Rethinking the Bitcoin PoW Mechanism: Does it Really Waste Energy?

This article was published on Ethereum Enthusiasts, original title: "Viewpoint | PoW is Efficient", author: Dan Held, translated by: A Jian.

Many people believe that Bitcoin's proof-of-work mechanism (PoW) "wastes electricity." I thought I would explain to everyone: everything is energy, currency is also energy, the evaluation of energy is subjective, and in the end, how does the energy consumption of PoW compare to existing regulatory systems? This article mixes the original thoughts of many people in the industry—what I have done is merely gather, refine, and combine these statements.

Work is Energy

"Work" is treated as a unit of energy, starting with the French mathematician Gaspard-Gustave de Coriolis: he defined energy as "work done."

Long ago, the work done in economic activities came entirely from human power, and this energy came from food.

About a million years ago, humans accidentally learned to use fire. As a result, the energy that humans could harness increased because we could not only maintain temperature by eating but also warm ourselves by the fire. Thus, this additional energy consumption improved our standard of living.

Thousands of years ago, our energy consumption further increased as we began to domesticate livestock. Animals could replace human labor, provided that these laborers also had food. We needed a large amount of food to meet this energy demand, but our lives became more prosperous as a result.

In the past few hundred years, we invented large machinery that could do work by itself. Initially, energy came from water flow and wind power, later becoming cheaper sources like coal and natural gas, and now nuclear energy (nuclear fusion / nuclear fission). Whether mechanical or biological, they all rely on consuming energy to do work.

Our economic life is not built on currency but on work and energy. Everything in our lives is closely related to the price of energy. Purifying water requires energy; transporting goods also requires energy; manufacturing products requires energy; cooking is no exception; even refrigerators and freezers require energy. In a free market, the unit cost of every commodity reflects the energy consumed to produce that unit of product. Because the free market encourages people to produce goods that are both high quality and low cost, the energy consumption of every commodity in production is minimized. Currency represents the work needed to produce goods and services, and can therefore also be seen as stored energy.

In the first half of the 20th century, industrial leaders like Henry Ford and Thomas Edison aspired to use "energy units" or "energy currency" (commodity currency / energy currency) to replace gold and the dollar. At that time, this concept was popular because it also possessed some characteristics of sound currency, including: being precisely definable as a statistical unit, easy to measure / hard to counterfeit, divisible into smaller units, and homogeneity (meaning each unit is indistinguishable from another). However, energy currency also has its flaws: it is difficult to transfer and store.

"To make men and women covet something, just make that thing hard to obtain."

------ Mark Twain

Fast forward to October 31, 2008, when Satoshi Nakamoto publicly published the Bitcoin white paper. The original intention of the proof-of-work used by Bitcoin (translator's note: it can also be translated as "work proof") was as a means to resist email attacks. However, Satoshi later applied it to digital currency. PoW mining, in short, is the use of specialized machines (such as ASICs) to convert electrical energy (via block rewards) into Bitcoin. The machines repeatedly perform hash calculations (both guessing and voting) until they solve a cryptographic puzzle and receive Bitcoin (block rewards). This cryptographic puzzle proves that the miner has expended a significant amount of energy (in the form of ASICs and electricity), demonstrating that the miner has done work. Bitcoin has a voting mechanism similar to capitalism, "taking risks with real money to gain voting rights," consuming energy and using ASICs to produce hash values (votes). ------ Hugo Nguyen

When Satoshi designed PoW, he fundamentally changed the mechanism of human consensus formation, transforming political voting into non-political voting (hash values), relying on the conversion of energy. Proof of work is also proof of consumption, or verification that energy has been consumed. So why is it important? Because it is the simplest and fairest way to verify physical world entities in the digital realm. PoW is related to physics, not code. Bitcoin is a super commodity, forged from energy, and is the fundamental commodity of the entire universe. PoW melts electrical energy into digital gold.

Bitcoin's ledger is only immutable when it is difficult to manufacture. Proof of work is "very expensive," and that's true, but it is a feature, not a flaw. To this day, the term "defend" still means building a thick physical wall around something valuable. The new world of cryptographic currency is less intuitive, with no tangible walls to protect our money, nor doors and locks. Bitcoin's public ledger is protected by the hash power it gathers, which is the total energy it spends on building the wall. This design is expensive and transparent, meaning that to dismantle this wall, one must expend the same amount of energy (an unforgeable luxury waste).

