This computer. All right, we're recording Great. Okay, so thanks for your time today, sir. Let's start with COVID, then maybe let's start like like when Where were you? And how did you first hear about this, this virus.
So I was just reading my daily content stream in January, and I had heard about the virus in early January, just from some of the biomedical stuff I'd seen, you know, novel, novel virus, but, um, I took it seriously, like many people did, but not enough, um, when China lockdown Wu Han, because that was an unfavorable signal, you know, you look at actions rather than words. And that was not something that the Chinese government, which is very serious about economic development would have done lightly. And then after that, you know, recognizing it's a contagious kind of thing. Um, and that it wasn't going to be a magical containment zones, I would keep it in China, in the age of international travel. And then reading all the early papers, I was like, this looks actually very serious. And it is not being covered enough. And it's being treated as if it's something that's like, you know, something that happens to foreigners, haha. And in fact, there were some remarks like this, you know, in the administration, they were like, Oh, well, this is going to hit the Chinese economy. And it'll show that America is on top. And even, you know, much of the reporting from the NYT and others was basically their, their foreign reporters were covering it. But domestically, they were pooping in the tech reporters thought it was, you know, I mean, they literally tried to cover up the whole thing which we can get to, um, to everybody's major detriment, right. And then, you know, we asked when, and how did you realize was more than just the flu? I mean, like, as soon as I had heard about it, I thought it was serious. The question was, basically, was this going to be like, a bola were so serious that it actually wasn't that contagious. You know, because all the early cases were actually fairly serious, like, you know, the Wu Han pneumonia, on the, you know, Ebola just for your audience to know or what have you. Like, if something is really serious as a pathogen, as many people probably know, by this point, at least, um, then the carrier of it dies before they can pass it on, you know, or they're incapacitated, and they're just lying on a bed. And so it's actually not that contagious. So sort of a sweet spot of being contagious enough, or barely enough rather to generate lots of viral particles to make you go back there, you know, like coughing up mucus, whatever, but not so much that it renders you completely non ambulatory. So you kind of want to turns you into a super spreader before bursting the host alien stuff. Um, so so it was pretty clear that it was serious. The question was, how contagious it was. And that's when like, kind of tracking the Johns Hopkins dashboard which had some of the best stats early on, and really just refreshing that every 12 hours and and triangulating off of noon, Journal of Medicine and whatnot. And, you know, I actually hesitated to tweet about it for a while, simply because it was going to be seen as you know, quote, paranoid or whatever, and then, you know, I put up a marker on January 30. Once I saw these pieces that were like, don't worry about Coronavirus, worry about the flu, infamous BuzzFeed, you know, piece which they just by the way, stealth changed the headline, they never terminate anybody, nobody resigned, nobody even apologized, no retraction. They just stealth edited the headline, like six weeks later to pretend that they had an all pump this article, right? Um, where you know, normally when you when you trust this stuff, you just lose money here, you literally could have lost your life, you know, it actually gets very serious. It's not a joke. And this is something we can come back to. Um, so January 30. I put up this thread, which, you know, I think has gone moderately viral is the one which was, you know, it was called going viral. Do you have you seen that one? Yes, I
remember it. And I can Yeah.
So it's like, you know, what, if this coronavirus is the pandemic that public health people have been warning about for years, you know, and the reason I phrased it that way, is that people were in such a Pooh poohing mode that had to kind of remind them that Wait a second, like, there's a lot of very legitimate, you know, science and academic research and, and frankly, historical records that show that pandemics do exist and things can go viral. And so just being in this Oh, you're so paranoid or whatever mode was just very bizarre to me. Yeah. Um, and so I said it would accelerate many pre existing Trans border closures nationalism, social isolation preppers remote work facemask destruction governments, and I kind of think that unfortunately all happened.
Yes, yes. Yeah, yeah, definitely. Yeah, absolutely. Yeah. So, I mean, and that's, you know, that I think the I mean, that sort of brings me to, I think, what is what is sort of like, the, the important question is like, how did I guess like, how did the I think we, it looks? It looks like he's, he's, I guess what I was gonna get at was, like, how did how did the it we went from? Don't, you know, don't don't wear masks? I remember the Surgeon General had don't wear masks. Right. And like, you know, and, you know, you know, that, you know, the the health care workers needed, which, which? Well, there will, there was there was two things there was the health care workers needed, which, actually, I suppose is like, mix make some sense, because they're the most exposed, but then there was also, you know, but they don't work. You know, are they are they are, they won't, they won't help? Right. And then, very quickly, like, it was like weeks, when like, it became mandatory. And
well, so it's funny, I was in the middle of a lot of that, right. And there was there's so much official misinformation flying around, you know, initially, just a few of them, the flu is more serious travel bans are overreacting, only Wu Han visitors are at risk. Avoiding handshakes, is paranoid, the virus is contained. Tests are available masks don't help. Yeah, right. All of those were things that have been promulgated over the critical months of January and February, but had basically been retracted or at least contested by like March 23. And people were fighting masks for actually quite a while. And, um, you know, the, then they just completely flipped as if they had never fought them. And then, you know, oh, my God, why aren't you wearing a mask? Right? And, and, you know, what I realized is that for lots of folks, they just want an authority to tell them what to do. And they get extremely uncomfortable, when they're asked to actually make their own decisions for anything genuinely important. Or to evaluate the evidence, it just made me realize how few people want to or care to evaluate evidence, or some point we can come back to later and how you actually have selves. And in a sense, you know, evaluating evidence in a given area is as much, of course it's much the skill but useful to articulate this, it's as much of a skill as being able to build a chair or engineer a printer, or, you know, like, like fixed eyeglasses or something like that, you know, you need to know carpentry or, you know, like, like, the physics of inkjets or optics, you have to have some domain knowledge. So being able to evaluate information in areas not something you can just say, oh, everyone should do. Um, you know, if this had been some nuclear disaster or something like that, I would have had to learn more nuclear physics. I knew enough biology that I could diligence love this stuff directly. was interesting to me, though, is that folks, folks who were regionalist that need training in the area, we speak about the stuff very authoritatively, right, to people who didn't have training in the area, which included lots of biotech VCs, and it was just bizarre kind of episode.
Go ahead. Oh, yeah. I mean, that's. So I mean, you sort of hit on the cert sort of question of expertise, right? I mean, yes. Like, like people should, you know, you know, you mentioned like, you know, being able to evaluate evidence, you know, but, you know, not everyone's a doctor, not even not it, you know, not everyone probably should study the hard sciences more than they do in my profession. But, you know, not everyone's going to be an epidemiologist and not everyone is going to be able to sort you know, the noise from the signal when it comes to when it comes to these things that that, you know, required an advanced degree. So, like, like, where does this go? And, you know, I mean, this sort of question like, what, what is what is x? What, what is trusted, reliable expertise in the future? And what how's that going to be different from, you know, what, what we know,
and what an excellent question and this is the thing that I've thought about most deeply this year, and I think I've got, I think I've got an answer actually on this. And I poked at it and so I'm short answer is Oracle's and advocates. Okay, okay. So, Oracle's are basically like crypto Oracle's Hmm. And, uh, you know, right now, if you think about various authorities and different things, the weather channel is like an authority and What the temperature was in, in Dallas on March 25 of this year? Okay. Um, you know, let's say Twitter is the authority on what somebody tweeted that at a given time. Zillow, or Redfin is an authority on like real estate prices in a given area. Okay? And, and so on and so forth, you've got each of these kinds of entities that are generating feeds of data. And an important observation is that, um, you know, when you talk about many articles today, they're basically just letters that are wrappers around the numbers. For example, sports articles are wrappers around box scores, and many financial articles are wrappers around ticker data. Mm hmm. And many political articles these days are wrappers around tweets. Yep. Right. And, and, you know, like, it's actually something, which is pretty interesting, because it used to be very hard to go and get a famous person on the record, you know, that used to be actually access used to be really very, very, very scarce. And now, I'm, everybody's on the record all the time. And so you've got all the six journalists things you can get who, what, when, where, why, and how, from Twitter? Because who said, What? Yeah, Twitter gives you the timestamp when there's the geoip, of where, right, um, which can be spoofed or whatever, but at least you know, like, it is usually not, I guess, um, why and how come from context how in the most narrow senses, they post from iPhone, or Android or something like that people have even written stories on that stuff. Um, and why comes from context, you know, so you can now interview someone and get the basic facts, or at least assertions in the absence of an actual interview. So the basic things have kind of been put up there. And one way of thinking about it is it's not just the sports and the financial and the tweet feeds. But there's all these other fields, like the ones I mentioned, there's the weather feed, there's a feed of Coronavirus stats, there's a feat of real estate price. And all of these feeds right now are siloed over thousands, really millions of databases, every single time that you go out and you block your Fitbit is writing some data to one of these databases, you know, you're you send an email and you're writing data to another database, etc. A long term vision is that many and eventually maybe I shouldn't say all but a lot of those databases, those fields go from online to on chain. Because on chain is the third level of deployment. Okay, so there's on disk, online, on chain, on disk, this are the files only you care about online is for the files other people care about. And on chain, it shows the files that people really care about where they want guarantees of let's say immutability, or a edit log, or, you know, basically, cryptographic receipts or proof backs proof of why that's the stuff where people aren't just going to trust you that they want like a higher level of guarantees. And you know, there's actually more use case for that than people have realized, for example, every scientific paper as Oh, come to, it shouldn't just have the PDF on chain, you need to have the code and the data on chain as well.
