In April of this year, in an explosion of hype, the DAO (Decentralized Autonomous Organization) set a record for crowdfunding. It raised 150 million enthusiastic dollars. That was a mighty wad of cash riding on a little more than a business concept, but that’s the kind of cookie that gets baked in the virtual world of the blockchain. To understand the DAO and its collapse, we first need to understand the blockchain. Let’s dig into this, beginning with what makes the blockchain special.
Think about money and the banking system. The vast majority of people “trust” the banks. OK, Joe Public may have a low opinion of the “too big to fail and too big to jail” Wall Street banks, but he still uses a checking and savings account. That is because he has no practical alternative – or to be precise, that was the case until the Bitcoin began to gain traction.
What is Bitcoin and its Much Admired Blockchain?
In simple practical terms, it amounts to a banking system without banks. There are exchanges where you can buy or sell Bitcoin for dollars and put bitcoin into your account. And there are now debit cards that allow you to spend bitcoin with any company or shop that processes cards. The payment is made in dollars (or euros, etc.), and the Bitcoin in your account (it’s called a “wallet”) is debited at the prevailing exchange rate. Even though the transaction is cleared by Visa or MasterCard, no bank is involved.
People use banking services because they trust banks to keep their money safe. People use Bitcoin because they trust the blockchain. It is a bank. The blockchain is an open ledger of bitcoin transaction, encrypted and unhackable (it has never been successfully hacked, despite thousands of attempts). And, incidentally, the transaction costs are really low (think cents), and it only takes about ten minutes to clear a transaction.
Now think of a blockchain system as a trust network; Bitcoin just happens to be a successful use of such technology. There are many other examples of trust networks in the world where the blockchain could replace an old-style trust network. For example, eBay is a trust network for buying and selling things. It acts as an intermediary between buyer and seller, assisting the two parties to come to an agreement. Recently, a blockchain alternative called OpenBazaar.org was launched. It provides a direct buyer-to-seller capability with no need for a website or middleman fees. It is made possible by the blockchain.
It was with this kind of idea in mind that the DAO was launched, with great fanfare and $$$$s of investment. And, to the chagrin of those who threw their dollars at it, it failed. The DAO was not based on Bitcoin but an equally secure cryptocurrency with its own blockchain called Ether. So the $150m that had been invested in the DAO was converted to 11.5m Ether to provide the foundation for the DAO. The idea for the DAO was that developers would be able to set up specifically targeted smart contracts that anyone could use. There had been lots of discussion of what would be possible using smart contracts on the DAO. For example, a peer-to-peer AirBnB business might be possible, a different kind of Uber might be possible and so on.
The DAO was hacked in the following way. Using (or should we say abusing) a smart contract, roughly a third of the Ether that underpinned the DAO (worth $50-60m) was transferred into a “child DAO.” It was still there in the DAO. It didn’t disappear, it just changed ownership.
A Hard Forking Decision
The hack created an immediate existential crisis for the DAO. The company that built it, Slock.it, decided to close it down and return the funds to investors. Whether it will be able to do so effectively is uncertain, and it will not become clear for a while. This is the conundrum.
Go back a few years to when the major Bitcoin exchange Mt. Gox was hacked. About 650,000 Bitcoin (worth about $400m) were stolen. It is important to understand that it was not the blockchain that was hacked but the computer systems at Mt Gox, allowing Bitcoin to be stolen from the Mt. Gox hot wallet. At the time, Mt. Gox, which subsequently went into liquidation, was by far the largest Bitcoin exchange, processing 70% of all transactions. It was the beating heart of the market, and it had a heart attack.
It would probably have been possible for the Bitcoin network to recover the stolen Bitcoins by reverting back in time to before the Mt. Gox hack took place. This is referred to as a “hard fork” of the blockchain. You revert to a point in time when the blockchain was “clean.” The Bitcoin community, an ensemble of bitcoin developers and miners, refused to do that. They reasoned as follows:
Nobody will trust the Bitcoin blockchain if it is possible by human decision to undo transactions (via a hard fork) no matter what the rationale.
That was the right decision. It was the only possible decision, and it has proven to be an excellent decision.
It was in the spirit of you-can-trust-the-network that the DAO introduced its smart contracts, saying that “the program code is the contract.” That is the point. It has to be that way.
Unfortunately, the DAO was hacked via a smart contract. In fact, it’s probably inaccurate to say it was hacked. If the code is the contract and you (the user) make that contract behave in a way it was never intended to, you are simply taking advantage of a clause of the contract. After all, “the code is the contract.” If that happens to put $50m of Ether in your control, then so be it.
This is the reason why the DAO has to close down. It appears that it is now going to implement a hard fork in an effort to rescue the Ether that were “stolen.” As a business idea, it has failed because it was unable to implement bulletproof trust.
Its smart contracts simply were not smart enough. That’s how the cookie crumbled.