’s incremental contribution to literature, right? So now you know what I’m describing here by the way is of the complexity of Wikipedia, but not greater than that. You’re basically talking about taking the scientific literature, putting it into truly reproducible research format and putting it on chain such that it’s permalink and then building a citation map back to each other. Could you do this? Yes. Is it a large project? Yes. Is it a finite project? Yes, because you know, the relevant scientific papers are only a subset of the literature, right? So you can focus on that area first. Once you do this, you have, at a minimum, the digital aspect is auditable.
You can look at a paper if you disagree with it, or you question something, you drill down, look at it, citations. So either it’s data is correct or incorrect. Maybe the data itself is fraudulent. There is a Benford’s law kind of thing, you can run against it. So you can see whether it’s internally consistent, like the stuff that accountants do. There’s very statistical tests that you can run to see if it looks like it was faked or what have you, right? So you can internally check it. You can plot it and graph it yourself and look at it. Sometimes you’ve rotated data. So you’re like, oh, that’s an artifact I can see on this axis that I couldn’t see looking at an X, Y plane, but it’s obvious in another plane. So you can analyze the data. You can look at the premises, you can analyze it a few different ways and then you might ask, well, okay, but what if they did like a fancy fake, all you can do is you can only rep reproduce some of it, not all of it.
And so my argument there is yes, it is true that first order, we don’t all have cloud chambers and incline playins and so on in our house, right? But we do have phones and laptops and computers. So we can reproduce the digital. That is actually independent replication on the digital is feasible. So at least the digital parts we can reproduce. So at least grant me that, right? A good chunk of what is done, post data collection. We can essentially replicate for free locally. That’s actually what Bitcoin is, by the way. That’s what all cryptocurrency is. It’s not scientific economics, it’s mathematical economics because you’re doing the math locally and you’re not trusting a central authority. You’re trusting the decentralized math. Right? Okay. What if they did fake the data? And one way of dealing with this is you go one step further. First you have to get this whole, truly reproducible research establishment built. I’ve kind of described the theoretical framework for it.
Then eventually what you do is you start making instruments, crypto instruments. Okay? So you might be aware that things like GoPro will just stream to the cloud. Okay. They’re attaching metadata to the images when they do that, but there’s also scientific instruments that do that. Like Illumina sequencing machines, I’ll just stream data up there. And there’s huge databases like NCBI that collect all of this biomedical data and they have metadata on every experiment. And why do you need that? Well, sometimes you’re doing some data analysis and it turns out that we thought was statistically significant like cancer patients versus non-cancer patients actually turns out to batch and versus batch two. And it’s a metadata which you can use to determine that because if it turns out that you actually are finding, oh, this is from instrument one, and this was an instrument two or batch one and batch two, as opposed to cancer patient, non-cancer patient it’s spurious correlation. Okay? So collecting of the metadata and analyzing the metadata is already standard in bioinformatics.
And then you would just add another aspect to it, which is every single individual data point at the time of its recording has effectively an audit trail on it, which shows which instrument uploaded it at which time. Why would you do this? Well, the thing is, have you ever seen, you’ve seen JPEG, you’ve seen an image, just an image of some kind. Have you ever actually printed it out on screen and seen the floating point numbers underneath it?
Shane: No. Never.
Balaji: Okay. You can do that. You can basically see the RGB numbers or those floating point or integers depending on the image format. You can actually see the numbers that underpin this thing. If they just flow on the screen, it’s incomprehensible to the human. So when the data is being collected off the instrument, it’s just floating points or integers. And if each one of those is being individually hashed, it’s hard to fake that. It’s still possible, but it’s harder to fake it. So you have it more rigorous. Because, what people do is they’ll take those millions of data points. If they’d want to fake something, they would fake it after collecting it. But if you’re auditing it and you’re hashing it at the time of collection and by the way, you don’t have to hash every single individual data point, you can do things like so-called Merkel trees where you group a bunch of things together and just have one hash.
But it’s difficult to falsify. Point being that you can build effectively a chain of custody for data. From the instrument, to the paper, to the chain of citations that leads you to the premise and conclusion that you’re supposed to base your life on today. That is what I think we need to build towards rather than trust me, trust me, trust me. The lower the trust in society, the less trust me works. And frankly elites don’t deserve a lot of trust because they’ve messed up so many other things from the Afghanistan intelligence, the Iraq intelligence to lots of the early official misinformation on COVID, Ben Bernanke saying The Great Moderation was there and so macroeconomic volatility was the thing of the past.
We’re talking about this in 2005, by the way, a few years before the financial crisis. Macroeconomic volatility is a thing in the past. Do you remember this one?
Shane: No, but I just want to ask a question before we go. How much of that misinformation, we’ll classify it as that, do you think is intentionally misleading in order to accomplish a goal versus just being inaccurate or with good intentions behind it?
Balaji: There’s several different things that accumulate. I’d say the single most important thing is everybody has in their head, a set of the taboos that they cannot violate with any public statement. For example, last year it was taboo to say, “oh, it could be a lab leak.” And then it became untaboo and then people could talk about it still somewhat, but it’s much less so than it was. So the lab leak is a great example where I shouldn’t say the truth to get the hypothesis out because that’s what people were doing. To even get the hypothesis out, these guys had to publish it pseudonymously and hope it wasn’t censored. And so that’s a huge part of it is there’s a school of fish kind of thing going on where, the way I think about it is the establishment in the west practices the school of fish strategy.
And it’s very similar. Here’s how it goes. Just repeat what everyone else is saying. If it’s proven wrong, well, everyone’s wrong together. That’s the establishment consent algorithm. It works until falsified by the outside world. And it’s a very simple consent algorithm. You’re just saying what the guy next to you is saying, and then no one can single you out. Huh? How can you single me out like I was singing who could have known? And the thing about this is of course somebody could have known. The minority report could have known. But the point is to deny the minority report credit, they’re a conspiracy theories or lunatic, whatever, until the credit can be stolen by some establishment Grande. There’s a great ad from FedEx many, many years ago, which was “When the CEO says it.” Are you familiar with this one?
Shane: No, it makes sense biologically though. We’re social creatures and we don’t want to stand out from the tribes. So we’d rather be wrong with everybody else than right in alone.
Balaji: That’s right. And that’s actually why betting reduces partisanship. We’ll see if this, a social science experiment that does replicate, but I have seen it in my own experience that if you just ask people what the answer is, they’ll give you the tribal answer. The moment there is some financial downside potentially on the table, they start trying to think about, okay, is this real or not? So this FedEx ad stolen idea, July 26th, 2007 on YouTube. That is a great example of kind of how the establishment kind of works. Some poor schlub goes out there, puts a concept out there. They’re taking all the risk, and you know what the establishment say, “Thank you very much.” If it does work. And by the that’s like the best version of it, where actually the good idea is at least stolen and used, right?
The worst version is a good idea isn’t even used and that’s actually even more frequent. And so the way I’d put it is, first, there’s this tremendous web of taboo and fear and so on. That basically means that independent thinking, which is what you need the most in our problem situation, especially an unexpected one is penalized over and above. It’s already rare and then you penalize it over and above that. So an enormous amount of the blame on this, I think goes towards the current culture about how they think about this, what is rewarded versus what is not.
Because otherwise the studies are like looking for quarters under the lamppost. If you have prejudged to a conclusion masks don’t work or masks do work, it’s something that the average Joe can understand that masks don’t work and now they do the science changed.
No, it wasn’t like some giant randomized controlled trial. Come on. And when you know that they’re lying about that, then you know they’re just invoking science as basically a priesthood of the credentialed. You’re a scientist if you’re a professor and you’re not if you’re not. Are you going to do science on your own? The vast majority of results in mathematics and physics and so on that we rely on happened before NSF. Let’s say like they predate the modern scientific establishment that kind of arbitrary set up. And people give them lots of credit. And I’m not saying there isn’t some good that came out of the centralization of science that happened post world war II, but there’s a lot of that came out of it.
And the bad that came out of it is you built this centralized funnel as opposed to lots of folks with independent power centers and funding sources, figuring out stuff on their own. Instead, you had a single thing which was like official science and official science became official religion, like state religion. And now we’re seeing that. Instead of beat that, you actually need a model that decentralizes science I’ve described one aspect of it, which is truly reproducible research, truly open access, on chain citation, and crypto instruments. So the combination of four things, which is you can take the code and the data and rebuild a paper. Everyone can access the paper for free anywhere at any time, you can look at the citation graph as a bunch of function calls, not simply citations, tracking it all the way back to the ancestral premises that your paper is based on.
And then you use crypto instruments so that you are hashing the data that is going on chain. So you give some diligence on the physical world, the uploading aspect. So now you can check quite a few things at home on your computer. It’s like the OSN concept, which you’re very familiar with, but in a different domain. Why do people do OSN? Well, these flight trackers and other things that are useful in peace time, you think it’s probably unlikely that people are going to fake the data sources in war time, so you have some familiarity with the thing. So those are kind of four big pieces, but there’s a fifth, which is everybody will always talk about and really fifth to sixth and the seventh and that is okay, well, that sounds great Balaji, but how’s my lab going to get paid?
And now I’ve got an answer for that. And the answer is it’s NFTs, it’s a DAO for your lab, it is a personal token for your lab members and for yourself and any lab that is doing reasonably interesting technology work absolutely will probably be able to get more funding online via cryptocurrency and NFTs than they will within the academic grants process.
It’s now known an incentive to effect, frankly, one cool result can get like a few hundred thousand dollars in NFT revenue and fund the entire year for postOk or what have you. Or you can even forget the entire academic track and just hire people who can do cool things. And you’re starting to see that with things like VitaDAO and molecule DAO and others. You’re starting to see DAOs because in the same way that a village would support a bard or a musician that would bring culture, it wouldn’t defend the village, but it would make the village worth defending.
Now you can actually support a real scientist, a person who that community trusts and who’s putting their work on chain and who is, showing all of their work and not just dumping their chest saying, “I am scientists. Listen to me, I’m part of the guild.” You’re effectively offering them an incentive to defect the public from the guild and to decentralize. And maybe they can make a million dollars a year for their lab outside, but to scrape for a couple hundred thousand in grants. So that’s a pretty good deal and they’ve got independence, and they can’t be canceled and they can publish on chain. You also get a few other things when you go on chain, you have unique identity.
So it turns out that when you’re mapping citations in academia like credit and so on is very important. So if you’re putting the paper on chain, you get precedents. So you get credit. If you’re digitally signing it’s also very obvious that you wrote it. This is actually important for folks, for example, with Asian last names, if you try searching like C. Li on PubMed. There’s like thousand people with the same name. It’s an important problem when you’re tracking academic citations, it’s hard to disambiguate certain names, digital signatures solve that. Seems like a small thing.
It’s actually a big thing for certain things where you can see who generated a certain fact, you can get a web of trust there, and you can publish pseudonymously, like Satoshi, like these guys at Project Evidence, like Newton. If you’re really pushing the envelope, you might need to publish pseudonymously. In today’s era, that means that people only are judging the facts. They’re not attacking the person.
So you combine those things and that’s like seven pieces that I think that crypto is providing for transition from Fiat information to crypto information. Truly reproducible research, truly open access, on chain citation, the crypto instruments concept, digital signatures for authenticating yourself, on chain monetization of your lab and pseudonymity if necessary. Put that together, that is actually a vision for rebuilding the upstream. That’s a vision for the crypto university.
Shane: I love it. Now that we’re on crypto, how are you thinking about Russia, Ukraine and the fear and the crypto circles that this will lead to the West finger pointing at crypto as an effective Russian tool and a crackdown without understanding the implications.
Balaji: So how do I think this is going to play out? Is that the Western establishment over the last decade, half decade is essentially in the long emergency of emergencies. And I think that that way we’ll look back on this is something where they have picked simultaneous fights with many different factions at the same time. Within the West itself, they’ve picked fights with tech and with the conservatives and with a lot of folks on the central left who have left for Substack and for newsletters and independent things.
They’ve also picked fights with parents, huge ones on education and on COVID restrictions to the extent that even San Francisco had the school board recall and with comedians and lots of folks in Hollywood who felt that some of the stuff was maybe merited, but a lot of the cancellation went overboard.
So many of their charismatic culture producers in Hollywood, in media, in technology they’ve alienated as well as a huge swath of conservatives. And then abroad, they’ve fought simultaneous battles with Russia, with China. They have managed to not fully alienate, but certainly antagonize India and Israel and manage to do that all at the same time. Basically Brooklyn’s wokest who I kind of think of as driving the establishment now only really like themselves and not really that much. They fight amongst themselves quite a bit. So it’s sort of the opposite of coalition building its coalition breaking, where these guys are, I think of them as nepotists rather than founders. They inherited a system from their ancestors that they could never have built.
And they just assume that this thing, which is at massive scale, they can sanction this person and sanction this person and attack this person, deep platform this person, and so and so forth. And it’s actually bananas how successful they were for six or seven years just using the power of the empire at the Zenith, bam, this guy’s canceled bam. But when you do that too many times, and everybody is a Nazi and everybody is deplatformed and everybody is canceled. What they did and we’re actually kind of fortunate in this, they’ve done it in a chaotic evil, rather than lawful evil way. That’s to say, do you know what I mean by the difference there?
Shane: I think so. Yeah. But you explain it just so everybody is on the same page.
