Free Thoughts, Ep. 212: Your World on the Blockchain (with Brock Cusick)


Aaron Powell: Welcome to Free Thoughts, I’m
Aaron Powell. Trevor Burrus: And I’m Trevor Burrus. Aaron Powell: Joining us today is Brock Cusick. He’s an attorney working in the foreign exchange
and derivative markets. Welcome to Free Thoughts, Brock. Brock Cusick: Hi Aaron. Thanks for having me on. Aaron Powell: What is a blockchain? Brock Cusick: A blockchain is a new kind of
database. That’s how I think about it. It’s really the core of what Satoshi [00:00:30]
Nakamoto, the inventor, of Bitcoin really invented. He … If you think about Bitcoin, it’s decentralized
money. We’ve had decentralized money before. Gold coins are decentralized money, and they’ve
been around for thousands of years. It’s really the blockchain which is a new
type of database that was his core invention. It’s really what’s [00:01:00] creating a lot
of change in the world right now. Trevor Burrus: It’s decentralized but what
does it allow us to do? Aaron Powell: Or just give us a [crosstalk
00:01:10] Just tell me a sketch of how it works. Trevor Burrus: First that, and then we get
to the application. Brock Cusick: The thumbnail sketch, the blockchain,
if you think about a database, a database is like a ledger, like an Excel spreadsheet. If you think it’s something regular most people
might be familiar with. [00:01:30] It’s just a list of entries with
arbitrary data that you can write to them. Usually if you’ve ever tried to work with
a spreadsheet on someone you have an issue of version control where you could put things
in your spreadsheet and then they could end up changing it. If you want to trust this spreadsheet is accurate,
the way I sort of trust my to do list at work, which I maintain in this spreadsheet, I don’t
share it with anyone, [00:02:00] because I don’t want anyone else going in there and
deleting things off my to do list that I need to do, because otherwise I’ll forget and they
won’t get done. When you have databases you usually have to
have a trusted person to maintain and run the database and verify that any changes made
to the database are appropriate. The blockchain is new in the sense that it
can be an open [00:02:30] network, which means anyone in the world can access it, anyone
in the world can attempt to write into it and yet it still maintains its integrity even
though it’s open to the entire world. That’s what really new about it and that’s
what really fascinating. The way it works is it relies on a number
of different underlying technologies to make this work. Basically it uses [00:03:00] cryptography
which is asynchronous. Let me back up a little bit. It uses cryptography which is very easy to
check but very hard to do to verify that transactions that enter into the database are appropriate,
so that you can send a transaction to the database, to the blockchain, and it can be
incorporated according to the rules. It’s arranged in such a way [00:03:30] that
if you want to send an incorrect or I could say a transaction that breaks the rules and
you want to fake it, the way you have to do it, because of the way the blockchain is set
up, is you have to burn an amount of electricity about equivalent to what the country of Yemen
uses in a day to falsify one transaction. Trevor Burrus: That’s what you mean by hard
to do is just computing power? Brock Cusick: Yeah. You can do it with computing power but [00:04:00]
what you can do is falsify a single transaction. The maximum payout of that falsification is
whatever the value of the transaction is. If you transfer $100,000 to someone, or $1,000
to someone, and then you want to transfer a false transaction, you want to move that
money to an account you control instead of the other person. That’s the maximum payout. [00:04:30] The cost is always the electricity
cost of the entire network. Trevor Burrus: Just to make sure that I’m
clear about his, because Aaron knows way more about blockchains than I do, and it’s a little
bit hard to put your brain around. If we use a metaphor, for example, you talked
about using a trusted ledger, held by a third party. If I transfer money to you via my bank, that’s
depending upon the fact that [00:05:00] we trust the bank. You trust the bank and I trust the bank, right? Brock Cusick: Exactly, right. Trevor Burrus: The blockchain takes the place
of the bank or another third party by encoding a series of, again I might be getting this
wrong, tell me if I’m wrong. Coding a series of indicators of a transaction
that is very, very distributed and hard to break, and therefore it verifies between you
and me without having to have a third [00:05:30] party. Is that kind of correct? Brock Cusick: No, you got it. There is the ledger of the blockchain is,
the blockchain is the ledger of all the bitcoin transactions that have existed since the software
was implemented in 2007. What it does is it’s run on thousands of nodes
all around the world who don’t know each other. They all independently verify every attempted
[00:06:00] transaction. If you want to send money, send bitcoin from
your account to my account, you have to submit a transaction to the ledger, you sign cryptographically
with the same signature, the same public key cryptography that you used to receive the
money in the first place. You sign it and say, “I received [00:06:30]
this money three months ago. When I received it, this was the public key
I used. Here I’m signing this transaction with the
exact same public key.” And you submit that to all the nodes globally. They verify that it’s the same public key. They verify it’s the correct digital signature. Then as long as your transaction that you’ve
submitted meets the rules of the bitcoin network, they record it to the blockchain. Aaron Powell: This brings up the nature [00:07:00]
of this ledger and the nature that all of the transactions, including the addresses
that the transactions were originating from and going to are stored in this public ledger
