There's a bunch of hype about Arweave, but I think it's going to zero. I think it's extremely overrated. Internet Computer already has the lead in the market that Arweave and Arweave AO are trying to be in, and all these centralized platforms like Amazon and Google already have cheaper storage costs. I'm going to walk you through why I think this, because there's been so much hype and the price has pumped so much in the last year, but in my opinion this is going to trend more toward zero than toward being a good investment. Here's why.
The price chart already had its bull run
Let's start with the price chart. It looks like a horrible time to buy right now, because it's up about 10x in the last year and it already has a roughly $2 billion market cap. But what is the real value behind Arweave? Arweave is a project that's been around a while. The price chart goes back to 2020. It had an all-time high of around $82 in the last bull run, so it's already had its first bull run. Rarely do projects have a second bull run that outperforms their first one. And utilities like Arweave — permanent storage — tend not to be valued that highly, because people tend to value hype and speculation and marketing instead. Arweave's basic idea is permanent information storage. But there are some critical flaws in this.
Critical flaw #1: the cost is brutal
The first critical flaw I see is how much it costs. Right now it costs about 0.74 AR per gigabyte. And the more the Arweave price pumps, the more the actual thing you want to use the network for becomes totally unusable. For example, right now if you want to put a gigabyte of data on Arweave, that's going to cost you a little less than $30. It says it's permanent storage — that's what it says on the face. But do I trust that it's really permanent? A couple of years, two, three, five years, maybe. But is the network going to last that long? How is it backed up? If one permanent storage node goes offline, is it going to stay online? And what is the actual infrastructure that it's on? We can't really see that.
By comparison, if you want a gigabyte of data fully on chain on Internet Computer, that's only going to cost you about $5. So right now you could put about five or six times as much data fully on chain with Internet Computer compared to putting it on Arweave. And unlike on Arweave, where you can't really do anything with the data — you have to build a separate app to pull the data in and build other infrastructure on top of it to process everything, because it's just storage — on Internet Computer the storage is fully functioning and you can build a complete application like OpenChat right on top of it. My website at jerrybanfield.com is hosted completely on chain. You can build an entire application totally on chain with gigabytes of data on Internet Computer.
And the question is, do you really need permanent storage? For some applications, like a social media application, you might. But the cost of it at this point is going to make it more expensive — it's going to make it absolutely unusable. That's the problem. Because it costs so much on Arweave to even put stuff on Arweave, and it doesn't seem to me that this is actually on chain. As far as I can tell, all the data is not on chain for Arweave. There are all these servers and they get paid in AR, but it's not all on chain the way it is on Internet Computer. So you have no idea what's actually happening with your data and how it's secured on Arweave, which means for crypto applications, is Arweave really that trustworthy? You might as well just use Amazon Web Services to put your data into an application.
For example, if you think $5 per gigabyte is too expensive on Internet Computer — which for some video applications where you're putting a lot of data into the app, like a Web3 game or a social media application, it could be — those builders are not going to pay Arweave storage prices at $30 a gigabyte. They're going to go for the cheapest solution, which would be Amazon Web Services or Google, where you can just dump terabytes. I have 16 terabytes — that's 16,000 gigabytes — and it's $45 a month on Dropbox. So Arweave can't compete with the centralized infrastructure on cost. And most people are not looking for permanent storage. They're looking for storage for a year or two, maybe five years. So Arweave, for the main thing it does, is not worth it in most cases. It's not the best in the market. This is the kind of coin-by-coin breakdown I keep going on my Money playlist, where I stack each project up against what it actually delivers.
Critical flaw #2: where is the team?
From there it keeps getting worse, because how is this actually governed? It's not even obvious upfront whether holding the AR token gives you any actual control over the protocol, or who's actually working on the protocol. I searched "Arweave team," and I get the Arweave website, but I don't see any actual people on there. There's an account listing who's behind Arweave — there are two founders of Arweave — but where's the actual Arweave team? If I look at the Arweave Foundation, where is the actual team? There's supposed to be an actual group of engineers and people working on this. There are no people, no transparency at all about the people who are working on this. And that to me is a huge red flag.
Because if you talk to developers in person who work for some of these crypto companies and blockchains, they often will take the payment and immediately dump it, because they have no faith in the protocol itself — they're just getting a paycheck. And if you were acting like that, you wouldn't want to have your picture listed on the website so people could come to you while the truth is you're dumping the coin immediately. There's no transparency about where the team of people working on Arweave actually is.
If you go to Internet Computer Protocol by comparison, you have the largest research and development team in blockchain. The employees have hundreds of patents. They publish thousands of articles on crypto and blockchain and the related technologies. You have this huge team of people, a huge amount of engineers and technical staff, people you recognize, people you can see on YouTube videos. Arweave has zero transparency about who exactly is building this. It also says it's totally decentralized — but the problem with that is, where's the money to fund further development of the underlying protocol? What you'll notice is that if things don't tend to get advanced, and if they don't get improved and constantly adapted and upgraded, then they tend to become irrelevant.