Balaji: So, lawful evil, it’s like the communist or the Nazis they’re evil, but highly organized. They are disciplined, they are persistent, they’re methodical and you can argue CCP is somewhat like that today. And it’s complicated of course, when you say someone’s evil, was every single person in Russia evil? No, of course not. Is every single person in China evil? No, but there’s aspects of regime you can describe as lawful evil where, when they decide to go and censor, it’s like a space shuttle launch.
It’s like, we’re going to censor this, we’re going to censor this. They are not lying to themselves about their intent to censor, they’re not self deceiving. They’re like, yep, this guy is falling through a trap door and this guy is falling through a trap door and we’re censoring these words and we’re censoring those words go time. Whereas in the West there’s all these lies that surround it. Like, oh, we’re not actually censoring. Because what they’re doing is they’re retrofitting speech and thought controls onto a previously free society.
Shane: Well, hold on, pause for a sec. We are censoring because we’re now dictating broadly speaking. We’re labeling people that have acceptable or unacceptable views in various forms.
Balaji: Yes. But they’re not admitting to it. And like explicit versus implicit. China’s actually set up an explicit system, The Great Firewall. They have an explicit thing where it’s like bands on social media. They have essentially an explicit set of rules on you cannot post and you can. And people understand that explicit set of rules. What the establishment is trying to do, is have a China style set of speech and thought controls without admitting that is what they are doing. Or they just double talk about it. And so what they really want is something where they can talk, but you cannot. The guys who own newspapers have freedom of speech and you do not. For example, this article over here in the Atlantic, In The debate of Freedom Versus Control of the Global network, China was largely correct and the US is wrong, right.
Or the NYT has tons of pieces on this like, Free Speech is Killing Us. These are Sulzberger employees. Sky Arthur. G Sulzberger. Six generation nepotist. Inherited the New York Times from his father’s father’s father’s father. Nobody ever writes about him, because they’re scared. Everybody writes about Zuckerberg, who for all his faults, basically Facebook himself and who I respect. I don’t respect Sulzberger, he’s just nepotist and who inherited the paper. And who’s basically a born millionaire. And by the way, on this topic, there’s so many articles… Just to talk a second, there’s so many articles in New York Times that talk about how non-diverse tech is or whatever. And do you know the actual succession contest at the New York Times company for who’s going to be the top publisher or the who’s going to have the top job. Who do you think it was between?
Shane: No. How does it work?
Balaji: Yeah. How does it work? Is it something the Rooney rule, the Rooney rule you’re supposed to go and interview people from diverse backgrounds for the top position, right?
Shane: Yeah. Yeah. In the NFL.
Balaji: You know the diverse backgrounds interviewed was three white male cousins for the top job, the New York Times. Read this article A.G Sulzberger, 37 To Take Over New York Times Publisher. So it’s actually like a caricature. I’m not the kind of person who thinks white is an insult, but these guys are. And they say it constantly, oh my God, it’s so white, it’s so male. And yet this article is written as a coronation of great leader taking over from dealer leader or whatever. So this meritless nepotist and his employees, these Sulzberger’s inheritance, that’s the New York times company and Sulzberger employees will write these articles about how free speech is killing us and how actually you’re not supposed to have free speech only Sulzberger employees are.
And how, if you say it’s misinformation, when a Sulzberger employee says that you’re supposed to bow. And they run these billboards that actually literally proclaim themselves to be the truth in 2017, Sulzberger ran this ad campaign. The Truth. You remember that in 2017, The Truth Is Hard To Find et cetera. Now, do you know what other newspaper called itself The Truth?
Balaji: It’s actually an obvious one. As an Intel guy. You might know this.
Shane: No, I don’t
Shane: Oh God.
Balaji: The Soviet newspaper is Truth. So the thing is Sulzberger innovation on is it’s not even com at least Pravda is free. He charges you 50 bucks for “The Truth”. And this is actually the same, we described that information, supply chain, it’s academia, media government. So this is the second leg, which is completely corrupt. These folks who will charge you $49 a year for access to the truth, where really what you’re just doing is propping up this family office. One of the things, by the way, you can sort of turn one of these corporate journalists to stone, by the way, when they’re pretending to be all speaking truth to power okay, when are you going to write the big investigative journalism report on your boss?
For example, recently, just to digress in this topic for a second, there’s this Buzzfeed journos went and doxed the Board Ape yacht club founders. Are they going to do that to Jonah Peretti who runs Buzzfeed? No. Are they going to do it to Arthur Sulzberger who owns the New York times? No. Are they going to do it to new houses that own Condé Nast? Absolutely not. Are they going to even mention them? Do you even know these people’s names? No, you do not. Why? You know the names of all the tech people, because media corporations are at war with tech companies and the owners of media corporations are, it’s not that they get good coverage. They get no coverage. They’re just blotted out. All this rhetoric about, oh my God, we need to investigate the powerful people. These journalists, you can just think of them as dogs on a leash.
They are hit men for old money and as I said, you can turn them to stone by just asking them who’s their owner. And you know what happens, they stop engaging on Twitter. You literally just turn them to stone. Why? Because whether consciously or unconsciously, they were either pretending to be decentralized, and unaccountable or whatever, and their own person. But when you mention their owner, they’re like, “I can’t believe you’d bring my boss into it or whatever.” And you’re like, “Oh, you have a boss. How about that? Oh, it’s so interesting. How courageous that you go and investigate your competitor, but not your boss.” Oh. And it just boom, the entire facade falls. You just hit it from the right angle, as we talked about, you rotated from the Z axis and you’re like, these are just corporate pons.
And actually the only journalism is citizen journalism. Because that’s independent. That’s Substack. That’s newsletters, that’s something where that person may be flawed, but they’re writing their own voice over their own time. They can’t be canceled and they actually have an incentive. That’s the opposite of the school of fish strategy. Remember our thing about how tribalism, if there’s an economic incentive, all those folks have to build their own newsletter distributions. So if they say exactly the same thing as everybody else, then they’re not unique. So they actually have an incentive to stand out versus to fit in.
Shane: Do you think that pushes people to more extremes in an effort to stand out? So I’m going to just push the right and the left even further.
Balaji: Yeah, Yeah. So, there is that aspect. However, I think it’s complicated because in the absence of the people who complain about filter bubbles are complaining that there’s more than one.
Shane: They just want to be the filter bubble.
Balaji: They just want to be the filter bubble. They want the, I love Lucy environment of 1950 where there’s just one telephone company and two superpowers and three television stations and everybody’s brainwashed the same thing. And by the way, that’s a very atypical environment through history. That level of centralization of power was actually very atypical. If you go further back in time, it’s highly decentralized in an obligate way because you didn’t have the ability to broadcast to the entire country at the same time, telegrams and all kinds of things. You could send a few bits of information, you go further back it’s letters. When it’s letters, you just have to run a decentralized system. But because this group over here in Montana might not even hear news that the war was over until, 20 days later.
Shane: So why would the political system get behind this? Because if I’m in power, this is a threat to my power. Why would I get behind any of this?
Balaji: Well, the good thing is the current political system doesn’t select for competent people. So you’re right that it’s a threat to their power in the same way that a new startup is a threat to a nepotist inherited corporation. But because they’re a nepotist and not a mediocrat, they’re not going to be able to wield that power and money in an intelligent way. And I’ll give some examples. Every single thing over the last 20 something years, since certainly I’ve been an adult and probably roughly contemporaries, 9/11 was a surprise. WMDs in Iraq, that was a surprise. The collapse of Bear Stearns and then Lehman Brothers, that was a surprise. The Snow and Revolations were a surprise at least to the public and probably much of the elite. Trump was a surprise, COVID was a surprise and Jan six’s Surprise and Afghani is a surprise and everything’s a surprise.
And of course, all the rise of technology where the stuff was dismissed as a fad and a bubble and it’s never … even in 2013 they were of articles like Facebook will never make a significant profit. And yeah, Facebook has had some recent troubles, but Facebook did make a significant profit. And in 2013 they were saying that this was not going to be the case, fad and bubble. Then just a couple of years later, they switched from fad and bubble to what? Threat to democracy. So the consistency is the negativity where at first they’re dismissing it because it’s never going to work and then they’re dismissing it because, or attacking it because it’s too powerful. But what is missing is essentially the world’s going to change and I’m going to get a share of that new world by taking some risk in it.
Because the moment you actually start, one of the things about the establishment is they can only perceive things that are around 50% popularity, at least double digit popularity. Anything that is too unpopular that’s at 0.1% or even less than that, they don’t have the right mindset that this could grow quickly. That’s why even all their discourse, for example, the rich, the poor, the middle class, the developed world and the developing world, that actually embeds as staticness in it. You know why? Because it’s not necessarily about the rich, the poor and the middle class. It’s about the ascending class and descending class. It’s not about the developed world and developing world. It’s about the ascending world and the descending world. That’s about rate of growth, not simply about statics. So just to elaborate on that. Your Brooklyn woke is actually rich, but in the descending class and a kid in India or Nigeria with a smartphone may be poor, but they’re in ascending class and they’re part of the ascending world.
Whereas you might be rich in San Francisco, but you’re part of the descending world, right? So rich versus poor and middle class, that’s static, ascending and descending is dynamic. And they think of things as static. Their fundamental frames on the world are incorrect. Because of this, once you actually start taking that seriously and once you start saying, okay, the world could change quite rapidly therefore I need to have radar that is alerting me to all the ways the world can change, what happens then? Then you have to start actually building up the skills of the venture capitalist or the angel investor or the startup founder, where you have effectively a dashboard where you’re looking at hundreds, thousands of possible and you have to score them in some way and figure out which of these rare events, this one is going to go to zero, but this one’s going to level up because you can’t invest in all of them.
So, it’s fundamentally a capital allocation thing and it’s hard. That’s why angel investing it’s actually much harder than it might seem. Those are actually the skills that you need to be a leader in this environment. Why? Because just to take the example, this is pure real politic. When did China censor the internet and build a great firewall? They did that early on. The golden shield project, they call it. When did they sensor social media? Relatively early on. When did they censor cryptocurrency? Relatively early on. The U.S. establishment is on the back foot on internet censorship on media, censorship and so on. And it’s a difference, the Chinese sort of bought censorship at the seed stage or series A stage. And the U.S. establishment has tried to institute censorship at the post IPO stage which is 1000 X or 10,000 X or a hundred thousand X cost. After 3 billion people have gotten smartphones and social media. Now you want to censor them.
Shane: It’s like keeping the genie in the bottle is easy, but once it’s out of the bottle, it’s really hard to put back in.
Balaji: That’s right. And the reason I use the VC analogy though, is because it’s about the cost of that level of censorship. That actually starts to rise dramatically. Seizing all the smartphones is infeasible. Turning off the internet is now infeasible. If they wanted to retain control and they were smart enough to realize that this was a threat to their control, they would’ve done what China did and Institute centralized choke points at various areas and take the PR hit and so on of doing that versus just try to retrofit that. Now to be clear, I’m not saying they should have done this. I’m saying that if they had been lawfully involved like CCP, that’s what they would’ve done. Now they’re pining for wistfully like this Jack Goldsmith article where he is like, we should have censored like China, but they weren’t farsighted enough to do that.
Balaji: They didn’t use their power to compound their power in a smart way. They didn’t use their money to compound their money in a smart way. And there’s different ways of thinking about this. One thing I actually, I was able to put this in words the other day, I think it’s actually, I thought it was insightful. At least I hadn’t thought of it in quite this way. Essentially before let’s say 1990 around there, but then you can push it back further, there’s advent of cable news and so on. But let’s say before the modern internet and especially before social media and then cryptocurrency, yeah, you had the nominal right to free speech, you had the nominal right to free markets, but you couldn’t just go and broadcast to everybody. You couldn’t easily start a company because the capital cost of doing so were so high.
Balaji: There’s that quote by Anatole France, which is, “The law and its majestic equality forbids rich and poor alike to sleep under bridges, to beg in the streets and steal loaves of bread.” It’s essentially saying, the law is fair. It forbids both the rich and the poor to beg. And of course it’s a sarcastic indictment of a truly just law would redistribute the money and so on and so forth. Leaving that aside for a second, if you just take that concept and you invert it, or not invert it, just kind of apply it in a different way, the majestic equality of the law allowed the rich and the poor alike the right to own a newspaper, start a company or speak freely. So in theory, everybody had the right to free speech and to start a company and so on. In practice only guys like Sulzberger or his father’s father had that right, and a parole could just go talk to their neighbor or complain or yelled at the television set they didn’t actually have broadcast authority.
Shane: That’s because distribution was sort of controlled, right?
Balaji: Yes, exactly. And one way I think about this is, remember the Unabomber?
Balaji: Yeah. So the Unabomber basically in the 80’s and early 90’s he sent these mail bombs around and why? Well, in 1995 he wanted The Washington Post and MIT to publish his op-ed. So, this guy at that time distribution, within our lifetime which is less than 30 years ago, distribution was so scarce that he had to kill for the distribution. That publication was so scarce that he felt that he had to kill somebody. So what that meant is there wasn’t truly free speech. There was a small oligopoly of a few dozen folks who effectively hammerlock on Hollywood and media corporations and so on and so forth and the average joke couldn’t just set up a website and a YouTube and just go and stream and so on and so forth.
And there’s a whole licensure apparatus by the way, around that. TV licenses to broadcast and you might meet a radio license. And one thing, somebody I had this little debate with them about a year ago, they were like, not sure how often it needs to be repeated, but you don’t get to be exempt from regulation because you built the same product using a different data structure. It’s basically like, “Haha grow up tech bro, your blockchain stuff doesn’t change anything you mean exactly the same rules that we had in the 1930s”
And my counter argument was, you don’t need to stamp to send an email, you don’t need a TV license broadcast on YouTube, you don’t need a radio license to do a podcast and you don’t need to use old regulations to constrain the future.
