July 12, 2010
When he awoke late, the morning after the Slashdot posting, Martti Malmi saw that the attention was not a hit-and-run phenomenon. People weren’t just taking a look at the site and moving on. They were also downloading and running the Bitcoin software. The number of downloads would jump from around three thousand in June to over twenty thousand in July. The day after the Slashdot piece appeared, Gavin Andresen’s Bitcoin faucet gave away 5,000 Bitcoins and was running empty. As he begged for donations, he marveled at the strength of the network:
Over the last two days of Bitcoin being “slashdotted” I haven’t heard of ANY problems with Bitcoin transactions getting lost, or of the network crashing due to the load, or any problem at all with the core functionality.
But while the Bitcoin software itself was working well, new users quickly ran up against the limitations of the Bitcoin ecosystem. Those who immediately wanted to acquire more Bitcoins than were available from Gavin’s faucet were left with only a few meager options, one of them a creaky, unreliable service that Martti had set up a few months earlier.
Jed McCaleb was one of the people who encountered this weakness. A native of Arkansas, Jed had been raised by his single mother, who made a living as a journalist. From a young age, Jed had been something of a math and science prodigy, and this allowed him to make it to Berkeley for college. Jed, though, had trouble sticking with things, and he soon dropped out of Berkeley and moved to New York. There he and a partner set up what became one of the main successors to Napster. His software, eDonkey, made it possible for individuals to trade large files like movies and it proved so successful that the Recording Industry Association of America sued Jed and his business partner. They eventually paid $30 million to settle the case and shut eDonkey down, but they also earned a few million along the way.
Despite being a soft-spoken introvert, Jed had a cool way about him that helped him make friends and girlfriends. When one of his romantic flings ended up pregnant, he and the woman, MiSoon, decided somewhat spontaneously to keep the baby and make a go of it. They used some of Jed’s earnings to buy an estate with a pool an hour or so north of New York City, just as they were expecting a second baby. In the sprawling, mostly empty house, Jed threw himself into an online game he had created called The Far Wilds, which had attracted only a few aficionados. He spent endless hours in a first-floor bedroom, which he had turned into a den. Books about neuroscience and artificial intelligence piled up around him—as did old food, attracting bugs that MiSoon initially tried to get rid of, but later came to accept as one of the side effects of Jed’s brilliant mind.
When Jed came across the Slashdot post about Bitcoin he was immediately intrigued. It seemed to fulfill many of the ideals behind Napster and eDonkey—taking power from authorities and giving it to individuals. But when Jed tried to buy some actual Bitcoins, he ran into the limitations of the few existing sites that sold them.
MiSoon was nursing their newborn son when she wandered into Jed’s study one night and encountered his frustration.
“There’s this really cool thing called Bitcoin—it’s like this nerd, libertarian thing,” Jed told MiSoon, in his hushed, intense voice. “But it’s so lame. I can’t buy any at night.”
Jed said he wanted to build a site himself where he could buy coins at any hour. When MiSoon arose the next morning, it was done. With some experience in amateur foreign-currency trading, Jed knew the basics of what an exchange required. But he had never actually set up a website before, having previously worked more on the sophisticated back-end software. His new Bitcoin exchange was something of a fun experiment.
He and MiSoon discussed possible names for the site. He mentioned an old domain name that he owned and was not using—mtgox.com. Jed had bought the site in 2007, for use as an online exchange to buy and sell the cards used in the role-playing game Magic: The Gathering—hence the acronym for Magic: The Gathering Online Exchange. It had operated for just a few months before Jed shut it down and the site had been vacant since.
“Yeah, you should use that,” MiSoon replied. “That’s kind of weird and easy to remember. Why not if you already have it registered?”
Seven days after the Slashdot post, Jed casually advertised his new site on the Bitcoin forum:
Hi Everyone,
I just put up a new Bitcoin exchange.
Please let me know what you think.
Mt. Gox was a significant departure from the exchanges that already existed, primarily because Jed offered to take money from customers into his PayPal account and thereby risk violating the PayPal prohibition on buying and selling currencies. This meant that Jed could receive funds from almost anywhere in the world. What’s more, customers didn’t have to send Jed money each time they wanted to do a trade. Instead, they could hold money—both dollars and Bitcoins—in Jed’s account and then trade in either direction at any time as long as they had sufficient funds, much as in a traditional brokerage account.