And in crypto investing, while many people say crypto is all about the money — "that's all we care about is the money, we don't care about the technology" — people with huge wallets and big investment funds do care about the technology, because they're not trying to gamble with their hundred-million-dollar investment fund on some hype and speculation most of the time. What they want is to find technology that's going to grow, and you need a team of people to research that. So why, with a $2 billion market cap, is there zero transparency about the Arweave team? That is a critical flaw, a huge weak point, a big problem.
How much is the network actually getting used?
Let's look at how much the network is actually getting used. If you compare the network usage to Internet Computer, Arweave has about 174 terabytes on the "weave," which I'm guessing is how much is actually being stored on there, while the network size is how much could be stored on there. You can see roughly how much data is being uploaded to the network — but you can't verify the authenticity of this data. Crypto is so nasty that there will often be manipulations to try and fake different data, uploading a bunch of junk data just to make it look like it's getting used. That's probably not happening here, but you don't know that.
What does look good is that there are consistently hundreds of gigabytes of data being uploaded to Arweave. This is probably because some apps built on other blockchains are then putting their data on Arweave. But then, how sustainable is that? You've got the data on Arweave and the website often hosted somewhere else. These apps that claim to be centralized are not, and some of that data is probably being put on Arweave. Right now there are about two hundred-some-thousand addresses total and around 300 transactions per second, which for a $2 billion market cap — I do like Arweave better than a lot of other altcoins. I reviewed Jasmine yesterday, and Arweave appears to have a lot more substance than Jasmine, even though they're at similar market caps.
That said, when you stack it up next to Internet Computer it does not look impressive at all, except in the quantity of data that's been uploaded to it. With a couple hundred gigabytes of data a day and the storage cost, you can see there are a couple hundred AR a day being paid to upload storage, so that's decent. However, many of the people using Arweave don't know that they could put their whole application on Internet Computer. If you compare Internet Computer's dashboard, you've got at least ten times as much in terms of transactions happening on the network, and about ten times or more as many active wallets and addresses on the network as well. Even though Internet Computer is only about double the market cap of Arweave, when you consider that on Internet Computer you can store data on chain plus do so much more — it's a Bitcoin layer-2, an Ethereum layer-2 — I think Internet Computer has way more growth potential than Arweave.
Now, on Internet Computer there are about four terabytes on chain right now, because it's not permanent storage — if you stop paying the cycle cost, it will drop off. So Arweave does have right now a little less than 50 times as much data. But then again, that's all Arweave does. It doesn't do anything else. So while it's specialized, it has a fraction of the accounts, a fraction of the transactions, and it's one tiny little thing: storing data. That's all. Its scope is so much more limited than Internet Computer that, when you compare the market caps at 50 percent of Internet Computer's, this should probably be a tenth or less of Internet Computer's market cap. If you want the other half of this comparison, I keep the deep ICP case going on my ICP Crypto playlist.
Now let's dig into the AO hype
Arweave is functioning as a "decentralized storage network," but I think it's drastically overvalued right now, because if you're not going to build everything fully on chain with smart contracts the way you can on Internet Computer, you might as well just use Amazon Web Services and do it cheaper than Arweave. One of the big reasons the Arweave price has pumped so much is that they started to hype up Arweave AO as a "world computer," which is funny. This actually shows you where things are going for Internet Computer, because right now you've got Arweave and some of these other machines hyping themselves up as a real world computer. They don't have the raw computation power. They don't have the technical capabilities to do that. They haven't been built with that in mind from the beginning, the way Internet Computer has, and therefore they have no real chance of providing it. They're just jumping on narratives. You can see how bullish that is for Internet Computer when you've got Arweave throwing up Arweave AO and generating hype around it.
So I have serious concerns with this. If you go to ao.arweave.dev, all you'll actually find is being able to boot up the testnet. The Arweave marketing related to this is also absolutely ridiculous and disingenuous, acting like this is some amazing thing that can do what nothing else can do — when it looks to me like it quite possibly could be Amazon. It could be a cluster of nodes connected together. That does not look too impressive at all. They're making it out like it could be this infinite AI computation engine, versus Internet Computer where it's all secured on blockchain and smart contracts.
You have to read the specs yourself
So I read through the developer specs, and this is why you really need to research stuff. You can't just read a tweet or two, or even watch one of my videos — you need to look at things yourself and then look at the bigger picture. Arweave AO is described as "an actor-oriented machine that emerges from the network of nodes that adhere to this core data protocol running on the Arweave network." So there are nodes that run on the network according to this core data protocol. Who are the people actually building this core data protocol? Because the core data protocol itself is probably going to be pretty challenging to build successfully into something that can actually handle all this.
Here's what somebody else said. One of the problems with Arweave: it's 25 minutes to create a transaction. How is this going to be a world computer? Don't even get into the consensus on where the compute is taking place. Do you want to wait half an hour for a transaction? When you see what people who've actually used this are saying, it's laughable — and it's not funny that people are dumping huge amounts of money into this and getting hyped up, when the people actually using it report 25 minutes to complete a transaction. How is that going to be a world computer? Absolutely ridiculous.
