The Infinite Machine

by Camila Russo

They were able to build on Ethereum in twenty-four hours what had taken them over two months to build on Bitcoin.

Augur works with its own internal token called Reputation, or REP. Users can place bets on future events—for example, will Donald Trump win the 2020 election?—and receive cryptocurrency-based shares depending on the outcome. To keep the system fully decentralized, the correct outcome needs to be defined by consensus, and that’s where the REP tokens come in. Holders will stake their REP to an outcome, either before or in a brief period after it happens. Those who staked their REP tokens to the correct outcome get their tokens back plus a portion of the fees users paid to participate in the process. All trading and payouts happen through Ethereum smart contracts.

forty-five days and they sold 11 million REP tokens for 60 cents each, or $5.3 million, to about 3,000 digital addresses distributed across the globe, without the help of banks or funds. Joey, Jack, and other developers kept 20 percent of the funds. This new method of funding open source protocols had proven successful once more.

Augur was also the first to build their REP coins using the so-called ERC20 token standard, which would become almost synonymous with Ethereum token sales.

“Standardized Contra...

“Although Ethereum allows developers to create absolutely any kind of application without restriction to specific feature types, and prides itself on its ‘lack of features,’ there is nevertheless a need to standardize certain very common use cases in order to allow users and a...

Reddit under a thread called “Let’s talk about the coin standard,” where Ethereans pitched in with their thoughts.

At around that time, Martin Becze had quietly been building a JavaScript client for Ethereum while living in a camper in his parents’ backyard in Indiana, until he’d been hired by the Ethereum Foundation as a contractor. He’d decided there needed to be a more formal process for proposing changes to Ethereum. So he copied Bitcoin Improvement Proposals (BIPs) and programming language Python’s Enhancement Proposals (PEPs) and created Ethereum Improvement Proposals (EIPs), a document designed for anyone to propose changes for Ethereum, which had to include a technical specification and the reason why it’s needed.

So in November, Ethereum developer Fabian Vogelsteller took Vitalik’s initial token standard draft, wrote a proper specification explaining each function and action, and created an issue on Ethereum’s GitHub EIP repository. He called it “ERC: Token standard.” ERC stood for Ethereum Request for Comment,

common practice among internet engineers and researchers who use RFCs, or Request for Comments. The standard was later known as ERC20, because it...

It was a very simple document consisting of just six commonly used functions for tokens, and two events to trigger after certain functions are invoked. Because it was so simple and because developers were very interested in deploying tokens on Ethereum, it quickly became popular with developers, who started using it as a blueprint they could just copy and paste for any token. If you’ve ever tried to create a website, you’ll know how much easier it is to just take a template off the shelf instead of creating your own. ERC20 did something similar for tokens.

Rune Christensen, a Danish student who had been teaching English and running a recruiting business in China, was also one of the few building an Ethereum project. Excitement for the technology and the potentially lucrative upside led Rune to Bitcoin in 2011.

He shut down his recruiting business and started heavily investing in the cryptocurrency. He made a lot of money in the 2013 spike, and then lost it all when it crashed a year later. That gut-wrenching volatility is what got him thinking about stablecoins, which Charles Hoskinson’s former company with Daniel Larimer, BitShares, was building. He became a BitShares enthusiast but eventually grew tired of the internal drama in the BitShares crowd as well as all the political talk, which was getting extreme, even if he was now too a libertarian believer. Rune felt the drama and politics got in the way of improving the product, allowing competitors to jump ahead of the project.

The budding Ethereum community seemed to be the exact opposite. The drama and politicking within the foundation and among cofounders had been kept out of public view; those outside the inner circle were an optimistic bunch of mostly millennial developers. They had been influenced by Vitalik’s penchant for T-shirts adorned with cats and unicorns. They cultivated the aesthetic, whi...

To grossly simplify the two cultures, if Bitcoiners were typically hard-core libertarian and carnivore, Ethereans le...

