Ask HN: Is blockchain tech already everything we need it to be?
13 comments
> Is it possible that blockchain has already reached maturity
Simply put, not at all. There are several massive developments that need to occur. These are mostly in the context of Ethereum but likely all blockchains would benefit from such changes:
- Complete transition to proof-of-stake on all popular web3 chains (i.e. Ethereum)
- Better decentralized staking pools, more accessible staking software/hardware solutions
- Scalability through rollups, data sharding, and zero-knowledge proofs
- Light clients and data availability sampling (eg: so your client can verify state of chain w/o downloading GBs of data)
- Better hardware: cold wallets, zk-proof generation, staking
- Better ways of auditing, formally verifying and interacting with contracts
- Account abstraction (EIP-4337)
- Post-quantum cryptography
I don't think its possible to predict how long this will take to mature, perhaps another several years (5-10). And very likely, by that point, there will be new research and new problems to solve. This can be seen with, for example, zero-knowledge research being something that has been around for decades, but succinct zero-knowledge proofs being only introduced in more recent years, and leading to sudden and profound changes in the direction of blockchain development.
Simply put, not at all. There are several massive developments that need to occur. These are mostly in the context of Ethereum but likely all blockchains would benefit from such changes:
- Complete transition to proof-of-stake on all popular web3 chains (i.e. Ethereum)
- Better decentralized staking pools, more accessible staking software/hardware solutions
- Scalability through rollups, data sharding, and zero-knowledge proofs
- Light clients and data availability sampling (eg: so your client can verify state of chain w/o downloading GBs of data)
- Better hardware: cold wallets, zk-proof generation, staking
- Better ways of auditing, formally verifying and interacting with contracts
- Account abstraction (EIP-4337)
- Post-quantum cryptography
I don't think its possible to predict how long this will take to mature, perhaps another several years (5-10). And very likely, by that point, there will be new research and new problems to solve. This can be seen with, for example, zero-knowledge research being something that has been around for decades, but succinct zero-knowledge proofs being only introduced in more recent years, and leading to sudden and profound changes in the direction of blockchain development.
> I also feel crazy in that there seems to be consensus that blockchain is suppose to do everything, or should be doing more than it does. In all of this, I am viewing the problem space from the domain of infrastructure, agnostic of application. Is blockchain tech already everything we need it to be?
This is where the line is drawn here and we start to see the problems in using 'blockchain for everything' which as silly as using one tool / language for every. single. application, which is what Ethereum originally proposed.
As I already said [0], since that has failed to scale on-chain, it cannot be used for many serious applications without overpaying for fees. If it cannot manage to scale and keep its fees low, then its possibilities are quite limited for anything on-chain.
Like Bitcoin, both of them have proven after years of existence, waiting for improvements and promises had no notable improvements for either of them for on-chain scaling.
One of the only useful applications for these blockchain tech is and has always been payments. That is why companies like Stripe [1] have gone and made a crypto payouts product with that which as always needs to use a layer-2 side chain since they also know that Ethereum itself is too expensive for on-chain payments.
Highly likely in the future Stripe will add a faster and cheaper on-chain layer one blockchain that can scale faster than both Ethereum and Bitcoin and is suitable for payouts.
[0] https://news.ycombinator.com/item?id=31031572
[1] https://stripe.com/blog/expanding-global-payouts-with-crypto
This is where the line is drawn here and we start to see the problems in using 'blockchain for everything' which as silly as using one tool / language for every. single. application, which is what Ethereum originally proposed.
As I already said [0], since that has failed to scale on-chain, it cannot be used for many serious applications without overpaying for fees. If it cannot manage to scale and keep its fees low, then its possibilities are quite limited for anything on-chain.
Like Bitcoin, both of them have proven after years of existence, waiting for improvements and promises had no notable improvements for either of them for on-chain scaling.
One of the only useful applications for these blockchain tech is and has always been payments. That is why companies like Stripe [1] have gone and made a crypto payouts product with that which as always needs to use a layer-2 side chain since they also know that Ethereum itself is too expensive for on-chain payments.
Highly likely in the future Stripe will add a faster and cheaper on-chain layer one blockchain that can scale faster than both Ethereum and Bitcoin and is suitable for payouts.
[0] https://news.ycombinator.com/item?id=31031572
[1] https://stripe.com/blog/expanding-global-payouts-with-crypto
If there's some way to actually solve the scaling problem [1] meaningfully, it might have a chance to be more than niche. Broadly applies to Ethereum and all other existing blockchain systems too. Ethereum can handle ~25 transactions per second at the moment. Commit a namespace entry? Transaction. Send coins? Transaction. Any smart contract interactions? Transaction. The actual Web 3.0, where all our name resolution tables, and instant messaging, weekly paycheques, grocery purchases, etc. are logged on some blockchain, would need, very conservatively, tens of thousands of of transactions per second.
I do not anticipate a resolution to the scaling problem any time soon [2]. But maybe they, and me, are not seeing the flaw in their argument. Should the scaling problem remain intractable, I think that the existing major blockchains have peaked. Both Ethereum [3] and Bitcoin [4] have been pinned near their maximum transactions per second for years now. Further adoption can only come by displacing some other transaction on-chain.
(I'm operating on the assumption that off-chain non-blockchain handling of blockchain data is not actually blockchain for this comment. No keys, no ownership as the BTC maximalists like to say.)
[1] https://en.wikipedia.org/wiki/Bitcoin_scalability_problem
[2] https://datafinnovation.medium.com/sharding-is-also-np-compl...
