Andreas Antonopoulos: The Case Against Reputation and Identity Systems

Andreas Antonopoulos makes the case against identity and reputation systems, arguing that with the invention of blockchains, multisignature transactions, and smart contracts, we can minimize default risk without reputation.

Also read: 13 Ways To Donate Bitcoin This Holiday Season

At the heart of Bitcoin is the well-known blockchain, a permanent record history cryptographically secure and provable. Its storage being decentralized grants the information it holds a high amount of trust in the minds of its users, whom in turn rely upon this record to engage in further trade and finance.

It is this characteristic of the blockchain that has brought Bitcoin developers and pioneers to lead engineer workshops like “Rebooting the web of trust,” with names like Vitalik Buterin, Peter Todd, Gregory Maxwell, and Joel Dietz as well as renown software engineers like Christopher Allen and legendary cryptographers like Jon Callas, the former CTO of PGP.

The Problems With Using Blockchain Reputation Systems

The capacity to store information about people’s reliability in terms of transactions they engage in, contracts they sign or promises they make, is what binds these two communities. Its purpose, to prevent future abuse or default risk.

However, reputation and identity systems are not only very complex both in theory and practice, but they present a grave potential danger to humanity; that of a world where every individual’s actions are judged by standards they don’t necessarily agree with and which mark them in a global, non-changeable record, turning the Web of Trust into a Web of Shame.

Furthermore, the whole purpose of a reputation system is to prevent abuse and default risk. So are we solving the wrong problem by focusing on reputation and identity systems rather than “trustless” smart contracts, decentralized escrow and multisignature tools? Andreas Antonopoulos thinks so.

In a recent interview with DisrupTek.info at Labitconf 2015, he illuminated this danger saying:

“In traditional finance we use reputation as a proxy to identifying the default risk associated with a specific identity and that is because identity is part of all financial transactions. The biggest problem I have with reputation is that it is not what we are trying to establish. Reputation does not matter; what matters is default risk.”

Reputation systems are kind of an archaic technology. It developed out of tribal communities as a way to prevent abuse and discourage theft. Tribes effectively had a kind of permanent memory because they were small enough to remember all members of their community. A capacity that is not scalable to more than about 150 people — something called Dunbar’s Number — which refers to how many people we can remember. We don’t have this on the internet.

“They are a relic of traditional financial systems because they don’t have better ways of managing default risk. Whereas in Bitcoin we do have better ways of managing default risk. Smart Contracts, Multisig, Escrow, and a whole bunch of tools. It’s very important to be able to do anonymous transactions and remove default risk, without having to attach reputation and identity, ” explained Antonopoulos.

One of the biggest problems with identity systems according to Antonopoulos is that they raise a barrier as to who can enter a financial market. As soon as you require to give a reputation score, such as a credit score to access credit and financial tools, you are excluding a vast amount of people from the market.

When it comes to something like bitcoin loans where today there is no scaled or established way to enforce payment of a debt, online reputation systems have risen to bring a higher quality of information to investors. Take for example BTCJam, the bitcoin lending platform, which you must link with social media accounts and recommendations to raise your score and have a better shot at a loan.

Even with all this information, participants still default on loans because they either innocently misinvest the funds or become the victims of fraud. When asked if investors need to accept the fact that some of their loans will never be paid, Antonopoulos said:

“I think the real issue with loans is concentration of risk, not the individual default risk.”

“If I have a loan for example – to loan out money to two thousand farmers, I use Kiva and Lending Club, in 1 investor of thousands, and my portfolio consists of thousands of loans, 2- 5 dollars per loan, so I have a default rate which is like 10%,” he continued.

“Which would be high in any other environment. I don’t care because its spread among so many investors and my portfolio is so diversified that I don’t have concentration of risk. Moreover, for all the bad performing loans I have so many more that are good.”

“So I don’t need to worry about collateralizing or reputation or any of those things. In fact, reputation and collateral force you to exclude financial inclusion.”

“In fact, one of the biggest barriers to financial inclusion is the lack of collateral and the lack of reputational metrics, so they are excluded from the financial system completely, whereas with anonymous and mass diversification of lending you can now do a lot more inclusion. Reputation to me is a huge barrier to global economic inclusion.”

This does, however, beg the question of the dreaded Sybil attack. Where a few parties create large quantities of fake accounts to maximize the amount of fraud they can inflict on a system.

Reputation or Default Risk?

In defense of reputation systems, Christopher Allen, organizer of the Rebooting the Web of Trust workshop, and Blockchain consultant with Alacrity Software says:

“I agree that avoiding default risk is important in financial transactions, but there are all kinds of other interactions that we make in life where bad actors (just following their own interests or actively subverting) cause problems. An example of a bad actor just following their own interests is the source of the tragedy of the commons and other forms of free loading. An example of a bad actor actively subverting is the troll or the malicious hacker — there was no default risk in the Sony attack, but they lost millions.

There is a quote from Noah Thorpe that I call ‘Thorp’s Law’ — ‘Any conversation about decentralization eventual evolves to a conversation about reputation’ (my corollary is to substitute devolve for the word evolve). This is substantially true — decentralization is hard, and reputation is even harder as it intersects with more complex human systems, drives and biases.”

He suggests interested parties check out the many papers and academic open source exploration of these topics on the Rebooting Web of Trust Github repo.

You can check out the full interview with Andreas Antonopoulos below: