Curious Insight

Technology, software, data science, machine learning, entrepreneurship, investing, and various other topics

The Cryptocurrency Movement

24th February 2018

I wrote the following thoughts in response to some questions from the CIO of the company I work for about mining Bitcoin. I decided to post my response (lightly edited) here because I think it summarizes my view on cryptocurrencies (and specifically Bitcoin) pretty well.

Currencies or stores of value require trust – trust that a unit of it will be recognized and accepted by others as a medium of exchange, trust that its supply is limited to prevent arbitrary devaluation, etc. All known forms of currency before 2008 relied on either centralization (fiat currencies) or physical scarcity (gold, commodities) to establish trust.

Bitcoin, and cryptocurrencies more generally, attempt to do something that has never been possible before – how do you create trust in a decentralized, digital system with no top-down control or ownership, in an environment where bits can be copied or manipulated at zero cost?

Decentralization has a lot of benefits, if you can pull it off. Under the right conditions, it enables humans to organize and collaborate in fundamentally new ways. In the long run, it may even disrupt political and social institutions by replacing networks with markets. But to do that, one first needs to solve the trust problem.

Bitcoin’s answer to this problem is “proof of work” – an algorithm for creating distributed trustless consensus. It gets around the double-spend problem (inconsistent ledgers) while also incentivizing validation of transactions on the ledger by using cryptography to require increasingly harder mathematical “puzzles” be solved to confirm a transaction.

Why increasingly harder? To prevent malicious actors from manipulating the ledger. Imposing a scaling cost on adding to the ledger makes it intractable to “re-write” large portions of the ledger. In would require compute power almost as big as the entire network.

In order to incentivize network participants to keep doing these “puzzles”, completing a puzzle rewards a small amount of Bitcoin. The reward itself has no intrinsic value, but belief in the network assigns it a value by the market. As the network expands, the “conventional” value of the reward increases, leading to more mining participation to keep up with the demands of the network in a self-reinforcing feedback loop.

However, there are problems with this model. Proof of work comes with a societal cost via consumption of other scarce resources (electricity). Since fiat money can buy compute power, and thus voting power in the network, it can lead to a de-facto centralization. The blockchain community is well aware of these limitations and a lot of time and effort is being devoted to solving them. Ethereum, for example, is planning to implement a different verification algorithm called proof of stake that theoretically eliminates these downsides. Bitcoin could follow suit eventually, or end up with an entirely different solution.

Could Bitcoin get much bigger than it is today? Yes, absolutely. Bitcoin’s market cap is proportional to the number of believers in the network. And compared to traditional financial markets, it’s tiny. All Bitcoin combined is worth less than $200 billion. By comparison, the worldwide value of gold is ~$8 trillion. Equity markets are \$100 trillion. Currency markets are bigger still. There are lots of good reasons why Bitcoin probably won’t ever get that big, but it might.

Given its potential size, does it make sense to try your hand at mining? Probably not. From an economics standpoint, the market is highly efficient. Participation has been commoditized thanks to easy access to specialized mining hardware. No barrier to entry, hence no moat. If the goal is simply understanding rather than financial gain, there’s nothing one could learn from mining that couldn’t be learned independently. Everything there is to know about how this stuff works is freely available online. The code is even open-source.

Many in the tech world view cryptonetworks today as analogous to the early stages of the internet. The implication, of course, is that the technology will be every bit as impactful as the internet has been, but it may take a while to see it materialize. They may well be right, but it's important to emphasize that it's still really early in the cycle. The internet went through a funding nuclear winter before it took off, and the same could still happen to crypto. The possibilities are exciting for sure, but personally I'm trying to temper short-term hype or extrapolations of value and take a long-term view.