[3] https://etherscan.io/chart/tx
[4] https://www.blockchain.com/charts/n-transactions-per-block
I do not anticipate a resolution to the scaling problem any time soon [2]. But maybe they, and me, are not seeing the flaw in their argument. Should the scaling problem remain intractable, I think that the existing major blockchains have peaked. Both Ethereum [3] and Bitcoin [4] have been pinned near their maximum transactions per second for years now. Further adoption can only come by displacing some other transaction on-chain.
(I'm operating on the assumption that off-chain non-blockchain handling of blockchain data is not actually blockchain for this comment. No keys, no ownership as the BTC maximalists like to say.)
[1] https://en.wikipedia.org/wiki/Bitcoin_scalability_problem
[2] https://datafinnovation.medium.com/sharding-is-also-np-compl...
[3] https://etherscan.io/chart/tx
[4] https://www.blockchain.com/charts/n-transactions-per-block
This seems to ignore Ethereum's rollup-centric approach to scalability. Many of those interactions you mention (such as sending coins, interacting with a smart contract) can be batched and, in a single L1 transaction, their integrity verified cryptographically with succinct zero-knowledge proofs (zk-SNARK/zk-STARK).
To the user, eventual UX is similar. Direct CEX on/off ramps to rollup, trading, DeFi, NFTs, smart contracts, etc. It is all "on-chain" by virtue of it being cryptographically verified on mainnet Ethereum (and, likely in less time than the finality of a single BTC transfer).
To the user, eventual UX is similar. Direct CEX on/off ramps to rollup, trading, DeFi, NFTs, smart contracts, etc. It is all "on-chain" by virtue of it being cryptographically verified on mainnet Ethereum (and, likely in less time than the finality of a single BTC transfer).
Everything about eth scaling is just kicking the can down some other chain. L2 and Rollups are the same sidechain paradigm from 2014 with newer, more fragile math.
Payment and State Channels are the only true way to scale. Off-chain txs. Lightning network works quite well today for Bitcoin and I look forward to seeing entirely new paradigms evolve in time for Eth.
Payment and State Channels are the only true way to scale. Off-chain txs. Lightning network works quite well today for Bitcoin and I look forward to seeing entirely new paradigms evolve in time for Eth.
Lightning Network is a Layer 2. State channels also exist on top of Ethereum, such as Raiden Network and others[1].
The Eth core devs and users are more interested in rollups due to them being general-purpose[2], i.e. suitable for applications and smart contracts to build on top of (for e.g. Uniswap on a rollup via zkEVM[3]).
[1] https://ethereum.org/en/developers/docs/scaling/
[2] https://vitalik.ca/general/2021/01/05/rollup.html
[3] https://scroll.io/blog/zkEVM
The Eth core devs and users are more interested in rollups due to them being general-purpose[2], i.e. suitable for applications and smart contracts to build on top of (for e.g. Uniswap on a rollup via zkEVM[3]).
[1] https://ethereum.org/en/developers/docs/scaling/
[2] https://vitalik.ca/general/2021/01/05/rollup.html
[3] https://scroll.io/blog/zkEVM
Ya, I know. kick that can down the road.
Not everything needs to be recorded for all eternity.
Not everything needs to be recorded for all eternity.
Blockchain trades efficiency for decentralization. In general, most people find centralized apps are more efficient --- cheaper, faster, simpler and more user friendly.
Blockchain addresses a problem that most people don't have --- a lack of trust in centralized solutions --- aka paranoia. For example, centralized banking handles transactions easily, efficiently and securely enough for most legal purposes.
Blockchain addresses a problem that most people don't have --- a lack of trust in centralized solutions --- aka paranoia. For example, centralized banking handles transactions easily, efficiently and securely enough for most legal purposes.
What I want to know is how realistic it is that you can create a blockchain app at scale without using "centralized" resources. Can you build a social media app clone that is used by millions without using AWS or Google Cloud or Azure? Because if not then what exactly is the point of decentralizing everything if you still depend on a few key players?
Can you build a social media app clone that is used by millions without using AWS or Google Cloud or Azure?
The only reasonable way to do this is to create/mint money to pay for the cost of the infrastructure. Success depends on the extent you can convince others to accept your currency.
The only reasonable way to do this is to create/mint money to pay for the cost of the infrastructure. Success depends on the extent you can convince others to accept your currency.
I may be oversimplifying, but if I look at Web3 as the existing internet, but without central intermediaries; then most applications can be built this way today. If we want a Clubhouse, or Zoom clone, these could be built atop existing WebRTC infrastructure. A Whatsapp clone can be built atop libp2p messaging (again I oversimplify). From the internet infrastructure view, the part that was missing has always been mechanisms of identity, ownership, and payment (trust mechanics in general). Today, blockchains handle these in a variety of formats, such that a project can just pick the chain that fits best with the application architecture. In 'Web 2.0', if I want auth, payments, file storage, etc - I would simply stitch together one or more infrastructure providers such as AWS. This feels very natural today.
Is it possible that blockchain has already reached maturity for the value it brings to the decentralized web? Are there any popular applications today that could not be built peer-2-peer in this context?
The big gap I still see in this view is who pays for services that are free in Web 2.0 (such as Facebook). But even in this case, blockchains provide a variety of incentive mechanics where increasing usage could appreciate a digital token, which in turn gives incentive for people to run the underlying infrastructures.
I also feel crazy in that there seems to be consensus that blockchain is suppose to do everything, or should be doing more than it does. In all of this, I am viewing the problem space from the domain of infrastructure, agnostic of application. Is blockchain tech already everything we need it to be?