S5 E8: Eric Voskuil on Libbitcoin Development, Cryptoeconomics Book & Austrian Economics

Eric Voskuil is the kind of Bitcoin developer whose understanding transcends computer science. Not only that he is a C++ wizard, but he is also an avid reader of Austrian economics and often references works by Mises and Rothbard.

As a matter of fact, Voskuil (whose correct pronunciation is “foss-kuyl”) refers to economics as an “axiomatic system” – a hard science that can be studied in a framework which works regardless of human subjectivity and behaviorism (more about this view at 33:25).

As he explains in this interview, Voskuil considers that Bitcoin transaction fees are an essential part of the network’s censorship resistance and decentralization. The increasing cost is not an issue, but a fundamental component which ensures financial sustainability and therefore base layer security.

Voskuil is best known in the Bitcoin space for the contributions that he has made to the Libbitcoin project. As he explains, Libbitcoin isn’t just another clone or port, but an entire rewriting of the Bitcoin Core software to bring certain optimizations. The software currently doesn’t have a graphical interface, but is very useful to miners that seek to maximize storage efficiency.

As he mentions during the interview around the 6:30 mark, one of his breakthroughs with Libbitcoin involves moving files that otherwise get stored in the limited RAM to the hard drive. This optimization saves operation costs, since the latter type of memory is more inexpensive.

“I think [that] not being a professional programmer or software engineer, Satoshi fell prey to one of the patterns which is common in software development: premature optimization” – Eric Voskuil

Yet Voskuil combines his passion for writing code with an affinity for Austrian economics. In conversation, he successfully blurs the lines between software engineering and market dynamics. And it was perhaps this inter-disciplinary approach of his that led to the publishing of “Cryptoeconomics“.

Interestingly, the book is a collection of unrelated articles that Voskuil has been publishing in the Libbitcoin-System GitHub repository for years. But it all came together and started to look more cohesive in recent months and thanks to contributions from people such as James Chang, a printed version is ready to ship whenever the pandemic ends.

Unlike Saifedean’s “The Bitcoin Standard”, “Cryptoeconomics” doesn’t have a central narrative or interlink phenomena to establish an ideological standpoint. Instead, it features a rational and ideologically-detached approach that some people may find blasphemous.

Furthermore, “Cryptoeconomics” is already available for free on GitHub. Therefore, purchasing the book is only a way to support the author and keep a physical copy that withstands any possible revisions.

Listen to Eric Voskuil on iTunes and Spotify!

Time stamps:

00:47 – Intro

02:00 – For how long has Eric Voskuil worked on Libbitcoin?

04:00 – Why are Bitcoin repository ports created?

6:30 – Breakthroughs made by Libbitcoin

12:10 – Eric Voskuil’s more advanced activity in Bitcoin

15:43 – Performance optimization

20:51 – Satoshi Nakamoto wasn’t a professional programmer

25:15 – Is Eric Voskuil a computer scientist or an economist?

28:10 – Austrian economics

29:30 – Credit and economic production

33:25 – Misean economics as an axiomatic system

38:54 – S2F (Stock to Flow) is not Austrian

42:20 – Price cannot be proven economically 

44:01 – The value proposition of Bitcoin as money

44:27 – Why store of value is subjective 

45:53 – Lunar fallacy

48:50 – Eric Voskuil’s Cryptoeconomics

50:52 – The Cryptoeconomics books

52:43 – Musing over censorship resistance for a year

54:15 – Economic rational explanation for censorship resistance

56:40 – James Chang’s contributions to the Cryptoeconomics book

59:20 – Cryptoeconomics vs The Bitcoin Standard

1:02:00 – Combining Libbitcoin and Cryptoeconomics

1:06:54 – Bitcoin took years to understand

1:09:20 – Bitcoin security

1:14:16 – Why Trezor offers the best security

1:22:37 – CB750 break pedal

1:24:20 – Is hyperbitcoinization possible?

1:26:38 – Cryptoeconomics on Amazon?

