Vitalik Buterin’s Academic Dishonesty

Vitalik enjoys criticizing other projects, notably Bitcoin. The differences between those two blockchains are vast, so a more simpler case study is his critique of EOS. EOS and Ethereum have a huge number of similarities. Both are attempting to use Proof of Stake, both have mostly centrally premined stake (lets ignore for now), both are trying to bring easier to use smart contracts to the world as alternatives to Bitcoin’s contracts.

No matter how you feel about EOS, it would be desirable for reviews by those pretending to be academics to be based at least a little on reality. The point is not to defend EOS, but to combat dishonesty. Vitalik is constantly caught criticizing the other project in long write ups while getting even basic information incorrect and most of his wild claims with no citations go unchallenged. The academic dishonesty of not even reaching out for help with something he hasn’t bothered to read about but delivering writing as a fact helps nobody.

Here’s an example of a complete lack of effort to verify very basic design (and constantly repeated):

(source https://steemit.com/eos/@dan/response-to-vitalik-buterin-on-eos )

To which Dan simply linked the publicly available white paper https://github.com/EOSIO/Documentation/blob/master/TechnicalWhitePaper.md#merkle-proofs-for-light-client-validation--lcv- with an entire section on this topic, part of which is shown here:

That’s the level of (no) expertise we’ve come to expect from Vitalik. There were additional corrections during that exchange: https://steemit.com/eos/@dan/reponse-to-vitalik-s-written-remarks

Let’s consider what is wrong here:
(source: https://vitalik.ca/general/2018/03/28/plutocracy.html)

Here’s a short summary of how much was wrong in 1 pagagraph:

  • in DPoS the reward is not constant, but can be changed. DPoS existed since 2014, long before Ethereum, first, in a blockchain network called Bitshares from same Dan Larimer of EOS. Bitshares witness reward and schedule has been changed by a voted-in committee as shown here http://archive.is/FuTtn so you would think even a google search was done on this topic. Even if those weren’t the case, the voters can change producers to alter block reward to another number.
  • If Vitalik asked anyone working on DPoS about incentive/penalty mechanisms for encouraging voters or honest behavior, he would’ve known the answer. A high level summary:
    ○ inflation —Voters have every incentive to reduce inflation. If producers can afford to bribe, they are overpaid and the inflation can simply be reduced.
    ○ time vested stake —Every voter is required to lock their coins for long periods of time when voting or using the platform. This is required to align incentives of the coin holders with the value of the platform so however their actions affect the platform, they won’t be able to jump ship and will face the consequences. This incentive is at least on level of ASIC miners equipment losing value if they choose to attack and devalue the network. Additionally, it’s not just those who want to vote that do this, but anybody hosting contracts or transacting who locked their stake for another reason and thus has incentives to keep that stake value safe by thinking about platform’s well being in the future for it not to lose value. The last thing they need is to vote malicious witnesses in.
    ○ Block Producers have incentives not to behave in any manner seen negative to value of the platform by the free markets. DPoS was the first solution to the “nothing at stake” problem by putting the block producer jobs on the line. Since job status highly depends on reputation and performance, once the job is lost for a good reason, it’s unlikely for kicked out producer to ever come back, which is equivalent to opportunity cost of salary for rest of time (lets call it infinity) in incentive not to cheat. This is significantly larger than anything slashing is capable of and why it’s deemed not necessary. Second reason is leverage can easily compensate for slashing of bonds, but not this enormous cost. Additionally, negative behavior by producers would also drop the value of the platform down likely larger than by 1% annually, reducing their income.
    Something Vitalik doesn’t seem to understand is that incentives and penalties can be more than a simple fine system the scale of which in Casper appears to be a completely arbitrary magic number anyway.
  • calling EOS, and not their own Proof of Stake, a plutocracy “governed by the wealthy” is complete favoritism when both are based on wealth. It will probably be easiest to show why Vitalik’s proof of stake will likely be far more like a plutocracy. Consider a common comparison table below:

The very act of introducing slashing requires bonds for weighting which creates direct plutocracy while achieving less significant punishments. The goal of approval voting, requiring no bonds, and weighting producers equally at larger number than naturally happens in Pareto weighted distribution are specifically to break the centralization that creates. Another visual graphic on this here: https://i.imgur.com/rhPiMiG.jpg

Who’s approach is better? Not as important as Vitalik’s incorrect or maliciously deceptive work being presented as academic analysis to promote his platform.

From http://archive.is/2KaMZ :

  1. It’s specifically 30 candidates maximum per account to create diverse top 100+ runner ups
  2. If 30 accounts are competing for 21 spots, all of them voting for each other doesn’t make them win. The difference creates votes that cancel other votes out with respect to rank. That’s how approval voting works.
  3. A BP voting without having stake has no effect. If stake is compromised, that’s an issue in any PoS.

Once again statement presented as fact wasn’t even fact checked. Maybe an article about voting should get details about voting correct.

Quick summary on other criticisms he’s often misleading on:

  • His critique often involves mentioning his own off-chain coin voting as example of low participation during bailout carbon vote, while DPoS chains show order of magnitude higher on-chain participation. The only thing his arbitrary poll shows that he used to make decisions is that his take on what’s considered acceptable “governance” is based on no credibility.
  • He claims DPoS nodes could be shut down or taken over. So can any nodes. With exact same methods. In fact, it might be easier to do to fewer nodes with Pareto based distributions like regular PoS has or pool operator nodes.
  • He likes to bring up Lisk which modified the system where only voting for top nodes gets users paid which promotes only voting for top nodes already in power. In fact, with 101 top nodes users are only allowed to vote same number of 101 times. This isn’t typical. That’s not part of majority of DPoS designs as well where often runner ups are paid to encourage runner ups to be online, compete, and be ready to take over for any reason.

When Vitalik discusses blockchain incentives, it’s important to remember this is the same person who considers his 70%-100% centralized premine “small” despite its obvious effect on both market value of incentives and control via stake. Vitalik’s (and Dan’s) distribution method of choice also literally incentivised centralization:

It’s difficult to imagine a less qualified person to discuss incentives or present facts than Vitalik with his history of delivering countless security failures on his live network and constant streams of misinformation.

Significant amount of work that Vitalik claims as entirely his own has already been done by others, but usually renamed and repackaged as their own.

Plasma is continuation of research on bitcoin’s channels & lightning network & sidechains / childchains (ARDR) / asset chains (KMD)

Sharding is continuation of research on bitcoin’s tree-chains

Casper is continuation of research on countless proof of stake and master node designs including something at stake solution of DPoS (2014). In fact, Dan Larimer helped Vitalik identify early Casper issues (source 2015).

And more: