One of the fundamental things that people tend to misunderstand about the blockchain technology is that they assume it is all about decentralizing finance. This is partly because Bitcoin, which was the first application of the blockchain, was released to be an actual currency that only you can access through your wallet (becoming the first major innovation in currency since we went off the gold standard and switched to fiat currency). Then, the application of blockchain technology vastly expanded with the release of Ethereum, which came out with the ability to run code on the blockchain (in an inefficient manner), in an entirely decentralized way. Running code enabled the creation of smart contracts, which is code that saves state on-chain and is executed on-chain, which is a fancy way of saying it is executed in a decentralized way, by many computers, sharing state, thus creating a single shared decentralized computer for arbitrary computation. This enabled the creation of tokens and NFTs built on top of Ethereum, alongside decentralized exchanges and a plethora of other projects. But one thing sticks out about these projects -- they are almost all about financial applications. People get their understanding of Crypto from frankly the only real applications that are used today, trading tokens and NFTs.

What most people don’t know are the projects that promise to open up decentralized compute, and decentralized storage, through simply making it cheaper by increasing TPS (transactions per second). The major barrier to Web3 is the fact that you have to pay the decentralized computer whenever you want to run code. And since it is expensive, the only ones that are sustainable at the present moment are exactly the financial applications. Nobody wants to run expensive code when you can’t make money off of it.

Ethereum can currently process 15 transactions per second (TPS — obviously inefficient and expensive). When you think of ‘Web3’, imagine Ethereum with a TPS in the millions. People will be able to use an entirely decentralized computer (impossible to stop) at any point for arbitrary computation, with a shared state amongst other people also using the computer. Websites will be hosted on the blockchain (instead of simply a front end interacting with a DB, it will be a frontend stored in a decentralized way that communicates with smart contracts) and take advantage of solutions like IPFS to be faster and cheaper to host than present day websites and applications. The promise of web3 is decentralized compute that will eventually be cheaper and faster and more secure than the present day internet (not to mention fully decentralized).

All you hear about in terms of owning your own data, democratizing access, being more secure, and everything else is mostly just fluff. The fact of the matter is that there are going to be more and more user facing applications that are hosted on the blockchain. There is no reason Yelp should exist as a business -- it could be recreated incredibly quickly and exist forever. Uber, Airbnb, Reddit, Facebook, Slack, Programmable spreadsheets, Emails, etc are all possible and will be recreated on chain. There is no “ifs” or “buts” -- all of these and more will be recreated on chain once the technology gets there.

Lets look at decentralized storage. We start with decentralized storage because it is significantly further along than decentralized compute. IPFS and Filecoin are the first projects that come to mind. IPFS is a peer to peer file sharing protocol, with an economic incentive (Filecoin, which operates on its own Chain, much like Ethereum). If you ever decided to pirate something, you connected to a peer which was seeding your download. You probably used something like BitTorrent.

Read the introduction to the IPFS Whitepaper.

IPFS is a distributed file system which synthesizes successful ideas from previous peer-to-peer sytems, including DHTs, BitTorrent, Git, and SFS. The contribution of IPFS is simplifying, evolving, and connecting proven techniques into a single cohesive system, greater than the sum of its parts. IPFS presents a new platform for writing and deploying applications, and a new system for distributing and versioning large data. IPFS could even evolve the web itself.

This means that IPFS took what technologies were applied to the most popular Torrenting software, updated it by 15 years, and added an entire economy on top of it. You can now upload a file for free, and have it exist forever, hosted in a decentralized way. Pretty cool.

Essentially, your files get split up into chunks, encrypted, and stored on “nodes” (aka disks that are storing your data which are ran by people who are participating in the network). Since you have the decision as to whether to encrypt or keep the data decrypted, you can store either public files or files only you can access, and people will only be able to access your public files through a hash that is organized like a link, with the possibility for directories. People storing your files only have chunks of your encrypted data, so they will never be able to access it. When you request your data, your data gets pieced together from these people’s machines and returned to you as the original file.

Then there is the Filecoin Virtual Machine. This chain will soon be able to run any code Ethereum can run, except more advanced (WASM instead of EVM) and with integrations to IPFS. They want L2s built on top of it (speed up computation significantly for the base layer through optimizations e.g. Optimism for Ethereum, bringing their 15 TPS to 4,000 TPS), for every single chain to be able to store and read data on filecoin (thus keeping other chains as layers for computation while establishing itself as a source for cheap decentralized storage), enabling the creation of economically incentivized CDNs for IPFS, volume discounts, auctions, perpetual storage, insurance for storage, Data DAOs, and more.