that goes back to the very beginning. Brock Cusick: Yes. Aaron Powell: Brings up the question of privacy,
because bitcoin often gets pitched as not just a better way to send money around but
one that’s couch more privacy [00:07:30] or less susceptible to having our transactions
snooped on, but now our transactions are all happening in public. I’m curious about that because one of the
early applications of bitcoin, on of the early big uses of it was the infamous Silk Road
where people were buying and selling drugs and all sorts of actually awful things too
with bitcoins. How private is it if you can log, if I sent
Trevor 20 bucks, [00:08:00] you can log in and see that here is my wallet, here is Trevor’s
wallet and here is the $20 flowing to it? Brock Cusick: The privacy features of the
base bitcoin feature set, just are like the core feature set have been definitely oversold
by a lot of people who I don’t think maybe understood exactly how the network works. Every bitcoin transaction happens in public,
every bitcoin transaction is verified by every node, which [00:08:30] is open source software
that anyone in the world can run. If the IRS is running a bitcoin node, and
they probably are honestly or the FBI is. They just being part of the bitcoin network,
they receive a copy of every single transaction that’s submitted to bitcoin. You have to register when you open your account
at bitcoin. It’s an open network, you just have to download
free software. It assigns you a random number. I think the numbered accounts in Switzerland. You [00:09:00] have a numbered account, it
doesn’t have your name on it, but once the authorities connect you to the numbered account,
all the transactions in and out of that account are public record. Trevor Burrus: Here is maybe a strange question,
but I think it’s one that a lot of people want to ask about bitcoin which is, what makes
bitcoin valuable? Brock Cusick: It’s rare. Every commodity in the world that has a fixed
quantity, there are some level of demand that meets that fixed quantity [00:09:30] and that
supply and demand curve tells you what the price is. Things with higher demand and lower quantity
are worth more, like gold. Gold is rarer than silver, so gold is worth
more than silver. Gold has other qualities that make it useful
as money. That’s true for any commodity, concrete, blacktop,
tons of wheat, whatever. What makes bitcoin rare though, that’s somewhat
[00:10:00] interesting, because if you think of bitcoin as a digital commodity, the entire
reason that Napster works and the entire reasons that bit torrent works is you can copy digital
files infinitely. Once something is a digital file you think
of that as being infinitely suppliable. You can make as many copies as you like and
you can share them around the world for free. Bitcoin is mathematically rare. If you are familiar at all with the network
you know that there are computers that are involved [00:10:30] in the network called
miners. The bitcoin miners, they are a global network
of computers that are competing with each other to find an exceptionally rare number,
and it’s so rare that this huge global network of computers which all together has more computing
power than google or Amazon or an of these companies. They only can find one of those numbers [00:11:00]
every 10 minutes. It’s very simple for anyone with a very small
low power computer, like a raspberry pie or your phone, or whatever … This number is
very simple for the small computers to verify how rare the number is. It’s extremely hard, in fact, it’s impossible,
mathematically impossible to fake how rare the number is. You can’t just make up a number and hope it’s
rare. It actually has to be rare. [00:11:30] That rarity. That rare number that’s found once every 10
minutes is recorded to the bitcoin blockchain along with the transactions that people submit
every 10 minutes. That’s the people that are verifying. You can submit, you could make up a bitcoin
blockchain yourself, you can download the current blockchain, make a copy of it and
then make a whole bunch of transactions and submit them, but you can’t fake that rare
number. That’s mathematically impossible. [00:12:00] If some third party out in the
world receives your copy of the block chain, and someone else’s copy of the blockchain,
they can look at those, the nodes that is appended to that block of transactions, and
they can see which one is rarer, which one required more electrical power to discover
and generate? That’s how bitcoin’s rarity is created and
from there just simple supply and demand means that it’s a commodity [00:12:30] and it therefore
has a market price. Aaron Powell: In finding and then the computer
that finds those, that finds that number gets rewarded. That’s what they are mining for is they are
mining for bitcoins. Brock Cusick: Right. One of the rules of the network, and these
rules are all agreed by, they are agreed socially by all the miners. Everyone who participates in the network is
essentially accenting to abide by the rules of the network by participating. [00:13:00] One of the rules is that whoever
finds this number, when you submit that block of transactions, you get to do two things,
you get to create a few bitcoins for yourself. And you get to keep all the transaction fees
that are in all the transactions that happened since the last rare number was discovered. On average these rare numbers are found every
10 minutes. That means you basically get to collect 10
minutes worth of the [00:13:30] transaction fees on the bitcoin network. Trevor Burrus: You mentioned this previously
but I might do like sort of fill in the history, you mentioned Satoshi Nakamoto. A lot of people, I was hearing about bitcoin,
because I ran in libertarian circles in 2010, 2011 and unfortunately did not buy any bitcoin
even though people were telling me to. At that point it was like a dollar or something
along those lines. [00:14:00] Where did this come out of, was
