YouSaid · the spoken record

Charles Hoskinson

lines on the record: 392
first: 2021-06-16
most recent: 2021-06-16
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“Yeah, we call it Plutus Fest, and we did the first one in 2018, 2019. I can't remember. And we were going to do one last year, but then COVID hit. So we'll bring it back and we'll probably do it annually at the University of Wyoming for their hackathon there. In fact, it just so happens that coincides with the Gogun Summit. So we're doing that, I think the third week of September. But yeah, it's great to do an annual conference. You can bring a lot of cool people together and you can do hackathons and awards and so forth. But to the question in particular, Plutus is like any other language. Plutus core, you can compile things into it. So it's entirely possible to write a scala to Plutus' core compiler or a TypeScript compiler or something like that. But, you know, I'm a big believer of separation of concerns. And we don't live in a single chain model anymore. So, you know, you have a chain where you probably want to have different execution environments and different chains. So you have different virtual machines.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“No, I think Rust is completely replaced in need for them. Go and Rust, those are the two twins of Doom. I mean, Google Created Go.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“And they're really good at it, you know. And that's a great gateway drug because you have quick check in Haskell, you have Scala check in Scala. You can also do model checking. You can also go and use a TLA spec and make it work with Scala and so forth. So it gets you a little bit of everything. And you can then move around that entire design space in a beautiful way.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so that's a great language. And it's easy to pick up, and it's easy to hire people into it. You just find these Eastern European guys who are Java programmers for 10 years, 15 years, and they got tired of making. $20 an hour. So they picked up Scala so they can make $35 an hour.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, it's great because it's like you want it to be like Java, it's Java. And it looks kind of like a Java program. You want it to be like Python and Scripted and you use a REPL. You can do that. You want to go hardcore dot dependent object types and do like weird proofs and stuff in the functional. You do all that. You have access to all of these things. And Martin Oderski is a brilliant guy. He's done some phenomenal work basically because he was one of the guys created the JVM and he's worked on compilers for over 20 years. He did a lot of really hardcore work on trying to build a concise, nice modern language that does a little bit of everything. And it's got great applications in data science and in AI. It's also heavily used in modern companies like Netflix uses Scala for all of their microservice architecture.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Like Idris or Agda or Isabel. And there you can really dial up the correctness and write all kinds of crazy proofs. But by the way, it's like the seven people who write your code, they go on vacation a lot. You'll never get anything done. So Haskell kind of felt like a nice middle ground between those two where if we needed to pull into the left, we could. If you wanted to pull into the right, you could as well. That said, it's really amazing to see what the hybrid languages have done. If I was a new student in computer science and I said learn any language to grow your career from Scala 3 is probably the language to go.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“There you go always, or in Russian. Everybody has their favorite tools and their favorite languages, but it just comes down to what problems are you trying to solve and what problem domain do you live in? If you're inventing new protocols based on science, you're going to take the time to write a paper, go through the peer review process, as you've done personally. And you know how hard it can be to get into a conference and go through that and get your ass kicked, then you also have to apply the exact same level of care to the engineering side in terms of the implementation of that, or else you will make a mistake and that mistake will probably be an exploit in the system that destroys the security properties of the system. So we really had no choice but to go to some notion of functional. The question was, what's the Goldilocks language? To use a hybrid language like Scala and F sharp or Clojure where you still have some connection to understandable things like.NET or the JVM? Or do you go to an overly academic lab?”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Language. It's like saying what's French versus Russian versus English. Everybody has their adherence. They say, oh, they have the best poetry here.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Haskell code or functional code tends to be significantly more concise. And I actually have a real life example of that. So if you take a look at Mantis, we implemented a full Ethereum node in Scala. It's only about 14,000 or 15,000 lines of code. And you compare that with like C++ Bitcoin. I think that's 120, 150,000 lines of code. So it's almost 10 times smaller. And so less code, less to read. And you tend to read code significantly more than you would write code. So it's always an advantage for a maintenance, understandability, documentation, and other things when you have more concise code bases and also it's a lot easier for you to apply stronger tools to a functional code base like static analysis or property-based testing or these types of things than in a parative code base. But the thing is it's almost like a religion. It's like, say, or”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Performance, and are you designing for developer accessibility? Are you designing for correctness? And are you designing for a high fidelity representation of the protocol? Okay, so Haskell was chosen as kind of the version one because we knew that the kinds of people who think about that are also the kinds of people that would have an