This isn’t mentioning any other projects like Arweave, in an attempt to keep this article brief. Mirror is storing and retrieving this article from Arweave. It seems like decentralized storage is starting to make major progress, and potentially can be much more popular in the future. Storage is currently the most expensive part of paying for “computation” on chains like Ethereum -- once their integrations with other chains set in, and you can store much more data off chain, computation will become significantly cheaper on existing chains seeking to be the computation layer.

Now let’s get into computation. As I mentioned before, Ethereum can only process 15 TPS. You look at Bitcoin, and it is even more horrendous, especially considering it doesn’t run code: 3-7. Ethereum is working on scalability through several technologies (Sharding and POS), alongside a variety of L2s (layer 2s; Optimistic rollups does this via game theory and incentive mechanisms while ZKRollups do this via a novel encryption method called ZKSnarks/ZKStarks invented around 2018). The infamous "Merge" is projected to happen September of this year (but will probably be delayed), which would transition Ethereum from a Proof of Work system to a Proof of Stake system, which will enable the use of Sharding to significantly speed up the base network layer to a hundred thousand TPS in 3-5 years, while reducing energy use by 99%. L2s are built on top the base layer, and will speed this up to be millions of TPS. Already deployed optimistic rollups speed the existing Ethereum protocol up from 15 to 4500 TPS, so you can do the math for what happens when Ethereum rolls out sharding (~100,000 TPS), while ZKRollups are currently being worked on, which appear to be a more promising solution. Danksharding is another promising solution (adding more storage space for each block), which might come before sharding, and increase TPS dramatically by natively supporting optimistic rollups/zkrollups.

This is also only Ethereum we are talking about. There are many other chains with novel consensus mechanisms and technology (Polygon has a theoretical TPS of 7200+, Avalanche supports 4500+ TPS on their main network and a potentially infinite TPS via chains built on top of it & has been live with many applications deployed on it, Solana supports up to 70,000+ TPS without additional modifications although they have been running into DDOS/centralization problems), and there are now an endless list of chains that support much higher TPS (Cosmos, Snarknet, Polkadot, etc). L2s are progressing incredibly quickly as well, and with recursive ZKSnark proofs, we can potentially be in a position to have nearly limitless TPS faster than you think.

Lets take the possibility that there will be multiple chains with TPS in the millions. Most people compare this to VISA TPS (1700) and claim we will be able to support cheaper transaction fees. This is true, but what is important to understand is that gas (what you pay for computation) is directly related to how much code you execute, and how much activity there is on the network. If the activity goes down, you pay significantly less for gas, in the fractions of cents, per transaction. This is with chains that support thousands of TPS, but not millions -- spending $1-2 dollars per month in exchange for owning your own data and interacting with the decentralized state machine and novel applications that are these chains enable seems more than reasonable.

This is only really scratching the surface of new projects that are slowly being released. Aleo has a model with zkSnarks where the client (e.g. you) actually performs the computation, the code is left secure and private, and you end up interacting with a chain that is able to store private state that only you can access. One of the major limitations with the blockchain as it exists today is that the state is entirely public for everybody to see, which leads to a limitation of applications -- it’s entirely possible that in 3-5 years there will be chains with private state, leading to entirely private transactions, alongside private arbitrary computation.

As a result, we are left with a potential vision for what the future of blockchain technology will look like. Cheaper, decentralized, and faster to access storage that will exist permanently leads to faster loading times for websites and a new method of sharing files. Cheaper and decentralized arbitrary computation, with potentially private state, leads to applications that simply weren’t possible in the past, and a recreation of many popular applications today on the decentralized way. All with a built in economy that is inherently more trustworthy than the existing economic model. When you talk to people that are “bullish” on crypto, this is primarily what they are talking about: what crypto can potentially be in 3-5 years, let alone 10-15 years. Sure, there isn’t a lot to show right now in terms of applications: only financial applications currently exist. As more infrastructure projects get built out, however, computation will be easier than using existing cloud providers, and users will pay for any computation they end up using. Sure, this doesn’t necessarily make sense for all applications, but it can make sense for a huge amount of consumer and enterprise applications. And although it doesn’t seem like it right now, you best believe that these incoming decentralized applications will figure out how to appropriately monetize via the blockchain.