this some sort of dark internet place where so many people were mining these bitcoins? And also if I’m correct I believe that at
beginning it was easier to mine bitcoins and it’s getting progressively more difficult,
correct? Brock Cusick: Yeah. I don’t remember exactly where Satoshi first
posted about it. I saw the post once but I don’t remember the
website. Satoshi, it’s a fake [00:14:30] name, it’s
a pseudonym, we don’t know who the person is, although there is lots of speculations
obviously. He just posted and shared the open source
software. And then from the open source community, ever
since the beginning, lots of people have been involved and then Satoshi himself stepped
out of development after a couple of years and hasn’t been seen since. It’s now just in the world so to speak. [00:15:00] Regarding your difficulty, you
are absolutely right. One of the rules of the network, the bitcoin
network is that it wants to discover the rare number once every 10 minutes. That its goal. Sort of like the federal reserve has an insulation
target, the bitcoin network has a 10 minute block time target. As more people compete to find this number,
more computing power is added to the network. [00:15:30] As that happens, the numbers get
discovered more often. If the average time for the discovery falls
below 10 minutes, what happens is every two weeks the network agrees to adjust the difficulty
upward, to get back to 10 minutes average block time. Aaron Powell: I think that answer, a couple
of follow ups, but I think they gave us an opportunity to then pivot into [00:16:00]
some of the other topics we are gonna discuss. There is that 10 minute target and then there
is ultimately a fixed supply of bitcoin, because the number of bitcoins you win for solving
it halves every so often- Brock Cusick: Few years I think. Aaron Powell: Yeah, but I think ultimately
it will stop with what is it? 21,000,000 bitcoins in circulation? Brock Cusick: That’s right, 21,000,000. Aaron Powell: Which is supposed [00:16:30]
to be in the year 2040 roughly? Brock Cusick: Mm-hmm (affirmative). Aaron Powell: First, when that happens, when
there is no longer any bitcoins to be mined and given that the bitcoin network is depending
on these miners to effectively to maintain it, to process transactions and maintain the
integrity of it. What happens then? Is there a reason for, because it’s this enormous
amount of computing power and it’s a high cost, because you have to pay for the electricity
for all of this on top of the equipment. [00:17:00] What’s the incentive for miners
to continue to mine even after there is no bitcoins to be extracted from the virtual
ground? And then also how does this play into … There
is a lot of controversy in the community right now, because there is supposed to be a fork
of bitcoin network coming up in middle November. And part of that has to do with the amount
of time it takes to process transactions has been going up [00:17:30] because the blockchain
is limited in size and it can’t keep up with the number of transactions as bitcoin grows. What’s happening in the bitcoin world right
now to address those issues and then how do the things that are happening impact this
question of what happens after 2040? Brock Cusick: It’s actually really great that
you ask those two questions together because they are very [00:18:00] related. Satoshi when he created bitcoin, he understood
back at the time that you need to reward the miners in order for them to provide the service,
because the miners aren’t in the bitcoin business for the benefit of themselves to have this
digital money. They are in it to make money for themselves,
it’s a business for them. You have to reward them somehow. Over the [00:18:30] long term, any sort of
payment network really needs to make itself profitable based on the transaction fees,
because people use the network and their usage is what cost money and so the most, the incentive
structure that’s most aligned with the users is the charge rate transaction fee. Back in the day, back when bitcoin was first
started, and hardly anyone was participating in it. [00:19:00] For the most part no one really
needed bitcoin, no one needed to pay money for this. It was all just sort of for fun. This is before it really had a market value. It wasn’t generating any transaction fees. What Satoshi did and he also did this to distribute
the 21,000,000 widely is he put all 21,000,000 bitcoins in a potential either pool [00:19:30]
for miners to reward themselves. It started back in the day, in the very beginning
that every time a miner found the rare number to mine a block, they got 50 bitcoins and
basically zero transaction fees, because no one was paying transaction fees. Over time these two, the whole design of the
network was to eventually have those two numbers switch places. Where the mining reward would be [00:20:00]
zero bitcoins, but the transaction fees would be high. That’s what we are seeing, which is nowadays
I think nowadays there is, I just looked it up and now I can’t remember, something like
200 bitcoins per day maybe 250 in transaction fees are paid. A bitcoin is worth $5,000 at the moment roughly,
the price moves around. That’s [00:20:30] a lot of money. Those are the transaction fees, and that’s
where we are now. We might even be generating more money from
transaction fees, than from the mining reward already. The idea is that by 2040, assuming bitcoin
remains popular, assuming people keep using it, the transaction fees only go up from here. That gets to your second point about the congestion
on the network. The reason that people are paying transaction
fees [00:21:00] is because they want their bitcoin transactions recorded to the blockchain. The miners get to choose which transactions
to include in any given block, a 10 block of bitcoin transactions. The miners are most inclined to include the
transactions that hey, a transaction fee because they can choose. You have to realize then that the transaction