We recognize that automobiles differ from horses, that airplanes are distinct from birds, that submarines and surface ships are not the same. And so blockchain started scarcity with the internet with information that change fundamentals, old ways of thinking do not directly apply. And so that’s like the rejection of the concept that an old rule should simply be applied thoughtlessly without concern for whether or not the technological environments change.
The point is distribution, exactly as you said, was totally locked down. Now it’s opened up and what’s actually happening if we think about it is enormous powers of speech and distribution were granted to people and now people are mad that they’re being used, that they’re trying to take them away. But on net, there’s been a massive decentralization if you were to graph it, number of people who can broadcast or whatever, it just went voosh like this. And now yes, there’s the attempt to retrofit these speech and thought controls on it.
It’s like a graph that went all the way up into the right in terms of practical speech and then they’re trying to jack it back down. But it’s going to come back up because the attempt to deprive somebody of the practical right, you just gave them that voice. That’s why people react very strongly to it being taken away. They just got it. It’s like the franchise, frankly. Just getting the franchise, people will react very strongly to being taken away. And so speaking rights are like voting rights in that sense.
In fact, arguably, if you can’t speak, you don’t actually have a true voice. Because the guy who writes a headline on the day of the election then controls a million votes and you have only one. That’s why they say our democracy in the possessive sense because somebody who runs a major newspaper or media corporation controls many more votes, it’s their democracy, not yours.
Because they’re basic points of aggregation. So that brings it back to your kind of original point, which is, we were talking about the Canada thing. It’s going to be very difficult for them to take away these rights because they’re poorly organized. They don’t anticipate future outcomes. They’re spending massive amounts of money without results. And the thing is if you’re spending a million dollars in everything, it doesn’t matter how rich you were born. You will eventually become poor.
If you’re a terrible capital allocator. If you’re spending $300 million on bus lanes like San Francisco, if you’re spending billions of dollars on planes and ships that don’t work like the US military, the F-35 or the Zumwalt or the probably stuff that you send, the LCS, rail gun, all these things are just super expensive boondoggles. If you’re doing that and somebody else is a hundred X more cost effective, even if they start from a lower base, they’re going to eventually beat you.
And I see this all the time with start-ups that they’re small, but they’re focused and they can get more work done and they can get more output because all that bureaucracy and cob web stuff at a large company ends up not being a moat, but a ball and chain, right? So what I think happens with Canada is a pyrrhic victory for the state, because remember they use this stuff on these truckers, but they pissed off the global tech and crypto elite.
It’s kind of a three layer thing, where you have the national elites who are oppressing these proles within their countries. Then you have the international, now a crypto Web 3 group who is at a technical level that is above, those politicians are just effectively trying to seize the technologies that this group above has created. And this group above does not want those technologies to be seized. That’s actually really the battle between the network and the state is where things are going to be.
And I think that Western governments will lose. It’s going to be super nasty and super gnarly, but on balance, they will lose because their ideology is not one that attracts the very top talent. And this is actually the same problem that CCP has in some ways, where they’re driving away Jack Ma and they’re driving away Pony Ma, and all the founders of tech companies are finding that their life’s work is either being expropriated or they’re being blamed for it. So lots of the best are trying to leave.
And the NYT ideology has repelled Barry Weiss and Jesse Single and Glen Greenwold and Matt Yglesias and also Marc Andreessen and all the tech founders and so on. And all that’s left are the basically middle manager types who are fine, I don’t have any beef with them per se, but they’re not founders. They’re not creative. They’re literally there to just follow orders. And they’re taking directive from NYT, but the really intelligent people aren’t.
And even actually some of Sulzberger’s loyalist employees are now into Web 3 and crypto. And it’s the tribute that vice pays to virtue right, right? Now, these people are in it. Why? Because they took all these shots at it, they couldn’t kill it over 10 years. Now they want to make some money from it. And the question is, is that good or bad? I think on net, it’s good because it’s them submitting to the frame. They’re not going to be able to ban it now, it’s going to be harder to ban it.
If enough members of the elite have defected and are making money from this, only the retro authoritarian chunk, which is only about 30% of the population, it’s not 75%. They start to lose that, right? Essentially in Canada, and I don’t know how you felt it on the ground, because you’re immersed in that social network, Iit’s like micro climates, so triggers like micro climates. So my perception might be different from yours.
But all the commentary in my micro climate was probably 95 to five against Canada. And the most positive I could find on the Canadian side was at best 50/50 for what Canada was doing or 40/60, but an insistent 40/60. I don’t know what your perception is on that.
Shane: Wait. How would you handle that? How would you handle what happened in Canada then?
Balaji: If I was Trudeau or something?
Shane: Yeah. You’re Trudeau, you’re in charge of the government, how would you have handled this differently?
Balaji: It’s a good question. Because okay, now you’re not just second guessing. You’re actually there to do something, and so and so forth. The upstream is what you actually want is to build an aligned high trust society where you have no bless oblige, you have the elites are from the people and part of the people and the people feel that the elites are on their side. And it’s like cheering the quarterback who came up from your small town, as opposed to this alien Davos class that seeks to rule you rather than middle managerial elite, rather than actually feel your pain.
People talk about representativeness, but there’s representativeness in terms of an arbitrary demographic cross section. There’s representativeness in terms of people truly feeling these people are generally representative, and what’s happened is that Goodheart’s law thing where those proxies for representativeness I feel are being taken away. Where people don’t fully feel, despite the mechanics of the whole system, that the government represents them on many of these issues, right?
It’s a little bit like asking there’s guy who’s in the last stage of bankruptcy and you’re like, “Okay, what should you have done in that last stage?” And my answer is there are things that should have been done five or 10, or 15 years ago to stop it from getting to that stage. Because if you are just the last stage, you’re going to fall off a cliff in some way because your financial situation is so precarious that anything can push you off. Go ahead.
Shane: You’re in an increasingly difficult position as a result of all of your previous decisions.
Balaji: That’s right. So what are those previous decisions? I’ll give two views on this, okay, because a very important question. One view is something I’m drawn to even if I feel it’s only partially explanatory and that is just that civilizations rise and fall. Just like human beings, they’re born and die. And they have their solid years and then they’re old age and so on.
And so we saw the rise of the West and now we’re seeing the fall of the West. And we saw the fall of Asia and now we’re seeing the rise of Asia. And the countries and the regions that were on top in the 20th century, which we’re broadly speaking the US and Western Europe and so on. They were the most free and the most capitalist and so and so forth are becoming the least free.
And the ones in the East that were communist or they were poor or what have you are becoming more powerful. A lot of them have flipped to becoming nationalist. And they’re the ones that are winning these conflicts. And certainly, I’m not a supporter of what Russia is doing right now, but the power level is shifting, right? We’ll see what happens with this Ukraine thing, just to talk about that for a second as of right now, and this could be completely fog of war. So I have very low confidence in this. I don’t speak Russian, every source I’m seeing is through translation.
But at least as of right now, last I checked, there’s reports that Russia just blitzkrieg the Ukrainian military. Just blow up all kinds of stuff, has taken all kinds of cities and so on and so forth. And it doesn’t look like there’s going to be significant troop sent by the EU or probably the US or EU because that would be a direct super power to super power confrontation. They want to avoid that and it’s all proxy war. So maybe they will fund the Ukrainian resistance or something, but Russia can win, they won in Chenia, they won in Syria because they’re more ruthless, right?
So it’s quite possible that Russia actually just beat the US or at least beat a client state of the US that thought the US would back it up. And it would be if this happens, I don’t know if this is going to be the case, it could be totally different. But if this does look like Russia fights a quick war, gets new territory, the world yells at it, they sanction it and then maybe Russia does a counter cyber attack or something like that, who the heck knows could be really crazy in the weak to come.
But if the dust settles and Russia is control of that territory, even after the sanctions and so on, then that’s a second US backed military that folded in days in six months. And so that’s a flippening of the 20th century where it’s not like… The US did have losses in Vietnam and so, but broadly speaking, the US generally won battles in the cold war, generally won economically.
You could argue that’s just fitting it retrospectively, and it came on strong the last 13 years or 10 years or whatever, but net-net, it was stronger, right? And now I feel like some reversals not withstanding the direction line is down into the right for the West. It’s down to the right as percentage of global GDP as potentially of population and military strength and so and so forth.
And the thing about that is if you say that people are like, “Oh, you’re anti-American. You hate the West.” That’s not actually it at all. It’s more complicated, I can give a few one liners on it. One is that woke America is to America as Soviet Russia is to Russia. It’s like a different thing. It’s like an ideological different thing.
For example with India, the India of my youth, if I said it wasn’t globally competitive, that’s just assuming the fact what I wanted or didn’t want the land of my ancestors was a basket case. Okay, that sucks. You know what? Let’s figure out what to do. Now, it’s rising, and that’s really surprising. There’s incredibly surprising to me that happened. But I’ve incorporated that into my mental model. That doesn’t mean I’m pro or anti something, I’m just trying to see the world as it is, right?
I prefer a situation where everybody was rising and everybody was doing well together, that’s just not the way the world is as distinct from all, right? So in this flippening, you’re seeing the West declining, you’re seeing the rest rising and it’s not exactly as clean as those countries that are doing well in the 20th century, are doing poorly this century and vice versa. But that’s like first order, just first principle component that’s kind of what it is.
One way of thinking about that is a lot of this countries are effectively vaccinated needed against socialism or communism, which took them down in the 20th century, right? How could you make China unproductive? Well, the mind virus of communism there would make these incredibly productive people unproductive, right? People have reckoned by the way that what happened to the Soviet Union, the Russian population might not have been 104 million. It might have been like 500 million without the red terror and the wars and all this stuff, which was in part caused by communism. You might say, “How are the wars caused by communism?” Well, there’s a good book called Stalin’s War. Have you ever read that?
Shane: No, never.
Balaji: Really good. Sean McMeekin, actually, all of his stuff is good. Like a really good historian, you’re like, “Okay, I didn’t actually understand that dimension to the past,” and you can obviously poke in their citations. And he writes in a very engaging style, I recommend it. But essentially he says, look, we normally think of world War II as Hitler’s war and that’s a valid perspective, but he thinks of it as Stalins War. Why? Stalin is actually involved in every theater.
The Soviet Union bordered Asia, it was basically near Japan as well over here. And something that I wasn’t actually aware of is Soviet Union was actually extremely involved on the whole Japan-US thing. After the Russo Japanese war of 1905, which was actually under the SAR, okay? The Japanese beat the Russians, and in fact that was the first time a non-white power had beaten a white power. Is this huge thing for anti-colonialism or whatever worldwide, right? Huge humiliations to Russians.
They took the Japanese very seriously. And even after the communist takeover, they inherited the geopolitical issue where before it was SAR versus Imperial Japan, afterwards, it was communist versus Imperial Japan. But they thought of Japan as like a very dangerous rival, right? An overriding foreign policy priority for the Soviets was to get their two big rivals like Japan and the US to slug it out between themselves versus their nightmare scenario, which would’ve been Germany attacking from the east and Japan attacking from the west, and taking the oil.
And so a good chunk of the book talks about how Soviet spies who are now all revealed by the way, it’s funny, there was a long period of time where it was considered paranoid to talk about Soviet spies in the US government. Then Venona came out, which you’re familiar with probably and showed that actually a lot of Soviet spies were in the government and now you can match them up with the Soviet archives, which McMeekin did during that period of opening.
And you can show that actually Stalin used spies to help push the US to confrontation with Japan and have them slug it out versus having a flanking attack on both sides. So the reason I was saying is communism helped cause those wars, right? Communism reduced the Russian population. Communism was this huge virus that affected Russia and China and much of that part of the world, Vietnam. In Cambodia, they killed all the people with glasses, probably set back industrial civilization by decades.
And now fortunately, that part of the world has shrug that off to greater or lesser extent. And some places, obviously, China’s still nominally communist. You can argue it’s genuinely communist because they still have state control over capitalism, but they’re nowhere near as communist as they were under Mao where they were. Right now you can argue they’re really a nationalist/socialists like nationalist/militarists versus communist.
Vietnam, all these other countries have gone into the China model because it’s worked. And so now they don’t have the mind virus, but Western Europe and the US does, which is not communism, but wokeness, which is all symbology. Not biology but symbology, it’s all online symbols and likes and artees. It’s all the word. Anything you can do on a screen, if you can’t push a button, it doesn’t exist. And you know what, it’s actually very optimized for that environment of shaming people online, of mass mobs, of hysterical online behavior.
It does work in that environment, it’s optimized for social media. But as the terrain shifts to pseudonymous cryptocurrency, as it shifts from the open planes to mountainous terrain, it’s like Napoleon in the winter so to speak. I know I’m mixing metaphors, but that army of wokeness, that ideology can win on centralized social media platforms English internet. But once you start going pseudonymous, when you start having cryptocurrency involved, once you start going international and having people abroad, it can’t win when the terrain changes so dramatically. It wasn’t optimized for that kind of conflict. It grew on Western social media.
Shane: Just to be clear, when you’re talking about winning, you’re just talking about popularity, you’re not actually talking about effective policy?
Balaji: Oh yeah, of course. That’s right. Yeah, sorry, sorry. Let me actually describe my mental model on this. So I have this thread on what I call the social war. So 1861, the blue and the gray, the Unions and the Confederacy, they were not just ideologically separate, they were physically separate. There was this boundary line, I think the Mason Dixon line that separated them, yeah?