When you say on to do, do you mean to hash it? Because when I think I mean, the thing, I think on chain means on chain proves it doesn't prove that something is true, it just proves that the data existed in a certain state at a certain time.
today, what on chain means is different from what it's going to mean in 2025, and 2030, and 2035 and 2040. And today, on chain may mean something like what you just said, which is settling for having a hash, you know, up there, right? And then that hash, maybe maybe it's a URL means a URL, plus a hash or something like that. And then you go, and you pull the data from ipfs or something. Um, and and that's fine, right? Just like the early. You know, internet apps, Amazon and Google, for example, are actually intentionally bandwidth limited and how they were engineered, right. You just type in a search query, and you get 10 blue links, right, right, click things on a website. It wasn't Amazon was doing streaming video in the in the 90s and early 2000. To my knowledge, at least right? Netflix was an either, you know, they waited almost 10 years before they did streaming video and it was gradual and then sudden. And so the the early ones are going to be things where Yeah, it's going to be a hash. But eventually you want all the data on chain. And just to give it one clarification, I say on chain, it doesn't have to literally be a blockchain, it just has to be something that gives appropriate guarantees. For example, there's gun Echo, which I'm an investor in that's, that's like a different kind of decentralized database. There's ipfs, of course, there's our we've not invested in that, but it's cool, and so on and so forth mirrors trying to do decentralized blogging, there's a bunch of these different approaches. And I think we're going to find that there's horses for courses, that is to say, different decentralized file system will be optimized for different workloads, just like you've got s3, and you've got glacier and you've got, you know, Ram locally and whatnot, you'll pay much more for storage that needs to be randomly accessed by smart contracts, or decentralized programs, and less for things that are archival. And you know, you don't need to access all that often. That's usually the case. Okay? So on chain just means decentralized with some cryptographic guarantees of some kind. Okay, so so once you can do this, so once once you realize, all right, we've got many of these articles or wrappers around tweets, or BOC scores, or financial tickers of some kind. And in fact, they're all feeds, and those feeds are going on chain. Okay? Now you start realizing, okay, that's cool, because that's actually like Twitter in an interesting sense, where you, you disintermediate, right, everybody has direct access to all the data that you need to put together an article. And, um, you know, we're close to that we're not yet fully there. You know, it's like, in the year 2000, they were sort of the theory that everything was going to be online. That was before Wikipedia, and GitHub and StackOverflow in Google Images, and Google Maps, and, you know, all of these various kinds of search engines and whatnot. Now, if I, if I just think of anything, if I think of purple traffic light, and I google that, I'm pretty sure I'm going to see a Google image of a perfect purple traffic light or a bunch of them, you know, it's actually kind of insane how, how many kinds of things like that people have actually taken images of, we just take it for granted. But it's, it's amazing thing that's happening, right. Um,
and, you know, so
so while today, it'll be sort of hackish it's a little bit of use your imagination to get all those fields online, or on chain, it's, it's happening starting with price stuff, right, so chain link is actually, you know, their, their their market cap is high, but you know, they're actually doing good stuff. I think they're, they're executing fairly well. I'm not I'm not like I may have some like small channeling holdings. I just want to give all my disclosures or whatever, just to be, you know, religious about that. Right. But, um, if some small channeling holdings, but but I've known Sergei nazarov, for a while, it's been a smart guy. And it's actually someone who's been interested in space for a long time. And chain link is doing a lot of the right things where they're, they're decentralizing the uploads of data, they've got multiple providers of data sets, they can do cross verification, lots of good stuff, right? So you start with the the tickers where they're just directly useful. Okay. And then eventually, you extend to other kinds of tradable data. And here is a key insight for me at least these Oracle's these cryptic, Oracle's are individually subsidized, frequently buy prediction markets, right? On, but in a way that I don't think people realize everybody's talking about prediction markets, focusing on their ability to predict the future. Okay. I'm agnostic on whether that'll actually happen. However, I think they're very valuable for verifying the past. So call them verification markets as opposed to prediction markets. And here's why. If you've got a horse race, okay, people go and they love to bet on a horse race, it does sound like a super high liquidity market. They just love to do it. Okay, just just humans being humans.
And, you know, maybe
there's somebody who's really good at predicting horses or whatever, but it's sort of meant to be kind of random. Yeah, you know, you do get out of that, as you get an extremely good record of which horse won the race? Yeah. Okay, that's something where because there's bets on it, there's actually an incentive to actually verify the data. Yeah. Okay. Verify who won. So that's super interesting. Where you starting in these verification markets for the past as opposed to prediction markets of the future. You've got a way to subsidize digital history, even if it's a zero sum game, that sub says that digital history who cares because there's a positive some externality for society? Which actually kind of good, right? Um, and if you also get predictions in a lot of it, wow, even better, okay. I'm just had an agnostic on that. It'd be great if it happens, but but it doesn't even have to happen. horse race, legal, it exists, let people knock themselves out. In general, by the way, you might call narrowly zero sum that there's a winner or loser in the game. But it's I think, also positive sum in a different respect. If you've seen, you know, there's a study from a while back betting reduces partisanship. Have you seen that?
Yeah, no, but it's a it's a it sounds familiar. It sounds like
Marginal Revolution, blogged about it several years ago. But essentially, the idea is what it sounds like, when when there's money on the line, even a relatively small amount of money, people kind of come down from their partisan echo chamber type things and give like the factual answer, for example, you ask a very partisan Republican. Hey, we're WGS found in Iraq. Okay. And they're going to give you the part as an answer. At least they were several years ago nowadays, maybe they won't care. But like several years ago, that was like a that was republican PC on. And so but then when you said, Here's 10 bucks, if you can get the at least the consensus answer, you might argue as to what is true, we can come back to that point. There's a huge shift in because now there's actually incentive to quote, defect from the tribe towards the truth, right. Truth over tribe. Okay. Um, and so, so betting, I think it's not the only thing that does it. But I think it is something that can reduce partisanship. So even if that zero sum kind of thing is there in a local sense, in a global sense, you've gone and you've started to resolve, you know, the screaming matches, that are there too common on social media, with an actual bet, where both sides kind of agree that the market is an arbiter of what's true. And you know, what's funny is, whenever you do that, you find that people converge so much more than you might think they do. Right? Like, you know, there's people who are like, actually be done, why should we done to me ask exactly where it is. They're like, Okay, I'm at like, point four, five, and I'm at point five, five, you know, and it's amazing how close in that that can sometimes happen. Of course, sometimes there's a big divergence, you know, some people think Bitcoin goes to zero or whatever, and that's fine. And that's actually even more informative. Okay, so TLDR is the prediction markets, maybe verification markets, and they may also be, um, you know, at least put some skin in the game markets, you know, like, like, kind of cool down various things to actually have a resolution to them. Where you get this is you get verified digital history of lots and lots of individual feeds, that cumulate into something that hopefully eventually approaches like the Google Images plus Wikipedia, and so on, and so forth. Right, you've got information streams, different things. And here's the very important thing. This collection of fields is what I call the ledger of record. Okay. There's also something called like the A Kaushik aka as aiic. Okay. Um, the the Akashic records. And this is a concept from, you know, like, you know, Indian culture, which is like the compendium of all events that have ever occurred in the past, present, or future in terms of all entities and life forms, right. And what's interesting is this ledger of record actually starts to move into that direction. And, you know, you're writing down on this gigantic global database, everything that humans have ever wanted to record, you've also made it unbelievable, you've got digital history. And now you can do amazing things. Like if Twitter's database was actually public, there's so many amazing data analyses, visualizations, clients, all the type of stuff you could make out of it. If you had an open state back end, and an open state back end is not open source code. It's the next step. Bitcoin is not just open source, such as the code is out there. It's open state and open execution. And every user is a root user. Okay? That's very important. Not every user is a root user to the Twitter database, you can't just, you know, select all from all tables, right? So every user is equal, when it comes to Bitcoin, right? Um, number one. And so that's open state, everyone can see the state of the database. And number two, it's open execution, where you can run a node and replay every single opcode. Right? So it's totally transparent. You don't have a, a secret ranking algorithm like, like, let's say, Twitter, or Facebook, whatever has right. And yet, it's still quite profitable for someone to deploy their code on chain, where the profit is just reduced to the necessary scarcity of the token. Okay, so you still have the pragmatism of, you know, hey, someone needs to eat and you need to fund innovation and so on and so forth. Without the corporate or state control, which is the when the amazing things about blockchains, right, okay. So what you have here is something very powerful, where you've got an incentive, just to recap, you have this ledger record, you have all these different feeds, they're incentivized by different verification market. So you've got individual subsidization that gets you a collective good, just like individual incentives, make Google Image images, get you google images, collectively, right? And
you have now something where you can run scripts over this to get data analyses and print things out. You can, when somebody gives a tweet, you can actually pull context around that, um, including other entries in the ledger record that happened in the same time or the same place, or by the same person. There's like 1000 different kinds of queries you can imagine running on this right all these streams, you know, happening, you know from time to time. And because it's event based, there's a technical point. But there's lots and lots and lots of interesting time series analyses you can do when you start cross correlating them understanding a cause to be, I think our understanding of like, even human events will get better, right? It's kind of like if you ever wanted, like the foundation like thing where you could predict the future, knowing the past is actually an important aspect of predicting the future, right? Like the psychohistory types. Okay. Um, now, here's where things get interesting. Right now, we have essentially three kinds of things that can be put on the ledger record, at least universally. And those are proof of what proof of when and proof of who, namely, proof of what is a hash proof of when is a timestamp, and proof of who is the digital signature. Now, you might argue with any one of those, and there's certainly caveats to it. Proof of who doesn't really say who the person is, what it does do is it correlates them to many other on chain events, of which the identity of Satoshi is like the first such example, right, where everybody understands the concept of prove your Satoshi by signing in with a private key. Okay. And why does that prove me wrong?
Sorry, go ahead. Real quick, right. That's, that's a joke.