fee is the market price [00:21:30] where all the users are competing to get into a given
block and to have their transactions processed faster. That’s what driving up the transaction fees,
that’s what’s creating the revenue from the miners, and therefore that’s what securing
the network. If you expanded the capacity of the blockchain
to the point where no one has to compete to get into a block anymore, no one’s gonna really
pay transaction fees, not very much. The revenue of the network goes down and then
the miners are gonna stop [00:22:00] putting as many computing resources or as much electricity
into finding the number. Now the rarity of the network itself is lower
and the security of the users is lower. There is a real balancing act there. There is a real balancing act that I think
the developers and the miners, and the users are all trying to edge forward without any
one of them getting too much advantage over the other, because if the miners are [00:22:30]
too successful at keeping the number of transactions that you can include low, you drive up the
transaction cost to the point where fewer people want to use bitcoin. On the other hand if the users are too successful
in getting the size of the transaction blocks expanded, then the transaction fees go down
the miners quit and the security of the network degrades. It’s a balancing act. Trevor Burrus: This [00:23:00] balancing act. It sounds like something that people are working
through, but is this a problem? Is this something that is concerning people? I guess connected to this other question? Because I hear people kind of criticizing
there are flaws in bitcoin that are gonna make it go down in value and not be the perimeter
one anymore. You have people creating other cryptocurrencies
like Ethereum, or things like Ethereum [00:23:30] to deal with perceived problems in bitcoin,
kind of the ones that you’ve discussed? Brock Cusick: I don’t think of Ethereum as
a real competitor to bitcoin. The reason I say that is because, like I said,
Satoshi Nakamoto invented the blockchain, to bring us back to the beginning, it’s a
new type of database, it’s open, it’s decentralized, it’s trusted without any central third party
being the person you are trusting. His [00:24:00] first use case was bitcoin
which is money, and the bitcoin network is optimized to be good at being money. Ethereum is optimized to be good as a medium
for computing. The developers at bitcoin, every time they
face a trade-off or an engineering decision, they are gonna, obviously they can make mistakes,
but they are gonna do their very best to make the decision that makes bitcoin better money. [00:24:30] If the Ethereum developers are
similarly focused on being the world computer, they are gonna make different trade offs. They are gonna make trade offs that make Ethereum
better as a computer. Will Ethereum always have a market price? You have the ether coins on the Ethereum network? Yeah. They should because you need to be able to
pay for computers on the Ethereum world computer to do their job. Ethereum needs to have an internal market
price. Ethereum’s [00:25:00] market price is a servant
to the Ethereum world computer, whereas on bitcoin the market price and it’s value as
money is the end in and off itself. I think that over the long term, because their
designs will never fully converge, you’ll have multiple blockchains, at least one per
use case so to speak. Trevor Burrus: What do you mean use case,
when it’s being [crosstalk 00:25:28]? Brock Cusick: There is the money use case,
the computing [00:25:30] use case. It’s possible another coin come along that
wants to be global internet money that would compete with bitcoin directly and maybe eventually
supplant and replace bitcoin. I don’t think it’s likely just because of
the lead bitcoin has and the market penetration it has, but I admit that it’s possible. I don’t think Ethereum is gonna be that, because
Ethereum has a different use case. Ethereum wants to be a world computer. The designers are gonna make trade-offs. Trevor Burrus: That’s get to [00:26:00] the
question which I think Aaron was gonna ask which is, if Ethereum is not money or you
are kind of saying it’s not as good for money, but it has value, that gets into the bigger
topic here which is what is the blockchain good for that’s not just money? Brock Cusick: That’s true. The blockchain as I said is decentralized
money has been around for a long time. The blockchain is [00:26:30] what’s new. And the blockchain allows ledgers recordings
that don’t require a third central party to be the trusted party. Obviously banks use ledgers, but they are
hardly the only type of company out there that uses ledgers. Lots of companies use ledgers, airlines have
sky miles. There is also ledgers of [00:27:00] useful
information like the phone book. Everyone in the world, maybe not in the world
but in each country, each jurisdiction, if you want to know someone’s phone number, at
least back in the day of landlines, you looked it up in the phone book. The phone company was the central trusted
party that maintained the phone book. You could trust that when you looked up somewhere
in the Yellow Pages or in the White Pages that that was the correct number. That’s an example of a ledger. That’s an [00:27:30] example of a ledger. If you heard any discussion out there about
blockchain for identity, what they are essentially saying is, we can now make a phone book without
a phone company. You can now have a central phone book where
instead of recording money transactions you record your phone number or your email address
or some other personal data that you don’t mind being public. Now you’ve got a blockchain that’s in the
center there that’s a trusted party and that’s a trusted ledger. You still have to find [00:28:00] an incentive
model to make it work, so that the people out there have an incentive to maintain the
ledger and to secure the ledger against being hacked. As long as that incentive model exists, you