Balaji: So the victory condition was clear, which was one group would invade the other group. And by invading their territory, they would also demonstrate their will over their group and get them to surrender and conquer them. Fast forward to 2016, 2020 blue versus red in the US is fractal. It’s not geographically split like it was in 1861. It’s not like Republicans are on one side of a line and Democrats are on the other side.
And it wasn’t exactly like that in 1861 either, but Republican majority was on one side, Democrat majority was on the other side. It was territorial enough, right? So here though, it is not obviously territorial. It’s something where it’s urban areas are mostly blue and the rulers are red, but it’s very cheek by jowl, very fractal.
And even if one side wanted to declare war on the other side, any bomb will blow up a bunch of their own people versus others, right?? So that’s why physical war in the sense of formations of the 1800s or 1900s is hard. But if you go forward in the digital space, these two factions are highly separated. And that is actually the theater of war.
Shane: It’s in words, it’s in actions and signaling.
Balaji: Well, it’s actually, when you cannot invade their territory, victory means invading their mind.
Shane: And then that comes back to distribution.
Balaji: It comes back to distribution, exactly. That’s right. There’s that saying war is politics by other means, right? If you think about it, let’s say there’s a bunch of drones coming at you, okay, a hundred drones. You can shoot each one out of the sky, but much better is to be able to hack them, hit enter and have them drop out of the sky or even flip around on their own sender, right?
Similar if there’s 100 soldiers coming towards you, you can shoot each one of them, but better is to be able to hit enter, propagandize them or drop leaflets on them in earlier era, get them to believe they can’t win. And then they just either drop their weapon or even around of their own government, do a strike at home or something like that, okay? That’s information war.
And so we know that’s always been a component of physical war, but now it’s like the whole thing, okay? It’s literally just fighting the information war online. And then eventually when you hit enter and print it out, it’s a riot here or it’s a protest there. This cloud swirling information war gets printed out into the physical world just like you might print out a document or 3D print something, this conflict gets printed out. But it’s fundamentally a digital conflict swirling here.
When you think about these Twitter threads, you’re talking about thousands of people. It’s literally like everybody got teleported into this massive online battle arena. And it’s like the medieval era where you have swords, clanking back and forth and people love battle. They just love it on some level, right? It’s stimulating, you know?
And they can just die a thousand times. And there’s actually real life consequences. And yes, it’s tearing apart society, but it’s engaging. There was that quote from 2016 or whatever, which was very ironic. Yeah. It’s like, “I feel bad for our country, but this is tremendous content.” The social media platforms and the mainstream media platforms are both things that have essentially profited from the fight club.
Shane: It reminds me of that scene in the Gladiator where he looks up and Maximus goes, “are you not entertained?” Right?
Balaji: That’s exactly what it is. I now understand the gladiatorial arena in a different way. I understand what the decivilization and the barbarising process looks like as you go from a civilization where there was common shared myths and common shared morality to one which is just tribe versus tribe. We’re actually seeing that happening, right?
Balaji: And it’s crazy. But coming back to the social war concept, once you map this and you realize, “Oh, okay, this is not a peace time state. It’s a war state.” And it’s being waged in the digital space because it can’t be waged in the physical space or it’s only printed out in the physical space. This is why, by the way, you see these blue versus red punch them up between Antifa and proud voice, whatever. What people are doing is printing out that dispute. They’re like, “Let’s go and rumble over here.” Right?
And they’re announcing a slug fest, okay? So then when you think about what’s happening here. Okay. The wind condition is conquering their mind. And so then that’s where unbanking, canceling, deplatforming, all that stuff. They’re tactics in a social civil war, right? They’re meant to make a node on their side, wink out or change color.
And now you can see why there’s so much focus on hashtags and bio and getting people to use certain language. If you can force them to say a word or not say a word or put a phrase on their site, okay. Or have some shibboleth, then you’ve invaded their mind. And so it’s like digital capture the flag. You are essentially by virtue of seeing people signal something, by virtue of their virtue signal, you are seeing that you have either, at least, even if it’s like an economic gun to the nape of their neck, even if it’s under duress, nevertheless, it’s still controlled territory.
Now, the thing about this is that flippening is, I don’t know, if you’ve ever seen the maps of Imperial Japan at Zenith or Nazi Germany at Zenith or the Soviet Union at Zenith. There’s a period where I was like, “Oh boy, these guys are really winning a ton of territory. Holy shit.” Right?
Now, what happened effectively was like a bandwagoning counter attack where everybody they had pissed off ganged up on them and beat them down. So what I think has happened here is the Western establishment. As I mentioned, they’ve taken on the right, the central left, tech, Russia, China, to some extent India, Israel and a bunch of other smaller countries, like obviously Iran and Hungary. And they’re mad at half of Britain because it Brexited, and they’re mad at France and they’re mad at various countries on immigration policies. They’re mad at lots of countries.
When you’re fighting everyone at once, even in this woke expansion. This thing just grew to this huge thing. It’s like when the Nazis were in 1940, right before Barbarossa, they were at their Zenith. Imperial Japan was at Zenith, but now, if they’re winning, they’re winning empirically. Okay. And one way is, as I should put my finger on this with the Rogan thing, yes.
They were able to grind out a cancellation of Rogan, okay. They ground out a victory, but it took a lot of ammo. It took a ton of ammo and crucially, there’s a big difference between this versus previous cancellations, which is they could only win it like 55/45, not 95/5. How can you see that? Because they couldn’t cancel the people defending Rogan. There were two numerous. It’s 95/5, then anybody who even speaks up in defense of the person being canceled after they finished with this guy, then they boom, gang up in that person and gang to that person. So nobody can even speak up in defense and everybody gets the signal. They understand how outnumbered they are, right?
When it’s 55/45, they just don’t have the numbers to swarm all these different people. There’s too many targets to shoot at and other guys have troops on their side and it’s really a social war by the way. That’s not a debate. It’s not the marketplace of ideas. It’s battlefield of ideas. Because there’s some ideas which by their nature are not about the feng shui of an apartment and where you lay out the things where it’s all, anyone can have this preference or that preference, but they are like monotheism. You have the belief or you don’t, they’re optimized for maximum virality. They’re the most mimetic mind viruses around like the memes that Dawkins would talk about, highly fit memes that are meant to spread and then block off the spread of any other competing ideas.
They’re literally nationally selected for this in our brains. Just like computer viruses, these are mind viruses. And you know what the wet market is, it’s Twitter. We’ve all got our brains dunked in a vat, which is Twitter. And so the most contagious possible things are selected there, where you call them viral. We call them viral, contagious is another word.
And so wokeness is one of them, but it was only selected in that environment. And when we change it to an environment of cryptocurrency and solemnity, there’s lots of roadblocks and stuff that arise and it makes it harder. I think Web 3 will be Waterloo for wokeness.
Shane: Talk to me about that.
Balaji: So Waterloo is like famous thing, Napoleon meeting his Waterloo, right? This is actually after this powerful army won all these victories eventually was defeated. And you can argue whether Waterloo is the right of analogy or Napoleon invading Russia and there’s Russian winter and terrain that’s defeating it. But fundamentally Web 3 is much tougher terrain for wokeness than centralized social media, and here’s why.
First is with pseudonymity, you can reboot under different identities and so cancellation is not permanent. There’s a very strong moral case for pseudonymity because it prevents both cancellation and discrimination. So you don’t have to give a single inch on the morality of it. If somebody says, “Oh, we need to count X, Y and Z group.” You can say, “Actually, I don’t want to implicitly discriminate. I don’t want to have unconscious bias because there’s a lot of papers on that.” You can cite the same science or quasi science, whatever. You can argue if it’s good science or bad science. If it’s bad science, you’re not going to be able to cite it. If it’s good science, then you can both cite it, right? So you can use the same premises.
So the same stuff that people cite for the purpose of say blind auditions or for band of box where you’re not supposed to be able to ask about somebody’s criminal history or for not asking about marital status or not asking about immigration status. I actually agree with some of these things where some of these things are irrelevant to doing the job. Like their immigration status, that’s just a stupid technicality. Of course you have to go through all the legal paperwork, but that does not affect whether they can calculate an integral or not, right?
You take that to its conclusion and you actually have this thing that you can get a left right coalition on for pseudonymity and you can make a moral case for it because that moral foundation is very important. You also have a moral case for decentralization. And what that means is that if a centralized thing is compromised, well, the backend is open state in crypto. That’s the thing, it’s not just open source, it’s open state and open execution. Do you know what I mean by that? It’s not just the code that’s public, it’s the data set and you can run a node and you can watch every single op code execute.
So if you’re concerned about a non-transparent algorithm, you can actually see it executing for yourself. What that means is people can’t just lock you out completely. If you want to, you could clone the equivalent of the entire backend of Twitter and then do your own chain if you had enough of a community and so on, right? So that’s another check. So pseudonymity is one, decentralization is another.
Third is censorship resistance. The conventions of the whole space, a huge amount of the whole thing is based on the fact that one person cannot prevent another from doing it, right? Fourth is a built in anti-establishment thing where the point of Bitcoin and the point of Ethereum is that they’re not inflation, right? They’re not Wall Street, they are outside it.
Now, you might say, “Oh, everything gets co-opted. And so on and so forth.” But that means a co-optor has to be smart. Those institutions have bled so much talent that they’ve just gone into cryptocurrency instead, right? Fifth is international. So all these internationals that are piling into Web 3, why? Because if you are in Brazil or India or Nigeria, you’re not going to get a visa to the US very easily.
You’re probably not going to be able to get onto wall Street, but you know what you can do? You can boot up on Web 3 and start trading. And so there’s this massive international constituency for this, which has a lot of intelligent people that have just come online and there’s a huge force. I cannot overstate how big that is, right? So it’s a huge army of folks where the wokes are not, they’re used to thinking about it in terms of domestic politics. They’re not actually used to thinking about people overseas, anything other than sympathy objects and statistics.
And I see this with discussion around China in the 2010s and India now, these are actors in their own right, right? Nigeria’s now coming up in terms of it’s tech scene, South America has now concepts of its own. El Salvador’s going Bitcoin, lots of south American countries are talking about Bitcoin. All these countries are now waking up and doing things other than being pawns in this great game. And so I mentioned decentralization, I mentioned pseudonymity, I mentioned sensor positions, I mentioned the international.
There’s also another interesting aspect, which is the monetizability vanishes if you centralize it. So the entire woke thing is gain control of an organization and then just start looting the treasury. And it’s very similar to any religion becomes a business model, becomes a scam. And it’s like John McWhorter and others, many others have talked about wokeness as essentially ultra Protestantism and look, I’m not anti Protestant or anti Catholic or anything like that. I consider myself ecumenical, but a religion that doesn’t admit it’s a religion or know it’s a religion yet makes moral claims that are at the level of essentially a Christianity fork.
And Tom Holland talks about this in his book, McWhorter talks about this in his book. There’s a real parallel of many things from original sin to climate change is hell and so and so forth. Even if you think some of these things have scientific basis, the cumulation of them is not something where if you think climate change is happening, one answer would be to do nuclear power. But no, instead it’s evidence of how we’re sinful and we have sinned and we need to actually destroy industrial civilization and your a sinner, and it’s an argument for social control. And it’s like the Puritan mindset of new England.
COVID is another excuse for state control and you must submit and all the type of stuff where it’s not solved pragmatically, it’s solved in a way or quasi solved in a way that is state controlled, right? So if you think of this thing as this religion, well, it’s going to have trouble evangelizing here because if it does centralize and control something, totally then the value, poof, vanishes. Because the blockchain it’ll just get marked all the way down to market. It’ll get marked down to zero so you grab it, you centralize it and then it’s like sand through the fingers.
What have you grabbed? You grab something which it dissolves, if you fully centralize it, right? And that is actually a new thing where you cannot totally capture it and seize it. It’s not like seizing an oil field where you have total control over it. If you have total control over it, if all Bitcoin miners were seized, then Bitcoin wouldn’t have value, right? And so they would’ve seized nothing.
Now, they may still do it. They may try to do it because it’s like a political thing and they’re spending money on it effectively, and they’re not trying to make you on it. Russia took a hit of 50% or something to their economy for this invasion. Some people do things that are noneconomic, that’s actually an important vulnerability, but the defense against that vulnerability is you have so many different chains now it’s going to be hard to take out all of them.
So you have multiple levels of decentralization and then you have one thing else, which is very important. And that is that Web 3 is splitting. On the right, it splits the nationalists from the libertarians. On the left, it splits the authoritarians for the libertarian left. And what I mean by that is the people who are voting against cryptocurrency are typically on the national security right or the top down socialist sympathetic left. The nationalists and socialists are against cryptocurrency, but the libertarian left and libertarian right are for it. And I think there’s more ability on both of those sides cumulatively.
And so you’re starting to see an important split develop on the left on this, where an important sense, Web 3 is outbidding the state. That is to say, it is providing more actual money and community and all of these things than the state’s pretense to provide socialist goodies. It’s not actually providing that, but Web 3 is, right? So finding a lot of people who are redefining socialism as an opt in Web 3 community.
I would think of that as the opposite of socialism, okay. But it’s one of these funny things where I have this one liner. There’s a really good thing when both sides feel they’re getting a screaming deal, that’s the basis for a sustained relationship. You want, that’s actually the opposite of the Justin Trudeau situation or whatever where both sides sides feel they’re getting a terrible deal, right? But both sides feel like got a great deal.
