Why crypto is a waste of energy, full of crime, and isn’t really decentralized

Can We Mitigate Cryptocurrencies’ Externalities

I’m David Rosenthal. I worked with James Gosling on CMU’s Andrew project in the early 80s. I was a DE with him at Sun later in the 80s working on window systems including X, and file systems. I quit to be employee #4 at Nvidia where Curtis Priem and I did the basic I/O architecture, then was an early employee at Vitria, the second company of founders of Tibco. Before I start talking about cryptocurrencies, I should stress that I hold no long or short positions in cryptocurrencies, their derivatives or related companies; I am long Nvidia. Unlike most people discussing them, I am not “talking my book”.

Cryptocurrencies’ roots lie deep in the libertarian culture of Silicon Valley and the cypherpunks. Libertarianism’s attraction is based on ignoring externalities, and cryptocurrencies are no exception.

Externalities

Bitcoin is notorious for consuming as much electricity as the Netherlands, but there are around 10,000 other cryptocurrencies, most using similar infrastructure and thus also in aggregate consuming unsustainable amounts of electricity. Bitcoin alone generates as much e-waste as the Netherlands, cryptocurrencies suffer an epidemic of pump-and-dump schemes and wash trading, they enable a $5.2B/year ransomware industry, they have disrupted supply chains for GPUs, hard disks, SSDs and other chips, they have made it impossible for web services to offer free tiers, and they are responsible for a massive crime wave including fraud, theft, tax evasion, funding of rogue states such as North Korea, drug smuggling, and even as documented by Jameson Lopp’s list of physical attacks, armed robbery, kidnapping, torture, and murder.

Alecus

The attempt to force El Salvador’s population to use cryptocurrency is a fiasco. They offer no significant social benefit beyond speculation; Igor Makarov and Antoinette Schoar write:

90% of transaction volume on the Bitcoin blockchain is not tied to economically meaningful activities but is the byproduct of the Bitcoin protocol design as well as the preference of many participants for anonymity. … exchanges play a central role in the Bitcoin system. They explain 75% of real Bitcoin volume … Our results do not support the idea that the high valuation of cryptocurrencies is based on the demand from illegal transactions. Instead, they suggest that the majority of Bitcoin transactions is linked to speculation.

Makarov and Schoar, Blockchain Analysis of the Bitcoin Market

“Transaction” Rate

Bitcoin is only processing around 27K “economically meaningful” transactions/day. And 75% of those are transactions between exchanges, so only 2.5% of the “transactions” are real blockchain-based transfers involving individuals. That’s less than 5 per minute. 

Nakamoto’s motivation for Bitcoin was distrust of institutions, especially central banks. When it launched in the early stage of the Global Financial Crisis, this had resonance. The key to a system that involves less trust is decentralization.

Resilience

Why do suspension bridges have stranded cables, not solid rods? The major reason is that solid rods would fail suddenly and catastrophically, whereas stranded cables fail slowly and make alarming noises while they do. We build software systems out of solid rods; they fail abruptly and completely. Most are designed to perform their tasks as fast as possible so that when they are compromised, they perform the attacker’s tasks as fast as possible. Changing this, making systems that are resilient, ductile like copper not brittle like glass, is an extraordinarily difficult problem in software engineering. Paul Vixie pointed out that rate limits are an essential part of the solution.

I got interested in it when, burnt out after three startups all of which IPO-ed, I started work at the Stanford Library on the problem of keeping digital information safe for the long term. This work won my Stanford CS co-authors (Petros Maniatis, Mema Roussopolous, TJ Giuli, and Prof. Mary Baker) and me a “Best Paper” award at the 2003 SOSP for a decentralized consensus system using Proof-of-Work. When, five years later, Satoshi Nakamoto published the Bitcoin protocol, a cryptocurrency based on a decentralized consensus mechanism using Proof-of-Work, I was naturally interested in how it turned out.

Decentralization is a necessary but insufficient requirement for system resilience. Centralized systems have a single locus of control. Subvert it, and the system is at your mercy. It only took six years for Bitcoin to fail Nakamoto’s goal of decentralization, with one mining pool controlling more than half the mining power. In the seven years since no more than five pools have always controlled a majority of the mining power.

Economies of Scale

In 2014 I wrote Economies of Scale in Peer-to-Peer Networks, explaining the economic cause of this failure. Briefly, this is an example of the phenomenon described by W. Brian Arthur in 1994’s Increasing returns and path dependence in the economy. Information technologies have strong economies of scale, so the larger the miner the lower their costs, and thus the greater their profit, and thus the greater their market share.

“Blockchain” is, unfortunately, a term used to describe two completely different technologies, which have in common only that they both use a Merkle Tree data structure.  Permissionedblockchains have a central authority controlling which network nodes can add blocks to the chain, and are thus not decentralized, whereas permissionless blockchains such as Bitcoin’s do not; this difference is fundamental:

• Permissioned blockchains can use well-established and relatively efficient techniques such as Byzantine Fault Tolerance, and thus don’t have significant carbon footprints. These techniques ensure that each node in the network has performed the same computation on the same data to arrive at the same state for the next block in the chain. This is a consensus mechanism.
• In principle each node in a permissionless blockchain’s network can perform a different computation on different data to arrive at a different state for the next block in the chain. Which of these blocks ends up in the chain is determined by a randomized, biased election mechanism. For example, in Proof-of-Work blockchains such as Bitcoin’s a node wins election by being the first to solve a puzzle. The length of time it takes to solve the puzzle is random, but the probability of being first is biased, it is proportional to the compute power the node uses. Initially, because of network latencies, nodes may disagree as to the next block in the chain, but eventually it will become clear which block gained the most acceptance among the nodes. This is why a Bitcoin transaction should not be regarded as final until it is six blocks from the head.

Blockchain Patent Filed 1990

Discussing “blockchains” and their externalities without specifying permissionless or permissioned is meaningless, they are completely different technologies. One is 30 years old, the other is 13 years old.

Because there is no central authority controlling who can participate, decentralized consensus systems must defend against Sybil attacks, in which the attacker creates a majority of seemingly independent participants who are secretly under his control. The defense is to ensure that the reward for a successful Sybil attack is less than the cost of mounting it. Thus participation in a permissionless blockchain must be expensive, so miners must be reimbursed for their costly efforts. There is no central authority capable of collecting funds from users and distributing them to the miners in proportion to these efforts. Thus miners’ reimbursement must be generated organically by the blockchain itself; a permissionless blockchain needs a cryptocurrency to be secure.

Because miners’ OPEX and CAPEX costs cannot be paid in the blockchain’s cryptocurrency, exchanges are required to enable the rewards for mining to be converted into fiat currency to pay these costs. Someone needs to be on the other side of these sell orders. The only reason to be on the buy-side of these orders is the belief that “number go up“. Thus the exchanges need to attract speculators in order to perform their function.

Thus a permissionless blockchain requires a cryptocurrency to function, and this cryptocurrency requires speculation to function.

Why are economies of scale a fundamental problem for decentralized systems? Participation must be expensive, and so will be subject to economies of scale. They will drive the system to centralize. So the expenditure in attempting to ensure that the system is decentralized is a futile waste.

Most cryptocurrencies impose these costs, as our earlier system did, using Proof-of-Work. It was a brilliant idea when