Yeah, sure. So So the reason that works, though, is because it wasn't just literally one event with Satoshi It was a bunch of events with provable timestamps, where you start to triangulate and be like, okay, only the person who only Satoshi would be the person who had all of these different on chain events. Mm hmm. So I'm saying so the various proof of wins, plus the fact that they were probably all controlled by the same actor lead to or you know, sign with the same key lead to a proof of who you see, I'm saying, there's an inference thing that you're doing there, which actually can start to get automated, right? You can basically be like, Okay, if n independent events that are all signed by the same thing, accumulate, you know, it's kind of like, you can, it's not 100%, but you can sort of tell whether something's a real Twitter account or a fake Twitter account by their pattern of replies, and so on, even with GPT three, you know, it's, it's, it's certainly getting good. And maybe it'll eventually pass this, but there's sort of a realistic pattern of human replies, which is still kind of hard to fake, like the timestamps and so on on that, right. And so, in the near future in ns name would be like that. Okay, so you have Marc Hochstein de eth, and that is also interacting with all this stuff in this crypto environment and digitally generating this digital trail right. Now, you might not want it to be Marc Hochstein, eth, you might want it to be my pseudonym dth so that you can be provable while still being pseudonymous, like separate out your real name and urine.
today, we have proof of what proof of why and proof of who, but as you know, people are working on lots of other proof of exes, right? proof of where proof of solvency proof of human proof, this proof of that, and here's the awesome thing, whenever one of those proof methods is developed, we've got an entirely new thing, which we can cryptographically prove in the lecture record, a new form of absolute truth that every chain can, you know, they might decide to adopt it over time, it might be like web browsers where you know, you want it but you have to stage in support or fears. Those are practical, important, but practical implementation details. The point being you get amazing leverage out of cryptographic research, right, you get it right, once you can do proof of x and proof of why and prove z. So perhaps by chaining together some of these primitives, you can start actually getting mathematically provable, unbelievable history. Now, again, there's limitations here, because that proof of wet proof of what when and who are really proofs of metadata, you know, they don't say for example, like when when when you have a tweet, if you had an on chain tweet, it would not tell you the string it was true. It would tell you that it's very difficult to falsify something at that time from that, you know, like, like a dress and and with that hash, right? That would be difficult to falsify. But doesn't tell you the string itself is true, right, that you need other methods for that, but we're getting somewhere, right? At least we're getting some on ambiguous facts unless we think that these are unimportant. Like, why would a timestamp be important? I'll give you three or four examples. One, several years ago, Elon Musk, um, you know, basically knocked down in your time story, but which purport to show that a Tesla had been stranded on the side of the road by publishing the device locks. Yeah. Right. And so the timestamps showed that the ride that was reported was not the ride that happened. Okay, number one. Second example, on the Catherine Wu actually reported this. She showed that in China in Hong sao I believe there was a court that established priority for one person's patent on or basically, one person had invented something before the other and they've proven it by hashing what they had done on chain. And the court was smart enough to recognize that that was actually a proof of priority. It was very interesting by the way that the state was outsourcing its determination of what was true to the international encrypted network. Very important macro point for later, but the network is becoming the new Leviathan That is to say that the force that hovers above the most powerful force in the world, right? Um, a third example is last year with the Brazilian fires Macron tweeted out and I'm not attacking McCrone here, I'm just saying, you know, he did the super
president France. Yeah, sorry. McCrone tweeted out a photo ostensibly of the Brazilian fires near times, reprinted uncritically turns out that when you look at the metadata on that photo, it was from a photographer who had died in 2003. So is at least 15 years old as a photo. Okay. And by the way, if you think about that, that's a good example of sort of, you know, a chain of factual inference, right? You can pull, you know, the meta, you do a google reverse image search, you find the image, you find it's in Getty Images, you find the photographer, you find the photographer's name and birthdate, your your cross referencing a few different things in a pro ledger of record, to show that this is unlikely to have happened. And one of the things to tinging on is, it's unlikely that you would have falsified a story about this journalist dying 15 years ago, in order to rebut a Brazilian fire thing that seems unlikely, right? One of the points here, by the way, is this by collecting everything that you can in a privacy preserving way, you know, because people want to be opting into doing this, clicking everything you can, gives you among other things, like ironclad alibis, and it's not obvious what piece of data will be critical in five years, you know, just like Who would have known that this, you know, Getty Images file or, or this this, you know, article on some journalists dying on as unfortunate, of course, you know, as this guy and his family were, but but who would have known that that would be so significant. 15 years later, and reasons say so significant. I mean, in the New Republic, there was a guy who's calling for going to war with Brazil, click over here's the New Republic or the Atlantic, there it goes calling for going to war with Brazil over these images of fires, which were which were not actually images, the Brazilian fires, right. So that's, that's yet a third example of where the metadata the timestamp actually prevented something quite bad. So and of course, we've seen, I mean, even during COVID, there are these images that are floating around of like, Chinese people ostensibly being locked in their in their rooms, or people falling down the ground and dying on their images floating around the summer about like mailboxes, all these images are just stripped of context. And you're told that they're coming from place extra place, why and maybe the person tweeting, it believes that we have no metadata, you have no way to trace it, track it back. So even just the metadata alone, giving an intra ledger record would be big. Okay. So, so that is the ledger record concept, and the concept of Oracle's that get us absolute truth. And you see, like, basically, we're taking a bunch of pre existing crypto things that exist, but we're stacking and combining them. Yeah. And and kind of, we are giving a roadmap, we're also forecasting where it's going to be 510 1520 years, right. Okay. Now, the next question, you might say, as well, okay, Balaji. That's great for a dispassionate record of machine readable facts. By the way, this is an important thing. It's pretty hard to editorialize in the ledger of record, because it's literally like JSON data entries. Okay, if you know, if you have some, like, if you got a stream of temperature readings, on it's like a huge disservice to every script, consuming it to suddenly add text fields that are editorializing interchange data format, in some way. So there's actually a small c conservatism in terms of your ability to editorialize. It's limited by all the clients that are consuming it or what have you. And there's also alternative fields that can be put up there where if you really muck up your your, your feed on and refuse to publish a fax or mess with people can switch to something else. There's some interesting forms of redundancy built in there. Okay. But of course, there's more to life than just raw streams of data. There's also opinions and so on. second person needs to the second thing advocates. Yes, as advocates, sit above oracles. Okay, advocates consume Oracle information and then publish on it. And they can be human or machine number one. And number two, they can sit on top of other advocates as well. Right? So that's your sort of, you have this raw stream, this river of data, and then you have people like refining it and refining it in different ways. But here's the critical thing. The critical thing is whether you're, you know, Indian, or Pakistani, Israeli or Palestinian, Japanese or Chinese Democrat or Republican, the Bitcoin blockchain is everybody agrees on the state of the Bitcoin blockchain. Okay. And that's an incredible thing where this super contentious thing that people fight with Whereas over namely, who has what property, you know, like, that is actually something that everybody sort of dispassionately exceeds to, because it's cryptography. And that's an amazing accomplishment that we should not underestimate, like, hundreds of billions of dollars is being managed in this way and has for now, 10 years. And lots of other properties coming online this way. These are the kinds of things that people historically fought about who holds what when, and we're starting to solve that reducing conflict in human affairs by reducing it to math, right. And
when we generalize this to not just all financial instruments, but as I've said, with this thing, many kinds of facts can be put on chain. Potentially, what we can do is factor out news and opinion, or fact and opinion in a in a literally mathematical software program where because fact is what is on chain, and opinion is an advocate who's linking to a block Explorer, right? Just like you'd link to Wikipedia, you link to effectively wiki data, or, you know, decentral, pedia or whatever we want to call it, right? The the ledger record the set of all fees, you link to that to prove your case, x happened at y time. Okay, citation needed, you know, on chain needed, right? And that's just very similar to what people already do, you're not asking for an enormous change in behavior we're asking for is those people who are publishing data to publish it on chain rather than just online, which is like, you know, channeling has shown that's already possible. We've got lots of fields doing that right now. And you're asking for writers to link to a block explorer of a different kind. Now, it'll it will be more fact focus, as opposed to transaction focused, but you can just imagine a different UX, you know, you've seen like, block chair and stuff like that their search engines on chain. So just imagine a search engine on top of effect chain. Right, a search engine that is, you know, not just looking for transactions, but assertions, designed assertions, right. So search engine on top of fact, chain is actually also by the way, a medium term way that Google and so on get disrupted me into long term because now the index becomes open. It's sort of like, if someone put common crawl on chain, and again, I say on chain, it could be ipfs or so nice in a decentralized data store. Um, now you could build an open source Google, okay. It wouldn't be trivial. You'd still need to throw compute at it. Right? I'm not saying it's easy. But But it starts to become more possible because you're sharing amortize the work among the community, right? Common crawl, by the way, common crawl loan help lead to GPT. Three. And so like having these open back ends is much bigger deal than people think, you know, having more eyes being on them would be would be really huge. Okay. So
the advocate is interpret is interpreting or is sort of a pining or editorializing about what you exactly the
reason I call them advocates is they're humans. So you assume they have an opinion. Right? You take the impartiality, rather than you having humans, feign impartiality. Right, you which, which I think is always just fading it Okay, especially today. Um, you instead have people be honest, and just declare that they're an activist for x or y or z. Okay. And I actually think that's going to be the flip, right? The flip is going to happen, where, you know, in the 20th century, due to centralizing technologies that favored mass media mass production, and you know, mid century, you had one telephone company and two superpowers and three television stations, right. And so technology was highly centralizing them. And so trust was, you know, because there's only one station, you had Fairness Doctrine, you had our three stations, when we had Fairness Doctrine, you get this huge consolidation and centralization. And so everything should have forced the middle of the road, the bell curve, right, the average, you know, the median or whatever, right? Um, and so trust came from basically pretending at least that one was middle of the road, etc. And perhaps people were more middle of the road in a genuine way. Now, today, trust comes from being authentic, and for not feigning that, right. And, and you're actually seeing this at every place, including it, you know, MIT, for example, where they're talking about moral clarity, as opposed to write, and, you know, what, I actually think that that's better. And then this sort of, like, like pretend neutrality, you know, um, and what's interesting is both, you know, not to be political, but both left and right, think that that's better than pretend neutrality, right. So that's interesting, where you for different reasons, of course, but, but that's actually good, right? And so when you're explicit about it, it makes you more trustworthy. You know, I am an advocate for x, discount what I'm saying appropriately. I have skin in the game on why W and z, I have these disclosures, and I'm going to try to convince you of my position with data and so on, but go ahead and discount it actually. Whatever grains of salt you want, right? And this is actually, you know, the,