Progressives, “Ha, ha.” Finally got a cut of all that tech money, at last libertarians learned how to share, right? And libertarians, “Ha, ha.” Finally got progressives to cut out that state regulation, at last they learned how to share risk. And Web 3 might just put the world back together again.
Shane: What advantages do established democracies have and what disadvantages do they have?
Balaji: Okay. So we have to get into the definition of what is a democracy, right? And ideally, you want a definition because right now, implicitly, it’s basically just whatever the Western establishment says a democracy is a democracy, right? Canada can go and freeze all these accounts, but it’s a democracy, right? Whereas a country they don’t like is a flaw democracy or not a democracy. And there’s different definitions.
One I’ve been flirting with, there’s a few interesting concepts. One concept I put out there is 100% democracy versus 51% democracy. So 51% is where we currently are in the West where it’s a minimum necessary level of approval. Okay. It’s a Fosbury Flop, you just barely clear the bar. And as distinct from 100% democracy or tilled 100% where you’re close to that, and you set that at least as an ideal where you basically want everybody to agree on something. And you’re like, “How could you possibly get that policy?”
And you’re like, “Well, first let’s agree that would be more desirable.” Right? And then once we agree that 51% democracy is the barest level, how’d you engineer 100%? Well, you would do it probably on the base of exit rights where you have a bunch of jurisdictions in the digital space to start and you freely enter and exit them just as people are entering and exiting dows, and then they have pockets of territory they hold around the world and then you see if you can gain egress there, right? And now before entering, you’re literally signing the social contract plus a smart contract. And you’re signing that on chain and it’s notarized and it’s videotaped so you know that you made full and free consent and you might be giving up some of your rights to enter this jurisdiction.
Maybe they can take some percentage of your deposit, for example, if you misbehave, whatever, but all those things are terms that are flagged up front that you consent to at the time of entrance, right? And that starts to actually get you to something which is like joining a company, right? The CEO is in charge and you make a tradeoff which says, “Okay, I’m having less freedom in return for more net benefit to myself,” and it’s also a net benefit to the CEO. It’s mutually agreeable and I’m agreeing to these rules. And if I don’t like it, at some point, I can leave or they can terminate me.” That works in the corporate world because you bring that thing, right?
So that’s one way of thinking about it, that you’re optimizing for the level of consent in society, the consent of the governed, as opposed to the procedure of an election that just generates 51% and swings wildly back and forth every four years and just makes people mad, right? A second concept is, is it even a democracy? Because leaving Canada say for now, but you take the US, okay, it’s got military bases all over the world, right? Did the people of Iraq vote for the invasion? That’s tens of millions of people. It’s actually a very important point, right? Insofar as the exercise of force is supposed to be bounded by some degree of the say of the people, can they vote the troops out, right? Can any of those Middle Eastern countries vote American troops out? Can any of the countries with American military bases around the world, can they have a referendum to vote them out?
Not to my knowledge, that’s not really being put on the table. And in fact, there’s a saying that for a long time, the US government makes sure that no party got too close to pushing for that, right? From either the left or the right. And our question is, is it democracy in the sense of, are most people in the government actually elected? No, they’re not. There’s something called the Plum Book and there’s a few elected positions like the president and the senate and so on, but here’s the thing, if those positions, like Congressman is changing every two years, president every four years, senator maybe every six years, if those positions are potentially changeable, how would the government continue to operate?
And actually, we realize a lot of that is ceremonial and the permanent government is what continues to operate all the way through that doesn’t change an election cycle. If the permanent government was changed out every few years on election, now maybe government services don’t work at all anyway, but services would just stop working because it’s like, imagine you could vote out all of Google every four years and there’s a new team, they just won’t even know what to do, right? That change of power would not be a continuous thing. There’s going to be a bunch of new people, there won’t be continuity.
So the actual government is what remains constant through elections and that’s the official bureaucracy of about 3 million federal employees, many of whom have career tenure. And if so, they were not elected and can’t be fired, okay? And of course, it’s the media corporations who, like Sulzberger, inherit it, all the way down. It’s the academics and they have tenure, they can’t be fired. They’re not accountable to electorate, right? They have enormous influence to all of this. So the actual permanent government is not actually accountable to election. So is it a democracy in that sense, right?
And then there’s the question of how much actual say you have, right? Because the guy who owns the distribution, as we just talked about, “Do you have the voice?” No, you really don’t. The guy who owns the newspaper has a million votes on the day of election and you have effectively one, which is why they favor the system that gives them informal huge amounts of power in a formal system, right? The newspapers aren’t written into the Constitution but their distribution difference is and that’s why they really like the current system, they don’t want to change, right?
So those are critiques of the thing as to you know, whether it’s “actually a democracy.” And then what’s a formal definition? So one is that 100% versus 51%, right? A second, and this is something I’ve been thinking about, is, it’s a fun definition, but if bitcoin is legal in a country, then it’s a democracy because it permits freedom to exit and freedom to transact, which there’s a great thread by my friend, punk6529, which said, “Freedom to transact is actually upstream of all of the rights.” I have put freedom to exit is upstream even of that, okay? Because you could agree to limits on your freedom to transact when you enter jurisdiction. But let’s say after that freedom to transact gives you, for example, the ability to buy a computer to broadcast online, right? To protest, you need to buy a sign.
Anything else you do, you’re going to probably need some economics, upstream of it, so if you can lock down freedom to transact, you don’t have these other rights. This is also a version of Hayek’s argument, which is that control over economics leads to control over everything else. So this is a fun definition, but it’s operationally a democracy if bitcoin is legal because then you’ve got freedom to exit with your money and freedom to transact. And if bitcoin is not legal or if it’s restricted, then it’s a limited democracy or flawed democracy or non-democracy because it is restricting that fundamental right of exit. What I like about this is very clean, okay? You can point to it, you can relate it to a global moral right, you can test it, an individual in that country can tell you how restricted it is and you can grade people on this pretty simple to understand thing, “How legal is bitcoin?” right?
The reason I say bitcoin specifically is it’s on track. We will see if it can actually get there because it’s got various kinds of issues, privacy and other kinds of things, but it’s on track to be the government of governments, global government. Because the world government people thought it was going to be like some UN thing or they thought world government is bad because you want to have decentralization. And this is like a pro-freedom global government that limits every government because it can’t spend too much and it can’t do certain things since people can, if they need to, exit the bitcoin blockchain, right?
One last observation, in many ways, the term democracy means American military ally to first order, right? That’s to say, if you look at the map, here’s an exercise, can you name a country that is a democracy that was not a former UK colony or does not have a US military base?”
Shane: I’ve never thought of it that way before.
Balaji: It’s a really good question, okay? Because you do the map that way. Germany, US military base. Japan, US military base. South Korea, US military base. India, former British colony, right? And you go down the list and there’s a few … You can say like Uruguay, I think, was neither, but it’s not that easy to find, right? The thing about that is if that is what it is to first order, if it is effectively a US military ally, then the whole thing takes on a different light. It’s a flawed democracy if it’s on the border, right? If it’s Hungary or it’s Israel or it’s India or if it’s a red state or a purple state that’s not as aligned with DC, the degree of democracy, as judged by DC, is how reliable they are relative to DC.
If you look at their sentiment, as expressed in these democratic rankings, versus how lockstep they voted in conjunction with DC and move their troops, it’s basically like one to one.
Shane: We’ve pursued a path of, it seems in at least the United States and Canada, where housing is an investment and not a consumption asset at this point and that implies that we want it to grow faster than inflation. How do you see the price of housing and what are the impacts?
Balaji: So I think of the last 10 years and maybe less several decades, but certainly the last 10, 15 years, as the state’s inflation versus technology’s deflation, so state is propping up mortgage prices. It’s bailing out GM, right? It is trying to keep the status quo intact, right? And subsidizing all of these jobs that would basically exist otherwise. And technology is trying to bring down car costs with autonomous cars and bring down shipping costs and automate this, disrupt that, bring down housing costs with drone construction and robot construction and all these things. It’s this tug of war between inflation and deflation, right?
And so what that’s resulted in so far is that all these things in the physical world have become more expensive, especially education and healthcare and housing, especially in, not everywhere, but in blue areas, housing became expensive. Everything digital has deflated. And the combination, the two evened out over the 2010, so people were like, “Whoa, where did the inflation go?” But if you look at the graphs, some prices went way up and some went way down. And in a sense, the Fed is upstream of all of that. It’s upstream of every political issue. Medicare for all, why? Because the Fed printed money and that led to healthcare prices going up in part. There’s other factors as well.
The Fed printed money and that led to all the YIMBY stuff, all housing went up. And the Fed printed money and that led to among other things. Education prices going up, and of course, the student loans and other things. And that was loan forgiveness. So that’s upstream of all this. And so where do I think it resolves? I think, eventually, this reaches a breaking point and we may get here by 2030 or 2040 or so on, right? But everything in the physical world can be thought of as a printout, okay? Let me explain. Obviously, you can print out a document, you do it online, you hit enter, you print it out of a printer, right? You can print out a book, okay?
Less trivially, you can print out food by hitting a thing on Uber Eats or Grab and there’s a process where it comes to your door. Now, moreover, I’m not sure if you’ve seen this, but there are robots that do robotic agriculture. There’s robots that will package things to the factory and there’s even sidewalk robots that will do robot delivery and will bring it to you. In theory, every single step there, you could eliminate any human and have the entire thing be effectively an electromechanical process that was catalyzed by you hitting enter that prints it out and mails it to you or brings it to you, right? Every individual step, I can show you robots. I’m not saying, “Let’s put them all together,” but you could have humanless growing and delivery of food, okay?
So you start thinking about printing out in a broader stage, right? When you think about a drone, it’s like hitting enter and it’s printing out effectively a flight path or a bomb path or something like that. And if you look, the GRASP Lab at UPenn has done this stuff. There’s like drone construction, there’s stuff like that that’s online, okay? You can watch drones assembling small buildings and so on. So in the not infinitely distant future, you will have the ability to just hit enter on a computer and have a bunch of robots snap together a prefab building with perfect speed, in the rain, in eight hours, perfect sight, all of that stuff, right? Some of them might be humanoid looking, some of them may be very non-humanoid looking, but robotic construction should hyper-deflate construction costs, right? And increase quality.
Everybody could have the fanciest pool and all this type of stuff, the kind of environment that only a king would have 150 years ago, you could have that if you had enough horizontal space, of course, right? So there’s a vision of totally hyper-deflating construction costs. There’s another vision of radically reducing your dependence on any specific place and having it all in the cloud so you can move between things and that also reduces construction cost. It’s like van life and things like that, but there’s nice versions of that like the digital nomad version.
And both of those, what they do is they start commoditizing the land and reducing the impact of location, location, location, right? And that should bring downward pressure on these housing prices, but I think that might come after there’s this crazy inflation upward where, as you mentioned, they’re trying to use this as investment. In China actually, Xi Jinping has talked about how he did not want houses to be used for investment. And actually, that’s part of their policy there, is to try to intentionally pop the bubble.
Rather than having a Bear and Lehman type thing, Evergrande was a like controlled demolition where, “We’ll see if it’s successful,” but they set up the so called three red lines years ago on things like debt-to-revenue ratio, so some ratio. Some ratio is like this where companies that were above that they wanted to put the regulatory brakes. They’re trying to deflate that bubble to bring the prices down because they don’t believe houses should be an investment. They want them to be utility. Then what happens, where does this Keeping Up with the Joneses thing go? Ideally, it just goes to NFT land, it goes digital. This is the ideal, okay? And we won’t get there for a while, but with robots, the physical world should get super cheap eventually. And then all status seeking goes to the optional digital world.
And you already see this, you already see this, you see this with phones. Because for the most part, a poor person and a rich person or a middle class person, they basically have the same phone and the status seeking is just pushing into whether you get like the gold case iPhone, but that’s just frills on the top. You can make it more expensive. The Apple Watch is a perfect example of this. Literally, the money is just pure Veblen good like signaling stuff at the top and it’s completely optional. The utility is baseline, right? So that’s, I think, a vision we want to shoot for where everybody has the utility of a house, you can construct it and then your positional red queen, elite on elite, stupid but human things, that’s all just digital NFT-type stuff which can happen online in a harmless way.
Shane: Do you believe in legacy wealth that is wealth that’s passed down from one generation to another?
Balaji: I don’t have a good answer on this, but I can tell you two things I don’t like. First is I don’t like nepotism because I feel it’s actually civilizational diabetes. The kid doesn’t feel like they’ve achieved something. There’s this video called Born Rich with this Johnson & Johnson heir. A lot of them actually fall into socialism because they didn’t work for their money, right? Their life experiences, they were just born rich and other people were born poor. Then also, there’s another aspect which is the guys who are working to become rich. These people who are born rich, they have money, but they don’t have a status, but if they advocate for socialism, they can gain status and also cut off these nouveau riche guys.
So while they think they’re doing good for the poor, they’re actually cutting off competition by the future rich, by advocating for various kind of socialist thing. Money is an abundance, so status is what’s remaining, right? So one is it leads to civilizational diabetes which is bad where there’s an overabundance of a resource, you didn’t work for it corrupts you, whatever, right? On the other hand, I also don’t think it should all be seized by the government because government is terribly poor capital allocator, at least in the West. It’s not doing good things with it.
