Vitalik Buterin: Ethereum, Cryptocurrency, and the Future of Money

00:00:00.000 | The following is a conversation with Vitalik Buterin,

00:00:03.620 | co-creator of and author of the white paper

00:00:06.360 | that launched Ethereum and Ether,

00:00:09.480 | which is a cryptocurrency that is currently

00:00:11.880 | the second largest digital currency after Bitcoin.

00:00:15.640 | Ethereum has a lot of interesting technical ideas

00:00:18.640 | that are defining the future of blockchain technology.

00:00:21.600 | And Vitalik is one of the most brilliant people

00:00:24.320 | innovating in the space today.

00:00:26.900 | Unlike Satoshi Nakamoto, the unknown person or group

00:00:31.560 | that created Bitcoin, Vitalik is very well known

00:00:36.160 | and at a young age is thrust into the limelight

00:00:40.080 | as one of the main faces of the technology

00:00:42.840 | that may redefine the nature of money

00:00:45.520 | and all forms of digital transactions in the 21st century.

00:00:49.180 | This is the Artificial Intelligence Podcast.

00:00:52.920 | If you enjoy it, subscribe on YouTube,

00:00:55.360 | review it with five stars on Apple Podcast,

00:00:58.080 | support it on Patreon, or simply connect with me on Twitter

00:01:01.400 | at Lex Friedman, spelled F-R-I-D-M-A-N.

00:01:05.620 | As usual, I'll do one or two minutes of ads now

00:01:09.760 | and never any ads in the middle

00:01:11.120 | that can break the flow of the conversation.

00:01:13.280 | I hope that works for you

00:01:14.680 | and doesn't hurt the listening experience.

00:01:17.560 | Quick summary of the ads.

00:01:19.160 | Two sponsors, Masterclass and ExpressVPN.

00:01:24.000 | Please consider supporting the podcast

00:01:25.760 | by signing up to Masterclass at masterclass.com/lex

00:01:30.760 | and getting ExpressVPN at expressvpn.com/lexpod.

00:01:35.760 | This show is sponsored by Masterclass.

00:01:41.440 | Sign up at masterclass.com/lex to get a discount

00:01:45.760 | and to support this podcast.

00:01:48.320 | When I first heard about Masterclass,

00:01:50.400 | I honestly thought it was too good to be true.

00:01:53.060 | For $180 a year, you get an all access pass

00:01:56.840 | to watch courses from experts at the top of their field.

00:02:00.920 | To list some of my favorites,

00:02:02.800 | Chris Hadfield on space exploration,

00:02:05.440 | Neil deGrasse Tyson on scientific thinking

00:02:07.520 | and communication, Will Wright, the creator of Sim City

00:02:11.840 | and Sims on game design.

00:02:14.500 | I love that game.

00:02:16.080 | Jane Goodall on conservation,

00:02:18.080 | Carlos Santana, one of my favorite guitarists on guitar,

00:02:21.640 | Garry Kasparov on chess.

00:02:23.760 | Obviously I'm Russian, I love Garry.

00:02:27.020 | Daniel Negrano on poker, one of my favorite poker players,

00:02:30.060 | also Phil Ivey gives a course as well, and many, many more.

00:02:35.060 | Chris Hadfield explaining how rockets work

00:02:38.300 | and the experience of being launched into space alone

00:02:41.420 | is worth the money.

00:02:43.140 | By way of advice, for me, the key is not to be overwhelmed

00:02:46.220 | by the abundance of choice.

00:02:48.420 | Pick three courses you want to complete,

00:02:50.420 | watch each all the way through from start to finish.

00:02:52.800 | It's not that long, but it's an experience

00:02:55.000 | that will stick with you for a long time, I promise.

00:02:58.160 | It's easily worth the money.

00:02:59.680 | You can watch it on basically any device.

00:03:02.260 | Once again, sign up at masterclass.com/lex

00:03:06.280 | to get a discount and to support this podcast.

00:03:09.920 | This show is sponsored by ExpressVPN.

00:03:13.120 | Download it at expressvpn.com/lexpod

00:03:17.580 | to get a discount and to support this podcast.

00:03:20.900 | I've been using ExpressVPN for many years.

00:03:23.820 | I honestly love it.

00:03:25.780 | It's easy to use, press the big power on button

00:03:29.160 | and your privacy is protected.

00:03:30.860 | And if you like, you can make it look like

00:03:33.400 | your location is anywhere else in the world.

00:03:35.820 | I might be in Boston now, but I can make it look

00:03:38.320 | like I'm in New York, London, Paris, or anywhere else.

00:03:42.780 | This has a large number of obvious benefits.

00:03:45.400 | For example, certainly it allows you to access

00:03:47.900 | international versions of streaming websites

00:03:50.360 | like the Japanese version of Netflix

00:03:53.060 | or the UK version of Hulu.

00:03:55.520 | As you probably know, I was born in the Soviet Union.

00:03:58.360 | So sadly, given my roots and appreciation

00:04:00.720 | of Russian history and culture, my website

00:04:03.640 | and the website for this podcast is blocked in Russia.

00:04:08.360 | So this is another example of where you can use ExpressVPN

00:04:12.440 | to access sites like the podcast

00:04:16.100 | that are not accessible in your country.

00:04:18.680 | ExpressVPN works on any device you can imagine.

00:04:21.900 | I use it on Linux, shout out to Ubuntu, Windows, Android,

00:04:26.640 | but it's available everywhere else too.

00:04:29.180 | Once again, download it at expressvpn.com/lexpod

00:04:34.180 | to get a discount and to support this podcast.

00:04:38.640 | And now here's my conversation

00:04:41.060 | with Vitalik Buterin.

00:04:42.940 | So before we talk about the fundamental ideas

00:04:46.100 | behind Ethereum and cryptocurrency,

00:04:49.440 | perhaps it'd be nice to talk about

00:04:52.060 | the origin story of Bitcoin

00:04:54.640 | and the mystery of Satoshi Nakamoto.

00:04:58.360 | You gave a talk that started with sort of asking the question

00:05:01.860 | what did Satoshi Nakamoto actually invent?

00:05:06.020 | Maybe you could say who is Satoshi Nakamoto

00:05:08.360 | and what did he invent?

00:05:10.060 | - Sure, so Satoshi Nakamoto is the name

00:05:15.060 | by which we know the person

00:05:17.240 | who originally came up with Bitcoin.

00:05:19.780 | So the reason why I say the name by which we know

00:05:22.200 | is that this is an anonymous fellow

00:05:25.600 | who has shown himself to us only over the internet

00:05:30.600 | just by first publishing the white paper for Bitcoin,

00:05:35.980 | then releasing the original source code for Bitcoin,

00:05:38.520 | and then talking to the very early Bitcoin community

00:05:42.560 | on Bitcoin forums and interacting with them

00:05:45.980 | and helping the project along for a couple of years.

00:05:48.900 | And then at some point in late 2010 to early 2011,

00:05:53.900 | he disappeared.

00:05:55.600 | So Bitcoin is a fairly unique project

00:06:00.100 | in how it has this kind of mythical,

00:06:04.440 | a kind of quasi-godlike founder

00:06:06.240 | who just kind of popped in and did the thing

00:06:09.120 | and then disappeared and we've somehow

00:06:11.240 | just never heard from him again.

00:06:13.080 | - So in 2008, so the white paper was the first,

00:06:16.800 | do you know if the white paper was the first time

00:06:18.820 | the name would actually appear, Satoshi Nakamoto?

00:06:22.360 | - I believe so.

00:06:23.700 | - So how is it possible that the creator

00:06:26.780 | of such a impactful project remains anonymous?

00:06:31.480 | - That's a tough question.

00:06:32.760 | And there's no similarity to it in history of technology

00:06:36.920 | as far as I'm aware.

00:06:38.240 | - Yeah.

00:06:39.160 | So one possibility is that it's Hal Finney

00:06:41.800 | because Hal Finney was kind of also active

00:06:46.320 | in the Bitcoin community as Hal Finney

00:06:49.480 | in those two beginning years.

00:06:53.480 | And Hal--

00:06:54.480 | - Who is Hal Finney maybe?

00:06:55.600 | - He is one of the people in the early Cypherpunk community.

00:07:00.600 | He was--

00:07:01.480 | - So he's a computer scientist, just one of the--

00:07:03.560 | - Yeah, computer scientists, cryptographers,

00:07:06.160 | people interested in technology, internet freedom,

00:07:11.160 | like those kinds of topics.

00:07:14.720 | - Is it correct that I read that he seemed

00:07:16.800 | to have been involved in either the earliest

00:07:19.160 | or the first transaction of Bitcoin?

00:07:21.720 | - Yes.

00:07:22.560 | The first transaction of Bitcoin

00:07:23.680 | was between Satoshi and Hal Finney.

00:07:25.440 | - Do you think he knew who Satoshi was?

00:07:28.440 | If he was--

00:07:29.280 | - If he wasn't Satoshi, probably no.

00:07:31.240 | - How is it possible to work so closely with people

00:07:35.040 | and nevertheless not know anything

00:07:37.000 | about their fundamental identity?

00:07:39.560 | Is this like a natural sort of characteristic

00:07:42.520 | of the internet?

00:07:43.480 | Like if we were to think about it,

00:07:46.800 | 'cause you and I just met now,

00:07:49.760 | there's a depth of knowledge that we now have

00:07:53.360 | about each other that's like physical.

00:07:54.960 | Like my vision system is able to recognize you.

00:07:58.360 | I can also verify your identity of uniqueness.

00:08:01.560 | - Yep.

00:08:02.400 | - Like it's very hard to fake you being you.

00:08:05.560 | - Yes.

00:08:06.400 | - So the internet has a fundamentally different quality

00:08:10.200 | to it, which is just fascinating.

00:08:11.680 | Can you maybe talk to that?

00:08:12.520 | - Yeah, no, this is definitely interesting

00:08:14.040 | as I definitely just know a lot of people

00:08:19.040 | just by their internet handles.

00:08:20.960 | And to me, when I think of them,

00:08:22.960 | like I see their internet handles

00:08:24.480 | and one of them has a profile picture

00:08:28.120 | of this kind of face that's kind of not quite human

00:08:31.840 | with a bunch of kind of psychedelic colors in it.

00:08:33.960 | And when I visualize him, like I just visualize that.

00:08:37.240 | - That, not an actual face.

00:08:39.400 | - Yeah.

00:08:40.760 | - You are the creator of the second,

00:08:43.560 | well, at least currently,

00:08:44.400 | the second most popular cryptocurrency, Ethereum.

00:08:47.880 | So on this topic, if we just stick on Satoshi Nakamoto

00:08:51.160 | for a little bit longer,

00:08:52.880 | you may be the most qualified person to speak

00:08:55.000 | to the psychology of this anonymity

00:08:57.640 | that we're talking about.

00:08:59.000 | Like your identity is known, like I've just verified it.

00:09:03.480 | But from your perspective,

00:09:05.160 | what are the benefits in creating a cryptocurrency

00:09:09.520 | and then remaining anonymous?

00:09:11.160 | Like if it can psychoanalyze Satoshi Nakamoto,

00:09:14.600 | is there something interesting there?

00:09:17.120 | Or is it just a peculiar quirk of him?

00:09:20.200 | - It definitely helps create this kind of image

00:09:24.360 | of this kind of neutral thing that doesn't belong to anyone.

00:09:27.920 | That you've created a project

00:09:31.400 | and because you're anonymous

00:09:33.400 | and because you also have disappeared,

00:09:36.840 | or as unfortunately happened to Halfini, if that is him,

00:09:39.920 | he ended up, I think, dying of Lou Gehrig's disease

00:09:42.840 | and he's in the cryogenic freezer now.

00:09:45.080 | But like if you pop in and you created and you're gone

00:09:51.440 | and all that's remaining of that whole process

00:09:56.360 | is the thing itself,

00:09:57.520 | then no one can go and try to interpret

00:10:02.520 | any of your other behavior and try to understand like,

00:10:05.720 | oh, this person wrote this thing in some essay at age 16

00:10:10.720 | where he expressed particular opinions about democracy.

00:10:16.720 | And so because of that, this project is a statement

00:10:19.800 | that's trying to do this specific thing.

00:10:22.320 | Instead it creates this environment

00:10:26.520 | where the thing is what you make of it.

00:10:29.240 | And...

00:10:30.560 | - It doesn't have the, yeah, right.

00:10:32.360 | The burden of your other ideas,

00:10:33.680 | political thought and so on.

00:10:35.240 | So now that we're sitting with you,

00:10:38.720 | do you feel the burden of being kind of the face of Ethereum?

00:10:43.720 | I mean, there's a very large community of developers,

00:10:46.760 | but nevertheless.

00:10:47.960 | - Yeah.

00:10:49.280 | - Is there like a burden associated with that?

00:10:52.120 | - There definitely is.

00:10:53.400 | This is definitely a big reason why I've been trying

00:10:57.600 | to kind of push for the Ethereum ecosystem

00:11:00.360 | to become more decentralized in many ways.

00:11:03.320 | Just encouraging a lot of kind of core Ethereum work

00:11:06.840 | to happen outside of the Ethereum foundation

00:11:09.120 | and of expanding the number of people

00:11:11.800 | that are making different kinds of decisions,

00:11:13.720 | having multiple software pollutations instead of one

00:11:17.200 | and all of these things.

00:11:18.280 | Like there's a lot of things that I've tried to do too

00:11:22.240 | and remove myself as a single point of failure

00:11:25.640 | because that is something

00:11:27.240 | that a lot of people criticize me for.

00:11:30.360 | - So if you look at like

00:11:33.080 | the most fundamentally successful open source projects,

00:11:36.320 | it seems that it's like a sad reality when I think about it

00:11:41.720 | is it seems to be that one person is a crucial contributor.

00:11:46.160 | Often you feel like a lioness for Linux, for the kernel.

00:11:51.160 | - Yeah, that is possible.

00:11:52.920 | I'm definitely not planning to disappear.

