Nick Szabo: If banks want benefits of blockchains they must go permissionless

Legendary cryptographer Nick Szabo says the only way to realise seamless, automated, and global financial integrity is if banks embrace permissionless blockchains, like the system Bitcoin runs on.

If banks were to embrace permissionless blockchains they could participate as well or better than the newcomers, Szabo told IBTimes UK.

"But their bureaucracies are so heavily invested in the expertise and importance of local regulations and standards that it's extremely difficult for them to cut the Gordian knot and implement seamless global systems.

"So they keep trying to re-inject points of control, and thus points of vulnerability, into blockchains, e.g. through 'permissioning'; but this nullifies their main benefits, which come from removing points of vulnerability."

Szabo reiterated the point: to remove vulnerability banks also have to remove individual human control and the individuals in charge or with root access. Banks naturally hate that loss to their power. But they don't have any choice if they want to gain the benefits of having an army of independent computers that rigorously, constantly and securely check each others' work, he said.

Szabo pointed out that proper financial controls are already somewhat decentralised, thanks to a "human blockchain" of accountants, auditors, etc. checking each others' work.

"There are half a dozen or so different entities involved when you do a stock trade on Wall Street, again checking each others' work. You typically see full vulnerability to unaccountable third parties only in pathological situations where naive newcomers try to do money on the centralised web, as with Mt.Gox.

"But the traditional 'human blockchain' is very labor-intensive and very local — each is based on local regulations and customs and thus splits the world up into mutually untrusting national silos. Permissionless blockchains cut through that like Alexander cutting through the Gordian knot."

Smart contracts

Szabo has a deep understanding of the way legal contracts and auditing systems work and has applied computing logic to these areas, famously coining the term "smart contracts" back in the 1990's.

IBTimes asked Szabo if he expected smart contract execution and other decentralised apps on censorship-resistant systems like Ethereum to challenge financial and legal institutions with a ferocity similar to that of Bitcoin.

"Yes, eventually moreso, since Ethereum's more flexible and general language can facilitate a much wider variety of commercial and other formal relationships.

"However, the Ethereum community needs to learn to reverse-engineer legal and other traditional patterns instead of trying to re-invent society from scratch. That will take some time to figure out how to do well. It will also take some time for the Ethereum platform to technologically mature, as it has with Bitcoin."

Financial contracts that have clearly-defined rules and can be collateralised are generally viewed as low-hanging fruit within the brave new world of smart contracts. Recently, UBS came out of the bushes with a smart bonds platform built on Ethereum.

Szabo said bonds are good example of a financial instrument suited to smart contracts because they could be backed by other assets that exist on the blockchain, if the bond issuer itself fails to make a payment.

"If it is secured by collateral that can be controlled digitally on or from the blockchain, that will be the most trust-minimised and most automated kind of interaction.

"These can be performed predominantly by 'dry code' (interpreted by computers rather than lawyers) and thus automated and trust-minimised (vulnerability-minimised) for seamless operations across borders.

"At the other end of the contractual spectrum, there are a wide variety of contracts where performances are inherently human in nature or where humans have tacit knowledge that isn't readily communicated to software. For example, paying an artist to create a work of art.

"More prosaically, most construction contracts involve quality aspects that probably won't anytime soon be specified well enough for a robot to be able to judge whether the work has been satisfactorily completed. This kind of thing will long remain the domain of 'wet code' — traditional law interpreted by lawyers rather than computers."

Block size debate

Bitcoin transaction volumes are expected to increase in the years ahead and the debate over how and when to increase headroom (size of blocks are currently limited to 1MB) was in the headlines recently.

Szabo favours Blockstream co-founder Pieter Wuille's proposal, a security-conscious approach using limiting-resource figures and scenarios which argues for about a 20% or less per year increase in block size.

Szabo provided the analogy of a "box size debate" over the number of dollar bills that could be fitted into an armoured car. One side of the debate says we should try to carry as many pieces of paper in the truck as possible in order to allow it to handle small denominations. They want to turn the armoured car into a big moving van, said Szabo.

The other side argues that truck should be smaller and heavily armoured. You can move more dollar bills in the moving van, much like Visa can do many more transactions per second than Bitcoin. To do that you have to remove some armour — you have to reduce the security.

"You'd have to hire more security guards and money-counters instead, and you'd have to depend more on the local laws to protect you. With block size too, as with the 'box size' of our money-carrying truck, there is a security vs. performance trade-off.

"If you reduce the redundancy of messages to allow for larger block sizes, beyond the growth of limiting-resource technology, that reduces the automated integrity that makes Bitcoin distinctive in financial IT.

"It creates avenues of attack and it would end up becoming more dependent on the human blockchain of accountants, investigators and lawyers for its integrity. It would be more labour-intensive and nationally siloed like traditional IT systems."

Visa and PayPal have much higher transactions per second than Bitcoin (analogous to Szabo's bigger truck moving more dollar bills), but they are more labour-intensive and often difficult to use across national borders.

"If that's the tradeoff you want to make, Visa and PayPal already exist, and within national borders they do what they do quite well. It's silly to try to turn Bitcoin into yet another Visa and PayPal," he said.

"That means that large-B Bitcoin (the blockchain) will evolve into a high-value settlement system: as it grows it won't turn into a big moving van carrying many more pieces of paper (transactions per second); instead it will be an armoured car that can carry higher denominations (higher value per transaction). That means we will need other systems to do small-b bitcoin (the currency) retail payments."

Szabo said such systems already exist, such as ChangeTip for example, which offers innovative ways to combine small payments with social networks.

"It's completely reasonable, even if some Bitcoin currency purists wouldn't like it, to have credit and debit card payments denominated in Bitcoin rather than dollars, and net settled on Bitcoin instead of on Fedwire.

"And in the future we will have vulnerability-minimised ("trust-minimised" in the technical lingo) and thus global retail payment systems closely linked to the Bitcoin blockchain, such as the Lightning network."