The Insanity and Brilliance at Ethereum’s Developer Conference
It’s 4am and I’m sitting in a hotel room in Shanghai, China. The second ever Ethereum developers conference is set to kick off in a few hours. The time-zone shift woke me up early, and I’m sitting in my bed playing with my newly modded phone. The Chinese government censors most large American mobile apps behind something colloquially referred to as “The Great Firewall” so you need to use all of these shoddily made, Chinese government approved, replacement apps instead. The preferred app for messaging is an app called WeChat, a bizarre Slack / Venmo / Tinder hybrid that 800 million people use to communicate every day. We have a group chatroom for the conference going, which I’m scrolling through when my phone buzzes. Someone posted a new message. “The Ethereum network is under attack”.
Ethereum is this insane thing going on in the tech world right now. It’s a new cryptocurrency that is picking up the pieces of the fractured Bitcoin ecosystem and trying to succeed in accomplishing what Bitcoin has failed to do, deliver the first batch of successful consumer applications built on top of blockchain technology. You can think of Bitcoin a bit like a shared google doc spreadsheet that anyone in the world can send data to, one global ledger of accounts and balances. Ethereum builds on this idea by providing a mechanism for users to add custom formulas into the cells. Anyone who has done any kind of Microsoft Excel programming knows how powerful spreadsheets can become with just a few simple formulas. Ethereum is the first working implementation of programmable money, a simple concept that may one day bring the entire global financial system to it’s knees. And it draws brilliant people from all walks of life into its orbit.
The 4AM attack on the network is bizarre in that it’s not really a malicious hack, but more like comical mischief. Ethereum runs as a distributed system. At any given time, there will be thousands of separate computers (also called nodes) running the Ethereum protocol and keeping the network online. The actual software that the computers use to run the Ethereum protocol comes in many different flavors, the most popular one is an application called Geth, written in Google’s Go programming language. 85% of all Ethereum nodes are running Geth at the time. The attack pushes a piece of data onto the blockchain that exploits a bug in the Geth software causing the program to crash. Because every live node syncs the latest version of the blockchain in real-time, just like that 85% of the Ethereum network goes offline. Some developers are able to respond by switching over to the Parity Ethereum client, a new client written in the high-performant Rust Language that is currently gaining market share, but for others switching over to Parity is too risky without having time to run extensive testing. The Ethereum network is essentially frozen until an update can be pushed to Geth.
The main developers of Geth are asleep in Shanghai. They might not even have their laptops with them, as it’s risky to take your work computer into mainland China if it contains any sensitive information. They’re also scheduled to go on stage and give a presentation in a couple hours. When they wake up, do they fix the bug and miss their talk or go onto the stage? Like I said, comical mischief. People slowly start to wake up and news of the attack starts spreading. In a back room somewhere developers start pounding on their laptops. In a matter of hours, the bug is identified, an update is shipped, the nodes download the latest version of the software and go back online. Crisis averted. The net result of the attack? The conference start time gets pushed back half an hour. The Geth developers come out on stage to a raucous applause. Just another day in the cryptocurrency world.
Ethereum has a certain resiliency to it. It reminds me a bit of being in Israel. When you live under the constant threat of attack, you develop a desensitization and rationality around the notion of outside threats, so you respond to them with logic instead of emotion. The term ‘anti-fragile’ (meaning something that becomes stronger every time it is put under attack) has been a buzzword in Silicon Valley lately, but Ethereum has thus far personified it. Nobody is overly phased by the 4am attack. It’s doubtful that such a large portion of the network will run the same client software at the same time in the future. This attack vector won’t work again.
DevCon is the single best tech conference that I’ve ever been to. It’s a firehose of technical information and ambitious product demos blasted at you non-stop over the course of three days. There’s an energy and an optimism that I’ve never felt before, and an underlying belief amongst the community that Ethereum has finally gotten it right and we’re t-minus one year away from cryptocurrency truly exploding into mainstream consciousness.
While other technical communities are bickering about who can find a more esoteric reason to censor a conference speaker, Ethereum is moving along with a laser focus on solving real problems, like how to scale the blockchain to thousands of transactions per second, how to make developer tooling more approachable, and how to keep the network running despite the constant barrage of hacks and DDoS attacks being levied against it. At DevCon you’re just as likely to find a web developer, a systems engineer, an academic, or an MBA. A transexual, an avid Trump supporter, a Chinese entrepreneur, a New York City venture capitalist, or a techie in a hoodie who has 5 million dollars worth of Ether. It’s a safe space for eccentric personalities. You still have your typical discussions at the bar about immutable databases and functional programming like you would at any tech conference, but the whole thing has an elevated sense of scope around it, like these discussions you’re having actually matter somehow.
The biggest news from the conference comes in the form of two competing visions for Ethereum’s eventual switch to a proof-of-stake model. A make-or-break release under the code name “Casper”. The proof-of-work model was one of the brilliant inventions of the original Bitcoin protocol. A system that manages to align the incentives of all actors towards keeping the network running honestly and investing in the future of the currency. Ethereum is switching away from this battle-tested methodology to something more ambitious and unproven. Something that, if it works, will consume less electricity and allow for faster transaction processing times.
Vitalik Buterin, the 22-year old wunderkind behind Ethereum, presents his “Mauve Paper”, a well-reasoned approach for how to move the Ethereum network to proof-of-stake. Vlad Zamfir, a theoretical mathematician, later unveils his competing vision for proof-of-stake. Vlad speaks at 100 miles per hour. “I feel bad for the Chinese translators” someone says. The entire conference is being live-translated into Chinese, or it least it was until Vlad took the stage. On cue I see a Chinese girl take her headset off, the translators had no chance on this talk. “I need a presentation about this presentation” someone else says. Two competing models, unveiled publicly in front of the community. The better model will win in the end.
I meet a guy from Germany who runs one of the biggest Ethereum mining rigs. He shows me a picture of his setup on his phone. I’m blown away by the scope of it. A massive farm of servers all working 24/7 to solve esoteric math problems to ensure that new transactions get mined into the Ethereum blockchain promptly. “How much electricity does that use?” Someone asks him. “A couple million dollars a month” he answers. But it’s still profitable, very profitable, because the Ether it produces is worth so much. He goes to Vitalik “You know, you better make sure that Casper is well tested before switching over. It needs a lot of testing.” He’s half-joking, but there’s a vague semblance of hostility in there. The longer it takes for the Casper fork to go live, the more money the miners will make. Moving to a proof-of-stake model renders their massive hardware farms irrelevant. Every month that goes by before Casper goes live has a multi-million dollar impact for miners running the network. No pressure or anything.
Alex Van De Sande unveils the next version of the Mist browser. It’s the most impressive product that’s demoed at DevCon. Mist is a next generation web browser that runs on top of the open source Chromium engine powering Google Chrome. It lets you browse the internet just like any web browser does, but it also hooks into your Ethereum wallet, meaning you can trivially pay for any kind of content as you’re browsing. No credit card needed, no bank account needed, no obnoxious sign-up forms. It will unlock an entire new class of digital business models.
Juan Benet unveils Filecoin. A cryptocurrency built on top of Ethereum that will pay users to store files, creating a distributed version of Amazon S3 for the world to use. A BitTorrent protocol on steroids. Juan announces that the team thought long and hard about how to build it, but eventually decided to build on top of Ethereum due to the incredible community emerging around it. The crowd goes wild.
I go to lunch with a couple of millionaires. They call themselves the next generation of “Oil Tycoons”. Crypto ballers who were savvy enough to buy up massive amounts of Ether during it’s initial offering to the public. Ether has increased in price substantially since it’s launch, something like 40x. Traditional investments don’t return 40x the money that you put into it 18 months later, but Ethereum just did. Again, crypto is insane.
