Blockspace: An Introduction with Chris Dixon

Bull markets are for earning; bear markets are for learning. As crypto craters, this is the ideal moment to learn about one of the industry’s foundations: blockspace. Though it may sound self-evident, understanding blockspace is fundamental to understanding the worlds of crypto and web3. It may also be good preparation for a potential future resurgence.

Just a few months ago, Chris Dixon, General Partner at a16z crypto and recent topper of the Midas List, said, “I think blockspace is the best product to be selling in the 2020s.” I remember hearing those words on the Bankless podcast and being unsure whether I grasped the magnitude of what he meant. 

This week, I asked him. 

Indeed, I asked him just about every question I could think of on the subject. Today’s piece is the result of that pestering and Chris’s patience. It is also, thanks to him, one of the clearest and most comprehensive discussions of blockspace and why it matters. 

Before getting to my conversation with Chris, a brief thanks to Alex Obadia, Tarun Chitra, Etienne Brunet, David Phelps, and Leo Zhang for sharing their perspectives on the space and helping improve my understanding. I am very grateful. 

With that, let’s jump in.

‍

OK, Chris, maybe we can start with the basics. What is blockspace?

Blockspace is space on a blockchain where you can run code and store data. Blockspace is different from traditional compute-space in that, until the advent of blockchains, software was always subordinate to hardware – and then, ultimately, to the owner of that hardware. If you write software for traditional computers, it’s the hardware owners who are in control. If Facebook writes some code and says any developer can come along and have access to a certain API, Facebook management can just change its mind and revoke access later. Because Facebook controls the hardware that the software runs on, it ultimately controls the software.

Blockchains are different in the way they're architected: the software is in charge of the hardware. If you write software for blockchains, you can write code that makes strong commitments; you can assure users and developers that the software will continue to run as designed. Specifically, blockchains use what's called a consensus mechanism to make these commitments. The various hardware operators that run the network come together every so often and vote on the state of the blockchain virtual computer. There's game theory around it that guarantees – that makes assurances under most conditions – that the software will continue to run as designed and that the integrity of the data will be maintained.

What we're seeing now is a wave of entrepreneurs and developers who are building new classes of applications that take advantage of this new computing property: that you can write code that makes strong commitments about how it’s going to behave in the future.

That’s an interesting articulation. I haven’t seen this topic framed around the relationship between software and hardware before, but I think what we’re really talking about here is the matter of control, right? 

Exactly. Let's hypothetically say that Google introduces GoogleCoin. Google claims there will only ever be 21 million coins. But the software that powers GoogleCoin runs on servers Google controls. Because Google controls its computers, they can just change the 21 million limit to whatever they want. The software is controlled by hardware, and the hardware is controlled by Google management.

Contrast this scenario with how Bitcoin works. Bitcoin promises that there will only ever be 21 million bitcoins; that scarcity is one of the factors that enables Bitcoin to have value. You can trust that Bitcoin will only ever have 21 million bitcoins because this rule is written in the Bitcoin blockchain – it's baked into its very architecture. Even if a whole lot of the people running the Bitcoin code – the so-called Bitcoin miners – are trying to subvert those rules, it's very, very difficult for them to do so. In the history of Bitcoin and Ethereum and other major blockchains, no one has been able to subvert those game-theoretic guarantees.

That’s what makes blockspace different. Developers and entrepreneurs who want to build on top of the blockchain ecosystem know what the rules are. They won’t get changed on them the way they get changed by traditional tech companies. When it comes to blockchains, instead of “don't be evil,” it's “can't be evil.” The rules of the system are baked into the code.

So, blockspace is a unit of compute and storage that lives on blockchains and, as a result, is free from the control of hardware owners. We’ve seen many different expressions of blockspace and its surrounding mechanisms beneath that foundational definition. What do you think are the most important design considerations? 

Blockspace exists on blockchains, and blockchains can be designed in various ways. The most important characteristic of a blockchain is its security properties. How reliable are the commitments it makes? Can you trust them? Can you trust the system can’t be subverted or hacked? That’s the most important feature.

