HNHacker News
TopNewBestAskShowJobs

vbuterin

753 karma · joined December 31, 2012

submissionscomments
vbuterin··on Ethereum will use around 99.95% less energy post merge
I think you're conflating _consensus_ and _governance_; the two are quite different. It's not PoW vs PoS that allows a chain to resist a coordinated attempt by elites to force a protocol change, it's users personally verifying the chain (and so automatically rejecting chains that violate the rules even if >51% of PoW/PoS nodes support those chains). So no, PoS is not "how our current financial system works". Our current financial system doesn't give people the ability to independently verify anything at all; it's even worse than the most centralized chains in that regard.

I would actually say PoS is more resistant to cabals and regulatory systems than PoW; PoW mining requires huge and visible capital investments and electricity consumption and it's incredibly easy for governments to detect and shut down miners in their own countries (not as true for GPU mining, but GPU-friendliness is difficult to sustain long term), whereas you can be a PoS validator with the most basic computer hardware from anywhere.

vbuterin··on The Most Important Scarce Resource Is Legitimacy
> It seems that nearly every "public good" in the Ethereum ecosystem got gobbled up by VCs and pivoted to for-profit endeavors

And this is because.... we don't have good public goods funding and so once the teams' capital needs get higher they have no choice but to pivot to VC/for-profit. I've talked to many of these teams personally, they are not greedy, but it's just unrealistic to expect people to turn away an opportunity for 100x more money. Turn away 2-5x, sure, but not 100x.

And this lack of public goods funding is exactly what I'm saying the ecosystem needs to focus on solving.

vbuterin··on The Most Important Scarce Resource Is Legitimacy
It is.
vbuterin··on Prediction Markets: Tales from the Election
> Doesn't this explain crypto-user's pro-Trump bias?

It does. But I was never surprised that there were people willing to bet for Trump. I was surprised by the lack of people taking the opportunity that was sitting there for over a month to bet against them.

vbuterin··on Ouroboros: A Provably Secure Proof-of-Stake Blockchain Protocol [pdf] (2019)
When I last looked at Ouroboros it still had the property that there's only "one confirmation per slot", so you need to wait ~log(n) slots before an attacker has a <1/n probability of reversing the chain. I personally think that a good PoS system should strive to have hundreds of confirmations per slot the way eth2's LMD GHOST does. That said, it's very possible that there's an improved version of Ouroboros that already does this and I just haven't caught up to it yet.
vbuterin··on Ouroboros: A Provably Secure Proof-of-Stake Blockchain Protocol [pdf] (2019)
> Second, any consensus protocol can only remain live as long as there are no more than f faults out of 3f+1 replicas. This is again irrespective of the network model

We actually know of consensus protocols that require synchrony including between participants and clients that go up to 99% fault tolerance for both liveness and safety; in fact Lamport himself described one in his original paper (remember that it's "written messages" and not "oral messages" in his vocab that's the relevant category; these days public key cryptography is cheap and uncontroversial so there's no need to care about the "oral" case). There's also a table on page 291 in the 1988 DLS paper (see https://groups.csail.mit.edu/tds/papers/Lynch/jacm88.pdf, the "authenticated byzantine" row and the "synchronous" column) that gives an overview of the fault tolerance levels in various different cases.

The original definition of consensus did not have a notion of passive clients needing to learn the result, so the "active participants synchronous, passive clients asynchronous" model common in blockchain land was not really analyzed well back then. It turns out that with a synchronous network, the passive client requirement is what brings safety and liveness down from 99% to 50%. And it also happens that the 50% fault tolerance protocols are less fragile in the case that the synchrony assumption breaks temporarily.

vbuterin··on Ouroboros: A Provably Secure Proof-of-Stake Blockchain Protocol [pdf] (2019)
There's a couple things wrong here:

> Proof of Stake not only replicates the same dynamics that Bitcoin was designed to eliminate (more wealth -> more power in system)

There is no legitimate sense in which PoS _is_ an instance of "rich have more power and get richer" but PoW is not. In fact, PoW is _worse_ than PoS in this regard, because PoW has economies of scale (if someone with a $100k investment earns $10k/year, someone with a $100m investment can earn significantly more than $10m/year). PoS on the other hand is much closer to a clean "what you earn is proportional to what you put in" design, which is realistically the best that you can do in an anonymous system where users can generate as many independent identities as they want.

