21 Inc. is Worse than Vox Thought

Recently I have seen several headlines surrounding a new tech firm 21 Inc. that has raised a considerable amount of capital. They specialize in bitcoin technology and are particularly focused on linking up the bitcoin protocol with networked home devices that aren’t personal computers, otherwise known as the internet of things.

Bitcoin is pretty new, so if you’d like a brief and probably somewhat incorrect explainer, I added one below or you can read about how it works here.

21 Inc. has a pretty interesting idea. They are going to develop small chips, designed for existing networked devices around the home, that can mine bitcoins when the device isn’t doing anything else. So, you have a coffee machine that has a little computer inside, that can be programmed to start brewing some coffee as soon as your bedroom light turns on in the morning. The little computer has a 21 Inc. chip installed which can mine bitcoins on the network all day, or maybe just when the chip isn’t trying to make a cup of coffee.

This is a cool idea! It’s not a powerful chip, so it can’t mine very fast, but it doesn’t require the owner to do anything so that’s fine. The owner just gets an email once a month notifying him that all of his appliances mined, say, 75 cents worth of bitcoins that week. It could even open the door to an arrangement where an appliance manufacturer gives you the coffee maker for free, and the bitcoins get sent to them instead of you! That’s legitimately awesome.

However, there is a big problem with this arrangement. As I noted in an earlier post, the bitcoin protocol is designed in such a way that mining a bitcoin is a non-trivial task, just like getting gold out of the ground is non-trivial. In bitcoin, this is accomplished by ensuring that it takes a large amount of computing power to verify and update the distributed bitcoin ledger and thereby mine a bitcoin. Computing power requires electricity, and electricity costs money. Timothy B. Lee at Vox notes that this is a pretty big flaw in 21 Inc.’s value to the customer.

The problem, as Lee sees it, is that bitcoins mined by 21 Inc. chips would use up a hard-to-measure amount of electricity, which would eat up a big chunk of the benefit.

The problem is that you wouldn’t know, and there would be no easy way to find out. It’s not like your power bill will have an itemized list showing how much electricity each device is consuming. You’d be forced to start measuring the power consumption of the devices around your house — exactly the kind of accounting hassle 21’s technology was supposed to eliminate.

Lee also points out that this would get worse over time, as larger mining operations upgraded their hardware to mine faster.

The 21 approach has another big problem, too: Bitcoin mining hardware tends to become obsolete quickly. All of the Bitcoin mining chips in the world compete to win a fixed pool of bitcoins. As new, more efficient chips come on the market, the number of bitcoins won by older, slower chips will go down over time.

When you buy a shiny new 21-powered device, it might only cost 70 cents’ worth of power to generate $1 worth of bitcoins. After a year, it might take $1.20 worth of power to generate $1 worth of bit coins. A year after that it might cost $2 to generate the same $1.

But this is where I think Lee underestimates the problem with 21 Inc.’s offer to the customer. On average, 21 Inc. chips would be likely to lose money right from the beginning.

All freely traded commodities like Bitcoin fluctuate in value, but in most cases the value tends to gravitate towards the production cost of the material. For example, in an economy where the demand for steel is relatively stable and the landscape of producers is competitive, the price of steel should eventually be pulled towards the cost of production for the most efficient steel producers, plus an acceptable profit margin. If the price is far above the production cost, the efficient producers will add capacity or less efficient marginal producers will enter the market. When the new capacity comes online, there will be an increase in supply and the price will get pulled down.

Conversely, if the price is too low, the least efficient producers will exit the market, reducing supply and putting positive pressure on the price.

I think the bitcoin economy eventually will settle to a market equilibrium like this (just replace steel producers with bitcoin miners), and then 21 Inc. chips end up as the highest cost production in the supply system. And they are a high-cost producer that never exits the market because the user isn’t going to pull the plug on the coffee maker when the price of a bitcoin goes down. So you will have a bitcoin price that is being pulled down to a marginally profitable level for the large specialized mining systems, which leaves the household chips at an equilibrium where the mined bitcoin value is less than the electricity cost on average.

The bitcoin economy is pretty wild right now, with prices all over the place, but I think it can go one of two ways from here. Either the bitcoin monetary system matures and users start using bitcoins primarily as a medium of exchange and not as a speculative instrument, or everyone gets sick of the wild price swings and the fad fades. If it’s the latter, the 21 Inc. chip is worthless. If it’s the former, the demand side of the equation stabilizes, the long run bitcoin price gravitates toward production costs, and the 21 Inc. customer is stuck with a chip that costs them more than they earn on the first day they use it, and it just goes downhill from there!