One of the fundamental things that people tend to misunderstand about the blockchain technology is that they assume it is all about decentralizing finance. This is partly because Bitcoin, which was the first application of the blockchain, was released to be an actual currency that only you can access through your wallet (becoming the first major innovation in currency since we went off the gold standard and switched to fiat currency). Then, the application of blockchain technology vastly expanded with the release of Ethereum, which came out with the ability to run code on the blockchain (in an inefficient manner), in an entirely decentralized way. Running code enabled the creation of smart contracts, which is code that saves state on-chain and is executed on-chain, which is a fancy way of saying it is executed in a decentralized way, by many computers, sharing state, thus creating a single shared decentralized computer for arbitrary computation. This enabled the creation of tokens and NFTs built on top of Ethereum, alongside decentralized exchanges and a plethora of other projects. But one thing sticks out about these projects -- they are almost all about financial applications. People get their understanding of Crypto from frankly the only real applications that are used today, trading tokens and NFTs.

What most people don’t know are the projects that promise to open up decentralized compute, and decentralized storage, through simply making it cheaper by increasing TPS (transactions per second). The major barrier to Web3 is the fact that you have to pay the decentralized computer whenever you want to run code. And since it is expensive, the only ones that are sustainable at the present moment are exactly the financial applications. Nobody wants to run expensive code when you can’t make money off of it.

Ethereum can currently process 15 transactions per second (TPS — obviously inefficient and expensive). When you think of ‘Web3’, imagine Ethereum with a TPS in the millions. People will be able to use an entirely decentralized computer (impossible to stop) at any point for arbitrary computation, with a shared state amongst other people also using the computer. Websites will be hosted on the blockchain (instead of simply a front end interacting with a DB, it will be a frontend stored in a decentralized way that communicates with smart contracts) and take advantage of solutions like IPFS to be faster and cheaper to host than present day websites and applications. The promise of web3 is decentralized compute that will eventually be cheaper and faster and more secure than the present day internet (not to mention fully decentralized).

All you hear about in terms of owning your own data, democratizing access, being more secure, and everything else is mostly just fluff. The fact of the matter is that there are going to be more and more user facing applications that are hosted on the blockchain. There is no reason Yelp should exist as a business -- it could be recreated incredibly quickly and exist forever. Uber, Airbnb, Reddit, Facebook, Slack, Programmable spreadsheets, Emails, etc are all possible and will be recreated on chain. There is no “ifs” or “buts” -- all of these and more will be recreated on chain once the technology gets there.

Lets look at decentralized storage. We start with decentralized storage because it is significantly further along than decentralized compute. IPFS and Filecoin are the first projects that come to mind. IPFS is a peer to peer file sharing protocol, with an economic incentive (Filecoin, which operates on its own Chain, much like Ethereum). If you ever decided to pirate something, you connected to a peer which was seeding your download. You probably used something like BitTorrent.

Read the introduction to the IPFS Whitepaper.

IPFS is a distributed file system which synthesizes successful ideas from previous peer-to-peer sytems, including DHTs, BitTorrent, Git, and SFS. The contribution of IPFS is simplifying, evolving, and connecting proven techniques into a single cohesive system, greater than the sum of its parts. IPFS presents a new platform for writing and deploying applications, and a new system for distributing and versioning large data. IPFS could even evolve the web itself.

This means that IPFS took what technologies were applied to the most popular Torrenting software, updated it by 15 years, and added an entire economy on top of it. You can now upload a file for free, and have it exist forever, hosted in a decentralized way. Pretty cool.

Essentially, your files get split up into chunks, encrypted, and stored on “nodes” (aka disks that are storing your data which are ran by people who are participating in the network). Since you have the decision as to whether to encrypt or keep the data decrypted, you can store either public files or files only you can access, and people will only be able to access your public files through a hash that is organized like a link, with the possibility for directories. People storing your files only have chunks of your encrypted data, so they will never be able to access it. When you request your data, your data gets pieced together from these people’s machines and returned to you as the original file.

Then there is the Filecoin Virtual Machine. This chain will soon be able to run any code Ethereum can run, except more advanced (WASM instead of EVM) and with integrations to IPFS. They want L2s built on top of it (speed up computation significantly for the base layer through optimizations e.g. Optimism for Ethereum, bringing their 15 TPS to 4,000 TPS), for every single chain to be able to store and read data on filecoin (thus keeping other chains as layers for computation while establishing itself as a source for cheap decentralized storage), enabling the creation of economically incentivized CDNs for IPFS, volume discounts, auctions, perpetual storage, insurance for storage, Data DAOs, and more.