Translator's note: This statement is actually not rigorous. Dismantling the PoW wall only requires the same number of hashes, not necessarily the same amount of energy, because the computational efficiency of machines (the number of hashes produced per unit of energy input) will continue to improve.

Energy Consumption

Critics of cryptographic currency always say: Bitcoin's PoW is garbage, and in a few years (like 2020), the world will be destroyed by Bitcoin! You may have noticed that most of the "apocalyptic" articles are based on an analysis by Alex De Vries. This person is a "financial economist and blockchain expert," working for PwC Netherlands and is the author of the Digiconomist website. His estimates have received much fair criticism, mainly due to his poor energy consumption calculation formula. The key metric he chose, "average electricity consumption per transaction," is also intentionally misleading to readers for the following reasons:

• Energy consumption is measured in terms of blocks, and the number of transactions within a block is uncertain. Packing more transactions does not necessarily mean consuming more energy.

• The economic density of a single Bitcoin transaction is also continuously increasing (with batch transactions, Segwit technology, Lightning Network, etc.). As Bitcoin increasingly becomes a settlement network, the economic value protected by each unit of energy is also growing exponentially.

• The average energy consumption per transaction is insufficient to measure the efficiency of Bitcoin's PoW mechanism; it should be defined based on the security of economic history. Energy consumption protects the Bitcoin that has already been mined, and as the inflation rate decreases, this ratio will become lower (translator's note: it is uncertain what "ratio" refers to here; if it refers to the ratio of the number of mined Bitcoins to its theoretical maximum, it should be continuously increasing). One Bitcoin "accumulates" all the energy spent during the mining of all blocks. A researcher, LaurentMT, has empirically demonstrated that Bitcoin's PoW will become increasingly efficient because the increased costs will be offset by a greater increase in the total value protected by the system.

Now we know what the correct KPI for the ROI (Return on Investment) of energy consumption should be. Let's take a look at the trend of energy costs for Bitcoin's PoW.

The growth rate of ASIC efficiency is slowing down. Because the efficiency growth is slow, competition among manufacturers will become more intense as profits decrease.

UCSD.edu

bloomberg.com

The total cost of mining will continue to shift from the upfront acquisition costs of ASIC devices (capital expenditure) to ongoing energy expenditures (operating costs). Because the physical location of mining centers is irrelevant to the Bitcoin network (mining farms can migrate), miners will move to areas where they can produce additional electricity at the lowest marginal cost. In the long term, this could create a more efficient global energy market, as Bitcoin miners will arbitrage electricity prices worldwide. The cost of Bitcoin mining will become the minimum value of excess electricity. This also holds promise for solving a problem with renewable energy (like hydrogen and biogas): power is predictable, and if not used immediately, it will be wasted. In the future, Bitcoin mining can help renewable energy sources with variable output power—energy producers can turn on mining machines on demand to convert excess power into Bitcoin.

Aluminum was previously a common means for countries with surplus renewable energy (like Iceland) to "export" electricity. Smelting bauxite (the ore of aluminum) requires enormous energy, and this conversion is one-way (isn't it similar to a hash function?). Concerns about the "anomalous" energy consumption surrounding aluminum have persisted for 40 years, starting in 1979 (including concerns about centralization). But all such enterprises are constantly seeking cheap energy and favorable policies on this planet. Because the aluminum manufacturing industry has developed for decades, the amount of electricity consumed per kilogram of aluminum has also been continuously decreasing.

energy.gov

"This global energy network liberates idle assets and allows new assets to emerge. Imagine a 3D-printed world map where areas with lower energy prices are lower in elevation, and areas with expensive energy are higher. To me, Bitcoin mining is like pouring a glass of water on the surface of this map; the water will flow into the valleys and crevices until it reaches a level everywhere."

------ Nic Carter

Bitcoin is the last buyer of all electricity, creating a platform that incentivizes the development of new utilization schemes around underutilized energy; these energies would not be developed without Bitcoin.

