Investing In the Permaweb

Today I’m excited to announce our investment in Arweave, a new protocol and blockchain pioneering an entirely new idea: permanent data storage. Arweave is the foundation of the Permaweb, a critical piece of infrastructure in the Web3 stack. We invested alongside our friends at a16z crypto and Union Square Ventures.

So, what is permanent storage?

The best way to understand permanent storage is to first consider the type of storage that everyone is familiar with today: contract-based storage (CBS). In CBS, the provider provides X bytes of storage for Y period of time with Z retrievability guarantees. The vast majority of the storage market today is contract-based, including centralized cloud providers like AWS, GCE, Azure, Amazon Glacier, and decentralized storage providers like Filecoin, Sia, and Storj.

Unlike CBS, permanent storage is what the name suggests: after paying an up-front fee, the Arweave network stores data forever. Like everything else in crypto, the Arweave protocol is designed to incentivize many distrusting, economically-motivated parties to store the data in the Arweave forever, and share data upon request.

Moreover, Arweave is backwards compatible with the IPFS, meaning that developers who have been building on IPFS can transition to Arweave seamlessly. This is very important. The consensus view among crypto developers has been “Filecoin is the only economic layer on top of IPFS.” That is no longer true. Arweave acts as a separate economic layer—one that makes fundamentally different guarantees—making IPFS itself more valuable. Moreover, this allows both Filecoin and Arweave to create mutual value for one another.

Before explaining how the Arweave protocol works, I’d like to dive into the problems Arweave solves, and the markets it solves them for.

The Permaweb

The web as we know it is fragile. As a result, huge swaths of the web have fallen by the wayside. And it’s not just frivolous content. 50% of links from US Supreme Court opinions are broken.

The concept of the Permaweb is foreign to most. It’s an entirely new concept that has never existed before. We don’t expect Arweave to replace the Internet Archive, which stores an imperfect history of the public web. Rather, we expect that Arweave can fill the gaps where the Internet Archive falls short, and we eventually hope that services like Internet Archive will leverage Arweave to make themselves more resilient.

One type of Permaweb demand will be comprised of businesses who don’t want to deal with managing their data, but know they need to store it forever. These are businesses that don’t want to deal with AWS, or be locked in with any SaaS providers who leverage AWS. They simply want to throw something into the cloud, and know that their data will be available forever at an immutable URL. There are businesses across many verticals that have this need: lawyers, journalists, courts, government agencies, and researchers, among many others. For example, TrueUSD is storing its audit records in the Arweave.

The other clear class of Permaweb demand is will come from Web3. As the Web3 stack matures and consumers adopt Web3 based applications, they will naturally want to store some— though not necessarily all—of their content permanently, in a universally accessible place with an immutable URL forever. You probably don’t want to store your 2:00 pm selfie forever, but you might want to store your annual family portrait forever.

How Arweave Works

On the surface, Arweave looks like Bitcoin. There are a series of blocks linked together cryptographically. Miners are compensated for mining blocks from two sources: inflation and transaction fees. The network comes to consensus using proof of work (POW).

However, unlike Bitcoin, Arweave consensus has a second component: proof of access (POA). In order to mine BTC, a Bitcoin miner only needs the last block. None of the prior blocks are necessary in order to mine the next block. As such, there is not an incentive for Bitcoin miners to store the entire history of the chain. The point of Arweave is to create incentives for miners to store history forever, and share it upon request. That’s where POA comes in.

In order to produce the next block, an Arweave miner must first find a suitable hash using traditional POW. Next, the miner must hash the current block with the list of transactions from a random historical block. The random historical block is selected by taking the modulus of the hash of the current block and the current block height. Therefore, in order to produce the next block, miners are incentivized to store as much of the history as possible. It would be very unfortunate for a miner to spend so much electricity hashing, only to discover that they don’t have the necessary block stored to fulfill POA.

Because storage is so much cheaper than POW hashing, miners are incentivized to store as much of the Arweave as possible. No one wants to be the miner who wins the POW competition, only to lose the POA competition. The Arweave team has written more about the incentive dynamics between POW and POA here.

The other big question is: how is it economically feasible to pay up front once, and store data forever? The answer is because of decreasing costs. Over the last ~50 years, the cost of storing data has fallen ~30% each year. There is not a compelling reason to think this trend will end. With perpetually decreasing cost of storage, one can model the cost of storing something forever as approaching some define asymptote.

The full Arweave yellow paper is available here. It explains the details of the consensus scheme, how data is stored, how users download data from the Arweave, the economics of the system, and more.

The Arweave Story

In 2017, Arweave founder Sam Williams began tinkering with ideas around storage. He recognized that many people were using the Bitcoin ledger as an immutable timestamp. He also recognized that the next natural step was to create a system that could store any arbitrary data forever. However, it was clear that Bitcoin was not designed for this. So Sam began experimenting with modifications to Bitcoin’s design to create a crypto-economic system to incentivize miners to store data forever.

I’ve been working with Sam for a few months now, and he is one of the most humble and committed founders I’ve had the fortune to work with. He is relentlessly focused on Arweave’s mission: to build the Permaweb. Whereas most people in crypto hype up their work before releasing it, Sam has done precisely the opposite. He quietly launched Arweave in June of 2018, and has continued to quietly build.

Since then, Sam has assembled a team of 12 people, including 4 PhDs, to build the permaweb. They launched Arweave on June 8th, 2018—precisely 69 years after the publication of Orwell’s Nineteen Eighty-Four—to close the memory hole plaguing society.

Unlike many others in crypto, the Arweave team has been quietly building and shipping, and a global community has formed around the project. There are over 400 miners mining the Arweave, about 100 independently developed applications, and about 100 active developers tinkering on top of the protocol.

The network is growing exponentially. The Arweave community submitted more than 47,000 objects to the Arweave in September, and the total amount of data in the Arweave has grown more than 5x in the last 6 months.

We are incredibly grateful for the chance to back Sam and the Arweave team, and to invest in the Permaweb.