Arweave, The challenging to grasp project.
Discussing Personal data as an undervalued vertical, betting on Arweave means betting on crypto as a whole and if Arweave can default on its promise of permanent storage?
As Qiao Wang from Alliance Dao says:
‘‘The key to building a successful tech company is to enter a market that doesn't exist yet. If you enter a market that already exists, you're competing with entrenched players who have more resources and experience. But if you can create a new market, you have a chance to become the dominant player. This is how you get Facebook, this is how you get Google, and this is how you get Amazon. They stepped into markets that didn't exist, and then they created and dominated those markets.’’
That’s how I felt with Arweave.
Arweave’s main value proposition is: ease your mind, pay once and store forever!
From one side it suits Web3 ethos to have our history saved forever and be verifiable and immutable.
And from the other side: It is more desirable to human beings; how much we lean toward convenience, the easy simple processes, and the joy/fulfillment we get from buying and owning things/services forever instead of subscribing for a period of time.
So Arweave…
The "Weave" expanded by 135% over the past year, reaching 125 TB today, with data stored across 3000 global miners
That’s unique data, the total storage capacity is more than that, before the 2.6 upgrade which happened a week ago; 200 replicas of the 125 TB existed which equal around 24PB of storage.
Being a zero-to-one change and a completely novel solution, it’s difficult to quantify how large the market for permanent storage is and will be
Betting on Arweave is betting on the growth of Crypto as a whole!
As per the last report by Messari; Web3 related data almost make up 3/4 of the total data stored on the Block weave
25% to NFTs metadata, 21% to historical Blockchain Data, 21% to web3 Apps, 5% to DeSocial
So the growth of Arweave increases if we see growth in NFTs, L1&L2s, DeSocial (lens and others), and web3 apps usage and hence their data size.
We have here 4 or 5 main verticals in crypto, considering that; the future of Arweave is related to the future of Crypto in some way
Extending the line, AR could be seen as some kind of index token that means you are bullish on different sectors, Some people included it in their metaverse index fund!
Or even some consider when investing based on narratives, Eg: NFT infra narrative as happened before, DeSocial narrative, etc.
Some of the current clients using Arweave solution are:
Aptos, ZkSync, Avalanche, Arbitrum, Meta plex, Solana Dapps store, Lens, Glass, and some others
Does Arweave have a market outside Web3 Apps? Do people really need permanent storage?
People often ask this, so let’s explore a vertical that I see overlooked IMO
Personal data as an undervalued Vertical to drive growth
Currently, personal Data makes up 10% = 13TB of the data stored on Arweave by some estimates done by Messari, with 20% CAGR to this sector.
Let’s think of this use case regarding its suitability for Permanent storage;
We haven’t faced the problem of the longevity of digital data as of yet, our experience with the internet has been very few decades.
We started to use smartphones lately, renewing our phones and transferring important data every few years, but as time passes, our % of special moments (pics, videos) with family & people we care about, increases, especially for the Zoomers who use and store digital data much more than previous generations from a very early age.
What if the data you have on a physical hard drive get lost, as those hard drives have a decay time, we don’t think properly about what happens if the phone’s memory crashes or our backup drive is damaged and we lose those important special moments forever?! We will see such thing soon enough.
That’s where the permanent storage of personal data through Arweave comes in place;
A product you buy to have a safe feeling once and for all regarding your data, to counter the laziness of transferring these important data every few years.
Let’s talk some numbers; It is very hard to capture customers who are not crypto-native, and send your brand message and utility to literally billions of potential customers.
Let’s just focus on the current crypto-related users (around 90M users registered on Binance) so that’s a conservative estimation of the current crypto users which is constantly growing nonstop.
Imagine just 1% of those (around 1M) buy 1 GB only through Arweave permanent storage? For storing their important Pics with family/close friends, Docs, or any special thing they wanna keep for themselves or their future generation.
1GB through Arweave costs around 7$ (changes algorithmically) which is affordable to a lot of people when they see how important to keep special data forever.
That’s 1PiB of data coming from this vertical alone! 8x the current data stored on the whole block weave (125TB)
Scaling!?? hmmm….
Sam tweeted recently a pic for an all-time high for Arweave transaction counts. 70m in Feb 2023
Let’s check some background for Arweave as a L1
We can consider Arweave as a L1 blockchain built and modularized just to be used for storage, from the pricing to the scalability of the storage transactions.
In terms of pricing;
For example, it currently costs 7$ to store 1 GB on Arweave forever (200yrs guarantee as a minimum)
While it would cost around $20M to store 1GB of data on the Ethereum blockchain
In terms of Scalability;
Although There's a hard limit of 1000 TXs/block (Every 2 mins), It has a very important feature! = Bundling
Some projects which can be considered as L2 of Arweave, offer bundling transactions together, the leading one is Bundlr, which reached 50k TPS recently.
It is the hard carry when it comes to processing transactions on Arweave. they managed to increase transactions by 4000% without sacrificing security/usability & is around ~3000x faster at uploading data.
Bundlr Network accounts for over 90% of data uploaded to Arweave and is a multi-chain solution, allowing Arweave to provide its permanent data storage functions to other blockchains.
Why is this unique and Arweave fits perfectly as a Data storage layer?
2 aspects here:
First; Gas fees!
there are no gas fees at all! regardless of any number of transactions processed! Unlike any other L1, Filecoin included, if the usage of the network increases (represented as transactions count) = the gas fee increases, so it is kind of limited to how many transactions can happen at any single time. otherwise, the fees would be insane.