00:11:54.960 | - That's an interesting tension

00:11:59.600 | that projects like this kind of desire a single entity

00:12:04.240 | and yet they're fundamentally distributed.

00:12:07.480 | I don't know if there's something interesting to say

00:12:11.000 | about that kind of structure

00:12:12.320 | and thinking about the future of cryptocurrency.

00:12:15.000 | Does there need to be a leader?

00:12:16.560 | - There's different kinds of leaders.

00:12:19.160 | There's dictators who control all the money.

00:12:22.640 | There's people who control organizations.

00:12:24.880 | There's kind of high priests

00:12:26.800 | that just have themselves and their Twitter followers.

00:12:30.520 | - What kind of leader are you, would you say?

00:12:32.640 | - Yeah, in these days,

00:12:34.520 | I actually a bit more in the high priest direction

00:12:39.800 | than before.

00:12:42.800 | I definitely actually don't do all that much

00:12:45.200 | of kind of going around

00:12:46.200 | and ordering Ethereum Foundation people to do things

00:12:50.080 | because I think those things are important.

00:12:52.600 | If there's something that I do think is important,

00:12:55.240 | I do just usually kind of say it publicly

00:12:58.200 | or just kind of say it to people.

00:12:59.920 | Quite often projects just gonna start doing it.

00:13:04.960 | - So let's ask the high philosophical question about money.

00:13:10.160 | What at the highest level is money?

00:13:13.080 | What is money?

00:13:13.960 | - It's a kind of game.

00:13:16.400 | And it's a game where we have points.

00:13:19.200 | And if you have points,

00:13:21.000 | there's this one move where you can reduce your points

00:13:22.960 | by a number and increase someone else's points

00:13:24.640 | by the same number.

00:13:25.960 | And these--

00:13:28.040 | - So it's a fair game, hopefully.

00:13:30.000 | - Well, it's one kind of fair game.

00:13:32.080 | Like for example, you can have other kinds of fair games.

00:13:34.800 | Like you're gonna have a game

00:13:35.920 | where if I give someone a point

00:13:37.280 | and you give someone a point,

00:13:38.400 | then instead of that person getting two points,

00:13:40.080 | that person gets four points.

00:13:41.720 | And that's also fair.

00:13:42.920 | But money is easy to kind of set up

00:13:49.000 | and it serves a lot of useful functions.

00:13:51.560 | And so it kind of just survives in society

00:13:55.240 | as a meme for thousands of years.

00:13:57.640 | - It's useful for the storage of wealth.

00:14:00.560 | It's useful for the exchange of value.

00:14:03.880 | - And it's also useful for denominating future payments,

00:14:08.400 | a unit of account.

00:14:09.640 | - A unit of account.

00:14:11.360 | So what, if you look at the history of money

00:14:13.800 | in human civilization,

00:14:15.160 | just if you're a student of history,

00:14:19.840 | how has its role or just the mechanisms of money

00:14:23.880 | changed over time in your view?

00:14:25.800 | Even if we just look at the 20th century or before

00:14:28.640 | and then leading up to cryptocurrencies,

00:14:30.560 | that's something you think about?

00:14:32.160 | - Yeah, and I think the big thing in the 20th century

00:14:36.320 | is kind of, we saw a lot more intermediation, I guess.

00:14:40.280 | The first part is kind of the move from adding

00:14:44.520 | more of different kinds of banking.

00:14:48.480 | And then we saw the move from dollars being backed by gold

00:14:53.480 | to dollars being backed by gold

00:14:57.520 | that's only redeemable by certain people

00:14:59.200 | to dollars not being backed by anything

00:15:02.360 | to this system where you have a bunch

00:15:06.280 | of free floating currencies

00:15:07.800 | and then people getting bank accounts

00:15:12.800 | and then those things becoming electronic,

00:15:14.680 | people getting accounts with payment processors

00:15:16.560 | that have bank accounts.

00:15:18.760 | - So what do you make of that?

00:15:22.080 | That's such a fascinating philosophical idea

00:15:25.080 | that money might not be backed by anything.

00:15:27.880 | Is that fascinating to you,

00:15:31.400 | that money can exist without being backed

00:15:33.360 | by something physical?

00:15:34.440 | - It definitely is.

00:15:36.880 | - Like what do you make of that?

00:15:38.440 | Like how is that possible?

00:15:40.360 | Is that stable?

00:15:41.280 | If we look at the future of human civilization,

00:15:43.280 | is it possible to have money at the large scale

00:15:46.360 | at such a hugely productive and rich societies

00:15:50.880 | be able to operate successfully

00:15:52.600 | without money being backed by anything physical?

00:15:55.280 | - I feel like the interesting thing

00:15:56.880 | about the 21st century especially

00:15:59.560 | is that a lot of the important valuable things

00:16:02.680 | are not backed by anything.

00:16:04.200 | Like if you look at tech companies for example,

00:16:07.040 | something like Twitter,

00:16:08.200 | you could theoretically imagine

00:16:11.400 | that if all of the employees wanted to,

00:16:14.200 | they could kind of come together,

00:16:15.840 | they would quit and start working on Twitter 2.0

00:16:19.760 | and then the value of,

00:16:24.760 | and just kind of build the exact same products

00:16:28.840 | or possibly build a better product

00:16:30.680 | and then just kind of continue on from there

00:16:33.320 | and the original Twitter

00:16:35.680 | would just not have people left anymore.

00:16:38.480 | There is theoretically kind of code and IP

00:16:43.080 | that's owned by the company,

00:16:44.240 | but in reality, good programmers

00:16:45.960 | could probably rewrite all that stuff in three months.

00:16:48.600 | So the reason why the thing has value

00:16:53.600 | is just kind of network effects

00:16:55.360 | and coordination problems.

00:16:56.880 | Like these employees in reality

00:16:59.240 | aren't going to switch all at once

00:17:00.640 | and also the users aren't all going to switch at once

00:17:04.360 | because it's just difficult for them to switch at once.

00:17:06.880 | And so there's these kind of meta stable

00:17:11.760 | and of equilibria in interactions

00:17:14.160 | between thousands of millions of people

00:17:16.200 | that are just actually quite sticky,

00:17:18.480 | even though if you try to kind of assume

00:17:21.440 | that everyone's a perfectly rational

00:17:23.040 | and kind of perfectly slippery spherical cow,

00:17:25.240 | they don't seem to exist at all.

00:17:26.840 | - That stickiness, do you have a sense,

00:17:29.920 | a grasp of the sort of the fundamental dynamic,

00:17:34.000 | like the physics of that stickiness?

00:17:36.280 | It seems to work,

00:17:37.480 | but and I think some of the cryptocurrency ideas

00:17:41.280 | kind of rely on it working.

00:17:42.960 | - Yeah, it's the sort of thing

00:17:45.760 | that's definitely been economically modeled a lot.

00:17:49.240 | One of the kind of analogy of something as similar

00:17:54.920 | that you often see in textbooks

00:17:58.080 | as like what is a government?

00:18:00.880 | Like if for example, like 80% of people in a country

00:18:04.680 | just like tomorrow suddenly had the idea

00:18:07.840 | that like the laws that are currently,

00:18:10.160 | the laws in the government

00:18:11.000 | that currently is the government

00:18:11.960 | are just people in some other thing is the government

00:18:15.600 | and they just kind of start acting like it,

00:18:17.120 | then that would kind of become the new reality.

00:18:19.600 | And then the question is, well,

00:18:20.720 | what happens if kind of between zero

00:18:23.800 | and 80 people or an 80% of people start believing that?

00:18:27.960 | And like, what is the thing you see is that

00:18:32.800 | if there is one of these kind of switches happening

00:18:35.520 | is kind of revolution,

00:18:36.600 | then if you're the first person to join,

00:18:39.080 | then like you probably don't have the incentive to do that.

00:18:44.040 | But then if you're the 55th percentile person to join,

00:18:47.280 | then suddenly becomes quite safe too.

00:18:49.480 | And so it's definitely is the sort of thing that you can

00:18:53.440 | kind of try to analyze and understand mathematically.

00:18:57.040 | But one of the kind of results is that the sort of

00:19:02.040 | like when the switch happens

00:19:06.560 | definitely can be chaotic sometimes.

00:19:09.080 | - Yeah, but still like to me,

00:19:10.480 | the idea that the network affects the fact that

00:19:13.520 | human beings at a scale of like millions, billions

00:19:17.080 | can share even the idea of currency, like all agree.

00:19:22.120 | That's just, I know economics can model it.

00:19:25.200 | I'm a skeptic on the economic.

00:19:27.040 | And it's like, so my favorite sort of field

00:19:31.000 | maybe recreationally psychology

00:19:32.880 | is trying to understand human behavior.

00:19:34.960 | And I think sometimes people just kind of pretend

00:19:37.640 | that they can have a grasp on human behavior,

00:19:40.140 | even though it's such a messy space

00:19:42.760 | that all the models that psychology or economics,

00:19:45.200 | those different perspectives on human behavior

00:19:47.440 | can have are difficult.

00:19:51.240 | It's difficult to know how much that's wishful thinking

00:19:53.640 | and how much it is actually getting to the core

00:19:55.760 | of understanding human behavior.

00:19:58.240 | But on that idea,

00:20:00.560 | what do you think is the role of money in human motivation?

00:20:05.560 | So do you think money from an economics perspective,

00:20:10.840 | from a psychology perspective is core to like human desires?

00:20:18.260 | Money is definitely very far from the only motivator.

00:20:21.980 | It is a big motivator.

00:20:23.500 | And that's one of the closest things you have

00:20:26.260 | to a universal motivator.

00:20:27.800 | Because ultimately in like almost any person in the world,

00:20:33.300 | if you ask them to do something,

00:20:35.900 | like there'll be more inclined to do it

00:20:37.580 | if you also offer them money, right?

00:20:40.660 | And that's like, there's definitely many cases

00:20:44.740 | where people will do things other than things

00:20:47.380 | and maximize how much money they have.

00:20:48.940 | And that happens all the time.

00:20:50.260 | But like a lot of those other things

00:20:53.140 | are kind of much more specific to

00:20:56.220 | and of who that person is and of where their situation is,

00:20:58.740 | the relationship between the motive and the action

00:21:00.780 | and these other things.

00:21:02.300 | - What do you think is the interplay of the other motivator

00:21:04.660 | from like Nietzsche perspective is power?

00:21:07.780 | Do you think money equals power?

00:21:09.780 | Do you think those are conflicting ideas?

00:21:11.900 | Do you think, I mean, that's one of the ideas

00:21:14.340 | that decentralized currency,

00:21:15.740 | decentralized applications are looking at

00:21:17.940 | is who holds the power?

00:21:20.780 | - Yeah, money is definitely a kind of power.

00:21:23.860 | And there's definitely people who want money

00:21:27.900 | because it gives them power.

00:21:29.300 | And then even if money doesn't seem to

00:21:34.300 | kind of explicitly be about money,

00:21:36.260 | a lot of things that people spend money on

00:21:38.620 | are ultimately about a social status of some kind.

00:21:43.340 | So I definitely view those two things as kind of interplaying.

00:21:47.780 | And then there's also money as just a way of like

00:21:52.780 | measuring how successful you are, like as a scoreboard, right?

00:21:56.620 | So this kind of gets back to the game.

00:21:58.540 | I mean, like if you have $4 billion,

00:22:02.140 | then the main benefit you get from going up,

00:22:05.140 | or one of the big benefits you get

00:22:06.580 | from going up to $6 billion is that now,

00:22:08.900 | instead of being below the guy who has five,

00:22:11.460 | you're above the guy who has five.

00:22:13.020 | - So you think money could be kind of,

00:22:15.260 | in the game of life, it's also a measure of self-worth.

00:22:19.900 | It's like how we-

00:22:21.660 | - It's definitely how a lot of people perceive it.

00:22:25.220 | - Define ourselves in a hierarchy of society.

00:22:28.700 | - Yeah, not saying it's sort of a healthy thing

00:22:30.780 | that people define their self-worth as money,

00:22:34.380 | 'cause it's definitely far from a perfect indicator

00:22:39.860 | of how much value you provide to society

00:22:43.500 | or anything like this.

00:22:44.380 | But I definitely think that as a matter

00:22:48.420 | of kind of current practice,

00:22:49.500 | a bunch of people do feel that way.

00:22:51.260 | - So what does utopia,

00:22:54.380 | from an economic perspective, look like to you?

00:22:56.860 | What does a perfect world look like?

00:23:00.300 | - I guess the economists say utopia would be one where

00:23:06.420 | kind of everything is incentive aligned

00:23:10.540 | in the sense that there aren't enough conflicts

00:23:13.460 | between what satisfies your goals

00:23:16.220 | and kind of what is good for everyone

00:23:20.380 | in the world as a whole.

00:23:22.740 | - What do you think that would look like?

00:23:26.700 | Does that mean there's still poor people and rich people?

00:23:29.820 | There's still income inequality?

00:23:32.380 | Do you think sort of Marxist ideas are strong?

00:23:35.420 | Do you think sort of ideas of objectivism,

00:23:39.900 | like where the market rules is strong?

00:23:44.220 | Is there different economic philosophies

00:23:47.300 | that just seem to be reflective of what utopia would be?

00:23:51.220 | - So I definitely think that existing economic philosophies

00:23:56.620 | do end up kind of systematically deviating

00:24:01.180 | from the utopia in a lot of ways.

00:24:03.540 | Like one of the big things I talk about, for example,

00:24:06.740 | is public goods, right?

00:24:07.860 | And public goods are especially important on the internet.

00:24:11.100 | Because like the idea is with kind of money as this game

00:24:15.740 | where I lose a few coins and you gain the same number

00:24:19.820 | of coins is that this usually happens in a trade

00:24:21.980 | where I lose some money, you gain some money,

00:24:24.380 | you lose a sandwich and I gain a sandwich.

00:24:27.780 | And this kind of model works really well

00:24:31.940 | when the thing that we're using money to incentivize

00:24:35.980 | this set of private goods, right?

00:24:37.340 | Things that you provide to one person

00:24:39.180 | where the benefit comes to one person.