Then someone else responds that the first guy has no idea what he's talking about, that data transactions are bundled for scalability. Look — data transactions are fast because they're not on chain. And that's drastically different from Internet Computer, where stuff is on chain. Do you understand that their existing infrastructure is not on chain? So whatever Arweave AO says it can do can easily be done elsewhere. Do you realize you could go on Amazon or Google right now and spin up a massive amount of computing power instantaneously, on networks that are proven and built all over the world, and do it way cheaper than Arweave? The only value Arweave AO could potentially offer is that it's somehow secure and able to be verified — and you do that by having it on chain. The data transactions on Arweave are fast because they are not on chain. And according to Crypto King, they're actually running on Amazon Web Services.
Why this pattern bothers me so much
This is what so bothers me about crypto. If I were a small group of people behind Arweave trying to make as much money for myself as possible, I would simply do it exactly like that, because the average person couldn't tell that you could just do it like that. You spin all these nodes up on Amazon, call it Arweave AO, do practically infinite amounts of computation that's all off chain, and generate a bunch of hype — but the truth is you're just using the existing infrastructure to do it. And then if you try to use a chain for something, the token transfer is miserably slow. So you can't actually make things with Arweave the way you can with Internet Computer, like OpenChat. On OpenChat I can go send a transaction right now that will almost instantaneously complete, and the entire application is on chain. It doesn't need to be secured with firewalls or security teams. It's all secured by cryptography, because the smart contracts are built directly on chain, and the entire infrastructure has been built from the ground up to be a fully on-chain, secure, verifiable computation platform.
Arweave AO, from what it looks like to me, is literally spinning up Amazon Web Services nodes in the background and running the computation on that. It's not on chain. There's no security. And this is what other people are sharing from their own experience too. Crypto King shares further that he wants things to be actually on blockchain. As I repeatedly mention, Brian Armstrong, the CEO of Coinbase, says that "on chain is the new online." ICP is the only place where you can do stuff online, fully on chain, at scale. Arweave AO does not appear to do any of that. A great summary I saw is that it's an ICP knockoff that's running on enterprise cloud.
It's literally a team of people — who aren't even transparent — sitting down and saying, you know, they can see where things are going. They can see that ICP is the direction things are going, that fully on chain is the direction things are going, and they're asking, "How can we capitalize on this and get Arweave to pump?" Well, we could make something like ICP. No, that's extremely difficult and takes huge teams of people to actually figure out how to do. Okay, what's the easy way? Put together something that combines Arweave data storage and uses Amazon Web Services or Google Cloud nodes to run the computation, put a protocol together to do that. An honest person would say, "Well, that's not going to be on chain or secure." And the answer is, "So what? The average person won't know that. They'll buy it and invest on the narrative when we can hype it up." But not all of us are going to be fooled.
One person notes that his wife is a full-stack software developer, and they got on the testnet last night — that's their experience of it. Based on the documentation, the vibes, and the core functionality I've read, the goal is to try to hype up the token. I am seriously skeptical as to whether they can even deliver something of value, because all we have is a testnet right now, and people are saying that if you look up where it's hosted, it involves Amazon Web Services. To me, this looks like it's going to zero. We'll see who's right, because I know some of you completely disagree with me.
How I track every one of these calls
Every single crypto I'm reviewing, which I'm doing live, I put exactly on my spreadsheet. I put the ICP price on there every day and I refresh it, and then I calculate, from the day I reviewed it — when I said it was going to zero and compared it to ICP — how it actually performs. In the short term the data is kind of meaningless and irrelevant. You really need at least a year to get some real quality data, so it's not just random chance and fluctuations and manipulations. So far it hasn't mattered in the first week for everything I've reviewed, but I'm going to keep updating this until we have hundreds and hundreds of cryptos, and we'll see — unlike anyone else in crypto — exactly how accurate what I say turns out to be in reality.
If you appreciate that kind of transparency, I obviously am holding almost all of my crypto in ICP, because I see it as so far the best technology in crypto. Literally, people are doing anything they can to act like they're doing something similar — but it's easy to fake things. Crypto is often executed like a one-night stand, where it's easy to lie to someone and misrepresent yourself and get yourself into a position for a quick win. But trying to make a happy marriage takes a whole lot more work, and a lot of honesty and transparency, most of the time. To me, ICP is like a happy marriage: this is going to go well for the long term. In the short term it can be a bumpy ride — you've got to work some stuff out — but most of the other cryptos are one-night stands: "I'll do anything to get your money today, and I don't care what happens tomorrow."
If you've watched my breakdown of why I think Aleph.im is going to zero, you'll recognize the same storage-and-compute pattern playing out here, and it's the same lens I used in my Humanity review and when exposing Taraxa. If you want something better than a one-night-stand token, you can come chat with me and the rest of the family directly — I read every comment, and the only way to reach me when I'm offline is on OpenChat, fully hosted on chain on ICP, where you can ask for a one-on-one call. You're always welcome to join the Jerry Banfield family and bring your own crypto questions.
I think Arweave is going to be a bumpy ride down, not up. We'll see who's right — but that's my honest read.