now have a different ledger, it’s not a money ledger. It’s an identity ledger. The blockchain, if you look out in the world
and you think about where are there lists of names or other very important information
that is [00:28:30] maintained by someone? They are basically just paid to maintain that
list. That’s a potential target for blockchain disruption. Aaron Powell: I want to eventually talk more
about these, the kinds of industries that might be disrupted by this tech and the kinds
of ways that this tech could invent new industries or really change the way that we do things. Before we get to that, we need to just go
back to bitcoin for a little bit and ask about [00:29:00] the recent innovations in and around
the bitcoin network that there are ways to do cool things on top of it or address some
of the concerns that we talked about with transaction cost or speed or whatever else
without embracing a different sort of coin or changing the underlying bitcoin protocol. Two of the things that get talked about are
[00:29:30] so called sidechains and then like level two protocols. Brock Cusick: Yep. Aaron Powell: What are those and what are
some of the interesting things happening within those? Brock Cusick: Those are both very interesting. They achieve different things. I’ll talk about level two first if you don’t
mind. The most common level two protocol that gets
discussed a lot is called the lightning network. As I mentioned before, the [00:30:00] bitcoin
network can record, has very poor performance. That’s the trade off that was made for the
decentralized trustless nature. It’s decentralized, it’s trusted, but it only
takes, it only takes in seven transactions per second globally. Everyone in the world who uses bitcoin has
to share those seven transactions per second and there is billions of people in the world. Obviously that’s a problem if everyone wants
to submit a transaction. Bitcoin [00:30:30] at the moment is impossible
to use as a way to say buy coffee, because there is just too many transactions out there,
they could never all be recorded. The solution that was proposed a few years
ago and is now finally almost ready for rolling out to the consumers is this level two protocol
called the lightning network. What you do and the reason it’s called level
two is, [00:31:00] instead of recording a payment to the bitcoin network. You submit a transaction to the bitcoin network
that opens a channel between you and somebody else, a payment channel, and you commit a
certain amount of bitcoin to it. Say a $1,000. As a metaphor that’s almost like submitting,
depositing $1,000 at a bank to open a checking account. You [00:31:30] now have a checking account
and it’s got $1,000 in it. Without the bank instead it’s recorded to
the blockchain and it’s decentralized and no one can, as long as they don’t steal your
digital keys that secure the cryptography that you open the account with, no one can
take that money or move it without your signature. Now you have this channel and it was submitted
to the seven transactions per second, and [00:32:00] you paid your fee, whatever the
fee is at the moment, maybe it’s a dollar, maybe it’s $5 to get a transaction submitted. You did have to pay a fee to the bitcoin network,
to the miners to get this transaction submitted. Once it’s open you now have this channel that’s
got $1,000 worth of bitcoin in it. Now whenever you want to transfer money to
that third party you don’t have to submit the transaction to the bitcoin network. You can just send them only along the channel
you have opened, just between the two of you by exchanging signatures. [00:32:30] You can say, “All right, we are
updating the debit and credit to say, “Instead of $1,000 to me and zero to you, it’s now
990 to me and 10 to you.” And you both sign that with your digital signature,
and it’s a good transaction. Now they have $10. Any time they want to close the channel out
after that they can submit a new transaction to the bitcoin network that says $10 to some
account they control and 990 [00:33:00] to an account you control. Or you can just leave the channel open. You can keep sending money back and forth
like this, just between the two of you without ever getting the bitcoin network involved. That’s why it’s called layer two. You are existing at a layer above the bitcoin
network. It’s a private channel, it’s only between
you and them. It’s more private now because these transactions
individually are not public records. The only public records is that you open the
channel of $1,000 in the first place. Trevor Burrus: [00:33:30] Just as an analogy
to this, if I’m understanding correctly, would it be like going to get a rental car and they
swipe your credit card, and they authorize it for more than maybe the eventual charge
you owe or is that a bad analogy? Brock Cusick: I think my first analogy about
the checking account was better. Another one could be say you open an account
with, say you do a lot of shopping on eBay, and you also sell a lot of stuff on eBay. You open a channel to eBay. And [00:34:00] you put $1,000 into it. You now every time you buy something off eBay,
instead of submitting a new bitcoin transaction, you submit a lightning network transaction
just between you and eBay on your private channel. Every time you sell something on eBay they
send you money on the channel. As long as you are still doing business with
eBay, you just keep the channel open. The useful thing here is eBay can now play
[00:34:30] the role of middle man, because it have a lightning network channel to everyone
else who buys themselves on eBay. Say your dad also uses eBay, and you want
to pay him $10 for lunch. Submitting that $10 transaction to bitcoin
would cost you $2 in fees and it would take six hours or a day to confirm. It’s not convenient. You could submit a transaction to eBay and
say, “Hey, I want to pay my dad $10, we both have lightning channels with you. I’m gonna send you 10 and [00:35:00] then
you send him 10.” eBay says, “Well, if I keep a penny I’ll do