00:24:40.740 | But on the internet especially,

00:24:44.740 | but also many, many contexts kind of off the internet,

00:24:48.100 | there's actions that kind of individuals or groups can take

00:24:51.540 | where instead of the benefit going to one person,

00:24:56.340 | the benefit just goes to many people at the same time

00:24:58.980 | and you can't control who the benefit goes to, right?

00:25:01.820 | So for example, this podcast, we publish it

00:25:05.460 | and when it's published,

00:25:07.780 | you don't have any fine grains control over like,

00:25:10.740 | oh, these 38,000 people can watch it

00:25:13.460 | and then like these other 29,000 people can't.

00:25:16.220 | It's like once the number goes high enough,

00:25:17.900 | then people will just copy it.

00:25:20.340 | And then when I write articles on a blog,

00:25:22.100 | then they're just like free for everyone

00:25:25.540 | and that stuff's even harder to prevent anyone from copying.

00:25:28.380 | So aside from that, things like scientific research,

00:25:32.940 | for example, and even taking more pedestrian examples

00:25:37.420 | like climate change mitigation would be a big one.

00:25:39.940 | So there's a lot of things in the world

00:25:44.420 | where you have these kind of individual actions

00:25:47.940 | with enough concentrated costs and distributed benefits

00:25:50.420 | and money as a point system

00:25:52.820 | does not do a good job of encouraging these things.

00:25:56.700 | And one of the kind of other things

00:25:59.380 | even tangentially connected to crypto,

00:26:03.340 | but kind of theoretically outside of it that I work on

00:26:06.340 | is this sort of mechanism called quadratic funding.

00:26:09.540 | And the way to think about it is,

00:26:11.140 | and if imagine a point system where if one person

00:26:15.140 | gives coins to one other person,

00:26:21.820 | then it works the same way as money.

00:26:23.580 | But if multiple people give coins to one person

00:26:28.500 | and they do so anonymously,

00:26:30.060 | so it's kind of not in consideration

00:26:32.100 | for a specific service to that person themselves,

00:26:34.820 | then the number of coins received by that person

00:26:39.020 | is greater than just the sum of the number of coins

00:26:42.540 | that have given by those different people.

00:26:45.300 | So the actual formula is you take the square root

00:26:48.060 | of the amount that each person gave,

00:26:49.580 | then you add all the square roots

00:26:50.780 | and then you had a square of the sum.

00:26:52.500 | - Yeah, and then you give that.

00:26:54.100 | And the idea here would basically be that if,

00:26:58.220 | let's say for example,

00:26:59.860 | you just started going off and kind of planting

00:27:03.180 | a lot of trees and there's a bunch of people

00:27:06.020 | that are really happy that you're planting trees

00:27:08.180 | so they go and all kind of throw a coin your way,

00:27:12.020 | then there is like basically the fact

00:27:16.380 | that kind of you get more than the sum,

00:27:17.940 | you get this kind of square of some of these,

00:27:20.380 | of square roots of these tiny amounts

00:27:22.660 | is that this actually kind of compensates

00:27:26.100 | for the tragedy of the commons, right?

00:27:28.260 | There's even this kind of mathematical proof

00:27:29.940 | that it sort of optimally compensates for it.

00:27:32.260 | - What is the tragedy of the common?

00:27:34.020 | - This is just this idea that if there is this situation

00:27:39.020 | where there's some public good

00:27:42.820 | that lots of people benefit from,

00:27:44.340 | then no individual person wants to contribute to it

00:27:47.180 | 'cause if they contribute,

00:27:48.140 | they only get a small part of the benefit

00:27:50.100 | from their contribution,

00:27:52.180 | but they pay the full cost of their contribution.

00:27:54.780 | - In which context is this,

00:27:56.720 | sorry, what is the term, quadratic what?

00:27:58.500 | - Quadratic funding.

00:27:59.580 | - Quadratic funding.

00:28:00.780 | Like what's, in which context is this mechanism useful?

00:28:05.700 | So obviously you said to combat

00:28:08.780 | the tragedy of the commons,

00:28:09.860 | but in which context do you see it as useful

00:28:12.380 | actually practically speaking?

00:28:13.220 | - Yeah, theoretically public goods in general, right?

00:28:15.580 | So like--

00:28:16.940 | - Like services, like what are we talking about?

00:28:19.100 | What's a public good?

00:28:19.940 | - Yeah, so within the Ethereum ecosystem, for example,

00:28:24.180 | we've actually tried using this mechanism.

00:28:26.740 | I wrote a couple of articles about this

00:28:29.540 | and on vitalik.ca where I go through

00:28:31.700 | some of the most recent rounds

00:28:33.020 | and it's been really interesting.

00:28:35.060 | Some of the top ones that people supported,

00:28:37.380 | there were things like just online user interfaces

00:28:42.380 | that make it easier for people to interact with Ethereum.

00:28:48.420 | There was documentation, there were podcasts,

00:28:52.540 | there were software and of clients,

00:28:58.820 | like kind of implementations of the Ethereum protocol

00:29:02.140 | of privacy tools, just like lots of things

00:29:06.020 | that are useful to lots of people.

00:29:09.260 | - When a lot of people are contributing,

00:29:11.140 | like funding a particular entity,

00:29:14.660 | that's really interesting.

00:29:16.260 | Is there something special about the quadratic,

00:29:18.220 | the summing of the square roots and the taken square?

00:29:22.300 | - So another way to think about it is like,

00:29:24.260 | imagine if N people each give a dollar,

00:29:27.140 | then the person gets N squared, right?

00:29:29.780 | And so each individual person's contribution

00:29:33.900 | gets multiplied by N, right?

00:29:35.220 | 'Cause you have N people.

00:29:36.420 | And so that kind of perfectly compensates

00:29:38.740 | for the kind of N to one tragedy of the commons.

00:29:41.980 | - I just wonder if the squared part is somehow fundamental.

00:29:45.340 | - No, it is.

00:29:47.380 | And I'd recommend you go to, on Vitalik.ca,

00:29:51.620 | I have this article called "Quadratic Payments, a Primer"

00:29:54.940 | and highly recommend it.

00:29:57.140 | It's kind of, at least my attempt so far,

00:29:59.420 | kind of explaining the intuition behind this.

00:30:01.340 | - Intuition.

00:30:02.300 | So if we could, can we go to the very basic?

00:30:05.980 | What is the blockchain?

00:30:09.820 | Or perhaps we might even start

00:30:12.620 | at the Byzantine Generals problem

00:30:15.620 | and Byzantine fault tolerance in general

00:30:18.060 | that Bitcoin was taking steps to providing a solution for.

00:30:23.060 | - So the Byzantine Generals problem,

00:30:28.220 | it's this paper that Leslie Lamport published in 1982,

00:30:33.220 | where he has this thought experiments

00:30:35.020 | where if you have two generals

00:30:37.500 | that are kind of camped out on opposite sides of a city

00:30:41.140 | and they're planning when to attack the city,

00:30:44.220 | then the question is,

00:30:47.380 | and if how could those generals coordinate with each other?

00:30:50.180 | And they could send messengers between each other,

00:30:53.380 | but those messengers could get sniped

00:30:56.780 | by the enemy on the road.

00:30:59.300 | Some of those messages could end up being traitors

00:31:01.660 | and if things could end up happening.

00:31:03.860 | And with just two mess generals,

00:31:08.860 | it turns out that there's kind of no solution

00:31:13.220 | in a finite number of rounds

00:31:14.820 | that guarantees that they will be able

00:31:17.660 | to end up coordinate on the same answer.

00:31:20.100 | But then in the case where you have more than two generals,

00:31:22.900 | so then Leslie analyzes cases like,

00:31:25.140 | are the messages kind of just oral messages?

00:31:29.740 | Are the messages kind of signed messages?

00:31:32.300 | So I could give you a signed message

00:31:34.420 | and then you can pass along that signed message

00:31:36.420 | and the third party can still verify

00:31:38.340 | that I originally made that message.

00:31:41.100 | And depending on those different cases,

00:31:43.860 | there's kind of different bounds on,

00:31:45.700 | given how many generals and how many traitors

00:31:50.020 | among those generals,

00:31:51.540 | and under what conditions you actually can agree

00:31:55.340 | when to launch an attack.

00:31:57.100 | So it's actually a big misconception

00:32:00.900 | that the Byzantine generals problem was unsolved.

00:32:03.740 | So Leslie Lamport solved it.

00:32:05.380 | The thing that was unsolved though,

00:32:07.180 | is that all of these solutions assume

00:32:09.700 | that you've already agreed on,

00:32:11.220 | and have a fixed list of who the generals are.

00:32:13.620 | And these generals have to be kind of semi-trusted

00:32:16.820 | to some extent.

00:32:17.660 | They can't just be anonymous people

00:32:18.780 | because if they're anonymous,

00:32:19.900 | then like the enemy could just be 99% of the generals.

00:32:23.940 | So in the 1980s and the 1990s,

00:32:28.940 | kind of the general use case for distributed systems stuff

00:32:34.260 | was more kind of enterprising stuff

00:32:36.860 | where you could kind of assume that,

00:32:40.820 | you know who the nodes are

00:32:43.900 | that are running these kind of computer networks.

00:32:46.220 | So if he wants to have some kind of

00:32:48.100 | decentralized computer network

00:32:49.500 | that pretends to be a single computer

00:32:51.260 | and that you can kind of do kind of operations on,

00:32:54.380 | then it's made out of these kind of 15 specific computers.

00:32:58.500 | And we know kind of who and where they are.

00:33:00.900 | And so we have a good reason to believe

00:33:02.540 | that say at least 11 of them would be fine.

00:33:05.020 | - And it could also be within a single system.

00:33:08.100 | - Exactly.

00:33:09.020 | - Almost a network of devices, sensors,

00:33:11.300 | so on like in airplanes.

00:33:13.180 | And I think like flight systems in general

00:33:15.940 | still use these kinds of ideas.

00:33:17.660 | - Yep, yep.

00:33:18.820 | - So that's the 80s.

00:33:20.260 | - That's the 80s and 90s.

00:33:21.540 | Now the cypherpunks had a different use case in mind,

00:33:24.580 | which is that they wanted to create

00:33:25.980 | a fully decentralized global permissionless currency.

00:33:30.100 | And the problem here is that

00:33:33.260 | they didn't want any authorities

00:33:34.540 | and they didn't even want any kind of

00:33:36.300 | privileged list of people.

00:33:38.260 | And so now the question is,

00:33:39.660 | well, how do you use these techniques to create consensus

00:33:44.660 | when you have no way of kind of measuring identities, right?

00:33:48.100 | You have no way of kind of determining

00:33:50.700 | whether or not some 99% of participants

00:33:54.420 | aren't actually all the same guy.

00:33:56.340 | And so the clever solution that Satoshi had,

00:33:59.180 | this is kind of going back to that presentation

00:34:02.700 | I made at DEF CON a few months ago,

00:34:04.220 | where I said that the thing Satoshi invented

00:34:06.740 | was crypto economics,

00:34:08.660 | is this really neat idea that you can use economic resources

00:34:13.660 | to kind of limit how many identities you can get.

00:34:18.140 | And if there isn't any existing

00:34:22.620 | decentralized digital currency,

00:34:24.380 | then the only way to do this is with proof of work, right?

00:34:26.900 | So with proof of work, the solution is just

00:34:31.580 | you publish a solution to a hard mathematical puzzle

00:34:36.580 | that takes some kind of clearly calculable amount

00:34:40.900 | of computational power to solve, you get an identity.

00:34:44.020 | And then you solve five of those puzzles,

00:34:45.980 | you get five identities.

00:34:47.580 | And then these are the identities

00:34:49.540 | that we run the consensus algorithm between.

00:34:52.140 | - So the proof of work mechanism you just described

00:34:54.820 | is like the fundamental idea proposed

00:34:56.940 | in the white paper that defines Bitcoin.

00:35:01.940 | What's the idea of consensus that we wish to reach?

00:35:07.180 | Why is consensus important here?

00:35:10.060 | What is consensus?

00:35:11.900 | - So the goal here in just simple technical terms

00:35:16.900 | is to basically kind of wire together

00:35:21.180 | a set of a large number of computers

00:35:23.740 | in such a way that they kind of pretend

00:35:25.940 | to the outside world to be a single computer

00:35:28.300 | where that single computer keeps working

00:35:30.580 | even if a large portion of the kind of constituents

00:35:33.340 | the computers that make it up break.

00:35:35.220 | And kind of break in arbitrary ways.

00:35:36.860 | Like they could shut off,

00:35:38.460 | they could try to actively break a system,

00:35:40.940 | they could do lots of mean things.

00:35:42.980 | So the reason why the cypherpunks wanted to do this

00:35:47.980 | is because they wanted to run one particular program

00:35:51.460 | on this virtual computer.

00:35:52.900 | And the one particular program that they wanted to run

00:35:55.100 | is just a currency system, right?

00:35:56.620 | It's a system that just processes a series of transactions.

00:36:00.780 | And for every transaction it verifies

00:36:04.300 | that the sender has enough coins to pay for the transaction

00:36:07.140 | and verifies that the digital signature is correct.

00:36:09.660 | And if the checks pass then it subtracts the coins

00:36:12.620 | from one account and adds the coins

00:36:13.900 | to the other account roughly.

00:36:15.380 | - So first of all, the proof of work idea is kind of,

00:36:20.500 | I mean, at least to me seems pretty fascinating.

00:36:24.580 | - It is.

00:36:25.420 | - I mean, that's a, it's kind of a revolutionary idea.

00:36:27.940 | I mean, is it obvious to come up with

00:36:30.340 | that you can use, you can exchange

00:36:33.100 | basically computational resources for identity?

00:36:37.940 | - It actually has a pretty long history.

00:36:41.780 | It was first proposed in a paper by Cynthia Dwork

00:36:46.460 | and Nixon Naylor in 1994, I believe.

00:36:51.180 | And the original use case was combating email spam.

00:36:54.740 | So the idea is that if you send an email,

00:36:56.420 | you have to send it with a proof of work attached.