I think just a, just put your cards on the table, it's a much more trustworthy, I think way of doing it right. And so that's why I call them advocates. And, you know, as I mentioned before, I think that, um, you know, currently lots of outlets monetize in the basis of page views, prestige and or profit pages being sheer number of views, prestige, being approval from one's in group, you know, pulser board wherever and profit being short term profit, but I think it's going to move to is a form of venture journalism, where these advocates will advocate for, among other things, technologically progressive causes, what I call investable content, will with you know, they'll write constructive criticisms of things and, and if they are able to fund entrepreneurs through this, let's say, just as a toy example, you have an entrepreneur who comes to your blog on fusion, your subject confusion, or what have you, you also have subscriptions or whatever to pay the bills on a daily basis, but you have a rolling fund that you've set up through angellist, okay, because you're decent, you know, influencer in the space, you have, um, you know, maybe a few hundred thousand dollars in a rolling fund, let's say $5,000. And this person, you know, is the, who's in your community on infusion is the best, you know, like, you know, scientists in the space that you've seen out of all the folks you've spoken to, and interviewed and talked about on the topic. And so you're like, Okay, I'm going to take a flyer on them, I'm going to put in 50 K, for you know, 5% of their company, or you know, if that's 1 million bucks, or you know, if it's a 5 million valuation, 1% of the company, whatever it is, okay? And, and then I'm gonna introduce them to all the other VCs and venture journalists and angels, and so on that I know, and put your human capital, you know, or reputation saying, Hey, I think this person is good. So now what happens? Let's say that that actually does that is a big win in five years, or 10 years, and it's 100 X, your 50,000 just become $5 million. And if you have to pay proceeds, of course, back to your Rolling Thunder, have you? But that's actually something where that's actually why a lot of journalists have become VCs. Not all. And it's almost like a boiling off, you know, boiling office. Like when you Yeah, you know, like the the hottest molecules escape, you know, from Brett. And so what sort of happened is, many journals, not all but a lot of journalists who were not very cynical or hostile, or what have you came over to tech, you know, because they're actually interested in the space, and so on and so forth. And, you know, there's ways in which being a journalist actually very similar to being a venture capitalist, I experienced this to a greater extent earlier this year, when I was kind of a closet full time journalist writing about, you know, kroehner, citizen journalists, for sure writing about Corona. And I got lots of tips. And so it was actually similar to being a VC, you're getting dozens, sometimes hundreds of messages in, you know, like a short period of time. And you have to filter them on the basis of like, your internal diligence, on, you know, their past track record, who knows them on, you know, blah, blah, blah, all these all these variables that kind of go into sizing it up? And then you decide, okay, do I include that, to amplify it, etc? That and it's, you know, it's funny as Twitter gives everybody a sense of venture almost, because any tweet could go super viral or not. And it's kind of random, you know, so you have that Aaron's customises jackpot kind of lifestyle, right? Um, a difference, of course, is that tweets don't actually pay you money, they just make money for twitter.com. Right? Look at I shouldn't say nothing against her.com is very complicated, because Twitter has, I think Paul Graham said it well as such high positive and high negative contribution, that it's hard to see where it nets out, you know, because you've learned so much from it, but oh my god, there's such like, terrible people. All right. Um, okay, so recursing back up the stack. What that means is these advocates would a, be honest, the, the reason they be honest, is on chain or GTFO. Right. Like, and, and by the way, this is actually pretty important, because this solves one of the big problems that I've been thinking about since the first half of this year, which is our information, some people talk about our supply chain, right, our information supply chain has been completely corrupted, in the sense that, like, you just, you know, that the process of going from, quote, a study, to a publication to report to a policy to like people, you know, the normal kind of bullshit that we would see around, I don't know, just is, is red wine, good for you or bad for you? Right? Right. That that kind of stuff, multiplied by 1000 was tons of the early and frankly, continuing Corona reporting in many ways, you know, and, and unfortunately, what that did was erode lots of people's trust and things when you know, for lots of reasons public health People would say x and then reverse themselves and very inconsistent things were given out and you know people were being told that you know this was serious and then public health officials were going into you know, restaurants and acting as if there's nothing there. And here's the thing like COVID-19 is a real thing you can see it under an electron microscope you can culture it it to my knowledge it satisfies Cox postulates which is to say you know, like a there's there's four postulates and biology it's like being able to culture it being able to hear it let me actually go and read them it's um it needs to be found in abundance and all organisms suffering from disease benign healthy organisms that needs to be isolated and grown in pure culture needs to be caused disease when introducing gel the organism and must be able to be re isolated, right? So the thing is, this an experiment that anybody can do and hundreds of people, thousands 10s of thousands at this point have done which is isolating culture, the virus improves the cause transmission pathogenesis, meaning you insert, right, so it's a real thing. And and yet, the behavior of these public health officials made people to dispute it, right. So our information supply chain is broken. And the traditional question situational, the two solutions that are on the table are a trust the state or B Trust no one, right? Both of those are horrible ideas. Because trust the state is just like, you know, trust people who are just making it up and wildly swinging from one thing to another, no masks, masks, whatever. Um, without any acknowledgement or humility or resignations or retractions, or accountability whatsoever. And and in a tone of holy sanctimony the whole way. Right. Then the other side, trust no one make it up as you go along. Well, that's also not going to land you in a great position. Because, um, you know, there's there, it's not just COVID Of course, it's it's something where, I mean, you're implicitly trusting someone every time you buy some food, you know, that it's not like contaminated or whatever of truly low trust society is a horrible place to live, because everybody's trying to cheat. You're right. Okay. So, I think the answer is, um, thinking from the very core principle of like, what is science? Okay? There's an awesome paper that came out last year, which helps to sort of didn't phrase it quite this way. It helps to sort of refound science, on computer science and stats in the following sense. Okay. Everything that we know about physics, um, you know, much of classical physics is in the language of partial differential equations. You know, for example, Maxwell's equations or the Schrodinger equation, you know, these are things where there's an assumed continuum, Nabi Stokes, etc. Right? Um, and the thing about that, though, is those partial differential equations are an inference based on data. Okay, ultimately, they're scatterplots, that people squinted that back before, they could collect a lot of data, by the way, they squinted at it. And they were able to put together Faraday's law and Amperes law, and so on and so forth. And then Maxwell, for example, be able to generalize all these into Maxwell's equations, right? And so if you go back far enough, people are actually doing like inclined plane experiments and recording individual data points. Okay. Now, an awesome concept is, what if everything like that was also on chain? What if our scientific supply chain was on chain, such that when you were citing a paper? It was like importing code? Right? That's totally possible. That's completely possible, because
the data collected in the experiment is, is, is recorded on chain?
Yes. Now, this is actually called reproducible research. There's a name for it, okay. I actually did it in grad school, it's like 20 years old, okay. Um, you can go and look at Jupiter j up YT are, you can go look at sweave, which is an R, and there's a package called org available in org mode that does this for different languages, okay. And the point is that the PDF is really an end product, what you really want to release also is the code and the data. In the same way. For example, when you refresh Facebook or Twitter, as an engineer, you're templating that webpage with variables from a database, okay, you're templating it with your personal that person's face and their their followers and their likes and whatever, right. And in the same way, you can think of a PDF for a paper as being templated with the data from the paper and the calculations to calculate, you know, the P values or the graphs or charts, tables, whatever in that, right. And the thing that's so important is when you can see the code, there's lots of things that are often not articulated in the text like particular parameter choices, or you know, how you pick the bins for the histogram, lots of these little things that actually often do matter to someone trying to reproduce it. When you look at the code and you look at the data, the paper is is really downstream of that and here's the thing. That's What every scientific study is it is collecting data and doing some calculations and then writing it up. Right. So it's the data. It's the code, it's paper. Now currently, this is mostly operative in reproducible research charted, like stats and bioinformatics and sort of where the datasets were big. And it was purely stats, and, and so on. But now, there's a paper that came out last year called AI Fineman, which is really good, which showed that you could take the Fineman lectures on physics. And they took 100 equations from the Fineman lectures on physics, and we're able to actually synthesize data as if it came from those equations, and then actually recapitulate the equations themselves using a form of symbolic regression. Okay. Um, and it's actually the algorithm they described such a very elegant paper very well done by Max Tegmark, Tegmark and his graduate student. And what's cool about this is it shows that the computer science and stats approach can actually be generalized to other domains. Okay. Now, um, why is why is that important? Well, it suggests that, you know, the early 20th century physicists went in and they impure realized every other domain, you know, basically, you know, you went into, I don't know, plastics or whatever, and you're like, yeah, you're doing your extrusion moles wrong. Let me write down VA Stokes. And we'll, we'll do this better, right. Or a famous example is in biology, where, you know, Luria and delbrook went in, and they just did something that a lot of biologists hadn't done, which they started counting. And, you know, they came up with, you know, their famous discoveries, like, Luria delbrook experiment, and so on. And the, the point is that physics and like, sort of continuous, you know, models, and so it was huge in the early 20th century. And now it's computer science and stats, why? Because every field has data structures, you know, if you're doing traffic, right, like, Okay, you've got cars, and you've got lanes, and you've got stop signs, and whatever, right? If you're doing I don't know, like, like food preparation, you have, like, shelves, and you have ingredients and whatnot. Every every space has data structures, okay. And if it's got data structures and algorithms, you record them, and they're in databases, so you get stats, okay, so you have algorithms and data structures, you that's computer science and stats. So it's a different way of just kind of going into any field, I mean, whether it's Walmart, or whether it's, you know, Barnes and Nobles, or, or a laundromat or something, they're gonna have customer records gonna have tables of data. And you can start doing scatter plots and histograms and correlation. So it's like a general skill that you can apply everywhere, right? But we can apply it not just all these applied commercial areas, but actually the pure science itself. And so here's what's super awesome. Now, when someone says, oh, quote, scientists say, you say on chain or GTFO, okay, show me the truly permalink. The truly permalink is the on chain link to the code, the data in the paper, which is uneditable, it's archived, you can't muck with it, okay. And it itself has links, which are import statements back to all the papers at sites that it actually uses in the text. By the way, this is something which any real scientist of which I, you know, just for your listeners, or whatever, I spent many years in academia before getting into crypto, I don't usually talk physics on on Twitter, but but I know a little bit.