easy time reading a paper like Waraboris. Working their way through all of this, and they would do a pretty good job writing a formal specification and then translating that into running code. Then once you have that, you have a blueprint that you can actually reason about, maintain. And if you really wanted to, you could then turn that into a Rust code base or into a Java code base. Going the other way around would be pointless and counterproductive. The other side of it is that”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“And so there's going to be a lot of semantical ambiguity between those two. And what I mean by that is that you might end up implementing a wrong thing. You might think that what you've built is the paper, but the computer's not going to tell you that because the paper is written in prose and maybe typed up in latex or something. But there's no proof chain, evidence chain that you can show that there's no ambiguity. When you look at a functional language, you're a little closer to math. And so as a consequence, the translation of the papers that we spent so damn long writing and writing proofs about and so forth to code is much smaller. Now the downside is these functional languages tend to be a bit more academic and they tend to have not necessarily the best Windows support and the libraries aren't so good and also they tend to be a little slower when compared as a whole on average to languages like C, for example. So it's really a question of, okay, what are you designing for for version one? Are you designing for”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“What does this need to look like? Functional programming is what we've chosen to use for Cardano mostly because we're living in the academic world. We've written 105 papers. And the problem is you have to translate that work into code. And the gap between an imperative language like a C++ or C and these academic rigorous papers is extremely large.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“But that's a great starting point. And it's not about falling in love with language. It's just falling in love with computing. It's about falling in love with having a dialogue with a computer and thinking about, well, how would I solve that? How would I interact with that?”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so there's the pedagogy of learning how to program and to express the theory of computer science. Like you have to learn how to write algorithms. You have to learn what data structures are. You have to be able to do analysis of these things. And that probably, I think the debate is over Python is probably the best language or JavaScript to get started with because they're very useful. The libraries are amazing. There's just tons of online materials. Even MIT is now teaching their introduction to computer science in Python. And they used to do Lisp. I mean, these guys were hardcore.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Exactly, but maybe we're the problem, not evolution, you know, because the markets, the market, and you can scream and cry and pout and stamp your foot and said this makes no sense. But that's the way the world works. There's plenty of mountain climbers that didn't want gravity to apply to them. It's the same situation here. There's plenty of people in these marketplaces that had the best of intentions, the best team, the best technology. And for whatever reason, they didn't get that adoption. So the question isn't the local, it should be the long term. And will the system over time converge to a state that actually is useful and meaningful to society and actually solve problems for it? And that's what we try to figure out is how do we perturb these things in a way to kind of push them in that direction.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“She got an IBM research, and Craig Jeffrey, the guy did homomorph encryption on her damn bonnet's there. There's all these amazing people on that team, and they have money in the VCs. So you'd say, okay, that's a contender. But if you look at market adoption, Ethereum Classic sometimes is above it. Other things are above it. So it's”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“In years past, you never got a chance to experiment with monetary policy. I mean, it's like every 20 years, 30 years you'd have some conference usually in a cool island like Jamaica and be like, okay, let's go talk about monetary policy and amend the Bretton Woods Agreement. And these would be nation states, invitation only. And now you have over 8,000 cryptocurrencies floating around all with their own monetary policy and their rules. And it's very Darwinian. A lot are dying, some are succeeding. Anomalies happen like Dogecoin and you say, God, is this temporary? Is this permanent? Why doesn't this horrible thing die? And then other things you think would be absolutely successful and just take off and be in the top 10 don't really get as much traction like El Grand is a great example of that. I mean, Sylvio, he's an incredibly bright guy. Every time I go to MIT, we have dinner. And, you know, his work is legendary and it's just beautiful and elegant. And he literally has all the people. He went and hired Tal Robin.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“That disrupts it have an incentive to push it back into a healthy productive, useful state, which is going back kind of to that complexity theory stuff that we began with and a little bit about how do you handle modern economics. Like, for example, we knew this coming into the COVID crisis that there would be catastrophic economic disruption throughout the entire world. In the developed world, it was print lots of money and hope to God it works. In the developing world, is try not to starve to death. Over a hundred million people were pushed into acute starvation, so acute hunger. It's a terrible situation. But every economist knew we were going into that. So the question is, how do you restart the system? How do you realign the system and so forth and make sure that it doesn't collapse at some point? So it's an imperfect, inexact science. And that's actually one of the things that makes our industry so much fun is that these are kind of like micro experiments for the macro.