Antifragile Cryptoeconomic Systems
This is a follow up to Cryptoeconomics 101. I’m not a cryptographer or an economist, just someone who realizes that there’s more going on here than tokens appreciating in value
Every discipline studied broadly by people today has started off niche and grown in relevance and importance as new reasons for people to care have emerged. It feels to me like cryptoeconomics is headed in that direction. Robust cryptoeconomic systems are more difficult to destroy than they are to build and that is why they have become increasingly important in the world and have the potential to be so disruptive to existing institutions.
Our physical world is generally easier to destroy than it is to build
Our physical world is much easier to destroy than it is to build. For example, in San Francisco it costs $1.1B and takes 4 years and hundreds of people to build a new high rise building, but it only costs a tiny fraction of that and a few hours for one person to destroy that high-rise with an IED.
That is just one example; another example can be found in national governments. It can take hundreds of years and the efforts of millions of people for a country to gain power in the world, but it only takes the whims of one leader to destroy that.
The concept can even be extended to traditional web companies. As a large web company today, you spend increasingly more to secure your centralized servers as you scale up your service. Global cybersecurity spending by large enterprises will reach $90B this year. Despite the exorbitant spending, we continue to regularly see security breaches that threaten to destroy traditional web companies. The successful attacks are generally executed by individuals with far less resources.
Cryptoeconomics is so important because it is about constructing systems that are actually much harder to destroy than they are to build, which could ultimately be much better for society.
What does it mean to be harder to destroy than to build?
Satoshi released Bitcoin into the world in 2009 by writing some code, publishing that code to the public and allowing anyone to use it, evaluate it and attempt to break it. Since then, millions of people have tried and no one has been able to break it. It has grown many orders of magnitude stronger since.
Difficulty of attack can be measured by a few different metrics: time, skill and money. On all three of these fronts, Bitcoin has proven to be defensible. If an attacker could break the cryptography layer of Bitcoin, he could destroy the system. However, even with access to all of the world’s computing power, breaking the cryptography layer of Bitcoin would take trillions of years.
Bitcoin is not easily susceptible to social engineering attacks because of its truly decentralized construction — no one person or group can change it. Bitcoin is the best example to date of a cryptoeconomic system that is easier to create than destroy from cryptography attacks, consensus attacks, social attacks, or any other attack one could throw at it.
Types of attacks on cryptoeconomic systems
There are three major attack vectors for cryptoeconomic systems broadly:
- ) Cryptography attacks. Cryptography is one important aspect of cryptoeconomic network security. Cryptography schemes enable secure and private information transmission. They enable secure communications while preventing an attacker from guessing the encryption keys of the schemes and unlocking the information. Cryptography is a separate discipline on its own, and if interested in learning more you should check out Dan Boneh’s Cryptography I (you can take it for free on Coursera). Most practitioners of cryptoeconomics aren’t cryptography experts themselves, but instead rely on the work of the researchers in the industry to use the best crypto schemes created. History has proven that cryptographic schemes can and will be broken, so it should be assumed that sometime in the future, the existing cryptography schemes that cryptotokens are based on will need to evolve. Developments in quantum computing may play a big role in that in the future.
- ) Consensus attacks. Cryptoeconomic systems rely on individuals who don’t know each other agreeing on truth without any trusted third party. Attacks that confuse the truth have the potential to destroy these systems. The most well known consensus attack is the 51% attack, which has long-been discussed as a potential attack vector in Bitcoin. If 51% or more of the hashing power is controlled by one party, that party could prevent transactions from being confirmed and reverse transactions sent. Right now, a 51% attack would cost billions of dollars in Bitcoin — clearly much more than it cost Satoshi to create the network. Game theory is important in thinking about consensus attacks and this is one of the big challenges of proof-of-stake mechanisms like Casper.
- ) Social engineering attacks. A social engineering attack is when a person or group of people sabotages a system via psychological manipulation rather than software engineering. This social attack can be either intentional or unintentional, but can wreak havoc on a network. Decentralization of systems is important in preventing social engineering attacks, and some argue that Bitcoin, which lacks an undistributed leader and is the most decentralized, is the most durable to social engineering attacks.
$100B+ dollar bug bounty program
People often say that cryptoeconomic systems like Bitcoin aren’t secure because there have been lots of coins stolen historically. It’s true that there have been lots of coins stolen, but that is because of poor security on behalf of individuals, not the systems themselves.
Tying economics (an incentive system) to cryptography has created a bug bounty program that now has over $100B in value at stake. Crypto tokens broadly now collectively offer the largest bug bounty program in the world by an order of magnitude, and that fact alone will continue to help them evolve and become more robust over time. The security of these systems does not require billions of dollars of investment either — it’s fundamentally built in from the start.
To be clear, there are other specific attack vectors beyond the three major vectors described above. Given the nascent nature of cryptoeconomics, it’s likely that others will emerge as well. But the conversion of economics and cryptography is worth paying close attention to in the coming years.