00:36:58.340 | And like this makes it reasonably cheap

00:37:00.820 | to send emails to your friends,

00:37:02.060 | but it makes it really expensive

00:37:03.380 | to send spam to a million people.

00:37:05.660 | - Yeah, that's a simple, brilliant idea.

00:37:08.580 | So maybe also taking a step back.

00:37:10.460 | So what is the role of blockchain in this?

00:37:14.460 | What is the blockchain?

00:37:15.980 | - Sure, so the blockchain, my way of thinking about it

00:37:19.660 | is that it is this kind of system

00:37:23.260 | where you have this kind of one virtual computer

00:37:25.580 | created by a bunch of these nodes in the network.

00:37:30.580 | And the reason why the term blockchain is used

00:37:33.980 | is because the data structure that these systems use,

00:37:37.180 | at least so far, is one where they,

00:37:41.460 | different nodes in the network periodically publish blocks.

00:37:46.540 | And a block is a kind of list of transactions

00:37:50.100 | together with a pointer, like a hash of a previous block

00:37:54.580 | that it builds on top of.

00:37:56.140 | And so you have a series of blocks

00:37:59.580 | that nodes in the network create

00:38:02.180 | where each block points to the previous block.

00:38:03.940 | And so you have this chain of them.

00:38:05.820 | - Is a fault tolerance mechanism

00:38:08.820 | built into the idea of blockchain

00:38:10.340 | or is there a lot of possibilities of different ways

00:38:13.100 | to make sure there's no funny stuff going on?

00:38:16.220 | - There are indeed a lot of possibilities.

00:38:19.460 | So in a kind of just simple architecture,

00:38:22.180 | as I just described,

00:38:23.180 | the way the fault tolerance happens is like this, right?

00:38:25.340 | So you have a bunch of nodes

00:38:27.500 | and they're just happily and occasionally creating blocks,

00:38:30.020 | building on top of each other's blocks.

00:38:32.940 | And let's say you have kind of one block,

00:38:35.220 | we'll call it kind of block one.

00:38:36.660 | And then someone else builds another block,

00:38:40.060 | honestly, we'll call it block two.

00:38:41.820 | Then we have an attacker.

00:38:43.900 | And what the attacker tries to do

00:38:45.420 | is the attacker tries to revert block two.

00:38:47.780 | And the way they revert block two

00:38:49.140 | is instead of doing the thing they're supposed to do,

00:38:51.340 | which is build a block on top of block two,

00:38:53.740 | they're gonna build another block on top of block one.

00:38:57.620 | So you have block one, which has two children,

00:38:59.700 | block two and then block two prime.

00:39:01.700 | Now, this might sometimes even happen by random chance

00:39:05.420 | if two nodes in the network

00:39:07.860 | just happen to create blocks at the same time

00:39:09.700 | and they don't hear about each other's things

00:39:11.700 | before they create their own.

00:39:13.100 | But this also could happen because of an attack.

00:39:16.020 | Now, if this happens, you have an attack,

00:39:18.740 | then in the Bitcoin system,

00:39:22.580 | the nodes follow the longest chain.

00:39:24.940 | So if this attack had happened

00:39:29.420 | and when the original chain had more than two blocks on it,

00:39:33.660 | so if it was trying to kind of revert

00:39:35.180 | more than two blocks,

00:39:36.580 | then everyone would just ignore it

00:39:40.460 | and everyone would just keep following the regular chain.

00:39:42.940 | But here, we have block two and we have block two prime

00:39:45.460 | and so the two are kind of even.

00:39:47.300 | And then whatever block the next block is created on top of,

00:39:52.060 | so say block three is now created on top of block two prime,

00:39:55.220 | then everyone agrees that block three is the new head

00:39:59.860 | and block two prime is just kind of forgotten

00:40:04.300 | and then everyone just kind of peacefully builds

00:40:05.980 | on top of block three and the thing continues.

00:40:08.020 | - So how difficult is it to mess with the system?

00:40:12.020 | So how, like if we look at the general problem,

00:40:14.180 | like how many, what fraction of people

00:40:18.420 | who participate in the system have to be bad players

00:40:22.180 | in order to mess with it truly?

00:40:25.460 | Like what's your, is there a good number?

00:40:27.220 | - There is.

00:40:28.820 | Well, depending on kind of what your model

00:40:31.460 | of the participants is

00:40:32.620 | and like what kind of attack we're talking about,

00:40:34.580 | it's anywhere between 23.2 and 50%.

00:40:39.580 | - Of what?

00:40:41.780 | - Of all of the computing power in the network.

00:40:45.100 | - Sorry, so 22 and 50.

00:40:47.660 | - Between 23.2 and 50%.

00:40:49.780 | - And 50% can be compromised.

00:40:53.820 | - So like once your portion of the total computing power

00:40:58.820 | in the network goes above the 23.2 level,

00:41:02.780 | then there's kind of things that you can,

00:41:05.220 | mean things that you can potentially do

00:41:07.540 | and as your percentage of the network kind of keeps going up,

00:41:10.340 | then your ability to do mean things kind of goes higher

00:41:13.540 | and then if you have above 50%,

00:41:15.100 | then you can just break everything.

00:41:16.860 | - So how hard is it to achieve that level?

00:41:19.820 | Like it seems that so far, historically speaking,

00:41:23.940 | it's been exceptionally difficult.

00:41:26.060 | - This is a challenging question.

00:41:28.620 | So the economic cost of acquiring that level of stuff

00:41:32.700 | from scratch is fairly high.

00:41:34.580 | I think it's somewhere in the low billions of dollars.

00:41:38.460 | - And when you say that stuff,

00:41:40.020 | you mean computational resources?

00:41:42.140 | - Yeah, so specifically specialized hardware

00:41:44.980 | and of ASICs that people use to solve these puzzles,

00:41:49.180 | to do the mining these days.

00:41:50.420 | - Small tangent.

00:41:52.340 | So obviously I work a lot in deep learning with GPUs

00:41:55.380 | and ASICs for that application.

00:41:57.500 | And I tangentially kind of hear that so many of these,

00:42:02.060 | sometimes NVIDIA GPUs are sold out

00:42:04.340 | because of this other application.

00:42:07.500 | Like what do, if you can comment,

00:42:09.940 | I don't know if you're familiar or interested in this space,

00:42:12.580 | what kind of ASICs, what kind of hardware

00:42:15.060 | is generally used these days

00:42:16.980 | to do the actual computation for the proof of work?

00:42:20.540 | - Sure, so in the case,

00:42:22.340 | and Bitcoin and Ethereum are a bit different.

00:42:24.220 | So in the case of Bitcoin,

00:42:25.340 | there is an algorithm called SHA-256,

00:42:28.580 | it's just a hash function.

00:42:30.060 | And so the puzzle is just coming up with a number

00:42:32.940 | where the hash of the number is below some threshold.

00:42:35.500 | And so, 'cause the hashes are designed to be random,

00:42:38.580 | you just have to keep on trying different numbers

00:42:40.660 | until one works.

00:42:42.180 | And the ASICs are just like specialized circuits

00:42:45.500 | that contain kind of circuits

00:42:49.100 | for evaluating this hash over and over again.

00:42:51.180 | And you have kind of like millions or billions

00:42:53.060 | of these hash evaluators

00:42:54.380 | and just stacked on top of each other inside of a box.

00:42:57.620 | And you just keep on running the box 24/7.

00:43:00.100 | - And the ASICs,

00:43:00.940 | there's literally specialized hardware designed for this?

00:43:03.260 | - Yes.

00:43:04.100 | - Wow, this is, we live in an amazing world.

00:43:06.300 | Another tangent, and I'll come back to the basics,

00:43:08.980 | but does quantum computing throw a wrench into any of this?

00:43:13.980 | - Very good question.

00:43:15.260 | So quantum computers have two main families of algorithms

00:43:20.260 | that are relevant to cryptography.

00:43:22.180 | One is a Shor's algorithm.

00:43:24.420 | And Shor's algorithm is one that kind of completely breaks

00:43:28.620 | the hardness of some specific kinds

00:43:31.580 | of mathematical problems.

00:43:33.100 | So the one that you've probably heard of

00:43:34.500 | is it makes it very easy to factor numbers.

00:43:37.340 | So like figure out kind of what prime factors are

00:43:40.380 | that kind of, that you need to multiply together

00:43:42.180 | to get some number, even if that number is extremely big.

00:43:45.980 | Shor's algorithm can also be used

00:43:48.060 | to break elliptic curve cryptography.

00:43:50.940 | It can break like any kind of hidden order groups.

00:43:54.940 | So it breaks a lot of kind of cryptographic nice things

00:43:58.300 | that we're used to.

00:43:59.740 | But the good news is that for every kind of major use

00:44:04.460 | of things that Shor's algorithm breaks,

00:44:07.260 | we already know of quantum proof alternatives.

00:44:10.740 | Now we don't use these quantum proof alternatives yet

00:44:13.100 | because in many cases they're five to 10 times

00:44:15.060 | less efficient, but the crypto industry in general

00:44:20.060 | kind of knows that this is coming eventually

00:44:23.060 | and that's kind of ready to take the hit

00:44:25.980 | and switch to that stuff when we have to.

00:44:28.900 | The second algorithm that is relevant to cryptography

00:44:32.380 | is Grover's algorithm.

00:44:33.860 | And in Grover's algorithm might even be a bit more familiar

00:44:38.780 | to AI people that's basically usually described

00:44:42.300 | as solving search problems.

00:44:44.380 | But the idea here is that if you have a problem

00:44:48.500 | of the form, find a number that satisfies some property,

00:44:53.100 | then if with a classical computer,

00:44:55.380 | you need to try out of N times before you find the number,

00:44:59.660 | then with a quantum computer,

00:45:01.580 | you only need to do a square root of N computations.

00:45:04.700 | And Grover's could potentially be used for mining,

00:45:09.700 | but there's two possibilities here.

00:45:12.940 | One is that Grover's could be used for mining

00:45:15.020 | and whoever creates the first working quantum computer

00:45:19.100 | that could do Grover's will just mine way faster

00:45:21.820 | than everyone else.

00:45:22.660 | And we'll see another round of what we saw

00:45:25.540 | when ASICs came out, which is that kind of the new hardware

00:45:28.020 | just kind of dominated the old stuff.

00:45:29.780 | And then eventually it switched to a new equilibrium.

00:45:32.420 | - But by the way, way faster, not exponentially faster.

00:45:36.220 | - Quadratically faster.

00:45:37.220 | - Quadratically faster, which is not sort of,

00:45:40.060 | it's not game changing, I would say.

00:45:43.620 | It's like ASICs, like you said, it would be.

00:45:45.740 | - Exactly.

00:45:46.700 | Yeah, so it would not necessarily break proof of work

00:45:49.700 | as a thing. - That's right, yeah.

00:45:51.020 | The other kind of possible world, right,

00:45:53.500 | is that quantum computers have a lot of overhead.

00:45:56.340 | There's a lot of complexity involved

00:45:58.820 | in maintaining quantum states.

00:46:00.340 | And there's also, as we've been realizing recently,

00:46:04.260 | making quantum computers actually work

00:46:07.180 | requires sort of quantum error correction,

00:46:09.140 | which requires kind of a thousand real qubits

00:46:11.580 | per logical qubit.

00:46:13.100 | And so there's the very real possibility

00:46:15.900 | that the overhead of running a quantum computer

00:46:18.020 | will be higher than the speedup you get with Grovers,

00:46:20.940 | which would be kind of sad, but which would also mean

00:46:23.260 | that the given proof of work would just keep working fine.

00:46:26.540 | - So, beautifully put.

00:46:28.500 | So proof of work is the core idea of Bitcoin.

00:46:32.140 | Is there other core ideas before we kind of

00:46:34.420 | take a step towards the origin story and ideas of Ethereum?

00:46:38.020 | Is there other stuff that were key

00:46:40.180 | to the white paper of Bitcoin?

00:46:42.020 | - There's proof of work,

00:46:42.980 | and then there's just the cryptography,

00:46:44.780 | just kind of public keys and signatures

00:46:47.020 | that are used to verify transactions.

00:46:49.900 | Those two are the big things.

00:46:51.540 | - So then what is the origin story?

00:46:54.580 | Maybe the human side,

00:46:55.980 | but also the technical side of Ethereum.

00:46:58.540 | - Sure.

00:46:59.380 | So I joined the Bitcoin community in 2011,

00:47:03.900 | and I started by just writing.

00:47:06.220 | I first wrote for this sort of online thing

00:47:08.140 | called Bitcoin Weekly.

00:47:09.820 | Then I started writing for Bitcoin Magazine.

00:47:12.820 | - Sorry to interrupt.

00:47:16.340 | I'm just trying to figure out

00:47:17.180 | if this funny kind of story, true or not,

00:47:19.420 | is that you were disillusioned by the downsides

00:47:24.340 | of centralized control from your experience

00:47:26.860 | with, wow, World of Warcraft.

00:47:28.580 | Is this true or you're just being witty?

00:47:31.420 | - I mean, the event is true.

00:47:32.860 | The fact that that's the reason

00:47:34.100 | I do decentralization is witty.

00:47:35.900 | - Maybe just a small tangent.

00:47:38.580 | Have you always had a skepticism of centralized control?

00:47:43.900 | Is that sort of--

00:47:44.740 | - To some degree, yeah.

00:47:47.180 | - Has that feeling evolved over time

00:47:49.100 | or has that just always been a core feeling

00:47:50.860 | that decentralized control is the future of human society?

00:47:55.460 | - It's definitely been something

00:47:56.860 | that felt very attractive to me

00:47:58.620 | ever since I could have learned

00:47:59.900 | that such a thing is possible.

00:48:01.500 | - It's possible even technically.

00:48:03.180 | So great.

00:48:04.020 | So you joined the Bitcoin community in 2011,

00:48:06.620 | you said you began writing.

00:48:08.540 | So what's next?

00:48:10.460 | - Started writing,

00:48:12.100 | moved from high school to university

00:48:14.140 | halfway in between that,

00:48:15.900 | spent a year in university.