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Russian colonel, and they installed a new satellite system, and it said, Hey, the Americans are launching missiles at us. You need to turn the key and launch all the missiles in the silo. And he said, oh, that's not right. That doesn't seem right. And he was reprimanded for not launching the missiles. So in both cases, a single person stood against the systems of superpowers between us and nuclear annihilation. So in general, we're really bad at building these types of things. So what you look for instead is say, can the system be self-correcting? It's not about avoiding a problem. It's more about can the problem be resolved? And that's how nature engineers think. It gives you an immune system. It gives you the ability to heal if a rainforest has a fire or some catastrophic event. The ecosystem will find a way to patch things up. So it's a better question of how do you align the incentives over the long term of a system where all the actors within this system, when an event occurs.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, it's a great question. And the problem with game theory and mechanism design is that you can develop this concept of a rational actor. And I don't think in my life I've ever met a rational actor. You know, there's a rational actor on Tuesday, but any other day of the week, who the hell knows? And there's even, I think there was a book Freakonomics, and there's a few of these things where it just shows again and again where people behave in ways that are against their best interest. So then you have these protocol designers and they say, well, we need an honest majority for this thing to work. And they say, oh, okay, well, we'll create this incentive model and rational actors will behave with that incentive model. And they say, well, the individual won't do that, but the firm, the government, the entity will. The problem with that is we have a lot of counter examples where the system was actually behaving in weird ways. Like we almost completely eradicated the human population twice in the 20th century, once during the Cuban Missile Crisis. And again, in the 19th century,”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Exactly. That's why you pay the guy who makes the break software for your train last and you make sure he writes the train every day.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Somewhere, it's running on your cell phone. It's running on your mom's phone. It's running your dad's phone. It's running on some computer in Mongolia that may have good internet on Tuesday, but not any other day. So when you live in that kind of environment, you really do need to think carefully about a whole new class of protocols. And then you need to think carefully about a whole new class of tools and techniques to test the reliability of those systems. And you need to separate the world and say, what is high assurance and cannot fail? Because if it fails, people lose money. And what is low assurance? And it's okay if that falls apart. The other thing I'll mention is there's a perverse financial incentives in our industry because the reality is when something blows up, the people who built those things that blow up usually get paid up front. So what they're focusing on is time to market, speed to market, and getting tokens out and getting them liquid. And then people come in, they buy it, but if there's inniscient bug in some DeFi protocol, it'll probably...”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Gargantuan systems. If you look at the Google Plex or what Amazon has or others, and they have so much value, so many users, so many things going on, and no person can keep that in their head. And so you're talking about systems may have 10 million lines of code, 15 million lines of code, millions of nodes connecting, faulty processes happening all the time, hackers breaking in on a regular basis. So when you're trying to model all that, trying to ask yourself, what formal guarantees and properties can I get to simplify this system as much as possible? So instead of the applications of formal methods slowing you down, in many cases, it actually massively reduces your debugging time and your ability to find where errors occur. In some cases, you can't find where errors occur in these massive concurrent systems. And you say, well, where are cryptocurrencies going? We're talking about just the same thing. But we're talking about a much more hostile operating environment where instead of it running in a pristine data center in California,”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Theory and Byzantine actors. So the balance there is saying, okay, what can't fail in that system? And that's the kind of stuff that you want to apply as heavy a tool set as you can because when that stuff fails, you either have the loss of billions of dollars, privacy, or potentially even life, depending on how these systems get adopted. But then other things, what can fail? Is it okay if the block doesn't get made every now and then? Is it okay if your latency goes up or your network suddenly becomes as synchronous and you disconnect from it and you have to restart the computer or something like that? That's probably okay. It's an inconvenience and burden to the user. And if you actually try to chase that tail, you'll end up spending 10 years chasing phantoms and ghosts. And meanwhile, the whole world moves on. So it's really figuring out those balance of the two. And what's really beautiful is that the formal methods tools have gotten so much better over the last 20 years in particular, mostly because of incredibly high investments from Microsoft and Google and big universities, because these guys are building these.