Another important feature is performance. This is related to the fees you pay when you transact on a blockchain. If you can make systems that are more performant, you can lower those fees. On a blockchain like Solana, for example, one of the nice features is the fees are low because of the way it's designed. Now, some people would argue that you make trade-offs on the security side to get that performance. But clearly, security and performance are both top priorities. 

Another consideration is the community around the blockchain. Some blockchains are home to communities focused on software development, building new applications, and creating valuable new internet services. Ethereum comes to mind as an example of a healthy developer community. Other blockchains focus more on speculation and gambling and, I would argue, are less healthy. 

So blockchains are, on the one hand, computers, so their security and performance properties matter; but they're also social networks, and they need to have healthy communities focused on building.
‍
When it comes to the topic of blockspace, a lot of the discussion centers around the question of scalability. We’ve all seen what happens to Ethereum when demand is high: we get congestion and steep gas fees. What do you think are the most interesting, promising ways to scale blockchains and the space they offer?

You’re right that the so-called scaling problem in blockchains is a hotly debated topic. Some blockchains, like Ethereum, have taken the view that the best way to grow blockspace is through what are called L2s, or “layer twos.” L2s are systems that sit on top of a “layer one” (L1) like Ethereum. If they're architected correctly, L2s inherit the security properties of the lower layer – so you still have the strong security guarantees of Ethereum – but they provide additional blockspace capacity on top where applications can be run with lower gas fees. There are a few prominent L2s right now: Optimism, Arbitrum, zkSync, Aztec, and Starkware. They all take different approaches, and they're all in various stages of development. 

L2s are one way to grow supply; another way is through system design. Solana, for example, is trying to get all of its scaling on layer one. 

Another way I think blockspace grows is just simply with more L1 blockchains. You've got a whole series of credible layer ones in development right now. You also have bridges coming online – ways for blockchains to interoperate, sending assets and messages back and forth. Imagine a future world where you have this fabric of blockchains, all connected, and you're moving seamlessly from one to another, depending on various technical trade-offs and community considerations.

There are philosophical and technical arguments as to which way is the best way to grow the supply of blockspace. I would personally bet on some mix of all three big methods I just outlined.

I love this image of a “fabric of blockchains.” Maybe because it sounds like another reality, it makes me think of one of my favorite pieces I’ve read on the subject of blockspace, “Consensus Capital Markets” written by Leo Zhang and Saneel Sreeni. In the piece, Zhang and Sreeni argue that blockspace will become the “central commodity of the metaverse.” Nothing seems to suggest the need for scale more than running a parallel reality on-chain. But, I’m curious, does this make sense to you? Is the metaverse a major player in the discussion of blockspace? 

First, let’s define the term “metaverse.” Metaverse is a catch-all phrase that describes a series of emerging technologies. It includes web3, new interfaces like VR, and just a general progression of the internet as it becomes more immersive and more central to our lives. Put simply, the metaverse can be thought of as the next wave of the internet.

A really important question with respect to the next wave of the internet is: will it be controlled by one big company, like Meta, in a centralized way, or will it be decentralized like the early web? In the decentralized case, control would be held by a collection of developers, creators, and other community participants, who all work together through common standards and systems, including blockchains. To the extent the decentralized vision wins, blockchains will be extremely important as a way to set the rules of the network, hold assets and virtual goods, and store other shared information. Blockchains are the first way you can have “state” data – in the computer memory sense – on the internet that's owned by a community as opposed to owned by a corporation.

To circle back to the question, I view blockspace as a new, emerging, critical computing resource alongside traditional computing resources like bandwidth, storage, compute, etc. If the web3 vision plays out, blockspace will probably be the most important new computing resource of the 2020s.

The analogies we’re using so far are grounded in traditional computing, but I wonder what you think of other framings. For example, some argue that blockspace should really be thought of as a commodity like land, oil, or grain. (The Zhang and Sreeni piece takes this tact.) Starting from that perspective opens up new ways of reasoning by analogy. For example, since other commodities have markets, will blockspace? To what extent will we see this new “substance” get financialized? 

Most blockspace will not be fungible, in my view, which will limit the financialization. 

While blockspace may be fungible within a single chain, between blockchains, there will be technical trade-offs in areas such as security and performance. More importantly, different blockchains have different communities around them. Therefore, it will mean something different to have a virtual good or a game on one blockchain versus another, in the same way, that it's different to post on LinkedIn versus Facebook versus Twitter. Different networks, different contexts and communities.

I expect we’ll see new innovations that help make the experience of bidding for blockspace more efficient and fair. There are already gas auction systems that financialize blockspace within single blockchains like Ethereum. But I think more broadly, there will be a tapestry of blockchains. Different blockchains will have different communities, and there generally won’t be fungible assets across blockchains.

You mention improvements around bidding, but I imagine there are many other innovations you expect to see. I’m curious to hear what excites you most these days. Where do you see the opportunity? 

We're continuing to see innovation in core L1 blockchains. For example, there are a couple of projects that spun out of Meta we're involved with that have interesting new distributed systems innovations. On the programming language side, there are interesting developments. For example, I’m excited about a new language called Move, which has some nice security properties. There's a lot of interesting stuff happening around zero knowledge proofs as ways to improve both the performance and the privacy properties of blockchains. We’re making active investments there. There's a bunch of exciting stuff happening at layer two, as mentioned above. Bridges are really important, too, to tie it all together. 

We still need improvements to the onboarding experience for new users. There are a whole bunch of user-experience friction points across wallets, custody, key recovery, and key management that need to be reduced. There’s also a constant need for better security and performance.

I expect blockchains to follow a pattern common to previous computing waves: there will be a reinforcing feedback loop between infrastructure and applications. As more applications are created, that creates greater demand for infrastructure. As the infrastructure gets better, that unlocks new applications. Economists call this “induced demand.” It's why, when you build another lane on a highway, often you end up with more traffic; people build more stores and buildings in the area, attracting more traffic. A similar dynamic will play out with blockchains. 

I believe there will be an unending appetite – and opportunity – for more infrastructure innovation and scaling over the next 10-to-20 years.

That feels like the perfect segue to talk about the statement that originally sparked my interest in this topic. On the Bankless podcast back in November of last year, you said, “I think blockspace is the best product to be selling in the 2020s.” What did you mean by that, and what should other builders and investors take away from those words?

Selling blockspace in the 2020s will be a good business in the same way that selling PCs and broadband was in the 1990s and 2000s and that selling mobile phones was in the last decade. Whenever there's a breakout computing wave, you get a reinforcing feedback loop that leads to exponential growth. When you’re in that cycle, it's generally very good to be selling one of the high-quality products that people are clamoring to get. I think that will be the case for high-quality blockspace in the next decade.

In the 1990s, there was a huge wave of investment in bandwidth, specifically in long haul fiber and switching gear. Then there was a huge collapse, and it all went under-utilized. I remember very clearly in the early 2000s, there were a lot of pessimists who said this infrastructure would never get used. At the time, you didn't have Netflix streaming; you didn't have YouTube streaming; you didn't have real internet video. The internet was basically email and some web pages. So people said: Why would you want to pay 50 bucks a month or whatever for broadband just to get faster email and websites? What the pessimists underestimated was that as more broadband came online, developers and entrepreneurs would invent all sorts of great things to do with it. 

Around 2005, that's when you had the launch of things like YouTube. That’s when you really started to see this flywheel kick in where the applications got better and the broadband got better. Then came the mobile wave with the iPhone in 2007, which further accelerated it. It all kind of hit. Cloud computing and social networking suddenly matured, and it led to a decade of really rapid technological improvement and expansion.

Now that it looks like we're entering a financial downturn – maybe analogous to the early 2000s. I wouldn't be surprised if we saw a lot of pessimists say: look at all the stuff we built that’s not very useful. But this is exactly the time when we could be entering an entrepreneurial golden period. Now is the time to go figure out: what is the YouTube or Netflix of blockspace? What are the killer apps that are going to drive this wave of computing forward? The apps may exist today already. They may be new things that don’t yet exist. We don’t know. That's what makes this period fun and exciting.

___

(a16z asked me to clarify that none of the above should be taken as investment advice and to point to their disclosures here.)