00:48:18.220 | Then at the end of that year,

00:48:21.820 | I dropped out to do Bitcoin things full-time.

00:48:26.380 | And this was a combination

00:48:27.820 | of continuing to write Bitcoin magazine,

00:48:29.700 | but also increasingly work on software projects.

00:48:33.420 | And I traveled around the world for about six months

00:48:36.500 | and just going to different Bitcoin communities.

00:48:38.820 | Like I went to first in New Hampshire,

00:48:41.620 | then Spain, other European places,

00:48:43.820 | Israel, and then San Francisco.

00:48:46.100 | And along the way, I've met a lot of other people

00:48:48.380 | that are working on different Bitcoin projects.

00:48:50.740 | And when I was in Israel,

00:48:52.580 | there were some kind of very smart teams there

00:48:54.900 | that were working on ideas

00:48:57.220 | that people were starting to kind of call Bitcoin 2.0.

00:49:00.780 | So one of these was covered coins,

00:49:02.420 | which is basically saying that,

00:49:04.220 | "Hey, let's not just use the blockchain for Bitcoin,

00:49:07.580 | but let's also issue other kinds of assets on it."

00:49:10.820 | And then there was a protocol called MasterCoin

00:49:12.940 | that supported issuing assets,

00:49:14.980 | but also supported many other things,

00:49:17.020 | like financial contracts, like domain name registration,

00:49:21.020 | and a lot of different things together.

00:49:23.380 | And I spent some time working with these teams,

00:49:28.380 | and I quickly kind of realized

00:49:32.140 | that this MasterCoin protocol could be improved

00:49:35.260 | by kind of generalizing it more, right?

00:49:37.220 | So the analogy I use is that the MasterCoin protocol

00:49:40.140 | was like the Swiss Army knife.

00:49:41.420 | You have 25 different transaction types

00:49:43.980 | for 25 different applications.

00:49:46.260 | But what I realized is that you could replace

00:49:50.060 | a bunch of them with things that are more general purpose.

00:49:53.300 | So one of them was that you could replace

00:49:56.380 | like three transaction types

00:49:58.020 | for three types of financial contracts

00:50:00.060 | with a generic transaction type for a financial contract

00:50:03.860 | that just lets you specify a mathematical formula

00:50:06.180 | for kind of how much money each side gets.

00:50:09.700 | - By the way, just a small pause.

00:50:11.220 | What's, you say financial contract,

00:50:13.620 | just the terminology, what is a contract?

00:50:16.460 | What's a financial contract?

00:50:18.060 | - So this is just generally an agreement

00:50:21.180 | where kind of either one or two parties

00:50:24.700 | kind of put collateral kind of in,

00:50:27.140 | and then depending on kind of certain conditions,

00:50:31.980 | like this could involve prices of assets,

00:50:33.940 | this could involve the actions of the two parties,

00:50:37.100 | it could involve other things.

00:50:39.100 | They kind of get different amounts of assets out

00:50:43.140 | that just depend on things that happened.

00:50:45.140 | - So a contract is really,

00:50:46.620 | a financial contract is the core,

00:50:49.700 | it's the core interactive element of a financial system.

00:50:53.820 | - Yeah, there's many different kinds of financial contracts.

00:50:57.260 | Like there's things like options

00:50:58.700 | where you kind of give someone the right to buy a thing

00:51:01.660 | that you have for some specific price

00:51:03.180 | for some period of time.

00:51:04.300 | There's contracts for difference

00:51:08.100 | where you basically are kind of making a bet

00:51:11.260 | that says like for every dollar this thing goes up,

00:51:13.980 | I'll give you $7 or for every dollar the thing goes down,

00:51:16.500 | you give me $7 or something like that.

00:51:18.700 | - But the main idea that these contracts

00:51:21.740 | have to be enforced and trusted.

00:51:23.940 | - Yes, exactly.

00:51:25.500 | - You have to trust that they will work out

00:51:27.580 | in a system where nobody can be trusted.

00:51:29.620 | - Yes.

00:51:30.460 | - This is such a beautiful, complicated system.

00:51:34.060 | Okay, so you were seeking to kind of generalize

00:51:37.940 | this basic framework of contracts.

00:51:40.900 | So what does that entail?

00:51:44.580 | So what technically are the steps to creating Ethereum?

00:51:48.620 | - Sure, so I guess just to kind of continue a bit

00:51:51.940 | with this MasterCoin story.

00:51:53.620 | So started by kind of giving ideas

00:51:56.180 | for how to generalize the thing.

00:51:57.900 | And eventually this turned into a much more

00:52:01.780 | kind of fully fledged proposal that just says,

00:52:03.740 | hey, how about you scrap all your futures

00:52:05.420 | and instead you just put in this programming language?

00:52:09.340 | And I gave this idea to them and their response

00:52:12.340 | was something like, hey, this is great,

00:52:14.740 | but this seems complicated and it seems like something

00:52:17.620 | that we're not gonna be able to put onto our roadmap

00:52:19.820 | for a while.

00:52:20.860 | And my response to this was like, wait,

00:52:22.580 | do you not realize how revolutionary this is?

00:52:24.740 | Well, I'll just go do it myself.

00:52:26.500 | And then I--

00:52:28.180 | - What was the name of the programming language?

00:52:29.900 | - I just called it ultimate scripting.

00:52:32.500 | - Great.

00:52:33.620 | - So then I kind of went through a couple more rounds

00:52:38.540 | of iteration and then the idea for Ethereum itself

00:52:42.420 | started to form.

00:52:44.140 | And the idea here is that you just have a blockchain

00:52:49.140 | where the core unit of the thing is what we call contracts.

00:52:53.020 | It's these kind of accounts that can hold assets

00:52:56.900 | and they have their own internal memory,

00:52:59.420 | but that are controlled by a piece of code.

00:53:02.580 | And so if I send some ether to a contract,

00:53:05.900 | the only thing that can determine where that ether,

00:53:10.180 | the currency inside Ethereum goes after that

00:53:12.660 | is the code of that contract itself.

00:53:16.900 | And so basically you're kind of sending assets

00:53:20.580 | to computer programs becomes this kind of paradigm

00:53:24.740 | for creating these sort of self-executing agreements.

00:53:28.860 | - Self-executing, it's so cool that code

00:53:31.460 | is sort of part of this contract.

00:53:33.180 | So that's what's meant by smart contracts.

00:53:36.540 | - Yeah.

00:53:37.380 | - So how hard was it to build this kind of thing?

00:53:40.540 | - Harder than expected.

00:53:42.380 | And originally I actually thought that this would be a thing

00:53:45.060 | that I would kind of casually work on

00:53:47.140 | for a couple of months, publish,

00:53:48.500 | and then go back to university.

00:53:50.060 | Then I released it and a bunch of people,

00:53:55.940 | or I released a white paper.

00:53:57.140 | - The white paper, the idea is there.

00:53:58.540 | - The idea, the white paper.

00:54:00.260 | A whole bunch of people came in offering to help,

00:54:02.620 | a huge number of people have expressed interest.

00:54:05.660 | And this was something I was totally not expecting.

00:54:08.260 | And then I realized that this would be something

00:54:12.580 | that's kind of much bigger than I had ever thought

00:54:16.100 | that it would be.

00:54:17.100 | And then we started on this kind of much longer

00:54:21.140 | development slog of making something that lives up

00:54:25.500 | to this much higher level of expectations.

00:54:27.980 | - What are some of the,

00:54:30.300 | is it fundamentally like software engineering challenges?

00:54:33.500 | - It was.

00:54:34.340 | - Is there social, okay, so there's--

00:54:35.380 | - And social.

00:54:36.220 | - So what are the biggest interesting challenges

00:54:39.820 | that you've learned about human civilization

00:54:43.100 | and software engineering through this process?

00:54:46.020 | - So I guess one of the challenges for me

00:54:50.820 | is that I'm one of the kind of apparently unusual geeks

00:54:54.860 | who has kind of never treated with anything

00:54:56.940 | but kindness in school.

00:54:58.980 | And so when I got into crypto,

00:55:03.660 | I kind of expected everyone would just kind of be

00:55:06.340 | the same kind of altruistic and nice in that same way.

00:55:09.980 | But the algorithm that I used for finding co-founders

00:55:14.980 | for this thing was not very good.

00:55:17.180 | It was sort of literally what computer scientists

00:55:19.060 | called the greedy algorithm.

00:55:20.420 | It's sort of the first 15 people who replied back

00:55:23.220 | offering to help kind of are the co-founders.

00:55:25.260 | - So you mean like literally the people

00:55:28.140 | that will form to be the co-founders of the community,

00:55:32.780 | the algorithm, I like how you call it the algorithm.

00:55:35.060 | - Yeah, and so what happened was that these,

00:55:39.860 | like especially as the project got really big,

00:55:44.860 | like there started to be a lot of this kind of infighting

00:55:47.380 | and there were a lot of,

00:55:48.540 | like I wanted the thing to be a nonprofit

00:55:51.140 | and some of them wanted to be a for-profit.

00:55:53.380 | And then there started to be people

00:55:56.340 | who were just kind of totally unable to work with each other.

00:55:58.780 | There were people that were kind of trying

00:56:02.860 | to get an advantage for themselves

00:56:04.340 | in a lot of different ways.

00:56:05.900 | And this just about six months later

00:56:10.020 | led to this big governance crisis.

00:56:11.980 | And then we kind of reshuffled leadership a bit.

00:56:14.900 | And then the project kept on going.

00:56:18.100 | Then nine months later, there was another governance crisis.

00:56:20.700 | And then there was a third governance crisis.

00:56:22.540 | And so is there a way to,

00:56:25.460 | if you're looking at the human side of things,

00:56:27.380 | is there a way to optimize this aspect

00:56:30.700 | of the cryptocurrency world?

00:56:32.220 | It seems that there is, from my perspective,

00:56:35.300 | there's a lot of different characters

00:56:37.300 | and personalities and egos.

00:56:39.500 | And like you said, I don't know if,

00:56:42.260 | I also like to think that most of the world,

00:56:47.260 | most of the people in the world are well-intentioned,

00:56:50.180 | but the way those intentions are realized

00:56:52.140 | may perhaps come off as negative.

00:56:56.340 | Like what, is there a hopeful message here

00:57:01.100 | about creating a governance structure for cryptocurrency

00:57:04.340 | that where everyone gets along?

00:57:06.860 | - I mean, after about four rounds of reshuffle,

00:57:09.300 | like I think we've actually come up with something

00:57:11.700 | that seems to be pretty stable and happy.

00:57:14.020 | I think, I mean, I definitely do think that

00:57:21.460 | most people are well-intentioned.

00:57:23.540 | I just think that one of the reasons

00:57:26.100 | why I like decentralization is just because there's like

00:57:29.780 | this thing about power where power attracts people

00:57:32.420 | with egos and so that just allows us

00:57:34.540 | a very small percentage of people

00:57:35.940 | to just ruin so many things.

00:57:37.820 | - You think ego has a, you think ego has a use?

00:57:42.260 | Like is ego always bad?

00:57:44.780 | It seems like some of the most--

00:57:45.620 | - It sometimes does.

00:57:46.580 | But then the Ethereum research team,

00:57:47.940 | I feel like we've found also kind of a lot of very good

00:57:52.940 | people that are just primarily just interested

00:57:58.260 | in things for the technology

00:58:00.100 | and things seem to just generally be going quite well.

00:58:05.100 | - Yeah, when the focus and the passion is in the tech.

00:58:09.740 | So that's the human side of things.

00:58:12.420 | The technology side, like what have you learned?

00:58:14.900 | What have been the biggest challenges

00:58:16.580 | bringing Ethereum to life on the technology side?

00:58:21.380 | - So I think first of all, just,

00:58:23.580 | you know, there's like the first law of software development

00:58:28.580 | which is that when someone gives you a timetable,

00:58:30.940 | I'm gonna switch the unit of time

00:58:32.100 | to the next largest unit of time and add one.

00:58:34.500 | And we basically fell victim to that.

00:58:36.940 | And so instead of taking like three months,

00:58:43.740 | it ended up taking like 20 months to launch the thing.

00:58:46.700 | And that was just, I think,

00:58:49.740 | underestimating the sheer technical complexity

00:58:51.980 | of the thing.

00:58:52.820 | There are research challenges.

00:58:55.140 | Like, so for example,

00:58:57.300 | one of the things that we've been saying

00:58:58.860 | from the start that we would do,

00:59:00.140 | one is a switch from a proof of work to a proof of stake.

00:59:04.100 | More proof of stake is this alternative consensus mechanism

00:59:07.980 | where instead of having to waste a lot of computing power

00:59:11.620 | on solving these mathematical puzzles

00:59:13.180 | that don't mean anything,

00:59:14.180 | you kind of prove that you have access to coins

00:59:17.900 | inside of the system.

00:59:18.900 | And this kind of gives you some level of participation

00:59:22.260 | in the consensus.

00:59:23.540 | - Can you maybe elaborate on that a little bit?

00:59:25.260 | I understand the idea of proof of work.

00:59:27.220 | I know that a lot of people say

00:59:29.740 | that the idea of proof of stake is really appealing.

00:59:31.980 | Can you maybe link on it a little longer,

00:59:33.700 | explain what it is?

00:59:35.380 | - Sure.

00:59:36.460 | So basically the idea is like,

00:59:40.260 | if I kind of lock up a hundred coins,

00:59:43.220 | then I turn that into a kind of quote virtual miner

00:59:47.180 | and the system itself kind of automatically

00:59:51.500 | and randomly assigns that in a virtual miner

00:59:54.540 | the right to create blocks at particular intervals.

00:59:58.220 | And then if someone else has 200 coins

01:00:00.340 | and they lock on the lock, there's 200 coins,

01:00:02.540 | then they get a kind of twice as big virtual miner,

01:00:05.620 | they'll be able to create blocks twice as often.