This isn’t mentioning any other projects like Arweave, in an attempt to keep this article brief. Mirror is storing and retrieving this article from Arweave. It seems like decentralized storage is starting to make major progress, and potentially can be much more popular in the future. Storage is currently the most expensive part of paying for “computation” on chains like Ethereum -- once their integrations with other chains set in, and you can store much more data off chain, computation will become significantly cheaper on existing chains seeking to be the computation layer.

Now let’s get into computation. As I mentioned before, Ethereum can only process 15 TPS. You look at Bitcoin, and it is even more horrendous, especially considering it doesn’t run code: 3-7. Ethereum is working on scalability through several technologies (Sharding and POS), alongside a variety of L2s (layer 2s; Optimistic rollups does this via game theory and incentive mechanisms while ZKRollups do this via a novel encryption method called ZKSnarks/ZKStarks invented around 2018). The infamous "Merge" is projected to happen September of this year (but will probably be delayed), which would transition Ethereum from a Proof of Work system to a Proof of Stake system, which will enable the use of Sharding to significantly speed up the base network layer to a hundred thousand TPS in 3-5 years, while reducing energy use by 99%. L2s are built on top the base layer, and will speed this up to be millions of TPS. Already deployed optimistic rollups speed the existing Ethereum protocol up from 15 to 4500 TPS, so you can do the math for what happens when Ethereum rolls out sharding (~100,000 TPS), while ZKRollups are currently being worked on, which appear to be a more promising solution. Danksharding is another promising solution (adding more storage space for each block), which might come before sharding, and increase TPS dramatically by natively supporting optimistic rollups/zkrollups.

This is also only Ethereum we are talking about. There are many other chains with novel consensus mechanisms and technology (Polygon has a theoretical TPS of 7200+, Avalanche supports 4500+ TPS on their main network and a potentially infinite TPS via chains built on top of it & has been live with many applications deployed on it, Solana supports up to 70,000+ TPS without additional modifications although they have been running into DDOS/centralization problems), and there are now an endless list of chains that support much higher TPS (Cosmos, Snarknet, Polkadot, etc). L2s are progressing incredibly quickly as well, and with recursive ZKSnark proofs, we can potentially be in a position to have nearly limitless TPS faster than you think.

Lets take the possibility that there will be multiple chains with TPS in the millions. Most people compare this to VISA TPS (1700) and claim we will be able to support cheaper transaction fees. This is true, but what is important to understand is that gas (what you pay for computation) is directly related to how much code you execute, and how much activity there is on the network. If the activity goes down, you pay significantly less for gas, in the fractions of cents, per transaction. This is with chains that support thousands of TPS, but not millions -- spending $1-2 dollars per month in exchange for owning your own data and interacting with the decentralized state machine and novel applications that are these chains enable seems more than reasonable.

This is only really scratching the surface of new projects that are slowly being released. Aleo has a model with zkSnarks where the client (e.g. you) actually performs the computation, the code is left secure and private, and you end up interacting with a chain that is able to store private state that only you can access. One of the major limitations with the blockchain as it exists today is that the state is entirely public for everybody to see, which leads to a limitation of applications -- it’s entirely possible that in 3-5 years there will be chains with private state, leading to entirely private transactions, alongside private arbitrary computation.

As a result, we are left with a potential vision for what the future of blockchain technology will look like. Cheaper, decentralized, and faster to access storage that will exist permanently leads to faster loading times for websites and a new method of sharing files. Cheaper and decentralized arbitrary computation, with potentially private state, leads to applications that simply weren’t possible in the past, and a recreation of many popular applications today on the decentralized way. All with a built in economy that is inherently more trustworthy than the existing economic model. When you talk to people that are “bullish” on crypto, this is primarily what they are talking about: what crypto can potentially be in 3-5 years, let alone 10-15 years. Sure, there isn’t a lot to show right now in terms of applications: only financial applications currently exist. As more infrastructure projects get built out, however, computation will be easier than using existing cloud providers, and users will pay for any computation they end up using. Sure, this doesn’t necessarily make sense for all applications, but it can make sense for a huge amount of consumer and enterprise applications. And although it doesn’t seem like it right now, you best believe that these incoming decentralized applications will figure out how to appropriately monetize via the blockchain.