"When will the energy used to produce PoW stop growing? Specifically, that will be when enough energy producers start directly producing PoW, and the marginal return of each kilowatt-hour used to run PoW equals the marginal profit of selling a kilowatt-hour to the grid—when the 'premium' of PoW drops to 0. I call this equilibrium point the 'Satoshi' point. I estimate that when the electricity consumption of PoW accounts for 1-10% of the world's total, this equilibrium point will be reached."

------ Dhruv Bansal

Some people complain that Bitcoin mining doesn't "do anything useful," like searching for prime numbers. While adding secondary reward projects for workers sounds like a good idea, it introduces a security risk. Splitting rewards could lead to a situation where "doing work for secondary functions yields greater value than doing work for primary functions" (Dergigi). Even if the secondary function is harmless (like generating heat), the expectation of receiving not only $100 but also an additional $5 (because you provided heat) for every X hashes means this "mining furnace" merely increases hardware efficiency, leading to increased block difficulty and energy consumption per unit block. Fortunately, Bitcoin will never encounter this problem because its security is protected by a pure proof-of-work algorithm.

Note: Bitcoin is already providing very useful things to society (if it weren't, mining wouldn't be profitable), and expecting miners to mine purely out of altruism is also irrational.

Relative Costs

Whatever you do, energy must be consumed (the first law of thermodynamics). To say that a certain use of energy is more or less wasteful is entirely presumptuous, as all users must pay the same price to use electricity.

"As long as people find uses worth paying for electricity, electricity is not wasted. Those who pay for electricity receive Bitcoin in return." ------ Saifedean Ammous

From a thermodynamic perspective, the entire universe is an ultimate closed system. The additional electricity consumption caused by Bitcoin is much less than that consumed by existing fiat currency systems; fiat currency systems not only require banking infrastructure but also violence and political machinery. Utilizing electricity to protect the skeleton of the financial system is a "positive-sum" result. Below, I roughly compare the existing financial, military, and political systems (notes at the bottom of the article).

The chart is inspired by @hassmccook and his original article: "Economic and Environmental Costs of Bitcoin Mining."

Type I Civilization

In the pursuit of cheap energy, we will create greater prosperity for the world. Bitcoin, by utilizing these new and entirely different energy sources, can not only bring us to the Kardashev Type I economic stage but can even bring us closer to a Kardashev Type I energy civilization (we have already achieved about 0.72 of the Kardashev index) (translator's note: Kardashev was a former Soviet astronomer who proposed evaluating the advancement of civilization based on the amount of energy it can utilize. Type I civilization can utilize all available energy on its home planet). With Bitcoin mining as an incentive, the time required to achieve Type I civilization could shrink from 200 years to just a few decades. Once we reach the Type I stage, there will be no need to constrain the growth of energy consumption, and everyone will enjoy a higher standard of living.

The pressure to find cheap electricity will drive people to build nuclear fusion reactors. Nature has pointed the way for us; planets derive their energy from the nuclear fusion of stars. Humanity is already on the path to building nuclear fusion reactors to mimic nature, and it is expected to take decades and $80 billion in research funding to unlock fusion technology. The fuel for fusion (mainly deuterium) is abundantly available in the oceans and can meet the Earth's energy needs for millions of years. Fusion has many characteristics of renewable energy, such as being a long-term energy supply and not causing greenhouse effects or air pollution. Fusion can provide very high power density and uninterrupted power transmission. On the other hand, the production costs of fusion will not be constrained by economies of scale. Hydropower and wind energy must gradually be laid out in less than ideal environments as the best locations are developed. But even with the mass production of fusion stations, production costs will not increase significantly because the raw materials (seawater) are very abundant and available everywhere.

"Water, water, everywhere, but not a drop to drink."

------ Samuel Taylor Coleridge

Nuclear energy and other cheap energy sources will solve many humanitarian issues, such as the shortage of clean water. We are surrounded by seawater, but desalinating seawater requires enormous energy. The cost of desalinating seawater is now higher than using freshwater, groundwater, water recycling, and water-saving technologies.

Humanity will not stop exploring, from high mountains to the ocean depths, from the minuscule to the vastness of space-time; to grow, we must not be bound by energy. We must reach for the stars.

A system that can settle $13.8 trillion in payments annually and provide all of us with cheaper energy is surely worth spending $4.5 billion on mining, right? I believe the answer is yes.