And that’s not the best option for a Data layer, as the data probably is chunked into Thousands to millions of small pieces, So we would experience pain when dealing with huge amounts of data in the future.
Second: Not a contract-based service & Data is treated equally.
In a contract-based service= Client has to Decide the period, how many copies for his data, Which SP to choose for storage that fits his criteria, etc. and each special deal contract should be adjudicated separately, this puts limitations on the network when dealing with a large number of clients and huge amounts of data.
While here, you just upload your data and it gets its unique ID and mixed with others in the same transactions, Cause all of the data gets the same treatment.
Can Arweave default on its promise of Permanent storage?
As a start, Arweave guarantees 200 yrs as a minimum for the most conservative scenario, and hopefully, for eternity, That’s what they promise.
We need to understand first how it works from a high-level view and then answer if their promise is valid or not.
Basically, we have Storage providers -miners- who compete with each other to store the data.
We have clients who wanna store their data permanently.
And we have Arweave as an intermediary who designed the network (and keep tweaking the edges still) facilitating the interactions between both.
Let’s visualize the journey for any data getting onboarded;
A client pays a certain (amount of money)#1 (algorithmically set by Arweave equations) for minimally 200 yrs of storage.
⅙ of the money goes to those miners (which he receives after 30 days)
⅚ of the money goes to the (endowment)#2
Miners keep storing the data with the initial money they got until Arweave mechanism detects that the (Revenue)#3 is lower than the storage costs, then it spends some money from the endowment to support them.
What if there is always new data being onboarded?
Miners will be good and happy, they getting ⅙ of whatever is being paid + some inflation rewards (it makes up a good chunk of their revenue streams, but it will end next year and they will depend mainly on the data fees rewards)
If there are no new fees coming in? the endowment should be enough to guarantee 200 yrs of storing your data.
That’s it, let’s explain quickly some of the highlighted words and then discuss risks.
#1 Cost of storage: it depends on 3 factors
#2 Endowment: Consider it the reserve fund, which holds the cost to store data for 200 yrs+. 86% of the transaction fees(in AR) goes there, the curve is up and to the right since it was created, Arweave didn’t use its Endowment funds as of yet, by some estimates, it won’t be used unless the weave reach the web surface size.
It acts as a way to reduce supply for long time so rest of tokens appreciate in value.
#3 Revenue: In order for mining of the Arweave to remain profitable and sustainable over time, the following basic principle must hold: the reward emitted by the network at any given block must be greater than the sum value expenditure required to maintain 15 replicas for the whole weave for that period
if Miners aren’t profitable at any given time (either because of fluctuation of token price or storage medium prices, the endowment compensates them until the equation changes again.
What if the above principle holds but there is much competition between miners making them not profitable?
=Not Arweave’s problem, they should compete and the ones with the lowest break evens win. There would be winners and losers in the game
The main point Arweave leans on, to offer its service for an affordable price: is what’s called Kryder ratio: (the cost of storage declining each year by 30% since 1979). They price storage based on a very conservative rate of 0.5% reduction every year.
Arweave is a delicate balance between game theory, mechanism design, and mathematic equations using historical data to promise the permanent storage, Although it is an experiment that may have issues, and that’s a risk people are taking by paying upfront, it is perfectly designed to achieve its purpose.
Now, Let’s stress-test it;
Risk1: new data fuels old data
What if there is no growth? no new data getting onboarded= no new fees, we start using endowment funds, people start to get bearish and the price falls, the endowment decreases even more in value than before = not enough to honor the 200 yrs promise.
And that’s mainly coming from using AR as the only token in the endowment, but it is a compromise with upside and downside that Arweave accepted it.
Maybe That’s why they focus on long term partnerships with protocols which always produce data at constant rate.
Risk 2: Centralizing risk on the whole protocol as a downside for not functioning at a contract level.
As all data is treated equally, when bad actions happen from some miners for example; it affects the whole data and the endowment, or when introducing network updates for some new rules. E.g; what happened directly after 2.6 upgrade, the number of miners dropped badly
Risk 3: They have carrots with no sticks
There isn’t any kind of slashing of collateral, rewards or kind of punishment for not honoring being a Storage provider on the network and profiting from it. Especially we are dealing with a precious asset like Data, So during hard times, there are no obligations for Miners to keep around. they might drop off fast and cause issues, pieces of data lost without getting replicated, etc.
Risk 4: Their supply side (miners) is a mess!
There is no such way to know the miners’ number, Miners are pooled together in some pool -At some point a certain pool had 50% of the total hash rate- not organized, there isn’t a team dedicated to help them, it just whoever got spare time to respond to their questions.
=that can lead to less distributed decentralized community of miners, risk of centralization by some Pools which is considered a single point of failure
How is it different from Filecoin from a fundamental perspective?
Literally, we are comparing apples and oranges;
Besides the popular difference between permanent vs temporary data storage,
They are different in their mechanism design, incentive structure, their addressable market to some degree, network features, type of data stored, and pricing mechanism, almost everything other than being both in the same market.
A final note;
Being a storage provider at Arweave is much harder than being one at Filecoin, it needs more I.T skills, a very steep learning curve, the community isn’t as organized as Filecoin, there aren’t the same efforts in terms of educating Miners and taking care of the community as I saw there.
They need to work on that. Caring about the supply side in the early stage of growth and making sure they are educated & loyal to the network is crucial to the process.