1:27:46 – Improving Bitcoin Privacy

1:29:34 – Take-aways from Vietnam conference

1:33:17 – What happens to Bitcoin if the internet gets regionalized?

1:34:49 – Why the Blockstream Satellite doesn’t solve problems

1:38:56 – The US government as a threat model

1:40:52 – Resilience vs privacy

1:42:37 – Outlawing Bitcoin development threatens devs?

1:51:44 – Bitcoin is not a gambling device

1:53:40 – Libbitcoin institute donations

Brief transcript covering the Trezor vs Ledger debate

“There were other attempts to solve that problem without spending the time, money, and effort to put a screen on the device. Those attempts were deeply flawed: they would tell you they had bank-level security or the same level of security as the Trezor, but they weren’t even close. When the first version of the Ledger came out, it had a solution to this ‘not wanting a screen’ problem where you put the USB stick in your computer and your run their software, and the software would present you the same prompt that this is the address you’re spending to. Well, of course you had to assume the software is the malware – that’s the assumption, that’s why you have the hardware because you don’t trust your computer.

So the malware is prompting you with the address, but in order to solve that security problem there was a credit-card looking thing that would come with some numbers and letters printed on it, that shipped with the device. And they were paired so that it took Base58 characters and there would be a prompt for you to enter 4 of the characters from the card. You would enter 4 characters and that would be essentially validated by the device. If you didn’t enter the right characters and you obviously didn’t have the card and the malware didn’t have the characters, it couldn’t spend your money. But if you divide 58 by 4, it comes by 14 and something. You realize that basically if you use the card a very very small number of times (like 14), then the malware can spend all your money because now it has all the mapping from the secret characters to the public ones. So that was a pretty gaping hole which you could discover from the documentation and that’s the difference between what Trezor did and what other people did.

They thought about these problems and said ‘okay, so there’s no way for us to do this without putting a screen on the device’, at least not in a cheap or reasonably easy way. There were a couple of other efforts to solve that problem by the same company later on, and they had more complicated implementations that had almost the same problem. And bootstrapping trust problem, and giving another example thing: a hardware wallet vulnerability where the secure element was used and this secure element is basically what you’re seeing in credit cards now.

This tamper-resistant thing in the card which basically has your private key in it, you put it in a device and that’s how the device knows that it’s actually your card. And the cost at the time, when I went to a hardware hacking cost in Portland and talked to some people over there who have experienced the same issue. So what was the cost to actually shave this card down, get the secure element and read the secret off the element? They said ‘well, it’s not entirely guaranteed you might destroy it once you try, but you might get it done for about a hundred thousand dollars’. I said ‘well, that’s not cheap’, but prices have come down quite a bit since then. I discovered it could be done in Seattle, where I contemplated going on and doing it at some point, but the interesting thing about it is that it’s fairly reasonable to assume that nobody’s gonna steal your card and spend $100.000 to shave it down and get $1000, right?

Maybe even 100 thousand because it might not work, right? But what if you put the same secret on every card? That’s what they did. So advertising bank-level security where every card has got a different secret so you’re only risking what you have the ability to spend with that one card… but what if everybody had the same card? That’s a big deal, right? So you just buy one of these cards, shave it down, get the secret for $100.000, and now you just put out some malware and own everybody’s bank account. That’s a big difference and that’s what I mean by not really thinking about the security architecture. In actuality, at the time there were 4 batches of secrets and it was advertised in documentation that that was the case.

So basically, by shaving one card you would own everybody’s bitcoin if you were able to put some malware on their computer. These are just things that were denied and discussed with me online and there’s a record out there somewhere of these discussions. And the company just kept doing it with another version and people ignored it… so I haven’t looked at these things in quite a while, but I never discovered anything like that with the Trezor and that’s what made me happy about it.

Donate to Bitcoin Takeover!

Domain renewal, content hosting, and local storage devices might be inexpensive, but they don’t come for free. By donating to the QR address above, you are supporting the Bitcoin Takeover project to keep on publishing content which remains stored on the internet for years to come.