And you’re putting in a foundation and so on is the third option, but those also tend to get corrupted and they just become basically part of the NGO industrial complex, right? So it’s hard to keep them on mission unless it’s something like the Clay Math Prize like the Millennium Prizes or something very scoped and focused like that which is difficult to corrupt. And maybe that’s part of the answer or a smart contract, putting it into a smart contract that does something, but it’s really hard because capital allocation, if you put in a smart contract and somebody passed away, that algorithm is probably not going to be smart enough to build to anticipate every way someone could game the system with their request. And so again, you’re back to the foundation thing. You need to have like a super AI in order to do that.
There’s a fifth model which is use it to build buildings or things. You spend it all down in your lifetime to have a legacy that lasts beyond life, whether it’s fellowships or … That’s why people endow buildings and things like that. I don’t think I have a good answer on what should be done with it, though I do not believe it should be seized by the state. I also don’t think it’s a great thing for the kid to inherit it and so I don’t know what the answer is,
Shane: What’s the biggest problem with our educational system and how do you solve it?
Balaji: Short answer is the educational system is geared for the assembly line past. It’s like the Bismarckian education system to make people conform and it’s one size fits all and it is bullying and it’s also really ideological indoctrination rather than education. Kids don’t learn basic skills like reading a bus schedule or filing forms for the government. All those things, they have to pick up on their own. They don’t learn about interest. They don’t learn about basic stats. They don’t learn a lot of practical skills. They don’t earn a single dollar for many years.
This is what they used to say and I don’t think it’s even true anymore, “America produce the most impractical 21-year-olds in the world, the most competent 30-year-olds,” because they get the educational system and they don’t know anything and all their concepts are wrong and then they’re getting corrected by the market and they’ve got a good head on their shoulders by 30, right? I think the answer is actually just totally root and branch, reimagining what it is, right? Short answer is you’re continuously learning. You’re learning from an early age, the learning is interspersed with earning. You gain cryptic credentials on chain for every act, so that you’re constantly building your portfolio, as you are learning things and earning.
Just like you might put on your CV, “I did this deal for X dollars. I shipped that product for Y dollars.” That’s actually a real time thing where you ask somebody for an on-chain endorsement, so you can like book that. It’s like a micro endorsement, right? And then the accumulation of that, that’s like your resume the same way, if a teacher is teaching a class and you get a question right, you book that as, “Okay, I got that question right,” right? So this concept like crypto credentials, crypto endorsements, so you’ve got this crypto resume, that is the spinal column.
So, a, you’re continuously learning, you’re not going into debt, you’re learning useful skills and you’re refreshing constantly. You have mnemonic media or something like that where you’re treating it as software that you’re not just loading into your brain one time at age 21 and then for the rest of your life, but rather as an ongoing maintenance thing. If you think about how silly it is, by the way that the concept is that you spend K through 12 and then four more years in school and then everything else is learning on the job, right? That’s like buying this really expensive car that you have zero maintenance budget for. You have no budget for oil. You have no budget for tire checks, right? You bought this expensive piece of capital equipment, you have no maintenance budget for it.
And then what happens is, it’s broken down by age 30 or whatever and you’re like, “That expensive college education, do you remember how to do integration by parts? Do you remember Bessel functions?” I remember it because I just like to do math, but most people don’t because they’re not using it, which I understand. It’s like really a creaking rusty thing in their head that you say, “I’m rusty on this.” That’s why that analogy is very good like rusting capital equipment I just bought early on there. Another you do is you’d spread the vacation out over your life, rather than have this thing where like the vacation, but people don’t appreciate what a vacation is. They have a vacation through college. College is amazing because it’s like … Or it was. I’m not sure what it is today. I think you’re being yelled at a lot to a greater extent, okay?
But college is like this amazing vacation that the young don’t appreciate because they haven’t worked yet. “And oh my, god, all I have to do is relax and learn and I can do that in the sun and just do some homework problems. Oh, my God, that’s the most amazing thing,” people don’t appreciate that until after they’ve actually done work for some somebody else or built something. So to summarize on finally decentralized. It’s funny, supposedly, state-controlled media is really bad, okay? But you know we have? We have state-controlled education system, right? We have a totally state-controlled K through 12. We have state-controlled academia, in large part, not just, state universities, but government funding for academia.
So like academia and education, K through 12 are controlled by the government, but state media is really bad. So if you do believe that state control is bad, why? Because I find very few people can give a clean articulation of why state media is bad, but state education is good beyond the mere Russell conjugation of it. Do you know what I mean by Russell conjugation? Like, “I dox, you leak, but the New York Times investigates. I sweat, you perspire, but she glows.”
Balaji: Okay, so same thing. It’s also called if-by-whiskey, so the same thing can be clad in a different tonal thing, right? So beyond Russell conjugating, you can’t usually give a good explanation of why state media is bad, but state education is good. And if you think they’re both bad, then you want to decentralize it, so it’s all individualized and it’s customized to the individual. Oh, one other aspect of this, by the way, used to be that families in the 1800s would have large families and they all worked on the farm. And that six-year-old would like work on the farm, and would be productive very early on and there wasn’t this extended adolescence period and so on. They were plucking the chickens or whatever and they’re earning their keep. You had more children because they were cashflow positive basically. You needed the labor force, right?
But you didn’t work them too hard because they’re your kids and you love them, right? So there’s an equity interest in a sense. This all changed when factories arose. And in the era of child labor, those factory owners, some of them were unscrupulous and would work the poor kids really hard. Libertarian theory breaks down when it’s children, right? These children don’t have the ability to speak up or the frame or something like that. So reasonably, all these child labor laws were passed that essentially began this long campaign of infantilization. And now it’s something where you’re on your parents health insurance until age 27 or something crazy like that, right?
And so that whole thing began on a reasonable basis and has gotten to a ludicrous basis, whereas people who are being educated as a doctor until age 30 or something before they start seeing a patient. It just doesn’t really intuitively make sense, so you’re going to need that much education to do that, right? And you might be able to install the software faster basically. And now what’s happening today, relating it today, is lots of kids are able to make money at home, okay? So the “child labor issue” where somebody is a factory and they’re being oppressed by a boss and the parent can’t see them and their little fingers are being worked off, all that goes away when they’re at home and you’re able to supervise them.
And basically the same buttons who’d press to play a video game, they’re pressing to work on replete or to build Web 3 stuff and they can work synonymously and you can supervise them and you can see that they’re not getting up to extreme mischief and that they’re earning responsibly. And now they can actually start doing real things much sooner and apprenticing much sooner. And like the internet, obviously, there’s unsafe things in the internet, but physical harm is less probable if they’re supervised. And so a lot of things come back to the future where now kids could work from an earlier age as much more of an occupation and remuneration rather than simply “education.”
Shane: How do you recognize talent? You must have some sort of insights or methods or different ways of thinking about it that help you do that.
Balaji: Well, first, it depends. There’s different kinds of talent. There’s raw intelligence, there’s persistence, creativity and various kinds of dimensions to it, right? And then if you were recruiting for a sports team or the military, you need their speed, their strength, their 40-yard dash time, their bench press, NFL Combine-type stuff, right? So if you had in theory like a dataset which had their NFL Combine, their standardized tests, their classwork and so it’s over their resume, all that stuff, that’s helpful, but I think the single best thing is basically just looking at their portfolio and what they’ve actually done. That’s like the materialized result.
Your brain is probably pretty well tuned for this. If you can load their portfolio and you’re like, “Hmm, not bad, right?” it could be a great open source project. It could be a great piece of design. It could be a creative blogpost. It could even be just a really clever tweet or something, but you can see from somebody who’s high point how clever they are in some area, and then from references, you can see how persistent they are. And so the first order, intelligence plus conscientiousness, those are important. And the third is ethics and integrity and whatnot. And that’s also something that you can pull from references.
In many ways actually, to first order your Twitter is your character. If your GitHub is your resume, your Twitter is your character. I’m not saying 100% and everybody slips up from time to time, but if you’ve got somebody who’s constantly starting fights with other people on Twitter, and I mean first strike, by the way, or if they’re constant cursing or they’re always angry at people, that’s somebody who probably bring that into the work environment, whereas somebody who is not like that is often better or at least can keep it under control. And again, I’m talking of percentage, obviously, everybody gets mad from time to time. Sometimes you might get in a fight, whatever, but it’s a percentage.
And so those three things, I’d say, their ability/intelligence is manifested by their best thing. Their persistence is manifested by or measured by reference checks and what they’ve done over time and then their integrity and ethics. Those, I think, are the three things I look for talent.
Shane: Speaking of Twitter, what did you think of Twitter suspending the account of a sitting president?
Balaji: So that was a very, very, very important event in world history, technology, history, etcetera. And the thing about this is there’s several different implications of it, right? One is it showed that the “most powerful man in the world” was no longer even the most powerful man in his own country, okay? And what that signaled, though people may not have seen it quite at that time is there’s a power vacuum, right? And the second thing it signaled is that American companies could de-platform. If they could de-platform the American president, they would certainly be able to de-platform any other president or politician around the world who would have much less appeal. And many other countries, by the way, don’t necessarily trust the US because after Iraq and other kinds of things, you could easily imagine the US drumming up some case against some country, “They’re a dictator. They’re harboring terrorists,” and then poof, the lights go out, poof, they’re all de-platformed off social media, poof, the troops go in, right?
Now, I actually don’t think the US is organized enough to do that nowadays, okay? But in theory, they could. Imagine what Russia did in Ukraine with the US in Iraq, except Iraq is on social media and like Saddam was tweeting, “I’m innocent. I don’t have it,” poof, misinformation, fact-check, de-platformed, etcetera, right? China actually had an important thing here where they’re blocking American social media, because if China had let Facebook and Twitter in, then absolutely, they would have de-platformed Xi Jinping and what have you.
You might say that’s good, but it certainly mean that China’s no longer in control of China. It’s American media and social media companies, right? So sovereignty, maintaining sovereignty, means control over your communication and transaction channels. That’s like one conclusion. On the other hand, there’s a different conclusion, a third conclusion you take from it, which is we need to decentralize. We need to have cryptocurrency, decentralized platforms for things, right? And so what I think is that the result is actually both, right? You go from these American platforms that straddle this intermediate for a long time of both being international and neutral and then America-based companies with American CEOs, right?
And actually, you take that and it goes into both national stacks and neutral protocols, okay? So the America stack eventually just becomes the America stack, right? They’ll say like all the major tech companies, every civilization zone will build its own national stack. China already has this. They have a full stack that competes with Google and Facebook and so on and so forth. It’s like a full stack of things. Russia has to some extent got the Yandex, right? On Korea has Naver. India has Flipkart. The Europeans have like some stuff and South Americans have Mercado Libre. They’ve got pieces of this, okay? Not the full ecosystem.
Over time, every language region, and I think actually that’s a true borders of the internet, by the way, it’s not the US and Canada, it is the English internet, the Chinese internet, the Russian internet, the Spanish internet, the Portuguese internet, right? Why Portugal? Because it’s got Brazil, right? India’s English speaking, but it’s like its own cluster. I think you’re going to see a whole parallel set of services arise that are the peers of the American tech companies, but in Web 3, in India over the next 10 years, one of the least anticipated things that’s going to happen because people think of individual Indians as competitive in Tech. I think Indians have done pretty well. But India was basically offline. Less than 5% of Indians were online even before 2009. Now it’s more than 50%.
And so they’re going to be a huge player in what has to come and they’re going to build a bunch of things and a bunch of things that build are going to be in Web 3 because China and big chunks of the US won’t be able to do that, so it’s a competitive advantage for India, okay? Or Indians really, because Indians, if they can’t build in India, they will migrate. They’re more willing than any other group and more able than any other group, just pick up and migrate. The Chinese are less able because they’re getting immigration restrictions placed on them. The Americans are less willing because they think leaving America is unpatriotic.
Coming to America is really good. It’s a nation of immigrants. It’s also a nation of immigrants, we don’t usually think about that. So Americans are less willing to leave, Chinese are less able to leave, Indians are willing and able to do so they’ll be able to go and build a lot of these services. So you get this disjunction into national stacks and neutral protocols. The national stacks are set up within these language zones that have the scale to build like local versions and those are going to have de-platforming and other things. It might be a consortium of Spanish language countries or of Russian language countries or of Arabic language countries that manage their kind of things. They’ve got the scale to build these platforms.
The Arab League determines what is being able to be posted on these or not. So it’s like local governance for these national stacks. And then neutral protocols are what the global international law is. International law becomes neutral protocols. That’s actually a really deep concept, I think. I wrote this article with Parag Khanna for foreign policy a couple of months ago, a great protocol politics, “The 21st century doesn’t belong to China, the US or Silicon Valley. It belongs to the internet, right?” So it’s not nation states nor tech companies, but rather decentralized protocols that are going to be the demilitarized zone like the law of the sea.
The law of the sea is the sea is a neutral zone, international waters are neutral zone and crypto protocols are like the law of the sea. And so countries, obviously, they want to have control domestically, but think about their phone system. It doesn’t just place domestic calls. It places international calls, right? So once you place domestic transactions in their digital currency, it will need to place international transactions and that’s where crypto protocols come in. So what happens is that de-platforming of Trump has set off a cascade of things that will lead to national stacks and neutral protocols.
Shane: If you had to bet on one country to dominate the future, and I know it’s not country specific, but if you had to pick one and go all in, which country would it be?
Balaji: One, China. Two, it’ll be I think, by 2040, there’s a chance, I’m not saying 100%, but I think that if the last century ended up, at least the second half, was USA versus USSR, right? I think this century may end up as being China versus Indians, not India. I think that’s where it might land up by 2040 or so. Centralized China versus the international India or international Indians, okay? This is just a historical analogy and it may be completely wrong and so I’m giving a scenario. So I’m putting not extremely high weight on this, okay? But at the beginning of the 20th century, you had the British Empire. It was the big deal. It was like the American Empire, but it’s rickety, but in 1913 was the big dog, right?