And, you know, the
the thing about this is any, any real scientist, once their data to be public, once their experiments to be reproducible, wants to build on the work of others wants to, um, you know, be correct, not simply prestigious, or whatever, right. And there's lots of real scientists out there. And these are people who are pushing for reproducible research and open access and so on. Right? So there's a genuine movement in academia for this is an organic thing that already exists, we want to do is push that further so that every paper is open access. So it's otherwise it's not reproducible. If it isn't every papers reproducible research again, it's not reproducible, if it isn't, and it says other papers that are and now anybody can reproduce things, because here's the critical thing, right? We have, you know, when you talk about expertise, or what have you, in your house, you don't have a wind tunnel, you don't have an inclined plane, you don't have a cloud chamber, but you do have his computer capable of doing billions of calculations per second. And anything we can reduce to math, you can replicate in your home. This is why I say it's kind of a funny way of phrasing it. That's why I say math greater than science. Okay, okay. Why did I say that? Well, first is math is greater than quarter. unquote, science, which is like the nonsense abuse of science that people put out there. And, but the math is greater than even normal science, because math is fully rigorous. And it is, it's completely reproducible. And, of course, there's, you know, there's limitations of computer circuits are not completely ideal, but they're very close. And it says that the calculations are extremely, extremely, extremely accurate. Yeah. And so this is actually the same logic as crypto or crypto did is it took science, you know, like, the Nobel Prize memorial in Economic Sciences, right, like economics is called science. It took economic science in turn into economic math, because it, decentralized it to households. And, you know, this is, by the way, sort of, one of the great things about Western culture is, it does have this decentralized renewal built into it, whenever a centralized institution just gets too big for its britches, right? Like the Protestant Reformation, boom, 95, theses decentralized were from the Catholic Church and get an alternative, you know, um, yeah, there's a lot of chaos that came out of that. I'm not saying there's all good, right. But there is that aspect of, you know, decentralized exit, reboot, a new refresh, when the existing thing is broken. And this field is just another iteration of that, where we, bam 95 theses, the data must be open or GTFO with the paper. And the reason this is so important, by the way, just to give it from another angle is, there's so many people who are stealing the prestige of Maxwell's equations for some sketchy data analysis. Okay, Maxwell equations have billions, trillions, I don't know, countless numbers of independent replications. Every time you hold up your phone, you know, like, You're, you're basically doing a replication, you know, the fact that, that you can connect to the towers and so on. they've, they've done the beamforming with the antennas and so on. Um, that's, that's, that's replicating it, right? Um, whereas some sketchy data analysis that came out last week simply hasn't had that many independent replications, or in language or crypto, that many independent confirmations? Yeah, yep. Boom. Right. So what we're doing here is we're taking a lot of the concepts that have been proven over the last 10 years in crypto, independent confirmations, open data, right, cryptographic verification, all these things where we seen how important they are, and now extending them to our full information supply chain. Okay, let me pause there. Gotcha. So, and download, I know, a lot of stuff.
And I want to be respectful of your time, we got to have a couple of things. But one last thing on Oracle's and advocate thing. So let's say 10 years from now, this this, this, this structure is in place where you have Oracle's and advocates and you have all these all these feeds coming on chain. In or let's say you mean, a counterfactual where we had this already in 20 2020. So how would the pandemic have gone differently? Like Like, what what's what does that senate What does that counterfactual look like?
Okay, so first is we would have been able to know, um, so like, I'll give, I'll give several things, right. Um, so first, you know, whether the videos and images and so on coming out of China were legitimate or not, because you have metadata and all these images. Yeah. Okay. That itself is big, because you can actually kind of proved because, actually, there's somebody in our community, Jill Carlson, who had a tweet, which she was like, I will date the pandemic, if you've been calling it that at the time, because it wasn't yet known that it was a pandemic, as opposed to an epidemic, right. She's like, I'll date this is the time that my trust and any video I saw on the internet just broke down. I didn't know if it was real or not, right? Because it exists in that Nether realm of could be real or not, and people just didn't act on it, because it was uncertain. Right? So that's, that's one. Number two very important. Why is COVID-19 so hyper variable? Why is it that Wu Han and Italy in New York were just wrecked by this, but their places had not been? And now they are today? Maybe it's just a number of trials? There's people have speculated things like, Oh, these are the countries that have this vaccination versus that. But if you think about it from a stats perspective, okay, um, what you really want is what I called the big table. Okay, the big table, you know, this is not even the biggest table, but the big table is something where every row is a person, and you have all the columns of data on them, like, you know, are they healthy? Or do they have COVID on you know, what's their temperature today, yesterday, etc. Now, of course, many of those cells are going to be hidden for privacy reasons, good reasons, etc. Okay. However, a lot of the data is being collected now in, you know, Fitbit, and things like that, right? It's siloed and broken up among this hospital and that medical record and whatnot. Okay? So with patient consent, because that is important, okay. You could allow people to opt in to share their data. Why? Because if you share your data you might give to get Okay, the thing about medical data is, of course, it's medical. So it's pretty it, but it's also statistical. So it benefits dramatically from aggregations. So there's ways of doing, for example, privacy preserving aggregations. Google's published papers on this. federated machine learning is one concept where basically, a lot of machine learning is about accumulating counts. You know, whether you're doing linear or logistic regressions, you're accumulating a bunch of counts, and then putting them into kind of estimate your coefficients at the top. And that can be parallelized across nodes. And then you can add on top of that methods for homomorphic encryption, where you are not even the cloud doesn't even see what each individual number is it seeing like a digest of it, it's adding them all together, and it's and digesting them, or saying them back down to be on digest locally. And so it's like it can add up and subtract and do other things with numbers where, where it doesn't even know what the individual notes are. So it's an even better approach, right. And there's folks are trying to commercialize that Craig Gentry had the first breakthrough on that many years ago, but it's improved, improved a lot since then.
when you start thinking in this way, you're, this is the second thing. So the first one I mentioned was with ledger records, you get metadata and photos, the second one, you actually get genuine biomedical data sharing, because rather than all these individual hospital feeds, you have on you know, like, datasets that researchers, anybody in the world can access, um, if those if those patients have granted the appropriate permissions, right. And so just just like your Facebook data, I mean, it's a pain in the ass really to set the permissions or whatever. But you might also get something where you opt in, and the pharma company will pay you a piece for for being part of the study, right? Um, you know, like, Hey, you got COVID, well, guess what, at least we'll pay, I don't know, x hundred bucks to you to look at your data to help other people also not get covered and help you, right? I think that's a reasonable deal. And I will take it but make sure it's consensual. Okay. So now what you've got is on a mechanism for doing data aggregation. Now, again, scientists already do this, it can be called meta analysis, where you're just like taking a bunch of angel studies, and like trying to get P values and stack the P values. I don't really like it that much. There's a better approach, which sometimes, because when you have data fusion, rather than calculating p values from individual studies, and then combining them, instead, you actually stack the rows of each study. Okay? This is often done in bioinformatics, where you find when you do that, by the way, as you often have so called bachelor study effects, which you have to correct for that is to say, like, there's just I don't know, maybe the humidity was different in the set of gene expression data versus this one versus this one. And, and so that actually dominates the effects there. And so you have to kind of correct and normalize for that. Okay. Um, there's various techniques to do this. But the point is that data fusion does work, it does get you done, right, it does get you more signal than any individual data set by itself, so long as the columns match up. And if you've got a blockchain or kind of decentralized database, you have an incentive for various hospitals to line up their columns, way of thinking about that is whether it's coin base, or it's finance or it's crack, and they all use the same fields in Bitcoin for the most part, right? It sort of forces everybody uses handfield. Right? So if you had a blockchain EMR, I know that sounds jargony it will eventually happen. It literally might take to like 2040 for it to happen. Okay. But that will eventually happen. And then you actually get defined columns and maybe different options. Of course, maybe just like there's a theorem verse, Bitcoin, whenever there may be different options for this blockchain EMR. So that's, that's a second huge thing you get where you get a clean data set. And now what you get from that, you can now have root cause, why is COVID hypervariable? I've got 30 hypotheses. Now I can actually test each of them, or at least observationally. Right? Currently, the only things that people can test observationally are these massive aggregates, they're like, oh, the Germans are doing good. And these folks are not right. Or these countries wear masks. And these aren't or these countries have this vaccine, and these done, and they're doing these like super aggregated things, national level aggregates. But if you know, some stats, there's something called Simpsons paradox, where I'm working with aggregates can actually obscure and sometimes even reverse the sign of the correlation, or the relationship when you're working with individual rows of data. So and that's just like one of many things where, like, working with big aggregates can be misleading, not always, but it can be you know, and so I see people fighting on Twitter, and you know, like, it's, it's something where look observational data at the country level, I'm not saying it doesn't have any importance. I am however, saying that it's a extremely coarse grained resolution relative to what we could do. Alright, so that's number two. And number three, actually, in some ways, because number number two is getting actually getting the data together to do the studies. And number three is actually find the cause of hyper variability. Um, number four, is
I, you know,
this is a little bit harder to prove, but I do think That we would get much better prevention as well, because quantified self would go on chain. And there's things like the you know, the health radar that can Kenza did. Okay? I think it's like, Can I say if I got that right Kansa health thermometer, okay. And they've got like a health map on k, i NSA health map, okay. And health weather.us. There's companies have done stuff like this, where, essentially, they've got a bunch of smart thermometers, they've got a million smart thermometers in the US. And they just kind of look for where temperature seems to be spiking. And that's a way to track kind of COVID sloshing around right. Now, if you had the ledger record where you have a bunch of fields like this, and nothing gets kins, it's a good company. You know, I know the founder a little bit could come I'm not an investor, but good company. But if you had a bunch of those, and you had, whereas data, and you had fitbits data, and you had the Apple Watch data, and you had so on, and it was on chain, and then some set of it was selectively decrypted, you know, to either researchers or people who would pay, you've got a pretty interesting data set for monitoring and prevention as well. Right. Um, so, you know, a lot of these things are fundamentally informational problems. You know,
Why is COVID what it is, I mean, for example, I'll give one more example of an informational problem. The, you know, majorna vaccine, I mean, anybody who knows, like, like bio knows, they, they essentially synthesize it in a couple of days, because they're, they're just using a template that, you know, that they had the coronavirus genome pretty quickly, and you could get in a couple of days, we could probably, by the way, have tested that vaccine and challenge trials in like February. And challenge trial is basically you've got you synthesize it, you got a vaccine, and you just have soldiers or somebody healthy, or, you know, folks volunteer, and they just, you know, that they get the vaccine, you know, whether it's oral or nasal, whatever, and then you expose them to Corona, and you have a hospital there, in case they actually get sick. And the reason that's so much faster than observational trials is and it's also much higher signal is, you know, who is exposed versus not, you don't need to do placebo controls, right. Now, of course, the ethics of doing this are some people might die from Corona on their hand, you also the hospital right there, you have the best of care, and frankly, far fewer people would have died if we had done that, because you would have instantly known whether you have no shortage of volunteers, okay, and function properly functioning, no shirt on tears. So you have, um, you know, the vaccine manufactured right away. The critical thing, though, from the vaccine is not even the manufacturing at least the mRNA one others are easier or harder, the critical part is the information of knowing whether it works or not, and what its side effects are, and so on. And so that information could have been gathered very rapidly in a competent society that did challenge trials on and, and we're probably going to want to do that some society will have to get competent, because, uh, you know, whether it's COVID, or another pandemic, something will probably come back at some point either mutated COVID, I mean, I don't know, I know, it sucks for the, for the viewers at home, but it's, it's maybe just goes away, you know, vaccine kills it. And you know, it's less mutable than the flu considerably. So so you know, those folks who hypothesize it, but it's also got a lot of shots on goal, you know, it's got millions and millions of people around the world that's infected, so is doing this huge parallel search of the genome for mutations that allow to bypass a vaccine or bypass treatment. And we can never, like, just write that off, it could absolutely happen, you know. So that gives a few ways in which better information in general could help us as well as three specific ways in which on chain data ledger of record could could help. Maybe maybe helpful, maybe, you know, I'm never sure how technical I should go. But I think it's useful to give this level of detail for some of the folks in the audience.