After spending time in the Bitcoin and BitShares forums, Rune found it was refreshing that the Ethereum community was focused primarily on building innovative tech, rather than on any particular ideology. Back home in Copenhagen, he decided to start an Ethereum-based stablecoin. He didn’t know how to code, but taught himself the basics in two weeks and...

“Introducing eDollar, the ultimate stablecoin built on Ethereum,” he wrote. “As someone who’s been obsessed with pegged cryptocurrencies for the past 6 months, I was delighted to find out that even with just my meager programming skills, developing for Ethereum is so incredibly easy that I’ve been able to come up wi...

The first to reply to his post on eDollar, later renamed Dai, was Vitalik, who goes by “vbuterin” on Reddit. He gave Rune a technical suggestion, which Rune’s project, called MakerDAO, ended up implementing.5 Rune says the system would have probably been hacked and a total disaster if he had released it as it was, but at the time, he was part of a cohort of Ethereans who believed smart contracts were invincible and unbreakable. “Build unstoppable applications,” the Ethereum.org website would later say. Not only that, but many believed anything was better if it was decentralized. The thinking was, if you picked any business idea and somehow added blockchain tec...

In 2015 DAOs were all the rage in the blockchain world. They fit perfectly into a futuristic, cypherpunk vision where digital money and blockchain-based platforms would replace old, dusty banks and all human intervention would be minimized.

Blockchain technology would take people out of the equation as much as possible and leave decision making to computer programs. The rules of an organization would be put into code and computers would execute all the decisions openly and predictably. The decentralized network would ensure that no single party could modify the code or shut down the program. The idea was that people can be corrupted, they can cheat and deceive, but code can’t. It was pure crypto-anarchy, because to some of its proponents, these DAOs would presumably be outside of the law. DAOs “don’t need regulation, you don’t want to regulate them, and happily you can’t,” Stan Larimer wrote. The only laws a DAO would be subject to would be the market’s. Developers

Later that month, a project called DigixDAO raised $5.5 million in one day. It claimed to be the first DAO to do a token sale. The Singapore-based team wanted to back their DGX tokens with gold bars stored in a vault, where one token equaled 1 gram of gold. A separate token, called DGD, would be used to vote in the DAO and DGD holders would be rewarded in the gold-backed DGX.

Augur, Maker, and Digix were some of the biggest projects in the budding Ethereum community. Developers were also hacking away to create the basic infrastructure needed for this world computer to work. Embark and Truffle were frameworks to develop, test, and deploy decentralized applications, or dapps. Ether.camp and TradeBlock were block explorers, or tools to track the evolution of the network. EthereumWallet.com and MyEtherWallet were digital ether wallets, to store, send, and trade ether.

One of the most ambitious projects was Slock.it. Christoph Jentzsch, a German theoretical physicist, had been working on testing and checking the compatibility among the different Ethereum clients. Christoph was known for being very good at his job, obsessive and meticulous, so when he jumped into a new venture, people paid attention.

Slock.it cofounders, including Jentzsch’s brother, wanted to make it possible for people to rent, sell, and share their property without having to use intermediaries. Instead of going to platforms like Uber and Airbnb to get matched up and pay, the whole process would be done through a lock placed on each item. The lock would be connected to the Ethereum blockchain and all transactions would be automated with smart contracts. This was at a time when not even digital wallets, the most basic interface for Ethereum users, were intuitive enough for nontechnical people, so it was still a long way off before something like Slock.it could work. But Ethereans were nothing if not optimistic.

But then 9/11 conspiracy videos got him to distrust governments and the banking system, and that got him into Austrian economists. He read Ludwig von Mises’s laissez-faire magnum opus Human Action: A Treatise on Economics twice. After dabbling in the extremes of the political spectrum, he finally just defined himself as a “hippie anarchist.” And that doesn’t jibe with an office job.

He sold everything he owned and bought physical bars of silver and gold coins to store his savings so that he could rely as little as possible on banks. He backpacked around South America, Southeast Asia, and India for two years (but always made the trek back home for Burning Man), and financed himself thanks to a friend who was selling his precious metals and transferring the cash.