it. If I get to keep a penny I’ll do it.” Now you can have instantaneous transfers between
you and anyone else in the world as long as you can make a daisy chain of lightning network
channels. You might have an account with eBay, and maybe
someone else has an account with Amazon, a lightning channel open to Amazon, because
that’s where they do their shopping. eBay and Amazon [00:35:30] have a lightning
channel between them because eBay pays Amazon for cloud services. It’s three parties but again, the cost of
a lightning transaction since it’s not submitted to the bitcoin network is only the cost of
the computer and the electricity to confirm a simple, do a simple cryptographic calculation. It takes a fraction of a second. The actual cost is probably not a penny but
a thousandth of a penny or a millionth of a [00:36:00] penny. You can do an infinite number of transactions
between you and someone else as long as you pay those very, very small transaction costs,
and it’s solved instantly, not 10 minutes or an hour from now. Aaron Powell: If that’s a level two, that’s
example of a level two network, then that’s a sidechain? Brock Cusick: A sidechain is completely different. The level two networks like I said, they are
good for payments. They allow bitcoin to be bitcoin but only
faster and cheaper. A sidechain adds [00:36:30] entirely new functionality. As we’ve discussed, the bitcoin network has
rules, and when you participate in the bitcoin network, you can only submit transactions
that follow those rules. If you submit a transaction that tries to
do something else or has data in it, it doesn’t make sense to the bitcoin network, the transaction
is just gonna get rejected. If you wanted to do something else entirely
you can set up a new blockchain with different rules. People have done [00:37:00] that, like Ethereum. Whenever you start a new blockchain like that,
there is a bootstrapping problem. We have to attract users because without users
the cryptocurrency has no value. If the cryptocurrency has no value, there
is no incentive to people like miners to participate or to lend resources to the network. It’s a real hard bootstrapping problem, and
it’s also filled with a lot of scams unfortunately, because people will start a new blockchain
[00:37:30] and they’ll try to build up a value for it, but instead of rewarding all the coins
to the miners they’ve kept 20% of the coins to themselves. It’s sort of a get rich quick scheme. A sidechain is like that, but instead of creating
a whole new network, you just create a new set of rules and you agree to what’s called
Atomic Swap protocol, which allows you to exchange cons on your network for [00:38:00]
bitcoin at a one for one ratio. The benefit of this is that they will then
[inaudible 00:38:07] of a one for one fixed exchange rate with bitcoin, the value of your
coins are the value of bitcoins. The bootstrapping problem of creating a valuable
coin is totally sidestepped. Your new blockchain can have completely different
rules. It can do completely different things in bitcoin. It might have blocks that are a gigabyte in
size [00:38:30] or might perform arbitrarily total incomplete transactions like Ethereum
does. What a sidechain does is it adds new functionality
to bitcoin by allowing people to move their bitcoin to the sidechain, doing whatever it
is that that other chain does, and then moving those coins back to bitcoin when they are
done and without [00:39:00] any market risk as long as you are willing to accept the market
risk of your bitcoin changing in value. Trevor Burrus: Aaron had mentioned previously
that we have other things that we can do, or you had mentioned that if you have anything
with a ledger where you need a third party like a phone book to verify a transaction
then it has possibilities of blockchain application, and you had mentioned that maybe Ethereum
is more [00:39:30] useful to these things. We kind of got on that a little bit. If you start really thinking about sort of
the world where the blockchain is running free, what sort of things that people may
be surprised, could have an application to the blockchain, what sort of areas that are
not explicitly money? Brock Cusick: There is first order effects
and second order effects. You might be familiar with the story where
lots of people predicted cars, no one predicted Walmart. Trevor Burrus: [00:40:00] Good point. Brock Cusick: Walmart is the second order
effect of consumers being able to drive to a store and the car has a large trunk. Now they can do book shopping. Walmart and Costco are second order effect
of the personal car. The first order effect is all the companies
out there are that are currently paid to maintain a ledger are under threat. Because now ledgers can be [00:40:30] anything
that is blockchain. To tie it back to a sidechain. Say I create a sidechain whose only purpose
is to write and create a phone book. You create a phone book and you list names,
phone numbers and email addresses, whatever you want to be public, since this is all purely
voluntary. You can transfer bitcoin to that sidechain,
[00:41:00] you can pay the transaction fee to record your name and your phone number
and your email to that ledger. It collects a transaction fee paid in bitcoin
because of the one for one exchange rate. You sign it with your digital signature, so
everyone knows it’s actually you. Then you transfer the change back to the bitcoin
network. You now have a sidechain that’s a phone book
or maybe the domain name service that records all [00:41:30] of the website addresses in
the world that’s a bitcoin powered sidechain. Any companies out there are under threat because
of that. That’s the first order effect. Second order effect I think is very interesting
because those are harder to guess, but I have a couple that I like to think about. One is how regulators are gonna be affected. [00:42:00] As Aaron mentioned when we started,
I’m an attorney, so I think about laws and regulators a lot. Right now there is a lot of regulations that