> can only be made secure against a maximum of ⅓ byzantine actors, compared to Proof of Work's superior ½.

This is also not true. The 1/3 bound is for safety-under-asynchrony (a form of safety for which PoW's security margin is zero). If you assume synchrony, then PoS protocols' safety approaches 50% much like PoW does. Protocols like Casper FFG combine the "best of both worlds", giving you both of those security guarantees (50% BFT if the network is good, 33% BFT if the network is terrible) at the same time.

vbuterin··on Zcash Open Major Grants
> * the fact that 20% of the block reward is given to the founders as a "fee". i.e. 6 people will end up with 20% of a "global currency".

This applied to the first 4 years of issuance, so it's actually only 10% of the total supply, and it went to not just founders but also advisors, investors, etc, and some was donated to the Zcash foundation, so the portion that was personal profit for 6 people was much smaller! The bulk went into... paying salaries for quite a lot of developers. Since then another dev reward was agreed on (for the next 4 years), but this is going directly to the company+foundation directly building Zcash with no "founder share".

TLDR: the share of the Zcash supply of which "6 people" are receiving through in-protocol issuance is much smaller than 20%.

> * the fact that it's a "privacy focused" crypto-currency and the CEO said "we can make it private, but not for criminals" i.e. backdoored. (https://twitter.com/zooko/status/863202798883577856?s=20)

Zooko did NOT mean backdoors; he meant things like KYC at exchanges and giving people the ability to voluntarily give "viewing keys" to law enforcement that prove which transactions they actually send/received.

See some follow up tweets:

https://twitter.com/zooko/status/863516385426755585 https://twitter.com/zooko/status/864342017605750784

> * the fact it's a crypto-currency with a CEO.

Fair point! Zcash governance is definitely more centralized in that way (though there is a lot of community engagement, as we saw in the [dev fund extension](https://electriccoin.co/reaching-consensus/)). We'll see how it evolves.

> * the fact that the ZK-SNARKs they use (however mathematically beautiful), requires a trusted setup, i.e. prover key(non I/O)

Also a valid point, though this is a big part of why they have been working on Halo, a ZK scheme which does not require trusted setups and still has the benefits of succinct proofs etc.

> * the fact the SNARKS are computationally expensive and no-one uses them. Thus opt in privacy and reduced anonymity set.

These days you only need a few seconds to make a SNARK, even on a phone! But the "opt in privacy" thing is definitely a fair point; I hope that they can escape their bitcoin roots somewhat and get rid of transparent addresses and Script and that whole mess entirely at some point, but we'll see.

vbuterin··on Crypto Fund II
I agree with this. Blockchains are pretty much a novel subset of cryptography, and I expect the most interesting protocols/applications that get called "crypto" in the next ten years are going to make heavy use of both blockchains and other new forms of cryptography (ZKPs, MPC....). There will continue to be a rift between "purist crypto" and "mainstream/enterprise crypto" but it's the purist vs enterprise divide that will matter, not blockchains vs cryptography.
vbuterin··on Ways to surf some of the decentralized web
You can't have decentralized DNS without some form of shared state, where the shared state must resolve the problem of "if I use key X to change the owner to Y, but then later on someone hacks my key and sends a message to change the owner to Z, how does the world know that Y came first?". This is equivalent to the double spending problem, and so a solution to this problem implies blockchains or some other architecture that can support a cryptocurrency.
vbuterin··on Ways to surf some of the decentralized web
> The decentralized web isn't using non-web projects or web abstraction layers. The decentralized web is everyone hosting their own webserver from home. It's easy, everyone has a fast enough connection, and without the motives and requirements of a business it's safe.