Yeah, for sure. That's great. So again, I got a whole bunch of things here. I wanted to hit on so let me so going, going back to going back to media so I do feel like I want to address this incident with the the Ricoh the recode reporter Yes. So why why respond on Twitter rather than directly to them?
Well, the thing is that what, you know, I've now got enough experience in this where I mean, you can tell very quickly, what somebody's slant or angle is with a story from a the outlet, be that reporters tweets or past publications and see, you know, the tone of their approach or what have you. Right. And, you know, I think a lot of journalists, I should say corporate journalists, as opposed to visitors, but a lot of corporate journalists are, you know, like little kids will play as if They're like spies by like covering their eyes and saying I'm invisible, you know, like that, right? Lots of corporate journalists think that their subjects don't read all their tweets. Mm hmm. Right. And here's the thing, like, anybody who's smart, reads all the tweets of any journalist who reaches out enough for the past articles, and you form a profile of who that person is and what they do, you know. And, like recode, like many of the tech outlets, is basically a glorified gossip column that doesn't know the first thing about coding, investing, or managing and just writes lots of hate pieces on tech essentially pivoted from being a gadget reviewer, to some kind of Gestapo. And so it was like, pretty obvious that, like, given those priors, plus, like the context of the approach that it was going to be a negative piece.
second thing is, why did I respond on Twitter rather than directly? Well, you know, it's, um, it's something where it's very, if somebody's telegraphing that they're going to punch you, and they're winding up like this, and like, I'm going to punch you, you know, it's pretty dumb to be like, okay, I'll wait for the punch, and, you know, then spit out my teeth, or whatever No, you answer is like, either get out of the way, block the punch, or, you know, be like, hey, this guy's trying to punch me and you point to the crowd, and so on, right. And so the entire model of like, you know, I'm gonna put them on the spot and hold them accountable. And, you know, have this evil corporation or person, you know, like, I'm going to ask them for response and put the microphone to them. And, you know, they're going to give it to me, that only holds in an era where your quote, subject is actually a subject, as opposed to a citizen journalist of their own, but their own distribution and their own following, you know, and the entire concept of a quote, subject, you know, as a journalist, and subject recalls the alternate definition of subject like, a king and their subjects, right, or, like, subject to or the subject of the experiment. I, and, like, you know, folks with us, we're not like, things under the microscope to be documented for, you know, the, the, oh, the doings of the Silicon Valley elite, or whatever, you know, this person wrote. Um, so, so these, these corporate journalists don't actually conceptualize their subjects as being fully human beings, they think of them as targets. And because of that, you, you don't just wait for their Bronze Age style of warfare, where they're going to, like telegraph the swing, and then punch you on. Instead, you just blow up their spot, you know, you first strike, you call out the story on you say what it actually is. And they're always like, ah, I can't believe you did that, because we did this, you broke their script, where you stand there, and they punch, you know, I don't think that's how it goes anymore. Right? I'm not here, we're not here to be grist for their shitty content mill, right. Um, in fact, we are going to deter them from doing this. And we're going to basically factcheck the crap out of these articles, which are written I mean, in particular, this one was particularly egregious, because this person, and this outlet had no biomedical background whatsoever. And attempting to like, show me not just me up, but like, all these biotech VCs, who have founded and built hundred million, like, just, it was it was like an ant crawling up an elephant's leg with with rape on its mind, okay, like just an absolute, um, arrogance and lack of humility, frankly, by people who don't know and don't know that they don't know. And so absolutely not, we're not going to wait for them to, you know, to slug us that that's that error is totally over.
So the one I want to move on from this in a second, the one argument I could see for, for taking the call or whatever would be, even if it's a it's going to be a headpiece, even if it's conceived as a headpiece, if you talk to them, there's at least a chance that you can, that you can influence the outcome of the story and that maybe they can actually rethink, they'll actually if they listen to you, they will rethink the premise.
But three, three thoughts on that. First is, I think pre 2013. That was a more reasonable strategy. Right where you actually had, um, you know, like, since 2013, since the beginning of the tech lash, right? Um, media companies and their culture is to treat tech companies as the enemy and people who they want to, quote, take down or hold accountable all the prices Is all the culture is built on dunking on the tech bros and tech brain and so on and so forth. Right? Um, and, you know, like, this is basically a bunch of woke whites in Brooklyn attacking people who are very largely if not majority immigrant, right, by the way. Um, but the thing about it is pre 2013 I think that would have been reasonable, right? Because these folks understood something of, you know, the older school reporter understood, understand something of carrot and stick, right. The older school reporter also, um, you know, wasn't as incentivized by Twitter to just demonize for clicks, because what Twitter does is it trains you to strip context, add exclamation marks, by that. So to be very clear, I put a ton of the blame on social media, I think Facebook gets way more crap than Twitter, Facebook gets 10 times as much crap than Twitter, but Twitter is 10 x worse than Facebook for this. And I think one of the reasons that is is that the bad actors on Twitter are the ones being on Facebook, but they're actually the bad actors on Twitter. Okay, blue checks have set a very bad example, on Twitter where the blue check was initially, for someone to realize that they were not being imitated, okay, or like not being mimicked, um, that there wasn't an imposter account, but it became a mark of prestige because the kind of accounts that are not being imitated are the ones that are often important in some way, right. And so the fact that blue checks model is extremely negative form of discourse, brought down the quality on Twitter and led to this feedback loop where to gain status was to be very negative, polarizing, etc. And by this is mathematical. Okay, you can go and create scatter plots that show that the more angry words magga words, woke words, whatever that you put into your copy, it boosts clicks, you know, Nick, Kristoff, who's you know, an older school journalist, I don't agree with everything he writes. But I feel that at least he has a degree of intellectual honesty. I'm actually wrote an article last year, titled, I think it's like the articles no one reads. And he disclosed that if he writes about some worthy project in, you know, the, like the developing world that gets like X number of clicks, but if he writes about Trump, that gets literally 10 X number of clicks, okay. Now, here's the thing, Nick, Kristoff, again, I don't agree with everything he writes. But he he actually wrote the article that inspired gates to go and do these flush toilets, or these are non flush toilets, these decentralized toilets, basically, in the developing world, I may be butchering that, but from like the Netflix special, okay, Bill Gates came up with funded people to come up with new models of toilets that you could actually go and put out there in places where there wasn't functioning sewage systems. And so Christoph did a lot of good with that article that didn't get a lot of traffic, but had like the right reader, mainly because this gets back to the point I was talking about earlier, where you're optimizing for the very highest, highest quality reader right? Now they're optimizing for the highest quality readers, Khrushchev being an unusual one. And one thing I found interesting is how many of these corporate journalists like refuse to admit that they're driven by pageviews, or are driven by clicks. And yet, of course, at a minimum, they're looking at their number of parties and the number of likes, which are absolutely a proxy for a number of pageviews. And they're looking at that all the time, and are hyper conscious of it. And so, so the answer is a pre 2012 may have been more reasonable. Not anymore. Be, um,
you have to also think about the iterated game of it, right? Like, if bad behavior is rewarded with supplication, you just get more bad behavior, right? It's literally like, I mean, it's like the, you know, being bullied in high school or whatever, once you hand it over some lunch money. And you're like, hey, bully, can I cooperate with you next time, maybe it's just 50 cents or whatever. They're just going to keep taking your lunch, buddy, right? However, unfortunately, if you can cause them more pain than they're trying to cause you, that's actually a deterrent. Right? And it's unfortunate that you have to do that. But, like, it's very clear if you read their Twitter feeds that these people think of technology as the enemy and are dehumanizing on a just fortunately, by the way, an industry that has about 20 or 30 points, less white than than journalism, you know, again, that's not their thing. I'm not the kind of person who like goes on about that all the time. But it is true. They're dehumanizing this this group of people who writes a code who builds the vaccines who does the drug development does the genetic test, build the robots? And and then they expect you to play by Marquis of Queensberry rules when they're just like firing with a new z. No, not not not reasonable anymore.
Yeah. Okay. We're running short on time. I can keep going. But if you if you want to, if you want to stop and I can
go a little bit longer.
Sure. Okay. Um, one more thing I want to go back to on COVID because I feel it's important to hit upon So and I don't want to misrepresent sharp political views, but I suspect there is at least a libertarian streak to your to your, to your worldview. But so there's there's a lot of people and I think I again, I think this is part this is a result of the sort of distrust of experts that we were talking about and that sort of the trust no one scenario but, you know, you know, how do you I guess, how do you feel about anti masters particularly that a lot of them are? Some of them are actually in the crypto community, some of them are definitely that sort of libertarian community, like, you know, like, like, you know, because because you obviously understand the virus very well. But like lately, like, what do you what do you what do you think of the sort of people who are who are refusing to wear masks and and skeptical about Coronavirus? It's being being a thing to this day.
So, Ah, man, it's a very complicated topic. So, the first of all, um, I think you should wear a mask and I think Coronavirus is real. Okay, number one, number two is on. I think that the reason that those people feel like that is because like public health officials and the press and quote unquote, scientists have done such a poor, contradictory, condescending non skin in the game way of communicating. When when you go and see, like Gavin Newsome going and having a dinner outside, write a quote outside, but actually indoors, when you see Nancy Pelosi going to her hairdresser or whatever. It shows them that the political elites who are imposing these restrictions on their businesses and their way of life, are not taking it serious, or not wearing masks. And therefore, these people who don't often don't have technical degrees, say COVID isn't real, right. so on. So I put a huge amount of the blame on our political actors, frankly, and that's, by the way, that's that's Democrat, that's Republican. That's a totally Equal Opportunity indictment of essentially the entire political class, who is filled with lawyers and people who can manipulate political realities. And they themselves are so used to saying one thing and then doing another that they don't realize Corona is actually real. Okay. Now, there's another aspect to it, which is another thing, there's also a real about Corona, at least as far as we can see, is it's got a hyper variable outcome distribution. You know, there's some people who know it does skew to be much more severe as people get older. That part we do know. But there's a huge variability in symptoms, where some people as we know, die. Some people are, like, have permanent lung damage. Some people have like weird neuralgia and like, post viral syndromes and, you know, so called long haul covid, they don't get better. And actually, quite a few of my friends have reported something like that. So that's something where and these are people who don't know each other, you know, individually, they're not malingerers. They're actually, like ultra fit people, many of them. And I know, like four or five people who've had COVID, in my circle of I don't know, 1000 people who I know, who have reported the site Penza, long haul symptoms, I haven't done the exact denominator shouldn't take that religiously. Okay. Um, and it may be something where I actually know 10,000 people, and I'm sort of overestimating the numerator versus denominator. So, like, that's why I take it very much as a with a grain of salt. Okay, but, but I do know that there's at least a, it's a seemingly non trivial number of folks with long haul COVID, both in personal experience and literature and in the support groups, which have hundreds of thousands of people who at least think they have long haul COVID. And then finally, of course, there's folks like Trump who seem to just shrug it off. And you know, like, you're seven years old? And is it for him, it really was just a fluke, quote, unquote, right. And so so that's, I think the second reason is there's a genuine huge variation in the symptoms. And for many people, they're going to heavily weight, the personal experience of the people near them. And if they had severe or non severe cases, how it's going to turn out, right, on. So to be clear, I'm not I'm not defending what I would consider an irrational decision to not wear masks or or, you know, go to the stuff. What I'm saying why it's happening is due to terrible leadership. And by the way, in other countries, like Asian countries, they haven't turned it into this silly, political clickbait exercise, where on the one side, you have people who say it's not real and others have you have people who give lip service to it and still act like it's not real. Instead, they actually are totally serious about it, where you get off the plane and they quarantine you in Asia for actually two weeks. You cannot leave your room. It's like an actual quarantine. They require a test before you get on the plane directly. Test after you get off the plane, they have central quarantine, if you ever test six, you're not like just running around your apartment. They have temperature checks in public locations, blah, blah. Like they're actually like for real about it, you know, and because they're for real, they have a functional society. And and by the way, so this is actually said, you know, a libertarian streak and so, I mean, like, it's funny. Um, I think that, like, if there's one political leader that I admire, it's Lee Kuan Yew, okay. And the reason is, Lee Kuan Yew was fundamentally a pragmatist. And there's, there's things that Singapore does that are libertarian ish, there's things they do that are conservative, it's just things to do. They're very progressive ish, or very statist, or nanny status, whatever. But the thing is that, um, I feel that a lot of these political ideologies describe axes, where you can set a slider, you know, and the western style this is one of the disadvantages of western style is like, ramming it all the way to one axis or another. And it's, like, argued as, like a holy thing, you know, and in practice, it's like, when you start a company, you know, you have, um, if some theory about how the world should work, but then in practice, you have to set a particular slider, like, how much equity Do you give your employees, for example, right, you can, oh, give the employees equity, or Oh, you know, like, you have to have some for later cohorts or the invest, like, that's an empirical decision, and it's gonna actually differ for different cohorts of people, even under different circumstances. And that's how I think about like government parameters, you know, like, um, how coercive versus how, like, you know, volitional you should be, I understand the ethical argument for volition, I think that's probably also gets from a utilitarian standpoint, the right answer much of the time. But I also recognize that in situations like, let's say vaccination, on that there's an argument for, you know, for coercion, because if you, if there's a segment of the population that doesn't want to get vaccinated for a vaccine that works, let me come back to this point, because that's a very important asterisk that last provides, okay, there segment that doesn't want to get vaccinated for a vaccine that works, that becomes a disease reservoir, where the virus can just mutate and churn or whatever and then reinfect, the rest of the population. It's not, it's not simply an individual decision, you're making the decision for the whole community right? Now, the flip side of this is that extremely important premise Stanley asterisk, a virus vaccine that works. And the problem one of the hugest problems is because you have these people who have gone to, like, you know, dinners and gone, I mean, like that, it's like, the first duty of a politician is optics or whatever, and they've just completely messed it up and made people not believe that this is this is real. Right? So so that the one thing Do one thing, they can do that one thing, you know, um, so because of that, there's anti backs sentiment on the left and the right, which, which seemed just like another stupid thing of America in 2019, is now like, potentially very serious thing for the vaccine rollout. I don't know how that's gonna go, maybe it'll be a non issue, and people will actually just take the vaccine, maybe folks will say, Hey, you know, vaccines are drugs? And why are you telling me that I'm supposed to be super skeptical of drug companies and drugs and have all these regulations and so on for them, but then super on skeptical of vaccines and just trust them, etc? Because a vaccine is a drug, right? And by the way, that contradiction can be argued both ways. You can argue both as anti Vax folks will tend to do to not trust vaccines, because they're from drug companies. Or you can argue a different angle, which is, hey, actually, maybe we should liberalize drugs, and so on as well, and give a little more leeway and actually push it harder on making them faster, right. So that contradiction holds in both directions. I just want to point out that contradiction. Moreover, there's a more rational antivax type person who says, Well, I believe vaccines work, but I'm not sure about this vaccine. And that's actually, you know, Kamala Harris said something like that. She's like, if Trump approves a vaccine that I'm not going to take it right, so people who just you know, custard Do you see that one? Right? Yeah. And it's like, I mean, look, I again, I can't I can't excuse that sentiment. I understand it because of the insane politicization of everything, right? But um, but you know, that's not someone who's anti Vax, that's anti this facts. Right. And, and so
I don't know, I don't know how it's gonna land up. Um, I think that one of the reasons that I haven't written about it as much recently is it got so politicized that you realize that people in America at least or the West you know, certainly in English, English language, the UK and other places New Zealand seems okay, Australia seems okay. They seem to be doing better at least I think it might be the Asian influence frankly, they're like other people. Your peer group is how you kind of take it right and but English language internet seems more interested in fighting with each other than solving this problem. We're taking it seriously. So I don't know what happens maybe it kind of fizzles like the election was was sort of how As this big, you know, fight clip thing that was going to happen afterwards and kind of fizzled on that. Hopefully, I mean, the thing is that because these are private companies, frankly, I've trusted much more than the public sector. And these are legit. I mean, you know, with Madonna and Oxford and Pfizer, we've got multiple vaccines on papers look good from what I've seen there. I mean, the sample sizes are not humongous, necessarily for the ones that have like, but, but I'm not criticizing anything, I'm just saying that, um, you know, we'll see what happens on it. I would take a vaccine from one of these companies, I think they're, they're very legit. Um, want to see whether other people do it. Because that'll be a big signal as to whether this fizzles like the election, the anti Vax sentiment, or whether it's a huge problem we have to deal with, because we don't want is on like, to be stuck with a choice between on a huge disease reservoir with the vaccine like spiraling and mutating within millions of citizens versus cores of vaccination that just pushes us into a different realm, where I hope I hope the country doesn't go, I hope the data is good enough. And this is where the information thing comes. If the trust hadn't been, you know, pushed down to such an extent with this bad behavior by politicians, there'd be a Trust Bank to draw from. When you say, trust us, this is going to work for you. Yeah, I know, it's a long answer. But it's a complicated topic.
Go ahead. Well, that's that that was really salient. Okay, so I want to make sure we hit on that. So what it what is next for you? what's what's what's your, what are you working on now? One, what's coming up?
Um, a lot of this stuff on truth and so on, that have been talking about here. I've been thinking about how to roll out there. There's other aspects of it as well, you know, probably it'll be like an open source project of some kind. Um, but we'll see. And, you know, I'll probably announce it when when I'm ready.
So, gotcha, gotcha, you, just let me know when you need to need to jump. So the,
I do a couple of barberia.
Okay, let's do it. Let's do a couple more. Um, maybe talk just talk a bit about looking back on on 21. And I earned calm and your time and your time at Coinbase. Like, what are your biggest lessons? What are your biggest accomplishments? from from from that period?
Yeah, so like, I think biggest accomplishments, like Coinbase earn, since you asked, it's actually been very successful. Uh, you know, if you go to Coinbase, it's like, well, first, it's heavily featured in the app and on the homepage, and so on, because it's working, right. It's, it's Coinbase is highest rated, individual offering, because everybody makes money when using it. And, you know, you can literally click buttons and make money, right, and educates people. It just just has a really, I think it's done extremely well. And I'm not sure if they've given up the numbers on it. But millions of people have earned crypto through it. 10s of millions of dollars of crypto have been earned through it. And, you know, stellar alone, put 100 million dollars worth of stellar into the program last year that was announced. I think you may you may have saw on that saw that. And and the quinby CERN the integration basically did hundreds of millions of dollars in sales in its first year. So there's this huge pile of crypto that we can use for onboarding that we're using for onboarding and it's gone. I think extremely well. In fact, it's gone so well, that coin market cap clone Durn and they have, you know, like coin market cap burn, and in fact, have the exact same language in the same font, like learn crypto and earn crypto, which is, you know, like the, the copy that I typed out, like several years ago. So it's actually that's, that's, you know, the thing, imitation is the sincerest form of flattery, right. So they've also seen the numbers. Basically, the reason that urn has been so successful is all these crypto projects. If they want to attract users, they often can't run ads on Google or Facebook, right? And if they did, even if they could, they'd have to liquidate their crypto to get USD to pay Google and Facebook, which would harm all their existing holders, right? To acquire customers. And then what do they do, they're basically going to redirect you to Coinbase or an exchange anyway. So you know, then there's another conversion at the end of the funnel, right? So instead by going directly to us, or to Coinbase, rather, on the Coinbase is a crypto aware channel, right? That could take $10 million in a cryptocurrency, award it to users who have value it as $10 million, and in fact hodl it rather than liquidating it, and that might make the price go up over the long term, though, crucially, every person who held it would have used the currency or the token or whatever, prior to buying any of it, right. So you also get something very useful for projects, which is namely a demonstration that a million people use the thing and watched videos and why was useful before buying it, which is actually like a track record of utility. So it solves another problem for them, which is like establishing utility, you know, you can see what's distinctive about it, it does have purpose, it's used for this VR world, or it's used for, you know, storing files, and then you buy on that very obvious basis is like an API key, right? So, so because of that, it's like the largest scaled customer acquisition channel in crypto. And one of the things that makes it work is on like, being able to determine whether someone's actually real or not, that's like a huge aspect of it, because otherwise, you get lots of fake accounts set up where people try to take the money or whatever. So so that's been, I'm very proud of that. And actually, I was also an investor in cameo comm, which is probably going to be a unicorn doing extremely well. And that was actually something where I got in touch with them. Because early in urn, we had some of the same folks on there. And, you know, everything earned would be a competitor, potentially, but you know, like, my thinking, then my thinking now and my thinking generally is a positive some person, and I felt like they, you know, they were going to do something quite different. I felt tasking, in the earn sense, is not a company. It's like an industry like socialists, you know, and so like, if you look at, get coin and stuff like that, I think you're gonna see tasking, be as huge by 2030, or 2035, as social media is, I think we're still in the early days, it's still building the input from it. And I think urn is one of the first to prove that at the scale of, you know, basically hundreds of millions of dollars. Again, I'm not sure all the stats have gone out there, but I'm giving up getting the stuff which has been public, you can see like the stellar deals and other deals are out there. That's what I'm proud of.