01:00:08.380 | So it tries to do similar things to proof of work,

01:00:13.100 | except instead of the thing

01:00:15.220 | and of rate limiting your participation

01:00:17.860 | being your ability to crank out solutions

01:00:21.380 | to kind of hash challenges,

01:00:23.420 | the thing that really limits your participation

01:00:25.180 | is kind of how much coins

01:00:26.220 | you're kind of locking into this mechanism.

01:00:28.980 | - Okay, so interesting.

01:00:29.820 | So that limited participation

01:00:31.820 | doesn't require you to run a lot of compute.

01:00:37.020 | Does that mean that the richer you are,

01:00:40.780 | so rich people are more,

01:00:46.540 | like their identity is more stable,

01:00:50.540 | verifiable or whatever the right terminology is.

01:00:55.220 | - Right, and this is definitely a common critique.

01:00:57.500 | I think my usual answer to this is that

01:01:00.620 | like proof of work is even more of that kind of system.

01:01:03.060 | - Yes, exactly.

01:01:03.900 | I didn't mean it in that statement as a criticism.

01:01:06.220 | I think you're exactly right.

01:01:07.300 | That's equivalent.

01:01:08.260 | The proof of work is the same kind of thing,

01:01:11.020 | but in the proof of work,

01:01:12.020 | you have to also use physical resources.

01:01:15.740 | - Yes, and burn computers and burn trees

01:01:18.060 | and all of that stuff.

01:01:19.340 | - Is there a way to mess with the system

01:01:22.220 | over the proof of stake?

01:01:25.100 | - There is, but you will once again need to have

01:01:27.980 | a very large portion of all the coins

01:01:29.660 | that are locked in the system to do anything bad.

01:01:31.740 | - Got it.

01:01:32.580 | Yeah, and just that maybe take a small change.

01:01:35.820 | One of the criticisms of cryptocurrency is the fact

01:01:39.780 | that I guess for the proof of work mechanism,

01:01:42.140 | you have to use so much energy in the world.

01:01:44.140 | - Yes.

01:01:45.340 | - Is one of the motivations of proof of stake

01:01:48.020 | is to move away from this?

01:01:49.780 | - Definitely.

01:01:50.620 | - Like what's your sense of the,

01:01:52.500 | maybe I'm just under-informed.

01:01:54.300 | Is there like legitimately environmental impact from this?

01:01:58.220 | - Yeah, so the latest thing was that Bitcoin consumed

01:02:02.700 | as much energy as the country of Austria

01:02:04.700 | or something like that.

01:02:06.380 | Yeah, and then Ethereum is like right now,

01:02:08.660 | only like half an order of magnitude smaller than Bitcoin.

01:02:12.180 | - I've heard you talk about Ethereum 2.0.

01:02:14.820 | So what's the dream of Ethereum 2.0?

01:02:18.340 | What's the status of proof of stake as a mechanism

01:02:22.300 | that Ethereum moves towards?

01:02:24.340 | And also, how do you move to a different mechanism

01:02:28.060 | of consensus within a cryptocurrency?

01:02:31.740 | - So Ethereum 2.0 is a collection of major upgrades

01:02:35.900 | that we've wanted to do to Ethereum for quite some time.

01:02:38.700 | The two big ones, one is a proof of stake

01:02:41.900 | and the other is what we call sharding.

01:02:44.300 | Sharding solves another problem with blockchains,

01:02:46.940 | which is a scalability.

01:02:48.780 | And what sharding does is it basically says,

01:02:52.020 | instead of every participant in the network

01:02:54.620 | having to personally download

01:02:56.020 | and verify every transaction,

01:02:57.780 | every participant in the network only downloads

01:03:00.060 | and verifies a small portion of transactions.

01:03:02.660 | And then you kind of randomly distribute

01:03:04.940 | who gets how much work.

01:03:06.980 | And because of how the distribution is random,

01:03:09.940 | it still has the property that you need a large portion

01:03:13.380 | of the entire network to corrupt what's going on

01:03:16.020 | inside of any shard,

01:03:17.380 | but the system is still very redundant and very secure.

01:03:21.940 | - That's brilliant.

01:03:23.500 | How hard is that to implement

01:03:25.700 | and how hard is proof of stake to implement?

01:03:28.580 | Like on the technical level, software level.

01:03:32.020 | - Proof of stake and sharding are both challenging.

01:03:34.500 | I'd say sharding is a bit more challenging.

01:03:37.220 | The reason is that proof of stake is kind of just a change

01:03:40.020 | to how the consensus layer works.

01:03:42.540 | Sharding does both that,

01:03:44.580 | but it's also a change to the networking layer.

01:03:47.260 | The reason is that sharding is kind of pointless

01:03:49.220 | if at the networking layer, you still do what you do today,

01:03:52.060 | which is you kind of gossip everything,

01:03:53.700 | which means that if someone publishes something,

01:03:56.260 | every other node and the client hears it

01:03:58.860 | from on the networking layer.

01:04:00.980 | And so instead we have to have kind of sub networks

01:04:03.380 | and the ability to quickly switch between sub networks

01:04:05.540 | and other sub networks, talk to each other.

01:04:07.740 | And this is all doable,

01:04:08.780 | but it's a more complex architecture

01:04:11.820 | and it's definitely the sort of thing

01:04:12.660 | that has not yet been done in cryptocurrency.

01:04:15.100 | - So most of the networking layer in cryptocurrency

01:04:19.380 | is you're shouting, you're like broadcasting messages.

01:04:22.620 | And this is more like ad hoc networks.

01:04:25.180 | - Yeah, you're shouting within smaller groups.

01:04:27.860 | - Smaller group, but do you have like a bunch of sub net?

01:04:30.100 | - Exactly.

01:04:31.300 | - Then you have to switch between,

01:04:32.660 | oh man, I'd love to see that.

01:04:34.580 | So it's a beautiful idea.

01:04:37.180 | So from a graph theoretic perspective,

01:04:39.900 | but just the software, like who's responsible?

01:04:43.140 | Is the Ethereum project, like the people involved,

01:04:46.660 | would they be implementing?

01:04:47.700 | Like what's the actual,

01:04:49.060 | you know, this is like legit software engineering.

01:04:53.780 | Who, like how does that work?

01:04:55.660 | How do people collaborate, build that kind of project?

01:04:58.340 | Is this like almost,

01:04:59.380 | like is there a software engineering lead?

01:05:03.700 | Is there, is it a legit,

01:05:06.180 | almost like large scale open source project?

01:05:08.500 | - There is, yeah.

01:05:09.420 | So we have someone named Danny Ryan on our team

01:05:12.980 | who has just been brilliant and great all around.

01:05:15.500 | And he is a kind of de facto

01:05:19.020 | kind of development coordinator, I guess.

01:05:22.180 | It's like, you have to invent job titles for this stuff.

01:05:25.020 | The reason is that,

01:05:27.020 | like we also have this unique kind of organizational

01:05:29.700 | structure where the Ethereum Foundation itself

01:05:31.860 | kind of does research in-house,

01:05:33.140 | but then the actual implementation is done

01:05:35.860 | by independent teams that are separate companies

01:05:38.380 | and they're located all around the world

01:05:40.900 | and like fun places like Australia.

01:05:43.860 | And so, you know, you kind of just need a bunch of

01:05:49.740 | kind of almost nonstop cat herding to just

01:05:52.420 | keep getting these people to kind of talk to each other

01:05:55.740 | and kind of implement the spec,

01:05:57.540 | make sure that everyone agrees on kind of what's going on

01:06:00.580 | and kind of how to interpret different things.

01:06:03.060 | - So how far into the future are we

01:06:04.780 | from these two mechanisms in Ethereum 2.0?

01:06:07.380 | Like what's your sense of the timeline,

01:06:10.460 | keeping in mind the previous comment you made

01:06:13.420 | about the sort of general curse of software projects?

01:06:19.420 | - So Ethereum 2.0 is split into three phases.

01:06:23.460 | So phase zero just creates a proof of stake network

01:06:26.460 | and it's actually separate from kind of proof of,

01:06:29.740 | the proof of work network at the beginning,

01:06:31.660 | just to kind of give it time to grow and improve itself.

01:06:35.100 | - Do people get to choose, sorry to interrupt,

01:06:36.860 | do people get to choose, I guess, which-

01:06:38.420 | - Yes, they get to choose to move over if they want to.

01:06:42.020 | Then phase one adds sharding,

01:06:44.020 | but it only adds sharding of kind of data storage

01:06:46.660 | and not sharding of computation.

01:06:48.980 | And then after that, there is kind of the merger phase,

01:06:51.940 | which is where the accounts, kind of smart contracts,

01:06:56.940 | like all of the kind of activity

01:06:59.060 | on the existing ETH1 system,

01:07:01.380 | just kind of gets cut and pasted into ETH2.

01:07:03.900 | And then the proof of work chain gets forgotten.

01:07:05.660 | And then all the things that were living there before

01:07:09.540 | just kind of continue living

01:07:10.980 | inside of the proof of stake system.

01:07:13.300 | So for timelines, phase zero has been,

01:07:18.300 | kind of almost fully implemented.

01:07:23.300 | And now it's just a matter of a whole bunch

01:07:26.940 | of security auditing and testing.

01:07:28.780 | My own experience is that right now,

01:07:31.780 | it feels like we're at about a phase

01:07:34.460 | comparable to when we were doing

01:07:38.300 | the original Ethereum launch,

01:07:39.900 | when we were maybe about four months away from launch.

01:07:42.620 | - So that's just a hunch, then you can-

01:07:44.580 | - That's just a hunch, yeah.

01:07:46.620 | - So how, you know, it took like over a decade

01:07:50.020 | for people to move from Python 2 to Python 3.

01:07:53.580 | How do you see the move from like this phase zero

01:07:58.580 | for different consensus mechanism?

01:08:01.180 | Do you see there being a drastic phase shift

01:08:05.060 | in people just kind of jumping to this better mechanism?

01:08:08.580 | - So in phase zero, I don't expect too many people

01:08:11.140 | to do much because in phase zero and phase one,

01:08:14.940 | the new chain, they get it deliberately

01:08:16.780 | and it doesn't have too much functionality turned on.

01:08:19.340 | It's there just like,

01:08:20.540 | if you want to be a proof of stake validator,

01:08:22.420 | you can get things started.

01:08:23.380 | If you want to store data

01:08:24.900 | for other blockchain applications, you can get started.

01:08:28.340 | But existing applications will largely

01:08:30.540 | keep living on each one.

01:08:32.420 | And then when the merger happens,

01:08:35.340 | then the merger is a operation that happens all at once.

01:08:40.100 | So that's kind of one of the benefits

01:08:41.900 | of a consensus system, that like on the one hand,

01:08:44.100 | you have to coordinate the upgrade,

01:08:45.580 | but on the other hand, the upgrade can be coordinated.

01:08:49.100 | - So what's CasperFFG by the way?

01:08:52.220 | - CasperFFG is the consensus algorithm

01:08:55.220 | that we are using for the proof of stake.

01:08:57.820 | - Is there something interesting,

01:08:59.340 | specific about CasperFFG,

01:09:01.420 | like some beautiful aspect of it that's-

01:09:03.820 | - There is.

01:09:04.660 | So CasperFFG combines together kind of two different schools

01:09:09.660 | of a consensus algorithm design.

01:09:12.740 | So the general two different schools

01:09:14.500 | of the design are, right?

01:09:16.740 | One is a 50% fault tolerant,

01:09:20.860 | but dependent on network synchrony.

01:09:23.220 | So 50% fault tolerant,

01:09:25.380 | but it didn't tolerate up to 50% of faults, but not more,

01:09:28.820 | but it depends on an assumption

01:09:31.300 | that all of the nodes can talk to each other

01:09:34.020 | within some limited period of time.

01:09:36.580 | Like if I send the message,

01:09:37.740 | you'll receive it within a few seconds.

01:09:40.500 | And the second school is 33% fault tolerant,

01:09:44.540 | but safe under asynchrony,

01:09:46.280 | which means that if we agree on something,

01:09:50.380 | then that thing is finalized.

01:09:51.940 | And even if the network goes horribly wonky

01:09:54.300 | the second after that thing is finalized,

01:09:55.940 | there's no way to revert that thing.

01:09:57.740 | And-

01:09:59.820 | - That's fascinating how you would make that happen.

01:10:03.100 | - It's definitely quite clever.

01:10:06.300 | I'd recommend the CasperFFG paper.

01:10:08.900 | If you just search like archive,

01:10:10.460 | as in like ARXIV and CasperFFG, it's right there.

01:10:13.140 | - That's an archive, the paper's an archive.

01:10:15.340 | - Yeah, yeah.

01:10:16.180 | - Who are the authors?

01:10:17.340 | - Myself and Virgil Griffith.

01:10:19.900 | - That's awesome.

01:10:20.740 | I take a small tangent.

01:10:23.460 | This idea of just putting out white papers and papers

01:10:26.180 | and putting them on archive and just putting them publicly,

01:10:28.940 | is that at the core?

01:10:32.060 | Is that a necessary component of cryptocurrency?

01:10:34.460 | Is that the tradition started

01:10:35.980 | with Satoshi Nakamoto?

01:10:39.780 | What do you make of it?

01:10:40.620 | Like, what do you make of the future

01:10:41.540 | of that kind of sharing of ideas?

01:10:42.900 | - I guess so, yeah.

01:10:44.060 | And it's definitely something that's kind of mandatory

01:10:48.980 | for crypto because like crypto is all about making systems

01:10:53.700 | where you don't have to trust the operators

01:10:56.220 | to trust that the thing works.

01:10:57.660 | And so if anything behind our system works

01:11:01.260 | is closed sourced, then that kind of kills the point.

01:11:04.660 | And so there is the kind of a sense

01:11:07.780 | in which the fundamental properties of the category

01:11:11.140 | of the thing we're trying to build

01:11:12.500 | just kind of forces openness,

01:11:14.820 | but also openness just has proven

01:11:17.060 | to be a really great way to collaborate.

01:11:19.060 | And then there's actually a lot of innovation

01:11:21.740 | and academic collaboration that's just kind of happened

01:11:24.780 | ad hoc in the crypto space the last few years.

