Last night I attended the EOS meet up in Los Angeles and got to hear directly from David Moss, SVP of Tech Operations. I would like to share my thoughts, but let me start by saying that I am a huge fan and active supporter of EOS. My hope is that these modest criticisms will only make the community stronger. Go team EOS!
The question I see most often amongst developers is that of budgeting for staking and leasing. In other words, you’ll need tokens in order to run your dapp, but we don’t know exactly how this will work from a technical or governance perspective yet. I asked David to explain it to me and he basically said that we won’t know any details until early 2018. So what happens if a developer creates the dapp version of Facebook two years after the launch of the mainnet when EOS tokens are trading at $500? They certainly won’t be expected to buy a bunch of them. In theory, they’ll just lease resources from current token holders, but how exactly does that work? Is it automated? Dan Larimer and David have both referenced leasing/renting EOS at various times, but developers need to understand the actual mechanics in order to confidently build the dapps we’re all supposed to be building at this very moment. We should probably stop making fun of Ethereum Gas until we sort this out.
Steemit and TPS
Steemit gets mentioned a lot. Personally, I find it to be a bit disingenuous for two main reasons. First, Steemit keeps getting mentioned alongside Reddit. Steemit has 35k active users, Reddit is the #8 website in the world and has half a billion monthly active users. I would love to see Steemit grow and become massively successful, but let’s be clear about our comparisons. Second, high TPS (Transactions Per Second) is not the sole magical solution to creating the decentralized future of our dreams. There is a chart that block.one likes to show that depicts the TPS of large blockchains and it makes it look like Bitcoin and Ethereum are smaller than Steem. Neither of those blockchains are (currently) built to handle high TPS so of course they’re smaller. It feels very apples and oranges to me.
Semantics and nit picking aside, high TPS is great — and it should enable a new class of applications built on top of blockchains — but ultimately such raw processing power is only cool to the crypto community. The metrics that really matter are actually all related to the health of the ecosystem: number of developers, number of active users, etc. It doesn’t even really matter if the EOS marketcap is $3 billion and the repo on github is going bonkers 24/7; if EOS becomes a super computer with no users then we’ve all failed (unless you’re only in it for the ICO). I really don’t want that to happen. It would be awesome if EOS would start to share more useful metrics about the growth of the ecosystem.
The Token Distribution
The token distribution has been somewhat controversial, but it certainly appears to be working. I personally feel like most of the common concerns are unwarranted. My main concern about it actually comes from looking at Steem and dapps like Steemit. Steem has a current marketcap of ~$280 million, but the dapps being built on it are arguably very much in their nascent stages and their value has yet to be demonstrated in a way that correlates to the value of Steem in a sane way, IMHO. What this means to me is that the vast majority of people treat Steem like an asset and not like a platform on which to develop. My fear is that EOS could follow this same path. Instead of being the Ethereum killer from a technology perspective, we may well be creating a coin that 99% of the population uses solely for speculation. In fact, I caught up with David after he spoke and told him that the very concept of staking/leasing encourages people to “hodl” their tokens and, as these tokens increase in value, might only further serve to attract even more currency speculators. What I’d like to know more about is whether or not EOS could face a similar problem to Bitcoin wherein the utility has gone to zero in an inverse relationship to its price. Staking/leasing might even contain a solve this issue, but in the meantime the EOS telegram is rife with “When lambo” and unsophisticated buyers from unvetted exchanges who have no idea there’s a decentralized operating system in the works. I would love an in-depth understanding of how EOS might mitigate against the pressure to become “only” a currency.
Delegated proof-of-stake vs. decentralization
Vitalik’s joke about EOS being Ethereum 2.0 appears to have caused a rally in the EOS token price, but if you watch the entire video he makes the argument that speed (e.g., TPS) has an inverse relationship to decentralization. In other words, decentralized systems are slow because all the computers need to verify all the transactions. When you try to speed things up, you have to use fewer computers, and that creates a centralization risk. I’ll leave this debate to Dan and Vitalik, but what interests me are the nuances of governance and dispute resolution on EOS. According to Dan’s own words, Delegated Proof of Stake (DPoS) — a key feature of EOS — relies on votes and fair token distribution in order to function. I have two thoughts about that. First, if it’s so crucial, why haven’t we attempted to ratify the constitution? Is the constitution independent of voting and token governance? Second, although the length of the ICO has no doubt opened the doors to more people, we have to assume there could be “super stakers” who’ve snatched up a lot of EOS and will come to dominate the system — just like what happened to mining in Bitcoin. What are the ramifications of stake centralization? Is it already happening?
Here’s a summary of what I’d like to know:
- How does staking/leasing actually work? Is there a potential for super stakers to centralize the leasing of EOS compute power, etc. and cause inequity in the system?
- When will we see more useful metrics for ecosystem health and not just stats related to engineering/crypto achievements?
- When will we see a proposal for the constitution? Can governance and/or staking and/or clever token distribution techniques be used to counterbalance the strong pressure to ignore the OS being built and treat EOS solely as an asset?
EOS is one of the most exciting projects I’ve read about in years. I really want it to work, but that means I’m concerned about some of the issues we’ve seen in other blockchain projects. It would be such a bummer if resource consolidation, messy governance, or the favorability of the asset over the platform lead to EOS not reaching its full potential. I hope we can discuss some of these details and work together to create something the world has never seen before. Thanks for reading!