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Maybe you can prove you'll never have a buffer overflow, or you won't divide by zero somewhere or something like that. Some event won't occur that'll cause a catastrophic failure in your system. But there's always this combinatorial explosion between what you contest and think about and what you can actually code. So the stuff on the left-hand side lives in a different cardinality, a different universe. There's something significantly larger there. And the tools on the right-hand side, we have property-based testing and these SAT solvers. We have all this great stuff here in formal methods LAN and computer science theory land, but there's only a small subset of things that they actually give you good answers about. So the balance of the two things is basically saying, well, what do you care about and what are you okay throwing away? That's the art of engineering and building these types of things. So, you know, cryptocurrencies, we deal with these complex distributed systems that have cryptography and game.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“So the engineer, they're obsessed with the domain of, well, what do you want to accomplish and who are you accomplishing it for? So they live in the world of people. If they're good engineers, and they say, okay, what's the experience? How are we going to use this? Why are we going to do this? What's the commercial application? What's the non-commercial application? And you collect all these business requirements. Once you've done all of that, the better job you do, the more self-evident it is of how do we apply the toys and tools of computer science and other such things to actually resolve that. And the point of theoretical computer science from the software engineering domain is it can tell you kind of where your guardrails are. It won't make perfect programs and there's no such thing as that, but rather it can give you a good notion and sense that your program has some desirable properties. Maybe you can prove that it can terminate if you're dealing with total programs or”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Did you ever watch Blade Runner 2049? Yeah, did you remember the whole relationship between joy and K? And like, did she really love him, the hologram or not? Was it like fake love or real love?”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Processing. We have better ways of handling speech to text, and we have better ways of computers kind of understanding the intent of what a human being is saying. And then the question is, well, how will a computer understand love or poetry or music? Well, it'll understand it the same way. We understand it. You have to go to a computer to grow up to a point where it can learn the way we learn or as close to it as possible. Then you just expose it to the things that we were exposed with. And then at some point, the computer will start creating things. So the question is, well, how do you quantify creativity? And I have no clue about that.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“I don't think it's human nonsense at all. I mean, there's even attempts to create algorithmically generated music and the question is, is love just strictly a chemical phenomena? Is there something metaphysical about it or transcendent of some sort of formal system? I mean, computer science is just saying, hey, we have this notion of computing. We have this brain that we've constructed, this formal system that we've built. And given that we have it, what can we do with it? And so some people worry about the roots of the tree of knowledge, the great Ygdrazole of computer science. They worry about the roots and say, how far can we grow them? And let's keep adding these new models of computation. And other people worry about the trunk of the tree and some people worry about the leaves of the tree. And the more advanced the field gets, the closer and closer it gets to the people who constructed it, us, you know, we have better image.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“I start tweeting in it. Yeah, there you go. So, you know, there's a lot there, and they're just fun to study and think about. And unfortunately, if you go down that rabbit hole, you'll spend way, way too much time. And there's diminishing returns. Now, the second question you asked was one on theoretical computer science to, I guess, engineer. Philosophy.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“And computers live in the formal world, and human beings live in the natural language world. And those bridges between those two are still not completely clear. And so a lot of the work that these guys were doing, the 20th century, 19th century had nothing at all to do with that, but gives you hope that perhaps a bridge can exist between those two worlds. And maybe there are some nice tools for that bridge to be built upon. And maybe that in some way will allow computers to better understand us. I mean, they've even created languages, natural spoken languages that are completely ambiguity-free, like Lochbahn and things like that.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“And so, this is just one of those guys that you can see through his work that he can think deeply about anything. It's like the Gawois of philosophy. And he chose to try to clean up a lot of the messes that Tarski and others couldn't resolve. He said, let's really get serious about the nature of truth. Let's really try to resolve paradoxes. Let's really try to build things in such a way that the work that we leave behind can actually be built upon and it's not thrown away every 50 years or 100 years. And it was the same for Tarski. He comes in. He says, we mathematicians love this concept of truth, but yet we've never really created a nice rigorous definition that doesn't have paradoxes embedded inside of it. So we had to invent meta languages and object languages and all these notions and so forth. So I really like those four if you think about them. And there's a lot of great lessons. And where it's relevant today is, you know, you have human beings and you have computers and they're trying to understand each other.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, exactly. Yeah, he's getting up there, man. But I mean, literally, when he was in high school, he wrote these papers and logic and Harvard contacted him and said, hey, could you teach graduate courses at Harvard? And he said, no, I really would like to finish high school first before I go and teach grad school at Harvard.