This attitude is IMO exactly why the decentralized web movement outside of the blockchain/crypto space has not come further than it has in the last two decades. I run my own website out of a VPS and even that is nontrivial: occasional outages due to SSL certificates expiring, the VPS getting shut down because my credit card expires so the auto-payments stop working, troubles with the server software itself, etc etc. Meanwhile, if you just dump stuff onto one of these decentralized layers or even just a blockchain, you publish once, it's easy and it's there forever.

I truly believe that for "fully decentralized" to be viable, it must mean "serverless" (as in no servers, meaning no single computers whose IP addresses or DNS names or equivalents are explicitly referenced in a line of client-side code, not meaning that AWS lambda stuff). That does mean you're making significant efficiency tradeoffs for the foreseeable future; for some types of applications it's worth it and for others it's not.

vbuterin··on The Hottest New Thing in Seasteading Is Land
I've talked to a bunch of these people. I definitely don't think abdication of social responsibility of any form is the goal. Those people don't bother with the effort of trying to cook up new jurisdictions, and just go to various existing small island nations that give them favorable tax treatment. These are people that genuinely think that governance of physical space is the next great frontier that needs to be innovated/disrupted.

That said, I actually agree with the critique that at least the Silicon Valley-based "let's rebuild new governance zones from scratch" community tends to have a way too uncritical love of capitalism, and doesn't appreciate the myriad ways in which shareholder voting, and fancy new-age analogues of shareholder voting like coin-holder voting, are terrible. I remember one conversation where we were discussing cryptocurrencies, and one person pointed out how "VC chains" (new blockchains funded by VC firms where a small number of VC firms hold the majority of the tokens) were not being received well around the world, and this person was met with blank stares. But this is a medium-sized ideological disagreement, very far from "these people are just selfish freeloaders".

vbuterin··on Final Steps
https://github.com/vbuterin/scalability_paper/raw/master/sca... is the newest one. And here's a google doc modeling transaction throughput with that algo: https://docs.google.com/spreadsheets/d/1wyul_iIKZiAeUprqA2HP...
vbuterin··on Final Steps
Have _I_ ever claimed that Ethereum will "change humanity"?

(for those who don't know, I'm the founder :) )

vbuterin··on Stellar Consensus Protocol: Proof and Code
> Let's say one day China had enough of Bitcoin, and used their essentially limitless resources to gain enough hashing power at will, to block transactions or rewrite them or what have you. Entirely plausible with Bitcoin

That's the Maginot Line attack, at Tim Swanson calls it. The more realistic attack is that China just hacks into five data centers and serves a warrant to another ten. An interesting property of the PoW incentive structure is that there is actually fairly little incentive to protect oneself against hacks, so I would not be surprised if it was fairly easy.

> By hacking enough organizations to take control of their nodes?

The key point in Stellar consensus is that even if enough nodes are hacked, then users can just stop trusting them and switch to other nodes, and so the network would "route around" the damage. With Bitcoin PoW, there's no way to exclude an attacker from participating; you have to accept their work just as much as everyone else's.

vbuterin··on Stellar Consensus Protocol: Proof and Code
You're saying "how do we know that that quora will sufficiently overlap?". David seems to be saying "how do we know they won't?". At this point, I think just about everyone can agree that the security depends on a set of empirical assumptions that we currently simply do not know how they will play out into the real world. So my plan at least is to simply sit back and see how it actually ends up evolving in real life; if it fails, the Stellar Foundation was centralized enough to temporarily take back the reins for itself the last time, so I don't see why it can't do it again.

That said, my _prediction_ is that it will work fine mostly, occasionally there will be concerns about people splitting off into islands, and a resulting second-order consequence is that people will start putting the Stellar equivalent of blockchain.info onto their trust list in order to ensure connectivity to the "main graph" (I had actually cited The Tyranny of Structurelessness in my own responses already :) ), and this will just have to be the social-network-consensus version of the GHash.io scare and we'll be fighting against people's private interests to be lazy to reduce the risk of that happening.

vbuterin··on Tendermint – The completely decentralized consensus engine
Validators are expensive to run. You need to put down a large security deposit in order to be a validator, and being a cheating validator will cost you the entire deposit.
vbuterin··on Tendermint – The completely decentralized consensus engine
> The purpose of a blockchain is to achieve consensus without trusting anyone. If you trust that group, you probably would get along just as well with them managing a centralized database of transactions for you.

See https://blog.ethereum.org/2014/11/25/proof-stake-learned-lov...

Specifically:

> This security assumption, the idea of “getting a block hash from a friend”, may seem unrigorous to many; Bitcoin developers often make the point that if the solution to long-range attacks is some alternative deciding mechanism X, then the security of the blockchain ultimately depends on X, and so the algorithm is in reality no more secure than using X directly – implying that most X, including our social-consensus-driven approach, are insecure.

> However, this logic ignores why consensus algorithms exist in the first place. Consensus is a social process, and human beings are fairly good at engaging in consensus on our own without any help from algorithms; perhaps the best example is the Rai stones, where a tribe in Yap essentially maintained a blockchain recording changes to the ownership of stones (used as a Bitcoin-like zero-intrinsic-value asset) as part of its collective memory. The reason why consensus algorithms are needed is, quite simply, because humans do not have infinite computational power, and prefer to rely on software agents to maintain consensus for us. Software agents are very smart, in the sense that they can maintain consensus on extremely large states with extremely complex rulesets with perfect precision, but they are also very ignorant, in the sense that they have very little social information, and the challenge of consensus algorithms is that of creating an algorithm that requires as little input of social information as possible.

vbuterin··on Tendermint – The completely decentralized consensus engine
I think something like that would happen about an order of magnitude less often than accidentally reusing k values or publishing your private key to the internet :)
vbuterin··on Tendermint – The completely decentralized consensus engine
> As your transaction goes into the blockchain, you can be pretty damn sure that won't be reversed, once enough computing power is behind it.

So, this by itself is not an argument. That amount of computing power was produced once, so you can be pretty damn sure that if the incentives line up it will be produced again just on a different history.

I have a few slides detailing the rather scary game theory behind PoW here:

http://vitalik.ca/files/stanford_presentation.pdf

Particularly note the "Low incentive to protect against hacks". If the major Bitcoin mining farms get hacked, raided by men wearing ski masks and black suits, etc in order to double spend, large 51% attacks become possible, and a coordinated global attack could earn >$10-50m via double spends, and even more with shorting at 10x leverage. But from the point of view of each individual mining farm, they only lose at most 10-50 blocks' worth of rewards multiplied by their hashpower - a total private incentive of ~$100,000. In fact, if everyone else gets attacked, it's in your interest to ensure that you get attacked too!

Also, the "bribe attack" variants all theoretically cost under $100,000. In fact, the largest countervailing incentive factor against all of those is the idea that each mining farm has a decently sized percentage of mining power and thus captures a substantial portion of the "public-good" incentive of the network not failing. Now, you can argue that Bitcoin has not succumbed to these and works fine in practice, but (i) right now, Bitcoin is still tiny, and (ii) if that's true, then naive proof of stake, with its supposedly fatal nothing-at-stake problems, will also work fine in practice.

> Nope, bitcoin blockchain forks very rarely. Where is this claim pulled from?

It forks back one block ~1-2% of the time and two blocks ~0.01-0.04% of the time. You can determine this from simple math plus network latency statistics, or check blockchain.info: https://blockchain.info/charts/n-orphaned-blocks?showDataPoi...

This might not seem like much, but if you are looking at the blockchain as an application platform then users are really not used to the idea of finality suddenly becoming non-final 1% of the time.

vbuterin··on Tendermint – The completely decentralized consensus engine
I think you underemphasize this point. My Slasher algo ( https://blog.ethereum.org/2014/01/15/slasher-a-punitive-proo... ) from way back then does use sampling; Tendermint does not, and thus gains a whole bunch of robustness properties at some cost to efficiency.
vbuterin··on Tendermint – The completely decentralized consensus engine
> you have total control over the seed determining the next set of valdators

There are many possible algorithms for determining this seed, some of which do solve the stake grinding problem. I personally like the NXT algo; its basic approach in simplified form is:

seed(genesis) = 0

seed(block) = hash(seed(block.parent) + block.proposer.address)