And then effectively a 30 years war arose starting in 1914. You can think of the next 30 years as just being like a giant conflagration even though there’s a break between World War I and World War II in the West, there’s invasion of Manchuria, other stuff that’s happening in the East, so it’s like this giant bar brawl across continents, right? And in this Final Four showdown, you had Japan, Germany, the US, the USSR, also France and the UK. And the guy who’s in number one, the number one spot, the UK, got knocked out, and reduced to, basically, number four or number three by 1950.
Certainly way down after the US and USSR. And that’s not necessarily what people would’ve predicted in 1913, but Tocqueville did, Alexis de Tocqueville did. He saw that both the US and Russia were rising.
1835. Okay. I mean, talk about like a hundred years before, but Tocqueville concluded the first two volumes of democracy in 1835 with this prediction. “Today, there are two great people on earth who, starting from different points, seem to advance towards the same goal. These are the Russians and the Anglo Americans. All other people seem to almost reach the limits drawn by nature and have nothing more to do except maintain themselves, but these two are growing.”
And he, basically, is pointing at population growth. Right? And he commented, “The American struggles against obstacles that nature opposed to him, the Russians grappling with men, the one combats of wilderness and barbers, some of the other civilization clothed in all its arms. Consequently, the conquest of the American are made with the farmer’s plow. Those are the Russian with the soldier sword. To reach his goal, the first relies in personal interest and, without directing them, allows them to strengthen and reason of individuals to operate. The second, in a way, concentrates all the power of society in one man. Their point of departure is different, their paths are varied. Nonetheless, each of them seems called by a secret design of Providence to hold in its hand, one day, the destinies of half the world.” Pretty amazing quote for 1835. Right?
Balaji: So, in a similar way, the civilizations that are on the rise today, after generations of being laid low, are China and India. And the US, I think, is in the state of the British Empire. The British Empire, you can also analogize it to the Soviet Empire in the eighties, where, in the Soviet Union, the strikes, solidarity and stuff like that, started in the early 80s.
And you just had rattling, nationalist rattling, of the whole thing. And then, constant pressure from the outside, and so on and so forth. And eventually, you had Gorbachev who tried, Perestroika and Glasnost, and tried to reform it.
And there were fits and starts by the way. Like they sent the Spetsnaz into the Vilnius tower in Lithuania to shoot people in the late eighties. So, there were times where the Soviet Union was trying to crack down or loosen up. It was just like this. And then, the whole thing just fell over post Berlin Wall or what have you. Right?
And I think that the Western system is like that, where the hyperinflation, if and when it does come, and I’m not saying it’s a hundred percent, but it’s possible, or military defeat, if this Ukraine thing goes through or whatever, we’ll see, something like that, or both, or all of the above, is something where it feels like this thing that’s a very rickety edifice that is cracking inside and outside and they’re trying these crackdowns, but it’s like this might just break apart.
And one scenario is financial succession, where individual cities and states don’t want to be saddled with the debts of the former US. And they decide to just break away. And you’re seeing this with these states, all these attorney generals and so on, that are constantly suing the federal government together. They’re building muscle to do things. The interstate compacts thing that was rolled out during the pandemic, also. If you look at interstate compacts, there was Western states compact, Northeastern states compact. That’s something that was actually in the Constitution.
I’m not sure if it’s clause, but the Interstate Compacts provision, or section of the law, was rolled out because the states didn’t feel the federal government was working on the pandemic hard enough. Right? So, you can analogize the US to the British Empire or the late Soviet Empire.
I’m definitely not a bull on the American Empire. I do think there’s individual pieces that will do, potentially, quite well, like Texas and Florida and Colorado and Wyoming. Those things that are leaning into the future rather than trying to fight it. Right? It’s possible even NYC, under Eric Adams, does fairly well. You can start to see that kind of leadership, but that’s different than a world empire that runs the whole thing. As an analogy, when you went from the USSR to Russia, all these people who were there for The Great Game… You know The Great Game? You’ve heard of that concept?
Balaji: It’s like playing risk on the world. Right?
Balaji: All those people who are in the USSR for The Great Game of global communist domination and moving pieces on the board in Cuba, all those people are like, okay, I mean, I’m out now. Right? Like an ambitious person, you were still in Russia if you love the culture and borscht, and they do have a glorious culture, but you’re in Russia for Russian culture or whatever. Right? After that because it was no longer an empire.
And it didn’t make pretenses to that. I mean, it does still try some stuff, but it’s clearly nationalist. It’s not universalist in its pretenses. And I think the transition from, just like the USSR became Russia, I think the USA, after whatever collapse, becomes America. That say, it’s like American culture. It’s baseball, it’s hot dogs, but it’s not global empire. Right?
And in fact, America was not always global empire. It was a commercial Republic, but it didn’t have hundreds of military bases around the world. That’s a mid-century thing. It’s more the late 1800s America, which was its own thing. And different people are partaking in this in different ways. Folks on the left, folks on the right.
There’s different views of isolationism. One is, the US is harming the rest of the world, so it needs to pull back. The other is, the US is helping the rest of the world and it should pull back, but that’s growing stronger and stronger. An empire is either expanding or it’s contracting. And so, I think that the US breaks.
And now, let’s talk about China and India. Right? So, with China, I think China plays the world’s best home game and India arguably plays the world’s best away game. So, China playing the world’s best home game, if you’ve just seen what they’ve done with the country… This is the thing. When I say this stuff, a lot of people are like, oh my God, you’re like a China shill. I’m like, look, I have no investments in China, to my knowledge, in the sense of there’s probably some crypto thing or some fund of a fund may have something. Right?
The thing about China is, if you just look at the cities there, if you look at Shanghai before and after, if you look at Shenzhen before and after, you look at the bullet trains, you look at the build out of infrastructure, they just built the entire country in 40 years. They made it the world’s manufacturing powerhouse. They, essentially, kept their heads down and took all kinds of abuse for many years. Like, oh, they’re making plastic stuff in the warehouses. You’re never going to get anywhere. You’re uncreative. All that stuff. And they just leveled up. And they built up a massive treasury. And they also went and built up Africa and so on. And they built up a huge military.
Look, there’s lots of things that one can and should criticize about the Chinese government. They’ve been extremely restrictive on human rights. And they’ve gone after everybody from Fallun Gong, to Tiananmen, to the Hong Kong protestors, to the Uighurs, to the tech people, to Taiwan. Long list. Right?
What they have accomplished for their people is far better than what Mao did. And that’s reflected in the polls, it’s reflected in the visuals. People can’t seem to hold two ideas in their head at the same time, which is, they’ve done bad things, they’ve also done good things for their people. They’ve built a lot. You have to actually respect a rival and be realistic about them before you can hope to compete with them.
If, instead, it’s like, ha, they’re so weak. They’re going to fall into a ditch. The US Navy is the most powerful thing in the world. China doesn’t even have a Blue Water Navy. This guy Zahan makes these kinds of points and so on, and I have no beef with him, honestly. I do disagree with a lot of his arguments. I agree with some things, that the US will withdraw from the world, that shale oil is important.
But on China, they ship things to everywhere in the world. Do I think they can roll out a Blue Water Navy? Absolutely. They are like the shipping operation of the entire world. They have probably more experience with that than anybody else. Do I think they can do logistics? Absolutely, I think they can do logistics. That’s a huge part of what a military is.
Do I think that they can innovate? People have knocked them on that for a long time, but if you follow the Chinese industry over the 2010s, their “Groupon” equivalent is way more innovative than Groupon. Their Facebook equivalent, WeChat, is better than Facebook. You can argue it’s not innovation per se, but coordination that enables innovation. They are simply able to coordinate across society to build things, in the physical world especially, but across things that Western society is not.
And immediately, people will say, oh, that’s because they’re an authoritarian country, and so on. And that may be part of it for sure, but the US was also able to build things when it was democratic in the mid 1900s. The 1950s is not considered a time when the US was not a democracy. You can argue it was a very different thing because distribution was actually very concentrated. And so, it wasn’t the same kind of thing it is now, which is this chaotic thing, but at least formally, the laws weren’t that different. You still have a president. A lot of the forms were the same.
And so, why was it that a democracy was able to build things then, it’s not able to build things now, why are you attributing to authoritarianism? I’d argue that it’s actually, in part, due to the fact that the entire culture of the US in much of the west revolves around the assembly line. And people would come in to work, and they would punch a clock, and they would take orders from a foreman. And even when they protested in a union, that was still organized, the same way of all folding in, listening to the foreman, taking instructions, execute in the physical world, that sub routine was there. And everybody was acculturated to doing that.
That is a complete opposite. What do you and I do all day? What are everybody in the west doing? We hit keys and keyboards and phones. We are not acculturated to working in a factory setting. The Chinese are. They’re basically the only people in the world, arguably, who are to first order. There’s still manufacturing in Germany and other places, I’m not saying none, but because of that, they have a culture that is, because it’s able to build in the physical world on the small, they’re able to build in the physical world on the large.
If you cannot build the physical world in the small, you’re not able to build in the large. This may be a super obvious point, but it’s basically something where the pushing out the manufacturing base has knock on consequences that are less obvious. And people think, oh, it’s just about money, but why is it that you can allocate billions of dollars and you can’t build a bus lane or a bull train?
Because nobody in the government, nobody on the ground, understands how hard it is to build things. They’ve never actually built something before. They all think they can outsource to some subcontractor. There’s 15 levels of graft. It’s like hiring somebody to do software and not knowing how to do software. You can easily get bit up massively by these people who will tell you, wow, it’s going to take a million dollars to frobnicate that. And you don’t know what that is, and so you can’t diligence it or call BS on it. So, if there’s no experience in any building, these people can’t call BS on it.
So, coming back up. With China, the Chinese state has just executed phenomenally well over the last 40 years. I do think that they have a huge weakness. And Ross Douthat has this concept of the Chinese decade, not the Chinese century. I think that’s roughly right. Their huge weakness is, they suck at English PR. I know that sounds like a cynical way of putting it, but they are absolutely terrible at making their moral case to the world. They’re actually really good at it on the Chinese internet.
The big difference between the Chinese internet and the English internet is the Chinese internet has root and the English internet does not. The Chinese internet is centralized and English internet, while they’re trying to centralize it, is still fundamentally decentralized overall. And that fundamental topological difference, where basically every node on the Chinese internet folds into CCP and they can just flip them off and on. That’s a topological difference of how the networks are constructed. Every peer to peer thing in the Chinese internet is, ultimately, very provisional, whereas the huge energy in the west is on Web 3, which is not provisional.
So, the Chinese state is basically ultimate centralization. Just like when we were talking about post-Trump, that censorship leads to both national stacks and neutral protocols. You have China, which is ultimate centralized manufacturing, top down, command and control, et cetera. And it’s going even harder in that direction under Xi Jinping.
The thing about that though is it’s obvious power. Obvious power causes fear, causes bandwagoning. All of China’s neighbors are as nervous about China as Russia’s neighbors are nervous about Russia, but to 10 X because it’s China. They’re massive.
One thing, by the way, about World War II, Nazi Germany was only like 70 million people. And the US plus UK plus USSR, the UK, I think, was like 50, and the US and USSR were like 150, so it was like 350 to 70. So, it’s like a five to one ratio. China versus the US is like four to one the other way. So, it’s like 1.3, 1.4 billion to like 330. And they’ve got the manufacturing base. And they’ve got the sense of themself as a civilization on the rise. And they’ve got the censorship dialed in.
They attack all these problems when they’re at the seed level, for the most part. Tech is one of the few that they’re attacking at the series E or series F level. So, they’re pretty formidable. There’s this book called the Kill Chain, do you know it? By Christian Brose.
Balaji: Okay. Worth reading. That and 2034. Basically, just looking at the Chinese military and, essentially, he’s like, in war games between the US and China, the US military has a perfect record. We’ve lost almost every single time. That’s how he opens his book. And this guy is on the Senate. Basically, he was McKean’s aid on the $700 billion budget. Senate Armed Services Committee. He saw the entire budget, including the Black Budget.
Because everyone will say, oh yeah, the F-35, that’s a failure, and Zumwalt, that’s a failure, the rail gun on it, the combat elevators, Fat Leonard, the woke military, all that’s a failure. Sure. Okay. Fine. But the black budget, we’ve got all this secret stuff. Dude, just believe me. There’s stuff there. It’s really strong. Behind the scenes. And maybe it’s not. And so, he makes the case that maybe it’s not. He actually went and left the government and joined Anduril because he wanted to do something that might work. And I think that, by the way, is the long term.
I think the future of defense is decentralized defense. China’s medium term problem is that they are terrible at English PR. And so, essentially, it’s just all about hard power for them. Hard power and money. And that’s explicit. It’s basically something where there’s a certain code that you need to follow if you’re in Chinese control territory or a Chinese Protectorate, and you can’t criticize Xi, but it’s sort of explicit.
It’s like, you cannot criticize a party. Keep your head down. Anything that’s politically controversial, keep your head down. It is kind of like the US, in terms of watching what you say constantly, but it’s just more explicit, as to, there’s a list of phrases you can’t say, and so on and so forth. So, China’s easy to understand. It’s a legible one.
India is more, I shouldn’t say more interesting. It’s just different. India, the state, is, in my view, it’s certainly improved dramatically. To an extent that I would not have believed, like the infrastructure, there’s traffic lights. Chennai is cleaner than San Francisco. It’s not close. There’s no question. Yeah. You haven’t visited recently?
Shane: No, man.
Balaji: I was shocked by this myself. That’s why I just call them the descending world and the ascending world. Go and check it out. Don’t believe me. You can look at Instagram. People do drives around the city. I’m not saying everything is completely spotless, but the sense of things cleaning up and roads getting paved. You couldn’t used to be able to drive in India without the dirt roads and bullock carts and all this stuff there. They’ve got modern highways now.
This is the kind of thing which sounds patronizing, but it’s not because it happened really fast over basically the last 10 years. This really impressive thing is happening and India’s just getting it together. And it’s weird, in a sense, the US and India are almost switching places or converging. It’s the US that is now the country that’s got a lot of smart people who just can’t seem to get it together as a society and they’re driven by ethnic and communal hatreds and so on, whereas India has a sense of itself and togetherness and, for its flaws, is able to put it together and is now exporting talent over the world and building up in this way.
But here’s the big thing. India’s strength, actually, may be that China is dangerously competent, and the US state is dangerously incompetent, and India’s in the middle. It’s a Goldilocks. It’s competent enough to provide a base for Indians to operate. It’s competent enough to keep the lights on, to keep the internet on. 4G LTE has got rolled out to almost a billion people across the country in the last five years, six years. Did you know that?
Balaji: Yeah. It’s this insanely important story. The implications of it are so profound. For example, most English speakers on the internet are going to be Indian.
Shane: No, that makes sense. That would be the case.
Balaji: But, basically, most followers on social media will be Indian. Most likes on social media will be Indian, on the English internet at least. The full implications of that, I don’t think people have fully understood. This is why, I think, if China has become a tech and manufacturing superpower, I think India becomes a tech and media superpower.
The tech part is relatively obvious in the sense of, they’re actually number three in tech unicorns now. They’ve exported a lot of tech CEOs. They’ve exported a lot of tech founders. Doing pretty well on tech and computer software. That part is, I think, easy for people to get. On the media side, that’s where people will immediately say, what are you saying, Balaji, Bollywood is going to take over the world? I don’t think so. It’s nice, but it’s not at the quality level. And that’s not what I mean.
I’m saying, Bollywood is fine. It just provides a proof point of production values that are possible there. But if you actually add up other things, so a lot of the animation for movies like Tenet, for example. All these Hollywood movies, if you look at the credits, it looks like a New Delhi phone book because it’s being done in New Delhi. A lot of the backend animation for things, computer graphics, as well as hand drawn pixels, have been done in India for a long time. And that’s ramping up.
You have, obviously, things like SAJA, South Asian Journalists Association. You have lots of Indians who have done well overseas in media politics, a lot of verbal stuff, law, those kinds of professions. And you have quite a few now in show business, like not just M Night Shyamalan, but like Hasan Minhaj and all these other folks. He’s from Pakistan, but let’s say, broadly, south Asian. And Aziz Ansari, these various folks. Mindy Kaling. Probably a long list I could go through.
And so, if you add all that up, there’s talent in media, but now we’re actually in a once in a century opportunity, which is, A, the US media has just lost a plot. And whether it’s Teen Vogue or Sports Illustrated, it’s all just wokeness 24-7, and you know exactly what they’re going to say before they say it. And they’ve abandoned their previous niches. Teen Vogue is not about teens and Vogue. It’s like Marx’s claptrap. And Sports Illustrated is going to lecture you about something. There’s a tone, where it’s almost like whatever outlet you look at, you know kind of exactly what they’re going to say before they say it. There’s not like a surprise or a differentiation.
Part of that has led to Substack, where now those things are actually different and they’re saying different things, but all those legacy media outlets, those niches still exist in some form. They’re going to be taken over, I think, by new players, number one.
Number two is, we have, years ago, I would say VR and crypto. Now, I’d say the metaverse and Web 3. But it is a thing. And I think authoring content for VR is a whole new media environment. And becoming a crypto creator is a whole new opportunity. Crypto creators are as big a deal over internet influencers as internet influencers were over offline because now you have full control over distribution, you can monetize much better. Basically, you can make millions and millions of dollars as an influencer in a way that you couldn’t before. It’s going to be a huge deal. And I think India is going to do extremely well in that because they’ve got a lot of people who are good verbally, as well as visually and whatnot. This is basically the reason why I think India’s underestimated.
But India as a country, it’ll probably have a strong enough military to be defensive. It’s got nuclear weapons and so on. It’s got it together enough that it can defend against China, but no one’s going to be worried about India invading them.
Do you see what I’m saying? It’s competent enough to be a base, but it’s not so competent, like China, that it’s a threat. And it’s got this diaspora that can migrate around the world. It can code. It can do media. It can do finance.
And one way I think about it is, as the US is sort of messing up. This is India, and then, more generally, India, plus Israel, plus this sort of international intermediate. There’s this liminal Anglosphere of folks who speak English, but have taken the UK CodeBase. And places like India, but also Israel, Singapore, Nigeria, they’ve taken that UK CodeBase and they’ve forked it in a different way than the US CodeBase.
So, there’s still a commonality there. These are still people which share cultural reference. If you have a Nigerian tech person, Singapore tech person, Israeli tech person, Indian tech person, you can immediately just understand each other very quickly. Of course, you can for other culture as well, but there’s English common law and stuff that applies.
And so, I have a feeling that group is going to become much more influential on the English internet, in part because they have the civilizational confidence of folks that are rising. And that offsets the woke group that’s falling, and it compliments the Texas, Florida, Colorado pragmatic folks within the west. So, that is kind of this coalition that I see rising on the English internet, which is 60/40 or 70/30 versus the Brooklyn woke establishment.
Shane: That was an awesome answer. What is the most overrated US big tech company?
Balaji: Overrated is a good question. See, I usually don’t say negative things about tech companies, but the really big ones have gotten fairly evil and they’re far past the founders. I guess, probably, Google.
Shane: And what’s the most underrated big tech in the US?
Balaji: Facebook. Why? Because Facebook still has its founder. And that’s the most important thing. Zuckerberg will take the hits of pushing them into new territory. Even with the 500 billion markdown or whatever that they just had, I’m still more bullish on them because Zuckerbergg has taken hits like that before when Facebook had to make the transition to mobile around the time of IPO. And I think what he’ll just do is he’ll be like, okay, all the parasites, all the people who are here for the paycheck, get out. And now, we’re doing Web 3 and we’re doing metaverse.
I’m not saying it’s a hundred percent. Who knows? Future’s unpredictable. Maybe he’s like, screw it. I’ve accomplished whatever I wanted to do in my life. I’m leaving. Possible. I doubt it because he is a will to power founder who has overcome so much. Because he’s got a strategy on Web 3 and the metaverse, and that’s more than you can say for most other big tech companies.
Google does have exceptional AI. They do. And DeepMind and stuff like that is genuinely, absolutely world class. But it has kind of become a researchy labs thing, where they’re doing these prestige things, but it’s not really in their products. Gmail, Google Docs, Google Search, those basically look the same as they did five years ago. There’re some improvements, and there’s probably things that, if you asked me, that they’ve added, that I’d say, okay, that is useful, but it’s incremental. There’s nothing that the boldness of like… Remember instant search was this insane thing that like holy cow, that level of innovation on core search, pushing that out there? Larry and Sergey are long gone.
Why is Google vulnerable? So, I’ll give a specific reason. Basically, I did a whole podcast with, actually, ASIMCO on this, so I won’t recapitulate the whole thing, but it’s how Web 3 goes after everything. And one of the things it goes after is identity, and that threatens Google login, but other social logins and so on. That’s now more obvious with ENS and whatnot.
Another thing is Google has relied on being able to index the open web. And that search index is a huge part of its defense ability. Social networks were a whole threat to Google because they couldn’t index them. That was how they thought about it. That’s why they were so concerned about Facebook and LinkedIn and Twitter. And they were right to be concerned because only those three companies can index their backends. And that’s why they can de-platform and so on. So, social networks are harder to index than even the open web, but the open web is hard to index. What’s easier to index than both social networks and the open web is crypto social networks.
A crypto social network like Mirror or DeSo or Capsule or one of these others that is arising, a crypto social network has every event on chain. Every like, every Tweet, every post, every reply is on chain. Now, that may be a big data set, and there’s ways of compressing it and so on and so forth, but rather than try to scrape the web to try and figure out what’s updated, you just get pushed a block every interval of time that has the updates on it. And now, you can build the index from that. That radically simplifies everything.
And so, new search engines, I’m already seeing them arise, I’m funding them, that are indexing this kind of content because the huge leap, the crossing the chasm, is a recognition that blockchains are not just suitable for storing financial data, but social data as well. And once you make that leap, you’re like, actually, a huge amount of content. I search probably much more on Twitter than I do on Google. There’s somebody else who mentioned they search much more on Reddit than they do on Google because the individually authentic accounts are more curated. You can sort of discount what they’re saying, but you get a sense of it. It’s a person as opposed to some weird spam bot thing.
And one of the things that Web 3 is going to add, as well, that is very underappreciated, is when you take that identity component I just mentioned, and the social component, and the indexing component, every post has a digital signature, so you know who authored it. That is a search ranking signal that is not really present on the web right now. Like who authored this webpage is not something you can easily pull out of the webpage, but who authored this on chain thing is a fundamental part. Do you see what I’m saying?
Balaji: And so, that means you can build a totally new social like iGen ranking of authors that’s similar to Facebook or Twitter’s thing for ranking posts, but open. And finally, Google, a lot of their index is very recency biased. So, they don’t do that much historical stuff. So, something, if A, people are searching on social metrics a lot, B, they’re searching recent stuff a lot, and C, it’s easy to index that, getting a huge chunk of the recent searches on your new blockchain based search engine. Block explorers are basically the stealth threat to search.
And the way you can see Google is totally behind the eight ball on this is they have not even rolled out a blockchain.com, or either scan competitor. Which has nothing to do, by the way, they don’t have to hold tokens or anything. That’s totally legal. There’s no regulatory overhead.
Old Google, under Larry Page, would’ve had that five years ago and it would’ve been the best out there, but this Google can’t even figure out, okay, if I’m regulatory constrained, what’s still the most aggressive thing I can do? Okay. Best block explorer of the world. They should have done that. They could easily have acquired their way to it. They just ignored the whole thing. And clearly, people are running a ton of searches on block explorers. So, Google, I think, is very overrated. It’s cruising for a bruising. It’ll take a while. It’s possible they could turn it around like Microsoft, but that’s where they’re right now.
Shane: Who would you say are the top three CEOs of public companies?
Balaji: It’s hard right now because every stock is down. Obviously, my friend Brian Armstrong at Coinbase I think has done really well. I think Tobi Lutke at Shopify has done really well. If you expand it to Web 3 protocols, many of which are of the scale of public companies, but they’re not stocks, they’re coins, then there’s quite a few people. Obviously, Vitalik has done extremely well. Anatoly and Raj of Solana have done very well. Sergey Nazarov of Chainlink has done very well.
And I’m an investor in pretty much everything, including public stuff, but so much of my head is in Web 3 now. I think of a lot of the Web 2 stuff as being… You know how people kept making money doing Windows programming in the 2000s and 2010s? And you could make money. And you can still make money doing that, but it’s just not where the energy was.
And so, similarly, all the Web 2 stuff is basically like enterprise software. And it’s like doing desktop programming. Look, it’s fine. It’s good. It’s fundable. Knock yourself out. I’ll probably use it and I’m glad you’re doing it, but it’s also not that exciting. It’s also not where the fundamental philosophical arguments are, which is what I think about.
So, when you say who’s a leader, I think, actually, all these public company leaders, with the exceptions of, I think, Armstrong and Tobi, probably others that I’m forgetting, but those two in particular, I know and I think they’ve taken good stances. As a leader, culturally, they can’t be because they’re so constrained.
I will say, actually, just on a pure performance basis, I do think Satya Nadella has done phenomenally well. Like turning around Microsoft under. The run he had. Forget about where the stock price is as of this moment, but the run he had was very difficult to do. Most people thought Microsoft was dead after that. And then, people will argue that all he did was convert to the Cloud and what have you, but he did do that. And it wasn’t being done beforehand. And that was very difficult to do, get up acquisition and so on. I think Satya is extremely underrated as a CEO. It’s very difficult to turn around a company like that.
Shane: Yeah. He got dealt a tough hand and played it really well, I thought, too.
Balaji: Exactly. Yep.
Shane: All right. Last question. If you could read one book over and over again for the rest of your life, what would it be?
Balaji: So, probably Timothy Gowers’ book, The Princeton Companion to Mathematics. I’ve recommended this before, but this is like the desert island book. It’s like a handful. It’s like a big thing, but it’s basically a self-contained guide to math by a Fields Medalist, which you could just get a lot of depth out of. There’s other books I like, but the thing about a math book is it’s highly compressed. Highly compressed. And it’s the best kind of information because it’s evergreen. Math is not like learning about some celebrity. That’s extremely low value, rapidly obsolete information. Math is eternal. And it’s very valuable. So, if you’re in lockdown, get The Princeton Companion to Mathematics and read that.
Shane: Amazing. Thank you so much, man. This has been a phenomenal, wide ranging, and great conversation.
Balaji: Cool. Great. I’m glad.
Press cmd⌘ + p or crtl + p to print this transcript. Sharing with one or two friends is ok. Posting on the internet is not.