On Cryptocurrency, Blockchain, & Cloud Computing

With Bitcoin’s price soaring, I’ve found myself spending a lot of cycles explaining why I’m still bullish on cryptocurrency and blockchain. I’ve also found myself in a number of conversations where I’m trying to convince friends that most of the talking heads from the financial world fundamentally don’t understand what blockchain is about or how it will change the game. This post is my attempt to channel those discussions and dispel several popular myths that are currently making the rounds on the Twittersphere. Along the way, I’m also going to try to convince you that the blockchain not only has the potential to revolutionize the financial world, but is also poised to have a massive impact on cloud computing. Without further ado, let’s dive into some background on both computing and blockchain.

Blockchain as a Cloud Computer

Every computer can be broken down into two fundamental components: compute and storage. When you use your computer to multiply two numbers together, your computer loads (compute) values from memory or disk (storage) into registers (storage), adds (compute) the values together, and stores (compute) the result in another register (storage) where it can then be stored (compute) in memory or on disk (storage). From web surfing to spreadsheet-crunching to gaming, all computer applications boil down to computation that manipulates stored values in interesting ways.

Cloud computing is no different. Amazon Web Services, the leader in hosted public cloud, offers tens (if not hundreds) of unique services that can all be broken down into compute and storage that runs on Amazon’s massive infrastructure footprint. For example, AWS CodeBuild allows software developers to build and test their code in the cloud and then store the built artifacts in a data store like Amazon’s Simple Storage Service (S3). Like most of its services, Amazon bases the pricing for CodeBuild and S3 on the number of minutes that the underlying virtual machine uses (compute) and the number of terabytes used per month (storage) because the company understands that most of its value-add can be decomposed into compute and storage. AWS made $4.6B in revenue in Q3 of 2017 alone which represents a massive year-over-year revenue growth of 42%, so the global market for cloud computing products is clearly vibrant and growing quickly.

The simplest way to think about a blockchain is a big, distributed cloud computer that no single person, company, or government controls. Individuals called miners connect their computers to the blockchain so that they can be used to process transactions (compute) and write the results of those transactions to a digital ledger (storage). The mechanics of executing transactions depend on the blockchain implementation and are typically heavily rooted in cryptography, but in a proof-of-work system like Bitcoin, each mining node is taking a block of transactions and hashing them (compute) together with the hash of the previous node and a value called a nonce to create a unique hash value that fits a set of constraints. Once a valid hash is mined, the new block is broadcast (compute) to other nodes and the transactions in that block are executed and stored (compute) written to both lightweight and full nodes (storage) across the blockchain network.

Mining is computationally expensive, so the blockchain is orders of magnitude less efficient than comparable distributed compute and storage solutions, but it’s crucial to note that efficiency is intentionally exchanged for a different property: no one entity owns the system, yet the system can still facilitate computational transactions that involve multiple untrusted parties without introducing a trusted intermediary. If you think about the number of transactions that we participate in on a daily basis where we incur some cost to engage a trusted intermediary, this is a big deal. Facebook can monetize your data in undesired ways, PayPal takes a healthy cut to move your money around, and the government can always compel Amazon to delete or hand over data that it’s storing on your behalf.

In order to incent miners to give their compute and storage to the blockchain, the creators of Bitcoin developed the concept of a digital coin or cryptocurrency that can be exchanged in to run transactions on the blockchain. Anyone that wants to write to the blockchain ledger has to offer a small number of coins for their transaction to be processed. The cryptocurrency cost of executing a transaction on the blockchain is linked to the demand for running compute on the network and inversely linked to the computing power connected to the network. The cost in terms of a fiat currency like the US Dollar is also obviously linked to the going exchange rate between the fiat currency and the cryptocurrency, thus the cost in USD to run a transaction on the Bitcoin blockchain has increased steadily of late: in late Q3 of 2017 the cost of writing 200 bytes to the Bitcoin blockchain ledger within 30 minutes was roughly $3-4 USD worth of BTC.

Because of this need to move currency from one party to another, the original application that was baked into Bitcoin’s blockchain was the exchange of the Bitcoin cryptocurrency between parties. Newer blockchains have built on the Bitcoin foundation and embraced the idea of more generic Smart Contracts that allow for arbitrary code to be executed directly on the blockchain. For example, the creators of the Ethereum blockchain have implemented a runtime environment on blockchain for Smart Contracts written in a language called Solidity that is Turing complete, which means that in principle, it can be used to solve any computational problem. The result is that blockchains like Ethereum look similar to cloud computing service in that they allow for arbitrary distributed compute and storage, yet they also display the interesting property of being able to facilitate computation that involves multiple untrusted actors without a single trusted third-party controlling the service. Efforts are underway to bolt this kind of behavior onto the Bitcoin blockchain via sidechains like Rootstock.

That was a lot to grok, but it’s really impossible to critique the current commentary on blockchain and cryptocurrency without at least a high-level understanding of how the pieces fit together. So with that all in mind, let’s dive into a few recent criticisms from high-profile individuals in the financial sector about both Bitcoin specifically and blockchain and cryptocurrencies in general.

“Bitcoin doesn’t have any intrinsic value…”

One extremely common narrative from people that come from the financial world recently is that coins like Bitcoin don’t have any intrinsic value. Just a few days ago, Nobel Prize-winning economist Joseph Stiglitz said that “Bitcoin is successful only because of its potential for circumvention, lack of oversight. It doesn’t serve any socially useful function.” JPMorgan Chase CEO Jamie Dimon has claimed “the only value of Bitcoin is what the other guy will pay for it.”

If you made it through my quick primer above, then you already understand why Stiglitz and Dimon are incorrect. Coins like Bitcoin and Ether can be exchanged for compute and storage on a massive supercomputer with some very compelling properties that allow you to do things like executing transactions between untrusted parties without a trusted intermediary. If you believe that an increasing number of applications will be written and deployed on blockchain to disintermediate our life (and if those currencies are architected in a way that limits velocity) then the value of these currencies will inherently go up as the demand for compute and storage on blockchain increases.

It’s worth pausing here for a moment to mention that there are a few very different kinds of cryptocurrencies floating around today. The first kind is typically called a utility token that has the intrinsic property of being exchangeable for some kind of service. Bitcoin and Ether are both utility tokens because they have the intrinsic property (coded directly into the token and blockchain) of being exchangeable for compute and storage. The second kind of token is often called a tokenized security because it functions more like a traditional security that just happens to be exchangeable on a blockchain. A tokenized security has no intrinsic value but may have value ascribed to it by extrinsic means. For example, a legal contract may promise a share of the future revenue streams of a corporation pro rata to the holders of a specific kind of token. I suspect that people like Stiglitz and Dimon are completely missing the power of blockchain as a cloud computer, so they’re mistaking utility tokens for a tokenized security that is linked to very little value.

“I’m excited about blockchain, but not Bitcoin…”

A second popular thread is that Bitcoin and similar technologies are interesting, but not in their current implementation. One flavor of this attack is that blockchain technology is compelling but cryptocurrencies are not. Another related flavor is that the existing decentralized blockchain implementations will be replaced by blockchain implementations that are controlled by governments and corporations. World Bank President Jim Yong Kim noted that “blockchain technology is something that everyone is excited about, but we have to remember that Bitcoin is one of the very few instances.” He went on to emphasize that the importance of blockchain is the speed with which it can facilitate transactions, drawing parallels to Alibaba’s infrastructure that can facilitate large transactions in seconds. Former Federal Reserve chair Ben Bernanke espoused a similar view when he talked about how Bitcoin would fail but blockchain was interesting and would help federal banks improve their existing payment systems.

Again, this line of thinking is flawed. As noted above, blockchain is an intentionally inefficient system because it trades efficiency for decentralization. It’s hard to see why a bank or government would want to implement an inefficient technology for no reason if decentralization isn’t a desired property. Banks already run on digital systems that can facilitate transactions between people, so what exactly is blockchain bringing to the table? Further, that decentralization can only be maintained if coins are woven directly into the fabric of the blockchain to compensate miners, so the idea of implementing a blockchain without a linked cryptocurrency doesn’t make a lot of sense.

“Bitcoin is an unreliable store of value…”

Another part of the conversation about the value of Bitcoin is centered on whether it will prove to be an effective store of value. A store of value is a mechanism that allows people to preserve facilitate the exchange of wealth and to preserve wealth across both physical space and time. To accomplish this the medium of storage must exhibit a few properties: it must be liquid, it must be scarce, it must possess longevity, and people must be willing to assign a value to it. Historically, stores of value have included things like precious metals, gemstones, livestock, real estate, and fiat currencies.

Some recent challenges to the validity of cryptocurrencies like Bitcoin as a store of value have focused on whether the currencies will continue to exhibit those required properties of a store of value. In the article that was linked above, Jamie Dimon states that “governments are going to crush Bitcoin one day. Governments like to know where the money is, who has it and what you’re doing with it.” In essence, his comments are an attack on the longevity of Bitcoin as well as its liquidity in markets as they become more regulated. Economist and author Raoul Pal claimed that Bitcoin was an unreliable store of value because the group of developers that controls the underlying codebase can change the code: “Even if they don’t change the formula, the fact that they could? That’s enough to say it’s not a long-term store of value.” Pal’s statements cast a doubt on Bitcoin’s scarcity (since software engineers could “print more money”) and whether people should trust the people at the helm enough to assign value to the currency.

The reality is that any government that bans cryptocurrency will miss out on the next great wave of innovation. Where would the US be today if the government issued a ban on all HTTPS traffic because it disrupted the way that intelligence was previously collected? The risk of rogue updates to the codebase is slightly more real, but it’s important to note that there are three groups of actors in each blockchain ecosystem that work as a set of checks and balances against each other: developers, miners, and cryptocurrency holders. The split between Ethereum and Ethereum Classic is a real-life case study in what happens when those groups move in different directions, and it will forever be a warning to the development communities for other blockchains.

Where To From Here?

None of this means that Bitcoin and other cryptocurrencies are destined to continue their meteoric rise. Blockchains have real challenges as they try to scale; when an app like CryptoKitties pushes your network to its limits, you have work to do. Cryptocurrency exchanges are still a major vulnerability of the system and market manipulation is possible at current volumes. For example, it’s widely speculated that the current price of BTC is being propped up by the fraudulent issue of Tether and that if USDT and Bitfinex implode, they will bring all cryptocurrencies along for the ride. All of these risks are real.

But as both a Software Engineer and a VC, I can tell you that I see a lot of companies making big bets on blockchain and using it as the Operating System for applications that were previously impossible to build and will change our lives. Those apps aren’t in production or operating at scale yet, so the analogies between the current environment and the dotcom bubble are reasonable: there may be a crash that is followed by a long period where apps are deployed, adoption grows, and the ecosystem justifies the valuation. Or, maybe, the current lofty valuation on cryptocurrencies is correct for a technology that has the potential to disrupt both the financial sector and cloud computing and near-term growth will continue.

When Clay Christensen introduced the concept of “disruptive innovation” in The Innovator’s Dilemma, he explained that incumbents can’t pursue disruptive innovation when it first arises because the opportunities aren’t profitable enough and the development of disruptive innovation would take scarce resources away from sustaining innovation that is required to keep up with the competition. As the disruptive innovation matures, it begins to capture share up-market, and the incumbent can’t react quickly enough.

Jamie Dimon claims that blockchain isn’t worth his attention because JPMorgan Chase moves $6T in money around the world every day while the daily trading volume of all cryptocurrencies is around $10B. Ironically, with a total market cap of roughly $370B, the basket of all of the cryptocurrencies in the world is now more valuable than JPMorgan Chase. Are major industries going to be disrupted in the next decade? Time will tell, but I’m betting on crypto.