One day his friend thought of using Bitcoin to make the transfers. Griff loved the idea of currency that wasn’t issued by any government and exchanged his real gold for digital gold. He went back to the United States to live with a girlfriend in West Hollywood, where he worked as a massage therapist—he was using the skills he picked up in Thailand—but he couldn’t stop thinking about Bitcoin. One day he decided to break up with his girlfriend, take the gains from his crypto investments, which at that point were worth several thousand dollars, and move to Ecuador. His plan was to become a cryptocurrency evangelist, create a Bitcoin ecosystem for the South American country out ...

The DAO had become the only thing the Ethereum community talked about. It was the biggest, most ambitious project attempted and so eleven of the most distinguished Ethereans came on as curators, including Vitalik Buterin, Gavin Wood, Alex Van de Sande, Vlad Zamfir, Fabian Vogelsteller, and Christian Reitwiessner. They all either had been or were currently part of the Ethereum Foundation.

Christoph was also getting worried. Suddenly millions of dollars were being poured into an experimental piece of code he created. What if something went wrong? To make things worse, the project started attracting attention, and he was getting constant calls and emails from journalists. It got to a point where he couldn’t handle it anymore and stopped answering emails. Griff, who was staying at Christoph’s mother’s house in the Mittweida region in Germany, went with him on walks to help him calm down.

With ETH at around $12, that was $150 million. Ethereum’s crowdsale had been groundbreaking with $18 million raised. This was on a different order of magnitude.

Görlitzer Park (Görli to locals) in the bohemian Kreuzberg district,

he was working on something unrelated to the mammoth project when he noticed a flaw that could potentially be used to drain funds from some contracts.

The problem was with how some contracts were set up to send funds. The code did what it was supposed to, but the order of the commands made it possible for someone really clever to withdraw more funds than they actually had. The glitch was that the computer was told to decrease the user’s balance only after sending the funds. And in between sending funds and updating the balance, a new call to the same transaction could be initiated, which would also send funds first before updating the balance. And this so-called reentrancy bug could keep repeating until almost all of the gas in the original transaction was depleted. This meant someone could ask for money multiple times before their balance was updated and before the computer noticed their account didn’t have any money left.

The fifty-ninth child DAO to appear was called “lonely, so lonely.” After a child DAO was created, there was a seven-day waiting period before the owner could pull their ether from the main DAO, and an additional twenty-seven days before they could withdraw those funds from the child DAO. In the initial one-week period, others could also jump on board, and that’s just what the attacker did with the “lonely, so lonely” split—or, as it would be later be known, the “dark DAO.” The countdown for the “dark DAO” to be able to pull funds from the main DAO had started on June 8. Ethereans had no way of knowing they only had a few days to fix the bug Christian had spotted.

“I’m pretty sure I know how to empty out The DAO,” Emin said, and they both continued to poke at the code.

“I still think that splitDAO may have a vulnerability,” he wrote to his student. “It violates the withdraw pattern by not zeroing the balances field until after the call. So I think it may be possible to have it move rewardTokens to a splitting DAO multiple times. This is happening on lines 640 to 666 (hah!) of DAO.sol. Am I wrong?”

Phil spent a few hours going through the lines of code and concluded there was no way to trigger the vulnerability.6 He replied to Emin and with that, the professor trusted his student and went back to bed. Stephan published his “no DAO funds at risk” post that day. The attacker could have been rubbing his hands together, rightly concluding that, despite that ominous red flag in the “666” line of the code, nobody was on to them.

Vitalik was in Shanghai staying at a friend’s apartment when at around 3 p.m. on June 17 he got a message saying something like, “I’m not sure if this is an issue, but it looks like The DAO is being drained.”

“That’s weird,” Vitalik thought. “No money is supposed to be able to come out of The DAO yet.” He went to look at the contract and saw about 100 ether was disappearing from it every second. At this point, the ether price had climbed so the total held in the smart contract equaled about $250 million. As he forwarded the message to Ethereum core developers and the Slock.it team, he didn’t want to believe this was malicious but very soon there was no other way around it. The biggest project ever attempted on Ethereum, a treasure chest that held rough...