And all this is barely even scratching the surface. You see a new UI for the Parity client, academic papers on formal verification of smart contracts, mobile apps that connect to the Ethereum network, and beyond. There’s a black hole of interesting topics to understand in the blockchain space. A friend tells me about the term “chronocracy”. The way you can have a “plutocracy” or a “meritocracy” a “chronocracy” is an ecosystem where your status is defined by how much time you have put into it. There’s so much to know about how the low level protocols work, about how the current financial system works, that the status hierarchy at DevCon is a proxy of who has put in the most time to understand it all.
You can’t talk honestly about the insane world of cryptocurrency without also acknowledging that the whole thing has a bit of a dotcom era vibe to it. I get the impression that companies are throwing a decade of best practices out the window because they used the word “blockchain” in their sales pitch. Over the past ten years Silicon Valley has evolved somewhat of a repeatable model for building successful technical ventures. A couple of smart founders, even if they’re just broke kids living in their offices, build a prototype of an innovative product that solves a simple problem they have. They hack on it until they start making money from it, realize that they’re onto something bigger than they ever imagined, get investors on board that believe in their vision and scale their companies from there.
Blockchain does it backwards, you have companies with 15 people and $10 million in funding that can’t even explain what they’re building. You see companies throwing parties before they’ve even launched their product, or talk about disrupting massive industries before they’ve even proven they can execute on a simple task. Founders who would get laughed out of the room in today’s startup climate are somehow swindling investors out of their money again because blockchain.
One company holds large amounts of Ether and builds open-source tooling for the Ethereum ecosystem. Normally open-source tooling is hard to monetize, but if the open source tools lead to better adoption among developers, and thus more people building on Ethereum, then the price of Ether will go up, and the company will make money. It’s a business model I’ve never seen before, but maybe it’s just crazy enough to work. Welcome to crypto.
I meet someone who is right out of college and a contributor to the core code base of the Solidity compiler. I mentioned that Ethereum is a shared spreadsheet with formulas. The Ethereum Virtual Machine (EVM) is what runs the code in these formulas which are normally referred to as “smart contracts.” The most popular way to write smart contracts is in a programming language called Solidity, but that Solidity code has to be compiled down to EVM bytecode before it can be executed, that’s where this compiler comes in. “If you asked me a year ago I never thought that I’d be writing compilers today” he tells me. It’s incredible someone could pick it up so quickly, but also a little disquieting. Can a bunch of millennials with little real world experience really hack together something that will eventually displace the U.S. dollar? He works for a company that sells smart contracts to Big 4 consulting firms. They’re one of the only companies in the entire blockchain space that actually makes money.
Even with the flashing red warning signs everywhere you look, it doesn’t diminish the point that people are right to be excited about this. The stewards of many large centralized technology products have begun to hit a plateau, and are failing to push us forward. I don’t want to simply settle for the world they’ve constructed when there are obvious ways to improve upon it. I just have this underlying feeling that something bigger is around the corner. You can see it everywhere if you pay attention.
I’ve been making coding screencasts on youtube recently and just eclipsed the 1,000 views / day threshold where you can start to think about making a bit of money from it. The monetization options youtube presents to content creators are outright pathetic. You can put ads on your videos, but if a user skips it you don’t get paid. If a user has Adblock you don’t get paid. And ads are a terrible way to monetize video content anyway. You can try to charge for the content directly, but youtube reserves the right to alter the amount you are charging at any time. No joke you don’t even have the right to choose how much to charge for the content you’re making on their platform. A much better model would be to have users pay content creators directly, even with tiny transaction values. This isn’t possible with our current financial system where purchases of digital content for less than $3 are not only outside the social zeitgeist but also not feasible for most credit card processors. But with Ethereum, sending small amounts of money directly to people who warrant it is exactly the point. On an Ethereum enabled video platform you could send a penny to the content creator every time you view more than x minutes of content. It’s a fundamentally better model for video content than the one youtube offers.