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Dude, not aristocratic at all. And he was one of those guys that just could have done anything because he was so, he's so brilliant. I guess he's still alive. I think”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Logicians started chipping away at this idea that you could actually construct a complete system of mathematics, first with Girdle and then later with the work of Turing and Church and others. They said, oh, you're not complete, you're not decidable. And so suddenly, Russell was left in this really bad position where his early life's work was basically forgotten. So he had to kind of reinvent himself. And so he went into ethics and he went into different fields of philosophy. And he became this Titan in analytic philosophy. And he was also a great pacifist. And he was just a phenomenal writer. If you read why I'm not a Christian or any of his other attacks on metaphysics, he said, look, I can only deal with the world I'm in in the senses that I have. And if I can deduce it, I believe it. If it's outside of that, I really can't make meaningful statements about it. And he said it in a lovely English prose that you would expect of a man of his stature. Now, Saul Cripri, it was like the complete opposite. This guy's really down her.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“I think you're right about that. However, Kripke has Kripkeenstein. He has his. His whole building on that, right? And he built a whole higher. And at least there I have You can then start getting a better understanding of basically how far a formal language can take you. And so that's the work of it. David Hilbert also did the same thing. And Russell, he got his career started working with Alfred North Whitehead. He was a logician. And there was this whole desire in late 19th century mathematics to formalize mathematics in a completely new and better way. And they started with geometry and Hilbert's geometry was like a complete system, although recently we've discovered there's a few holes in that. But for the most part, it was complete and the axioms are independent and they're consistent. And they said, oh, well, now we can do this for all of mathematics. And Russell and Whitehead wrote this huge set of books, like two big books Principia Mathematica, a thousand pages. And the conclusions one plus one equals two. So they linked set theory and arithmetic and logic all in these beautiful ways. Then little by little, as we entered the 20th century,”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so yeah, there's both a deeply human element to both Russell and Saul. And then there's this, of course, amazing work that they did in the late 19th and 20th century. And you can't really talk about Russell or Saul without also mentioning Wittgenstein and Tarski. Because when you actually look at these guys and you put them together, what they were attempting to do was increase the level of precision we had in analyzing both formal languages and also language in general. And so evicstein makes no sense at all to me. So there's amazing people out there that somehow can parse that.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Okay, well, that's going to be fun. So, we're going to set up an auction system or something like that with him, and I think maybe end of this month or next month, we'll figure out how to do NFTs on Cardano with Wolfram Universes. So as soon as we can sell one, I'll give you one. And then we can all just claim that we're living in some sort of Wolfram simulation.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, and it's not a new concept. I mean, like the Santa Fe Institute's been doing that for a long time, and you can use it for economic modeling, and you can show that in certain cases there's this concept of simple rules evolving into a complex system is somewhat more predictive than trying to build a complex top-down model for things. And I guess there's some analogies, these things in AI where you start with some simple things and then somehow it just figures stuff out in its environment over time and much better than if you actually tried to model it with prologue or something like that. So that is exciting because you get to just do a few things, let the thing run, and then see what happens. And that's a lot of fun. In fact, it got so exciting. Wolfram came to us and he said, hey, let's do an NFT marketplace. And I said, what do you want to do? And he said, I got all these universes to sell.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, exactly. And so I have no clue what type of language it would have to look like, but it would probably not be anything we're used to at the moment. It'd have to be something else. And we'd have to have some more fundamental resolution of the relationship these formal systems and how they get extended. So that's like a 22nd century question.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“Well, also the question is it's a computation question. Does P equal NP because you'd have to be able to emulate all of the long-term?”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source
“There are a lot of really lovely papers. Like one came out of MSR, the autodidactic universe. Did you have a chance to see that? All you got to read it is like April of 2021. It's just brand new. But basically, the idea is that the universe is like some sort of giant self-learning system that can self-evolve almost like gnomic. And then you have Wolfram running around saying, hey, we can come up with these like very simple rules and we can reconstruct all of reality at some arbitrary point. So I have absolutely no idea what's right. I look at it kind of like a formal system and I say, well, if you're stuck within the system, it's hard to actually understand that you're inside the system. You know, it's kind of like an object language to a meta language, there's this thing that is outside of it, but because you're constrained and limited by the simulation, you can't really understand the nature of the thing that's outside of it. It's almost like Minecraft. You can build these redstone computers within Minecraft and simulate and emulate things within it, but you really can't go.”
2021-06-16 · Lex Fridman Podcast · #192 – Charles Hoskinson: Cardano · IDENTIFIED FROM THE TRANSCRIPT · source