The revenue revenue is great the revenue model for earn, like, so. So stellar, stellar contributes the crypto to give to people for learning about stellar. But then like, they pay they pay Coinbase a fee for, for for the privilege?
I think it depends on the project. When I was there, we didn't charge fees directly, because it was just win win win for all parties. Like it was a win for Coinbase. Because our customers are happy is a win for the customers because they've got free crypto, right. But who are they not really free, they have to earn it, but but it was like relatively, you know, fun to earn, and as a win for the project, because they got highly qualified users, you know, they could target engineers or what have you, right. So, um, but you know, we never ruled out the possibility of charging a fee. And I think that'd be on a project by project basis. So you know, contact Coinbase, whatever.
Okay, cool. Anyway, I interrupted, you go on so many
lessons learned. I mean, I think, um, you know, the turnaround of 21 was, was challenging, but I'm proud of where it landed up, I wrote that whole blog post on it. So I won't, you know, go back and rehash everything, but the post called the turnaround, um, I think, what was what did I learn about that? Ah, it's, it's, it's interesting to think about what I would have done differently with the information I had at the time. Um,
it's challenging, because, you know, like taking over a hardware business, and like trying to figure out something and like pivoting it, and so on, when there's no clear door, one of the things that made it particularly hard was that the Bitcoin roadmap sort of silently changed in the middle from, you know, Gavin Anderson had written like, the billion transaction roadmap, if you remember this in late 2014, right, remember that? Yes. Yeah. And then that changed, of course, to the whole small block roadmap, we're not litigating that. But the point being that I'm, like, mid pivot to realize that your pivot is kind of blown up by other factors, because micro payments became increasingly impossible on on chain, Fine, whatever, you know, like, that's, that's now in the rearview mirror. But you know, they did. That was something which, I mean, I had known this before, and sometimes you just have bad luck. And sometimes you have good luck. Um, but, you know, there's this concept of, again, not new to me, but just something I might have stressed even more of, like, really trying to minimize every possible risk, and just do the minimum number of risks at the same time. That's one concept. I mean, the second thing that's interesting about it is, like, I actually think a lot of the ideas that we put out at 21, prior to focusing on one of them, which turned into earn, which became very successful, um, you know, like paid API keys and, you know, machine payable, URLs, and so on and so forth. I think a lot of that stuff is going to come about, um, a lot of this tutorials and stuff we had all work If you swap in basically another coin that is more suitable for, like small payments, or more programmable payments. And so and it turned out that paid email, you know, we did a whole spreadsheet on it. So systematic paid email was one that was kind of the least bandwidth requirements. And where we could do a lot of batching, where people would do a bunch of paid emails, you know, in a centralized way. And then once you mail the cash out, and the cash out was an on chain transaction, and you might answer 20 paid emails and only cash out once. So we've reduced the on chain volume, which was like the number one engineering constraint, right. Um, I think a lot other things will work, what I would probably want to do is like republish them with like, open source edits of them to make them work with, you know, either aetherium, or one of the other coins that's out there, right now, I just have to kind of set up the dev environment to do it. So I haven't, I haven't sat down and tried out every point, but maybe, like salon or something like that, that that might be better for lots of fast payments. So that's, that's me what I would have done. Not sure if it would have done differently. But um, it's, that's something I've thought about is just seeing if you can mitigate risk even more.
All right, so we're over time. So I'll just basically, we'll close it out. Just anything that we haven't talked about, that you think, is important for the audience to know that we should talk about.
Um, I think crypto is what comes after Silicon Valley. I think crypto is sort of in the range that, you know, web 2.0, and so on was in the early 2000s, where there's a core group of true believers, and the desktop people have sort of written it off. I mean, the thing that's interesting is I talked to a very senior Microsoft executive A while back. And he said something offhand, that I thought was fascinating. He said, No one was making money on the internet except Google until this guy after 2008. I was like, what a fast. I mean, that's, that's true for when you're talking to Microsoft scale money, right? And that's actually why bomber was able to ignore the internet for so long, because he thought it was just Google or Google Plus apple. And Apple was really doing hardware. So it wasn't really an internet company, you know, and Amazon was fine. But you know, Amazon is a little bit of a late bloomer, where it kind of only became the Goliath that it is, or last 10 years. You know, I obviously was big before that, Amazon, AWS is big. But, you know, the first AWS exit was Instagram in 2012. Right, the cloud is not that old, you know, on and in fact, AWS was the kind of business that everybody got warned not to do. Because it was so capital intensive, it had been the graveyard of empires, like Ben Horowitz barely survived with loud cloud. And, you know, for Amazon to do that several years later, was an enormous risk for them in the mid 2000s. And, and so we sort of forget how relatively recent, that time period was, where angel investing wasn't something everybody was doing, where VC was still something where, you know, as all 30 somethings investing in 30 something well, it's like 50 somethings investing in 30 somethings 20,000,030 somethings right. And, and the entire yc, web 2.0 SAS put everything online, believe in the internet, the internet is still real, it's not a hype, it's not a fad, it's here to stay. That whole thing, basically, you know, it's not the only thing, but the initial sort of yc thing was just take every desktop app and put it online. Right. And so of course, Google did that with docs and PowerPoint and word. Um, but, you know, in a real sense, if you look at many of the, you know, SAS apps, they're doing that they're taking some offline desktop functionality or business process and putting that online, right. So major theme, and and now what we've got is something similar, where crypto is sort of off to the side of tech overlaps with it. But it's still really not taken seriously by lots of people in tech. And what's what's interesting about it is, it's the same kind of potential, in some ways, also that open source offer, which is to say, crypto offers open state and open execution. And if you understand what that is, if you understand that, hey, not just access to the source code, but access to the database, and access to reviewing the history of every opcode is actually really important. Um, you're like, wow, okay, if open source was this big, how big your Open Season open execution? How big is a tab and database for every user root user? And so I think that crypto becomes the next Silicon Valley for for those reasons. Then a couple more, which is crypto is remote first, by default. crypto is international by default. crypto is post American by default, right? crypto is encrypted. It is, you know, that's why I had this article earlier this year, Bitcoin becomes a flag of technology, where a lot of the implicit values of Bitcoin or the implicit values of tech become explicit in Bitcoin and just to like, quote from that article for a second. Basically there's like 10 values, which are, uh, you know, it's internationalist. Bitcoin is internationalist. And so it's tech. It's capitalist, it's decentralized. It's hyper deflationary. It's networked. It's encrypted. Digital, it's volatile. It's ambitious. And it's quietly revolutionary, meaning you're not standing out in the street corner. Revolution comes from a million individual voluntary actions in the privacy of one's home or, you know, one's phone. Right? And, and yet, the change still still happens, right? And, you know, I mentioned Why is tech decentralized? Well, there's no joke king of tech. Right? You know, it's funny, like, Ben Feldman says, Google AI, which is like, Man, it's amazing that it's just amazing. The Tech has monopolies. It's amazing how many of them there are.
Hahaha right? Meaning, Oh, you got so many monopolies Well, okay, maybe they're not actually monopolies. And you can of course, argue that for any given vertical, whatever it but it is decentralized, where there's no like one person who speaks for all tech, it's like, and there's a bunch of different VCs, angels, lots of different theses. There's a community, but it is, it diverges in many ways. And what crypto is is sort of like the muscular version of that, you know, it's it's the, it's a version that puts those values into code on and has done so fortunately, in a, in such a way that's very hard to corrupt. And I think about TEALS. One liner is not really one liner is observation 10 years ago, which I didn't really understand at the time. But I think about a lot in terms of how precious it was. And he said, you know, the race between politics and technology may turn out to be very close, it may come down to the wire. And the fate of humanity may depend on one person who makes a world safer capitalism. And that's Satoshi. Right? And like Satoshi, I think had just shipped Bitcoin, like two months before teal wrote that in 2009. And I'm really amazed. I, you know, that that's actually Till's most pressing line, because, you know, to see that in 2009, maybe it's because he had seen the financial crisis and, you know, the, the printing of money and so on, that was happening. But also, it seemed the rise of Facebook, probably, you know, those are the kind of the data points that are in his head at that time. But, but boy, it really seems to hold today when you have on the one hand, all of this surveillance and censorship and, um, you know, like corporate D, platforming, and just general polarization negativity and whatnot happening. On the other hand, you have, you know, supersonic airplanes finally happening in Mars landers and you have crypto and you have GPT. Three, and you have, like, you know, even though it was held back by the FDA, you have these fast vaccines, you actually have a lot of good things happening. And so it's like this, you know, really neck and neck race where the collision, I think, you know, one of the big collisions is going to be mmt versus BTC. This decade.
perfect way to end it. Thanks, man. I'm gonna hit stop on the record.