Bitcoin is Doomed

The Bitcoin network is secured by mining, which consumes electricity. That means there is a real-world cost to keep the miners running, which has to be paid for somehow. Currently, the Bitcoin network “prints” new coins in order to pay miners, in the form of a reward in every block. This has worked fine for the first decade, but since the mining reward is set to gradually decrease and eventually stop at 21 million coins, we’ll eventually have to find another way to pay for the network’s security.

People typically respond to this issue with two strategies: fees or inflation.

Fees: Since every Bitcoin transaction pays a fee to the miner who puts it in a block, we can imagine this stream of revenue eventually growing to sustain miners’ power costs. However, this means either the cost to send a transaction would go way up (a “fee market”, limiting who can transact) or the blockchain would grow to include tons of data (“big blocks”, limiting who can run a node in the network). Fee revenue also isn’t guaranteed as there are emerging technologies which reduce the demand for on-chain transactions, e.g. the Lightning Network or sidechains.

Inflation: Bitcoin could be hard-forked to have new rules which would throw out the existing 21 million coin limit and continue to mint coins for miners, say, an additional 1% per year. This would be tricky since hard forks have been shown to be difficult to coordinate in Bitcoin, and it weakens trust in the currency since the whole point is that the rules are immutable and there is a known, limited supply.

Since the default state is to continue running as-is, what will happen? Will miners eventually turn off their machines since they can’t make a profit, lowering the security of the network? That would open up the chain to be 51%-attacked, causing BTC-holders to exit the unsafe currency, lowering the profits of miners even more, and repeat, in a so-called “death spiral” that could kill Bitcoin sometime in the next decade or two.

A Solution

There is a third way to pay for the mining costs. It would let people mine with no extra cost, and it’s distributed across the world population. Mining creates a byproduct, one that every human needs and already pays for: heat.

If you mine to heat your house, the hashpower is free. The cost of providing a secure hashrate can be subsidized by the world’s existing demand for heating. If every home which was already paying for electric heating was generating that heat with mining rigs instead of simple heating elements, not only would they pay a little less to heat their house, the Bitcoin network would stay secure. Amazingly, this would not even increase anyone’s power bill assuming they already use electric heating (if gas heating is cheaper for you, or you live in a warm place and don’t need to heat your home, this doesn’t apply).

Residential heating scales the Bitcoin network far enough to be secure: Bitcoin currently consumes an estimated 51 terawatt-hours of electricity per year through mining and provides substantial security, whereas electric heating of homes in the US consumes an estimated 207 terawatt-hours. This 207 Twh figure only covers space heating in the US (a mere 4% of the world population), so extrapolating for the rest of the developed world, plus all commercial and industrial demand for heating, we can see that this source of “free energy” is enough to keep Bitcoin going and then some.

Even though it won’t be profitable to mine, you get the heat you were already planning on paying for at the same cost, so it makes economic sense to keep hashing away.

Naturally, people meet this crazy idea with some skepticism so let me address the common criticisms I’ve heard when talking about this:

This covers the marginal costs of mining, but we still have to worry about the capital costs of buying hardware. You might not break even if you buy mining rigs for your home.

I believe the hardware costs can be viable after considering a few factors: 1) Moore’s Law is ending so chips will only get cheaper, and 2) the cost of the hardware weighed against the cost of building a home is negligible.

  1. State-of-the-art mining ASICs made by Bitmain use 7nm transistors. That’s really small. Silicon atoms are only 0.2nm. While we’re not quite maxed out yet, the amazing manufacturing technology of the human race has come many orders of magnitude and is now approaching the physical limits of our universe. As it gets exponentially more expensive to develop processes for smaller, more efficient transistors, chip manufacturers will catch up to the leader, leading to more competition and lower costs (not to mention a smaller lead in the performance of the state-of-the-art vs. last year’s model). This end of Moore’s Law will likely affect the tech world in many other significant ways, but that’s another story.
  2. A single Bitmain Antminer S9 unit consumes ~1300W, which happens to be roughly the same amount of power for a typical space heater. At the time of writing, these units sell on the order of $300 each, which means today the cost of hardware to produce enough heat for the typical home is in the hundreds or low-thousands of dollars, not too bad of an additional cost when building a new home. Also, future heating systems will probably already be internet-connected, and Bitcoin nodes can run on a $30 Raspberry Pi, so the rest of the mining system is also cheap.

You’ll end up consuming more electricity than if you were only heating since chips aren’t 100% efficient.

This actually isn’t true due to the law of conservation of energy. Usually, when we measure something as not being 100% power efficient, it is because some of the power is lost as heat. When you’re actually trying to generate heat, you get 100% efficiency. In other words, a 1300W mining rig generates 1300W of heat.

There may be some tiny amount lost as electromagnetic radiation (which would eventually become heat) instead of direct thermal energy, but this is likely so small that it’s even hard to measure.

Mining rigs are LOUD! Nobody would want that in their home.

Many modern homes are built with central heating, so you’ll only have to hear the noise in your garage. And if that’s too loud, they even make soundproof boxes for indoor generators which would probably do the trick.

For everyone else, it’s not too different from a space heater with a fan. While a mining rig might be louder today, manufacturers aren’t designing for the in-home use case, and they can likely build quieter models in the future.

If everyone can mine for free, doesn’t this break the security model of Bitcoin and open it up to attacks?

Everyone can mine for free (plus the one-time hardware costs) - but only up to a finite limit: their natural need for heating. This is nice because this demand is distributed evenly across all humans (well, humans who live in colder places) - it doesn’t matter if you’re rich or poor, you probably consume the same amount of power for heating.

Every watt-hour you spend on mining past what you actually needed for heat is just wasted.

While searching around to see if anyone else had discussed this idea, I found that Satoshi called it way back in 2010: Bitcointalk screenshot

A huge benefit of mining-as-heating is that it nullifies one of the biggest criticisms of Bitcoin: that it is “wasteful” or “bad for the environment”. Another benefit is that the hashpower becomes very physically distributed, unlike today where much of the hashrate lives in a few big datacenters (which could be destroyed or taken by force).

I’m not sure how exactly this will come into the world, but I believe there are incentives for companies to produce heating units for the heating use-case, for instance a consumer in-home mining company like Coinmine. Depending on who captures the value (the manufacturer of the heater, or the owner of the building), they can either be marketed as “a cheaper heating unit”, “a heating unit which costs less to operate”, or “a heating unit which sends some BTC to your wallet as you use it”. The latter seems ideal since it also helps get BTC into some people’s pockets, which is a great way to accelerate adoption of the currency.

I think Bitcoin will be just fine for the rest of our lifetimes.

Thank you to Judd Keppel for reading the first draft, and everyone I talked to who brought up good points.

BTW, I’m Matt Bell (@mappum). I’m a developer working on Bitcoin proof-of-stake sidechains built on Cosmos, at a company called Nomic. This is my first blog post, after many years of saying “Hey I’m going to write a blog post” and never actually doing it.