affect us as consumers on a day to day basis without us really being aware of it, because
we are not regulated directly. We are regulated indirectly by having say
our banks be regulated. There are certain transactions you can’t engage
in because the bank won’t let you. Or the bank will, if you engage in certain
[00:42:30] types of transactions or with certain types of counter parties or your transactions
are a certain size like you withdraw over $10,000 in cash from your account, it all
gets automatically reported to various regulators. That law is something that would be very hard
to enforce against consumers directly. But it’s easy to enforce against the bank,
because it’s one bank, it’s a big target, they’ve got a compliance department, who can
check on these things [00:43:00] and could be held accountable. You have all these laws. The bitcoin network doesn’t have a compliance
department. It doesn’t report things automatically. If the United States tried to force it to
change its rules the servers might just move to Peru, or China, or Singapore, or wherever. It’s very hard to nail down. It’s very hard to regulate. Regulators are gonna have to approach things
very differently. [00:43:30] Any time we have … And that’s
true for not just banks, but if you think about currently the debate in privacy with
… The NSA has servers installed at AT&T’s headquarters, so they can listen to all phone
transactions. That’s only possible because AT&T cooperates
with them. If AT&T, if the phone book becomes decentralized
and all the communications switch from [00:44:00] phone conversations to encrypted voice channels
and everything about AT&T gets decentralized, now there is no office for the NSA to put
its servers at. It would have to take a different approach
to achieving national security, which I think is good. I think national security is a useful thing
to have, but it’s gonna force them to adapt, and that’s gonna be a second order effect. There is probably gonna be a lot of gnashing
of teeth and wailing, because no one likes [00:44:30] change. Aaron Powell: I’m gonna ask about one potential
use that I have no idea about but my boss John Samples who heads our First Amendment
Project at the Cato Institute when I was telling him we are doing this episode asked me to
ask this. He was curious if blockchain or related tech
could do anything about the current widely believed in problem of fake news. You talked about distributed identity [00:45:00]
systems and verified NNC systems, is there some way that we can kind of address these
questions of what’s authentic, what’s not? Brock Cusick: Yeah, I think there is actually. Let’s take an example of the identity blockchain,
the phone book blockchain. We’ll build in a couple of more features. You’ve got a phone book blockchain and you
record your name and your couple of pieces of identity that [00:45:30] are yours. Maybe an email address, maybe a Twitter handle. That’s your online social identity and is
recorded to the blockchain. Now we have already discussed the concept
of the lightning network. The lightning network allows thousands of
small cheap, instantaneous transfers of value. What if the phone book blockchain, or sidechain
had a lightning network but not for money? For trust, so that every time I interact with
you [00:46:00] on Twitter, every time we exchange you reply, every time you reply to me or we
send a DM, there could be an anonymous sort of little up-vote or tick to you or to me
saying, “This person is a real person. I’ve interacted with them for the last six
years. I’ve had conversations with them I think they
are a real person. I think they are trustworthy.” I could even write reviews [00:46:30] like
Amazon. You can write someone five stars. Maybe you get a limited number of likes every
week from the blockchain. Instead of the blockchain giving out bitcoins
it gives out likes that you are allowed to reward the people. Over time, in a very organic way you can verify
to the blockchain that this is my identity, and I have [00:47:00] 500 people who’ve all
been interacting with me for the last five years, versus an account that maybe registers
a comment on a news article or maybe they submit something to Reddit, but that account
doesn’t have that sort of history. It doesn’t have that weight, the social weight,
because it’s fake. It would be exceedingly hard for the, not
impossible. This is just like bitcoin, it’s not impossible
to fake a bitcoin transaction. [00:47:30] It’s just exceedingly expensive
for the amount of value you get out of it. Imagine if every time [inaudible 00:47:39]
whoever wants to submit or a political operative wants to submit a fake news article. They can’t just have fake a news article. They have to fake six years or seven years
or 10 years worth of social interaction. It’s now cost prohibitive. Trevor Burrus: [00:48:00] That’s a really
good example of where people, where this can go. A lot of people think that this is confined
to bitcoin. We are still, to a bunch of people it’s strange
and so new. It seems like this could get even bigger in
terms of how disruptive so to speak or I wouldn’t even call it disruptive, I would call it progressive
that we can apply this to many different things. How big can this general blockchain technology
get? [00:48:30] The second question is this. It seems at some point the governments of
the world might start getting upset about it, because they might be- Aaron Powell: Especially considering that
this all originates among, the early people behind this and the early conversations when
bitcoin was still a very small thing was quite Gold Scourge and this is libertarians and
anarchists, the cipher punks and so on. That there’s always been a strong anti- [00:49:00]
authority, anti-government strain to all of this. Brock Cusick: Very strong anti-authority. Not just anti-government, they are anti any
form of authority. They don’t like Microsoft, they don’t like