01:11:27.860 | So like, for example, we have this forum

01:11:30.860 | called ETH Research, that's like E-T-H-R-E-S-E-A-R

01:11:34.620 | and then dot C-H.

01:11:36.500 | And there we publish kind of just ideas

01:11:40.900 | in a form that's kind of half formal,

01:11:42.940 | like it's halfway in between,

01:11:45.460 | like it's a kind of a text write up

01:11:47.740 | and you can have math in it,

01:11:49.220 | but it's often in a much shorter than a paper.

01:11:52.620 | And it turns out that the great majority of new ideas,

01:11:57.620 | like they're just kind of fairly small nuggets

01:11:59.420 | that you can explain in like five to 10 lines

01:12:01.620 | and they don't really--

01:12:02.820 | - They don't need the whole formality of a paper.

01:12:04.940 | - Exactly, they don't require the kind of like 10 pages

01:12:07.540 | of any filler and so--

01:12:09.100 | - Introduction and conclusion is not needed.

01:12:11.740 | - Yeah, and so instead you just kind of publish the idea

01:12:15.220 | and then people can go comment on it.

01:12:17.340 | - That's brilliant.

01:12:18.180 | - Yeah, this has been great for us.

01:12:19.980 | - I think I interrupted you.

01:12:21.820 | Was there something else on Casper FFG?

01:12:23.900 | - Yeah, so just Casper FFG is just kind of combines together

01:12:27.580 | these two schools.

01:12:29.100 | And so basically it creates this system

01:12:32.820 | where if you have more than 50% that are honest

01:12:37.820 | then and you have a network synchrony,

01:12:42.660 | then the thing kind of goes as a chain.

01:12:45.100 | But then if network synchrony fails,

01:12:46.660 | then kind of the last few blocks in the chain

01:12:48.700 | might kind of get replaced.

01:12:50.980 | But anything that was finalized

01:12:53.260 | by this kind of more asynchronous process

01:12:58.260 | can't be reverted.

01:12:59.380 | And so you essentially get a kind of best of both worlds

01:13:02.420 | between those two models.

01:13:04.300 | - Okay, so I know what I'm doing to them.

01:13:06.100 | I'm gonna be reading the Casper FFG paper.

01:13:08.660 | Apologize for the romanticized question,

01:13:11.460 | but what to you are some or the most beautiful idea

01:13:16.460 | in the world of Ethereum?

01:13:19.020 | Just something surprising, something beautiful,

01:13:22.140 | something powerful.

01:13:23.540 | - Yeah, I mean, I think the fact that money

01:13:26.260 | can just emerge out of a database

01:13:28.020 | if enough people believe in it,

01:13:29.260 | I think is definitely one of those things that's up there.

01:13:32.300 | I think one of the things that I really love about Ethereum

01:13:37.260 | is also this concept of composability.

01:13:39.620 | So this is the idea that if I build an application

01:13:42.540 | on top of Ethereum,

01:13:43.860 | then you can build an application

01:13:46.460 | that talks to my application

01:13:48.220 | and you don't even need my permission.

01:13:50.100 | You don't even need to talk to me, right?

01:13:52.580 | So one really fun example of this is there was this

01:13:55.220 | kind of game on Ethereum called CryptoKitties

01:13:58.180 | that just involved kind of breeding digital cats.

01:14:00.900 | And someone else created a game called CryptoDragons,

01:14:05.260 | where the way you play CryptoDragons is you have a dragon

01:14:08.100 | and you have to feed it CryptoKitties.

01:14:10.100 | And they just created the whole thing

01:14:15.620 | just like as an Ethereum contract

01:14:18.220 | that you would send these tokens

01:14:21.340 | that are defined by this other Ethereum contract.

01:14:24.100 | And for the interoperability to happen,

01:14:26.060 | like the projects don't really need to,

01:14:29.220 | like the teams don't really need to talk to each other.

01:14:31.180 | You just kind of interface with the existing program.

01:14:35.180 | - So it's arbitrarily composable in this kind of way.

01:14:37.700 | So you have different groups that could be working.

01:14:40.420 | So you could see it scaling to just

01:14:43.060 | outside of dragons and kitties,

01:14:44.860 | it could be, you could build like entire ecosystems

01:14:47.660 | of software in this kind of way.

01:14:48.860 | - Yeah, I mean, especially in the decentralized finance

01:14:53.900 | space that's been popping up in the last two years,

01:14:56.220 | there has been a huge amount of really interesting things

01:14:59.460 | happen as a result of this.

01:15:00.940 | - Is it a particular kind of like financial applications

01:15:04.260 | kind of thing?

01:15:05.380 | - Yeah, I mean, there's like stable coins.

01:15:07.740 | So this is a kind of tokens retain value equal to $1,

01:15:12.740 | but they're kind of backed by a cryptocurrency.

01:15:17.500 | Then there's decentralized exchanges.

01:15:20.100 | So, well, as far as decentralized exchanges goes,

01:15:23.740 | there's this really interesting construction

01:15:27.740 | that has existed for about one and a half years now

01:15:31.140 | called Uniswap.

01:15:32.580 | So what Uniswap is, it's a smart contract

01:15:35.340 | that holds the balances of two tokens.

01:15:38.860 | We'll call them token A and token B.

01:15:41.020 | And it maintains an invariance that the balance of token A

01:15:44.100 | multiplied by the balance of token B

01:15:45.860 | has to equal the same value.

01:15:47.660 | And so the way that you trade against the thing

01:15:49.900 | is basically like you have this kind of curve,

01:15:52.860 | you know, like X times Y equals K.

01:15:54.940 | And before you trade, it's at some points on the curve.

01:15:58.460 | After you trade, you just like pick some different,

01:16:00.220 | any other points on the curve.

01:16:02.300 | And then whatever the delta X is,

01:16:03.860 | that's the amount of A tokens you provide.

01:16:05.700 | Whatever the delta Y is,

01:16:06.780 | that's the amount of B tokens you get or vice versa.

01:16:09.820 | And that's just, and then kind of the slope

01:16:12.460 | at the current points on the curve kind of is the price.

01:16:16.180 | And so that just is the whole thing.

01:16:19.820 | And that just allows you to have this exchange for tokens.

01:16:24.820 | And even if there's very few participants

01:16:29.260 | and the whole thing is just like so simple

01:16:32.620 | and it's just very easy to set up,

01:16:34.500 | very easy to participate in.

01:16:36.220 | And it just provides so much value to people.

01:16:40.140 | - And the fundamental,

01:16:44.860 | the distributed application infrastructure

01:16:47.900 | allows that somehow.

01:16:49.500 | - Yes.

01:16:50.340 | So this is a smart contract meeting.

01:16:52.340 | This is all a computer program

01:16:53.700 | that's just running on Ethereum.

01:16:55.460 | - Smart contracts too are just fascinating.

01:16:58.100 | - They are.

01:16:59.540 | - Okay.

01:17:00.380 | Do you think cryptocurrency may become the main currency

01:17:04.060 | in the world one day?

01:17:05.380 | So where do you think we're headed in terms of

01:17:08.500 | the role of currency,

01:17:09.820 | the structure type of currency in the world?

01:17:12.820 | - I definitely expect fiat currencies to continue to exist

01:17:18.540 | and continue to be strong.

01:17:19.580 | And I definitely expect kind of fiat currencies

01:17:21.780 | to also digitize in their own way

01:17:23.460 | over the next couple of decades.

01:17:24.980 | - What's fiat currency by the way?

01:17:26.580 | - Oh, just like things like US dollars

01:17:29.300 | and like dollars and euros and yen and these other things.

01:17:32.060 | - And they're sort of backed by governments.

01:17:33.940 | - Yes.

01:17:34.780 | But I also expect sort of cryptocurrencies

01:17:37.900 | to play this sort of important role

01:17:40.740 | in just making sure that people always have an alternative

01:17:44.140 | if fiat currencies start breaking.

01:17:46.740 | So like if, or if you're in, you know,

01:17:50.580 | some very high inflation place like Venezuela, for example,

01:17:54.460 | or if your country just kind of gets cut off from,

01:17:59.460 | cut off from other financial systems

01:18:04.900 | because of something the banks do,

01:18:07.140 | like if any kind of,

01:18:09.300 | if there's even like some major trade disruption

01:18:11.980 | or something worse happens,

01:18:13.220 | then like cryptocurrencies are the sort of thing

01:18:16.700 | that just because of their kind of global neutrality,

01:18:18.980 | they're just kind of always there

01:18:20.700 | and you can keep using them.

01:18:22.220 | - It's interesting that you're quite humble

01:18:24.220 | about the possibilities of the future of cryptocurrency.

01:18:27.380 | You don't think there's a possible future

01:18:29.820 | where it becomes the main set of currency

01:18:33.900 | because it feels like fiat,

01:18:35.820 | it feels like the centralized control

01:18:39.420 | by governments of currency is limiting somehow,

01:18:42.660 | maybe my naive utopian view of the world.

01:18:46.100 | - It's definitely very possible.

01:18:48.940 | I mean, I think like for cryptocurrencies

01:18:53.380 | being the main form of value to kind of work well,

01:18:58.380 | like you do need to have some much more price stability

01:19:04.220 | than they have today.

01:19:05.540 | And I mean, there are now stable coins

01:19:07.340 | and there are kind of cryptocurrencies

01:19:09.700 | that try to be more stable than existing things

01:19:12.140 | like Bitcoin and Ether,

01:19:13.100 | but that just is to me kind of the main challenge.

01:19:18.100 | - Do you think that's a characteristic

01:19:20.660 | of just being the early days?

01:19:22.380 | It's such a young concept.

01:19:24.980 | 10 years is nothing in the history of money.

01:19:27.420 | - Yeah, and I think it's a combination of two things, right?

01:19:30.860 | One is it's still early days,

01:19:35.340 | but the other is a kind of more durable

01:19:37.180 | and a kind of economic problem,

01:19:38.580 | which is that like demand for currency is volatile, right?

01:19:42.220 | Because of like recessions, booms,

01:19:45.620 | changes to technology, lots of things,

01:19:47.900 | and if people's demand for how much currency

01:19:49.860 | they wanna hold changes.

01:19:51.340 | And if you have a currency that has a fixed supply,

01:19:54.780 | then the change in demand has to be entirely expressed

01:19:58.580 | as a change in value of the currency.

01:20:00.500 | And so what that means is that kind of the volatility

01:20:02.780 | of demand becomes entirely translated into volatility

01:20:06.780 | and kind of prices of things that dominated

01:20:08.660 | in that currency.

01:20:10.100 | But if you have a currency where instead

01:20:12.980 | the supply can change and so the supply can kind of go up

01:20:15.700 | when there's more demand,

01:20:16.700 | then you have the supply kind of absorbing

01:20:20.460 | more of that volatility and so the price of the currency

01:20:23.020 | would absorb less of the volatility.

01:20:24.860 | - On that topic, so Bitcoin does have a limited supply,

01:20:27.620 | a specific fixed supply.

01:20:29.740 | - Yes.

01:20:30.580 | - What's the idea, and Ethereum doesn't,

01:20:33.820 | but can you clarify just in the comments you just made,

01:20:38.540 | is Ethereum qualify to the kind of currency

01:20:41.620 | that you're talking about and being flexible in the supply?

01:20:44.460 | - I mean, it's a bit more flexible,

01:20:45.980 | but kind of the thing that you would really want

01:20:48.540 | is something that's kind of specifically flexible

01:20:53.540 | in response to how valuable the currency is.

01:20:56.860 | And I'd recommend you look at stable coins as well.

01:21:01.060 | So like things like DAI, for example, it's like-

01:21:04.780 | - How do you spell that?

01:21:05.620 | - DAI.

01:21:07.060 | - And what's stable coins?

01:21:09.460 | Is it a type of cryptocurrency?

01:21:11.020 | - It is a type of cryptocurrency.

01:21:12.900 | It's a type of cryptocurrency that's issued

01:21:16.100 | by a smart contract, one of these Ethereum computer programs

01:21:19.860 | that where the smart contract holds a bunch of ether

01:21:23.700 | and then it basically, like that people deposit

01:21:27.500 | and then it issues DAI and the reason why people deposit

01:21:30.740 | is because they want to kind of go high leverage

01:21:32.740 | on their ether.

01:21:33.700 | And so it kind of pairs these two sets of users,

01:21:36.100 | one that wants stability and one that kind of wants

01:21:38.140 | extra risk together with each other.

01:21:40.580 | And it basically creates some,

01:21:45.580 | or gives one set of participants a guarantee

01:21:48.900 | that they'll be paid, that they have this asset

01:21:51.900 | that can be later converted back into ether,

01:21:54.620 | but specifically at kind of the $1 rate.

01:21:57.180 | - And it has some kind of a stabilizing network effects.

01:22:01.180 | - Yeah, it has this.

01:22:02.660 | Yeah, it has many kinds of stabilizing mechanisms in it.

01:22:06.020 | - That's fascinating.

01:22:06.860 | Okay, this world is awesome technically,

01:22:09.700 | just from a scientific perspective, it's an awesome world

01:22:12.700 | that I often don't see from an outsider's perspective.

01:22:15.460 | What I often see is kind of maybe hype

01:22:19.500 | and a little bit, if I may say so, like charlatanism.

01:22:23.160 | And you don't often see, at least from my outsider's

01:22:27.020 | perspective, the beautiful science of it

01:22:28.980 | and the engineering of it.

01:22:30.860 | Maybe, is there a comment you can make

01:22:34.220 | of who to follow, how to learn about this world

01:22:39.220 | without being interrupted by the charlatans

01:22:44.300 | and the hype people in this space?

01:22:47.180 | - I think you do need to just know the specific

01:22:49.740 | kind of just people to follow.

01:22:51.820 | Like there's all the kind of the cryptographers

01:22:55.900 | and the researchers, and there's just like,

01:22:59.260 | even just the Ethereum research crew,

01:23:01.340 | like myself, like Dan Kradt, Danny Justin,

01:23:04.260 | the other people, and then the academic cryptographers.

01:23:09.260 | Before this today, I was at Stanford,

01:23:14.500 | and Stanford has the Center for Blockchain Research,

01:23:17.340 | and Dan Bonet, that's a really famous

01:23:20.420 | and great cryptographer, was running it.

01:23:24.580 | And there's a lot of other people there,

01:23:26.540 | and there's people working on zero-knowledge proofs,

01:23:29.780 | for example, and Zuko from Zcash

01:23:34.140 | is one other person that I respect.

01:23:38.100 | So I think if you follow the technical--

01:23:40.900 | - You just crawl along that--

01:23:42.220 | - Yeah, yeah, you just kind of crawl.

01:23:43.060 | - Start with the Ethereum group,

01:23:44.220 | and then look at the academics, Dan Bonet and so on,

01:23:46.700 | and then just cautiously expand the network

01:23:49.580 | of people you follow.

01:23:50.540 | - Yeah, exactly, and like if someone seems

01:23:52.780 | too self-promotional, then just like remove 'em.

01:23:56.860 | Is there books that are, there's these white papers,

01:24:01.100 | and we just discussed about ideas being condescend

01:24:04.380 | to really small parts.

01:24:05.900 | Is there books that are emerging that are kind of

01:24:09.180 | good introductory material for cryptographers?

01:24:11.820 | - So for historical ones,

01:24:13.580 | there's like Nathaniel Popper's "Digital Gold,"

01:24:15.900 | which is just about the history of Bitcoin.

01:24:18.220 | There's one, I think Matthew Leising announced

01:24:20.900 | that there's one about the history of Ethereum.

01:24:23.660 | For technical ones, there's Andrea Santanopoulos'

01:24:26.500 | "Mastering Ethereum."

01:24:28.980 | - Great, so let me ask you sort of,

01:24:32.100 | sorry, to pull back to the idea of governments

01:24:37.100 | and decentralized currency.

01:24:38.540 | You know, there's a tension between decentralization

01:24:44.180 | of currency and the power of nations,

01:24:46.940 | the power of governments.

01:24:48.220 | What's your sense about that tension?

01:24:52.220 | Is there some rule for regulation of currency?

01:24:56.540 | You've, yeah, is there, like,

01:24:59.260 | is the government the enemy of digital currency,

01:25:02.940 | of distributed currency,

01:25:04.220 | or can they be like cautious friends?

01:25:08.860 | - I mean, I think like,

01:25:11.780 | the one thing that people forget is that

01:25:14.940 | it's clearly not entirely an enemy,

01:25:16.620 | because I think if there hadn't been

01:25:21.620 | so much government regulation on

01:25:24.460 | a kind of centralized,

01:25:25.900 | issuing centralized digital currencies,

01:25:28.980 | then we'd be seeing things,

01:25:31.740 | people like Google and Facebook and Twitter

01:25:33.380 | just kind of issuing them left and right.

01:25:35.140 | And then, like, if that was the case,

01:25:37.340 | then decentralized currencies would still appeal

01:25:39.260 | to some people, but they definitely would appeal

01:25:41.180 | to less people than today.

01:25:42.740 | So even in that sense, I think it's

01:25:45.820 | clearly been more of a,

01:25:49.740 | just kind of set the stage for the existence

01:25:53.340 | of the sector in some ways.

01:25:56.140 | But also, and I think some of both,

01:26:01.140 | there's definitely things that governments

01:26:03.820 | can do, in some cases, have done to

01:26:06.740 | hurt the spread of, and of growth of blockchains.

01:26:11.460 | There's things that they've done to help,

01:26:13.380 | and they've, in some cases, definitely done a good job

01:26:18.100 | of going after fraudulent projects,

01:26:21.060 | going after some of the projects that have

01:26:24.060 | some of the kind of craziest and most misleading marketing.

01:26:27.060 | There's also the possibility that governments

01:26:31.580 | will end up using blockchains for a lot of different things.

01:26:35.500 | Governments, yeah, they do a lot more

01:26:38.500 | than just regulating, right?

01:26:39.820 | There's also, they have the kind of

01:26:43.300 | identity records, and they have property registries,

01:26:49.860 | even their own currency is secured,

01:26:53.700 | lots of different kind of things that they're operating.

01:26:57.100 | And there's even blockchain applications in a lot of those.

01:27:00.780 | - Yeah, and they can leverage technology

01:27:03.740 | to do a lot of good for our societies.

01:27:06.100 | It is a little unfortunate that governments

01:27:09.140 | often lag behind in terms of their acceptance

01:27:13.180 | and leverage of technology.

01:27:14.420 | If you look at the autonomous vehicle space,

01:27:16.380 | AI in general, they're a few years behind.

01:27:20.980 | It'd be nice to help them catch up.

01:27:24.100 | That's an always ongoing problem.

01:27:27.820 | You met Vladimir Putin to discuss

01:27:31.500 | the centralized currency here.

01:27:33.300 | You're born in, where were you born?

01:27:35.620 | - Kolomna, it's a city about 115 kilometers south of Moscow.

01:27:39.700 | - In Russia? - Yes.

01:27:40.980 | - Yeah, I grew up in Moscow.

01:27:44.540 | I mean, Vladimir Putin is a central figure

01:27:48.020 | in this part of the world.

01:27:49.060 | So what was that like meeting him?

01:27:52.300 | What was that experience like?

01:27:54.900 | - He's taller in photos than in person.

01:27:57.020 | - Yeah, he's, yeah, that's right.

01:28:01.220 | He's five seven, I think, five eight, maybe.

01:28:04.020 | - Yeah, and that's, unfortunately,

01:28:07.180 | we didn't actually kind of have too much of a chance

01:28:09.700 | to talk to him.

01:28:10.540 | Like I managed to see him for about one minute

01:28:12.740 | at the end of this meeting.

01:28:14.620 | And I did get a chance to see a lot,

01:28:16.980 | some of the other end of government ministers

01:28:19.100 | and he recommended some.

01:28:20.940 | And some of them are actually interested

01:28:25.420 | in trying to use blockchains to,

01:28:29.820 | like for various government use cases,

01:28:31.900 | they're gonna have a little bit of corruption

01:28:33.020 | and other things.

01:28:33.980 | And I have, it's hard to tell from one conversation

01:28:37.580 | kind of what things are genuine

01:28:39.540 | and what things are just like, oh, blockchain is cool,

01:28:41.580 | let's do blockchain.

01:28:42.700 | - Right, but when I listened to like Barack Obama

01:28:47.700 | talk about artificial intelligence,

01:28:50.620 | there's certain things I hear where,

01:28:54.060 | okay, so he might not be an expert in AI,

01:28:56.340 | but he actually studied it carefully enough

01:29:00.460 | to think about it.

01:29:01.420 | Even if it's just reading a Wikipedia page,

01:29:06.540 | like he really thought about what this technology means.

01:29:09.380 | Did you get a sense that Putin or some of the ministers

01:29:13.860 | like thought about blockchain,

01:29:15.740 | like thought about the fundamentals of technology,

01:29:17.620 | or like understand it intuitively,

01:29:19.780 | or are they too old school to try to grasp it?

01:29:23.340 | - Some are old school, some are more new school.

01:29:26.860 | It depends, it's definitely,

01:29:28.820 | it depends on who you talk to.

01:29:30.660 | - I mean, that's an open question for me with Putin

01:29:33.020 | because Putin has said some stuff about AI.

01:29:34.900 | - I don't know, as I said,

01:29:36.020 | I've only talked to him for about one minute, so.

01:29:38.140 | - But sometimes you can pick up sort of insights.

01:29:41.940 | As a quick comment, they're about,

01:29:44.460 | maybe you can correct me on this,

01:29:45.820 | but they're about 3,000 cryptocurrencies

01:29:48.180 | being actively traded.

01:29:49.900 | - Yes.

01:29:50.980 | - And Ethereum is one of,

01:29:52.540 | you know, a lot of people believe

01:29:53.580 | that Ethereum will be the main cryptocurrency.

01:29:56.140 | I think Bitcoin is currently still the main cryptocurrency,

01:29:58.660 | but Ethereum very likely might become the main one.

01:30:03.420 | Is this kind of diversity good in the crypto world?

01:30:07.180 | Do you see it sticking around?

01:30:08.940 | Should there be a winner?

01:30:10.660 | Like, should there be some consensus globally

01:30:12.980 | around Bitcoin or around Ethereum?

01:30:16.060 | Like, what's your sense?

01:30:17.740 | - I definitely think the diversity is good.

01:30:19.460 | And I definitely think also

01:30:21.380 | that there's probably too many people

01:30:23.540 | trying to make separate blockchains right now.

01:30:26.900 | The number should definitely be greater than one

01:30:28.900 | and probably greater than two or even five.

01:30:31.780 | - Not 3,000.

01:30:33.180 | - Not 3,000, yeah.

01:30:34.740 | And also not even like 40 high quality platforms

01:30:37.860 | that try to do the same thing.

01:30:39.420 | And there's definitely this range

01:30:41.300 | from just like one person who just like wrongly thinks

01:30:46.300 | that you can create a cryptocurrency in like 12 hours

01:30:49.500 | and doesn't even think about kind of the community aspects

01:30:52.860 | of maintaining it,

01:30:53.780 | going to people actually trying

01:30:58.020 | but only creating a really tiny one,

01:30:59.780 | to like scammers,

01:31:01.420 | to people like making something

01:31:03.700 | that's actually successful.

01:31:06.140 | And then there's a lot of different categories of blockchain

01:31:10.860 | and of project in terms of what it's trying to do

01:31:12.980 | and what applications it's for.

01:31:14.660 | And I think the experimentation is definitely healthy.

01:31:18.900 | - If you look at the two worlds,

01:31:21.700 | it might be a little bit disjoints,

01:31:23.100 | but the distributed applications, cryptocurrency,

01:31:27.060 | and then the world of artificial intelligence.

01:31:29.380 | Do you see there's some overlap between these worlds

01:31:33.020 | that both worry about centralized control?

01:31:35.860 | Is there some overlap that's interesting

01:31:38.220 | that you think about, do you think about AI much?

01:31:40.700 | - Yeah, and I think definitely I would have thought

01:31:44.620 | about things like the AI head of control problems

01:31:48.340 | and a lot of the problems and all of those things.

01:31:50.620 | - Do you worry about the existential threat of AI?

01:31:53.260 | - That's definitely one of the things I worry about.

01:31:55.820 | I think block, there's a lot of kind of common challenges

01:32:02.700 | because in both cases, what you're ultimately trying to do

01:32:05.980 | is you're trying to kind of get a simple system

01:32:09.660 | to direct a more complex system.

01:32:11.820 | Like in the case of this strong AI,

01:32:15.220 | the idea would be that the simple system is people

01:32:17.660 | and the complex system is, well, whatever thing

01:32:21.380 | the people end up kind of unleashing on the universe

01:32:24.780 | that'll hopefully be a great thing.

01:32:27.140 | And in the case of blockchains,

01:32:30.380 | kind of the complex, well, the simple thing is

01:32:33.420 | the algorithm, which is a piece of static

01:32:35.180 | and fully open source code.

01:32:36.460 | And the more complex thing is just all

01:32:40.660 | of the different possible kind of human actors

01:32:43.300 | and of the strategies that they might end up use

01:32:46.460 | to participate in the network.

01:32:48.020 | - Do you think about your own mortality?

01:32:51.140 | Like what you hope to accomplish in your life?

01:32:54.580 | - Oh, I definitely think about ending my own mortality.

01:32:58.900 | - So that's, if I gave you the option to live forever,

01:33:01.780 | would you?

01:33:02.820 | - Depends a lot on what the fine birds is.

01:33:05.100 | But I mean, if it's one of those things

01:33:08.020 | where I'm gonna be kind of like floating through empty space

01:33:10.700 | for 10 to the 75 years, then no.

01:33:13.260 | If it's a forever worth of,

01:33:17.620 | and of having a fulfilling life

01:33:21.060 | with friends to spend the time with,

01:33:27.460 | with kind of meaningful challenges to explore

01:33:31.580 | and interesting things to be working on,

01:33:33.860 | then I think absolutely move.

01:33:37.540 | - That's a beautifully put.

01:33:38.940 | Live forever, but you'd have to check the fine print.

01:33:42.780 | I think there's no better way to end it.

01:33:45.500 | Vitalik, thank you so much for talking to us.

01:33:48.180 | So exciting to follow your work from a distance.

01:33:51.140 | And thank you for creating a revolutionary idea

01:33:54.460 | and sticking with it and building it out

01:33:56.460 | and doing some incredible engineering work.

01:33:58.020 | And thanks for talking today.

01:33:59.300 | - Yeah, thank you.

01:34:00.820 | - Thanks for listening to this conversation

01:34:02.460 | with Vitalik Buterin.

01:34:03.820 | And thank you to our sponsors, ExpressVPN and Masterclass.

01:34:08.620 | Please consider supporting the podcast

01:34:10.420 | by signing up to Masterclass at masterclass.com/lex

01:34:14.900 | and getting ExpressVPN at expressvpn.com/lexpod.

01:34:19.900 | If you enjoy this podcast, subscribe on YouTube,

01:34:23.940 | review it with five stars on Apple Podcast,

01:34:26.340 | support it on Patreon,

01:34:27.740 | or simply connect with me on Twitter @LexFriedman.

01:34:31.780 | And now let me leave you with some words

01:34:33.980 | from Vitalik Buterin.

01:34:35.940 | The thing that I often ask startups on top of Ethereum is,

01:34:40.180 | can you please tell me why using Ethereum blockchain

01:34:43.300 | is better than using Excel?

01:34:45.540 | And if they can come up with a good answer,

01:34:47.620 | that's when you know you got something really interesting.

01:34:51.540 | Thank you for listening and hope to see you next time.

01:34:55.060 | (upbeat music)

01:34:57.660 | (upbeat music)

01:35:00.260 | [BLANK_AUDIO]

Vitalik Buterin: Ethereum, Cryptocurrency, and the Future of Money | Lex Fridman Podcast #80

Chapters