No grind capabilities at all. No matter how many times you try, you will always generate the exact same result. The randomness basically comes from absentee proposers, so the only way to influence the outcome (not proposing) is costly as you sacrifice your reward and all other influence over the block beyond signing / not signing it.

vbuterin··on Consensus in Bitcoin: One system, many models
I strongly agree that byzantine fault tolerance theory by itself is a vastly insufficient model for explaining the workings of a cryptoeconomic consensus system. Cryptoeconomic consensus systems actually have to deal with five separate problems:

1. What mechanism is used to process a batch of state transitions during each round? (blocks in bitcoin, a BFT algorithm in Pebble, etc)

2. What mechanism incentivizes correct participation, as opposed to incorrect participation or non-participation, during a round? (block rewards, slasher penalties, etc)

3. What mechanism disincentivizes nodes from replaying a given round or set of rounds after the fact with different state transitions? (ie. the short-range double-spend problem) - block rewards in BTC, slasher deposits in PoS, being voted out in DPOS

4. Assuming a replay does happen, what mechanism do nodes use to determine which of the branches is correct? (PoW scoring in BTC, other scoring algorithms in PoS)

5. What mechanism invalidates long-range forks? (PoW for bitcoin, social checkpointing for PoS systems)

BFT consensus algos are equipped to only solve (1), and even there only part of (1) (specifically, given N participants, how does the next block get produced; the neglected part is how to pick the N participants given a free-entry economic system in the first place). It's an economics problem more than anything else, really.

vbuterin··on Uber introduces surge pricing in downtown Sydney during hostage siege
I'm all in favor of price surging and all that, but it saddens me that Uber spends so little time thinking about PR. Particularly, there's an obvious fix that will probably win them a lot of support: during surges, only charge the 20% fee on the base price. That way they can very justifiably claim to be earning no profit from surge incidents and are increasing fares solely to create enough availability.
vbuterin··on State of Ethereum: August Edition
No, the block count is the only thing that matters. If an attacker has X hashpower, the probability that they will succeed in a 51% attack after N confirmations is (X / (1 - X))^N ; you can derive this formula yourself via generating functions. The block time appears nowhere in that analysis.
vbuterin··on State of Ethereum: August Edition
I briefly mentioned Ethereum 1.5 at the end of my blog post; I'll soon cover some clever ways of creating quasi-scalable dapps using the ETH blockchain as a usually-inexpensive security backend.

Believe me, consensus and scalability are the two largest problems on my mind. We don't have nice solutions (yet :) ) because the problems are fundamentally hard.

vbuterin··on BitsharesX Catapults to No. 3, Leaves Ripple Behind
> Accordingly, I'd argue that this is a more accurate representation of market cap (view by total supply): http://coinmarketcap.com/currencies/views/market-cap-by-tota...

That one is going to be fun if we (Ethereum) decide to premine 2^200 ether into a few standard-library contracts (eg. ecrecover, bloom filters, data structures) to compensate people a few microcents per call for the mismatch between their gas cost in a naive EVM implementation and their actual execution cost...

Edit: although I suppose you could argue that currency units which are existent, but which cannot be accessed for 1000000000000000000000000000000 years because they are locked up in a contract that no one can open up at anything more than a trickle rate, actually inexistent. Will get complicated in any case.

vbuterin··on Inside a Chinese Bitcoin Mine
They also only need one connection; they can simply split the data across all the workers. They will need to send shares from all the workers if they're not solo mining, but even those are less than 100 bytes each, and if that becomes too much of a nuisance they can just solo mine.
vbuterin··on We’re tackling aging
Okay how about I'll send you a contract to transfer all of your property to me and then you'll be free to spend the rest of your life in the Amazon rainforest. If you get malaria, don't worry, it's natural. Just learn to leave things be.

Exactly the same argument applies here.

vbuterin··on Stellar
There is absolutely nothing wrong with premining, if it is done honestly and transparently and especially if your coin is using a consensus algorithm that has no need for regular mining in the first place. Let's evaluate the fairness of the distribution model on its merits. This one is quasi-centralized (relying on Facebook accounts), but I predict it will be among the most egalitarian.
← PreviousPage 2 of 9Next →