Google, they don’t like the banks. They just want to be very independent. That’s true and the government is, we have
I think as a preview, look at the fights we have constantly over encrypted communications. That is [00:49:30] an ongoing permanent fight
with the government. So far America is a fight that privacy people
are winning, but it’s not easy. Now multiply that by 1,000 because bitcoin
and all these other ledgers, they don’t just encrypt communications, they potentially encrypt
everything. They encrypt finance. They encrypt money. Now let’s combine [00:50:00] a few potential
blockchains that are all decentralized. If you combine decentralized encrypted money
and decentralized encrypted communications, and a decentralized phone book which could
also record say a mailing address, a P.O Box. You probably won’t want your home address
on the public record, but a P.O Box could be okay, or any sort of centralized receipt,
[00:50:30] place for receipt of packages. You combine all that with self driving cars. I can now communicate with you over an encrypted
channel, we can agree to exchange goods and services for money on that encrypted channel. We can send each other the money over the
encrypted channel, either bitcoin channel. Then I can call up a self driving car, give
it a box and the box is [00:51:00] a form of encrypted, because the box itself is brown
paper, it’s opaque, and the self driving car will drive it to you. There is now, you have basically the entire
economy could become encrypted. Trevor Burrus: This is libertopia we founded. Brock Cusick: Yeah, the physical economy,
not just communications but the physical economy and the finance networks could al become encrypted
at certain levels of sophistication if blockchains become [00:51:30] sufficiently developed and
sufficiently mature and sufficiently widely accepted. There is no guarantee we get there obviously,
but that’s a future that people are gonna think about and people are gonna worry about,
especially on the regulatory side. Trevor Burrus: Do you think governments can
stop this? Brock Cusick: They can make it hard. They could make it really hard. In fact, they can, here is an example, everyone
involved in bitcoin has, everyone involved [00:52:00] in bitcoin has some level of exposure
to the market price of bitcoin. The miners are mostly in China, because China
has very cheap electricity. Their hydropower out in the Himalayas is almost
free. China has all the miners. You might think the US regulators have very
little leverage over the bitcoin network because the miners are all in China and the nodes
are global. Bitcoin only has a market value because [00:52:30]
rich people, and by rich I mean the developed world, wants to buy. If bitcoin is banned in the US and Japan,
and Western Europe, what do you think is gonna happen to the market price? Trevor Burrus: Plummet. Brock Cusick: Go through the floor. It’s gonna go through the floor, if no one
can buy it. In that sense, the regulators do have some
leverage over the miners, at least indirectly because they want to sell the bitcoin for
the highest possible price [00:53:00] to the greatest number of buyers. They want the transaction fees volumes to
be high. There is some leverage there. Trevor Burrus: Should people buy bitcoin with
that in mind? Brock Cusick: Whether they could attack the
networks directly really depends on how widely they are adopted. Trevor Burrus: With that in mind and other
concerns and also positives that you’ve raised, should people buy bitcoin? Do you think it’s a good investment? Brock Cusick: [00:53:30] It’s a gamble. If it becomes the global reserved currency
for all computing resources, it’s worth a hell of a lot more than $6,000. If it gets banned by all the world’s governments
it’s worth nothing. Aaron Powell: With those potential threats
in mind and with the interesting innovations that we’ve talked about, the sidechains and
the level two and the way that this can then [00:54:00] interface with other things in
our lives, what do you think the next say five to 10 years in this space look like? Brock Cusick: Next five to 10 years are still
building out. I think what’s really fascinating, it’s hard
for me to guess. I predicted that a recent upgrade to the bitcoin
network called Segregated Witnesses, just got adopted this summer. I thought it was gonna get adopted two years
ago. I even made a bet with a friend of mine for
[00:54:30] 50 bucks that the SegWit would be adopted and that micro-transactions would
be a thing like a big thing by last October. Obviously I was completely wrong. I completely missed how big a faction within
the bitcoin network was resistant to SegWit being adopted. Without those protocol upgrades, the seven
transaction per second limit [00:55:00] is gonna prevent bitcoin from ever being widespread
adopted. It could end up being a replacement for gold,
but all of the wild eyed fantasies of libertopia will not come true because the process throughput
is too limited. On the other hand, SegWit’s been adopted now,
the lightning network finally has everything it needs to go and I’ve seen indications that
the technology needed to support [00:55:30] sidechains should be available within the
next six months. I haven’t seen any organized resistance to
it. If those two things actually go through and
we actually have lightning network in sidechains by say March, the next five to 10 years are
gonna be prototyping out the technologies, the consumer facing technologies on the level
two protocols and on the sidechains that really add functionality and throughput [00:56:00]
to bitcoin and really allow it, give it the growth head groom it needs to actually become
a global currency. Aaron Powell: Thanks for listening. This episode of Free Thoughts was produced
by Tess Terrible and Evan Banks. To learn more visit us at www.libertarianism.org.

, , , , , , , , , , , , , , , , , , , ,

Post navigation

Leave a Reply

Your email address will not be published. Required fields are marked *