Twitter put the final nail in it’s coffin when it made the decision to permanently ban Milo Yiannopoulos — a popular conservative personality — from it’s platform without explanation. When social networks that once marketed themselves as a public utility for free speech start banning accounts for wrong-speech, sanitizing their content in an effort to become more “advertiser friendly” it marks the beginning of the end. Whatever replaces Twitter will use blockchain technology to prevent censorship and hopefully create a better public interface to your identity than the one Twitter has proven itself capable of providing.
The Apple App Store has failed us as well. Monetizing mobile apps is brutally hard for developers, and the entire consumer mobile application space has ground to a halt. Instead of needing to raise tons of money, needing to scale to a massive user base, or needing to force obnoxious ads on your users to be viable, it would be a much better model if a user that opens your app x times in a month automatically sends a dollar to the developer who made it. Then developers could optimize for a pleasant experience instead of an adversarial one. There’s nothing forcing us to live under a centralized entity tithing 30% of revenue and setting constricting rules on monetization options other than our own lack of ambition. We can do better than this.
Seed-stage funding for nascent startups is a pool of perverse incentives right now, but cryptocurrencies can provide a better approach. An Ethereum startup, Augur, raised $5 million in a pre-sale of reputation tokens that can be used to generate revenue for users that manually resolve bets on the prediction market that it’s building. The idea of selling digital assets directly to your users without the need for venture capital to bootstrap online ventures is very powerful. This will undoubtably become a popular way for companies to construct their ecosystems in the future.
And this isn’t even mentioning the centralized institutions whose monetary systems are failing us as well. I can’t pretend to understand the first thing about global finance, but there’s plenty of conversation at the conference about how Social Security is fucked, about the financial systems of Venezuela, Greece, and Syria, how the Euro is in trouble, and how digital currencies can come to the aid of many who global finance has wronged.
There is understandably a lot of skepticism about Ethereum and cryptocurrencies. 95% of cryptocurrencies are pump-and-dump schemes, and Ethereum is constantly berated by the greater community for choosing to hard-fork it’s software back in June to revert a transaction that hacked 14% of the net worth of the ecosystem from a poorly coded smart contract. It was a controversial decision that intentionally betrayed the techno-libertarian idealism that fueled Ethereum’s creation by manually reverting a computationally valid transaction.
There are people I know who work in finance who can’t reconcile any optimism about the future of blockchain with the volatility of the currencies, exchanges being hacked, and the irrational hype around some of the companies in the space. I understand their skepticism, but I disagree with their assessment.
Until you see it for yourself, it’s hard to truly grasp the scope of what is being built. If Ethereum works it will fundamentally change society, and not just the one in the US, Ethereum is a global project. For all the discourse about “Silicon Valley internet money” I think people miss that this has little to do with Silicon Valley, it’s a worldwide effort. The idea of international digital money having value sounds crazy on the surface, but is it any crazier than gold having an 8 trillion dollar market cap? Some metal that looks pretty and sits in a vault somewhere? These digital tokens can actually power applications, churn out arbitrary computation that can power all sorts of novelty. I would honestly rather have the digital asset than a piece of metal. At some point, our monetary system evolved from paper money to credit cards. I can only imagine the hostility when credit cards were introduced. “Plastic cards representing money that lives on a computer? Are you crazy?” But somehow it works. Is cryptocurrency really any crazier?
Ultimately though, my optimism about Ethereum comes from an underlying belief that I have about software. That it’s not about the code, it’s about the people. And the 700 people that flew halfway around the world to present their visions for the future of the internet are not the people that I want to bet against. For a week, the Hyatt on the Bund is the most interesting place to be in Shanghai. A crazy mix of techno-libertarian idealism, serious software development credentials, money, energy, and complete vitriol from the non-believers. And despite the hate, Ethereum moves forward, obsessed with the idea of a more decentralized web. Bet against it at your own peril.