These advances made it significantly more convenient to buy and sell Bitcoins, but also brought new dangers that threatened to betray some of the currency’s basic principles. Satoshi had designed Bitcoin to eliminate the need for trusted central authorities. It was supposed to be a new money that people could hold on their own, without a bank, secured with a private key that only the user knew. Mt. Gox customers would be moving back to the old model in which a single institution—Jed’s company—held everyone’s money. If Jed offered good security measures, this might prove safer than holding coins on a home computer. But Jed was not a security expert, and if he did somehow lose the private keys to the exchange’s digital wallets, his customers had little recourse. Unlike the banks that Bitcoiners had bashed, Mt. Gox had no deposit insurance and no regulators overseeing the safety and soundness of Jed’s operation. The choice was between security and principles on one hand and convenience on the other.
When a forum member asked why they should choose Mt. Gox over the alternatives, Jed responded in his characteristically modest but confident way.
“It is always online, automated, the site is faster and on dedicated hosting and I think the interface is nicer.”
Even Jed, though, was surprised at how quickly people trusted his setup and sent money to his PayPal account. During his first day in business, July 18, twenty Bitcoins were traded at five cents each on Mt. Gox—an inauspicious opening. But within the first week he had his first hundred-dollar day of trading, and by the end of the month Mt. Gox had overtaken Martti’s service and the other existing exchange in trading volume to become the largest Bitcoin business around.
These weeks marked a subtle but dramatic transition for Bitcoin. Until this point, there had been occasional transactions, but mostly between aficionados making them out of a desire to help the network. After the Slashdot story, the difficulty of mining new Bitcoins ramped up quickly with the surge in the number of people racing to win coins. Satoshi had determined that as more computers joined the network, the mining of new Bitcoins would become more difficult, ensuring that it would always be roughly ten minutes between releases of new coins. The week after the Slashdot story, the difficulty of mining new Bitcoins jumped 300 percent. Gavin Andresen, who had initially started mining Bitcoins to help the network, now found it all but impossible to win new coins with his four-year-old Mac laptop.
Suddenly, if a person wanted Bitcoins, he or she had to buy them. And people were showing a willingness to do just that and part with real money for these unproved slots on a digital spreadsheet. The growing popularity of Bitcoin was hard to miss. One new forum member wrote:
What I like about Bitcoin is that it is a community with a solution that we are actually trying. I don’t know many people in real life that are even close to as radical in their thinking as I (and many others on these forums) am. Surprisingly, however, I am able to talk with my real life friends about Bitcoin much longer than my normal rants about “what should be,” because Bitcoin actually exists.
IN LATE JULY Martti launched the first foreign-language forum, in Russian, and within a few weeks it had hundreds of postings. The English forum grew much faster. In one month, the forum had gained more new members—370—than it had since coming online in November 2009. Craving more conversation, the expanding herd of dedicated Bitcoin followers found their way to the chat channel Martti had set up. Now, the Bitcoin channel on Internet relay chat, or IRC, became a sort of twenty-four-hour global coffeehouse where the new users could gather and marvel at this experiment they were all taking part in.
Around midnight on September 26, one new Bitcoiner wrote: “gosh I can’t sleep ! I keep thinking about this great stuff. To me Bitcoin is the ‘cyberspace gold.’ I’m just amazed.”
The next afternoon another new user spoke of spending ten hours reading everything he could find about the network.
“I did the same thing when I first heard about Bitcoin,” Gavin wrote back.
The appeal of Bitcoin varied from person to person, but most were in love with the basic idea of a digital cash that each user could control and move around the world with nothing more than a private key. The users, at this point, were mostly young men whose lives were untethered to anything other than their laptops, in constant communication with people on the other side of the world. For them, moving money around the globe with a paper check or an old-fashioned wire transfer seemed absurdly backward.
Satoshi chimed in on the forums to note that the Bitcoin software was designed to do more than just move coins. The software also had the capability to attach specific instructions to each coin so that the coins could behave in a particular way, according to the users’ wishes. A coin on the blockchain could, for example, be programmed to move from one address to another only if it was signed off on by three or four different private keys, enabling its use in the types of legal transactions that currently required cumbersome and expensive middlemen.
“The design supports a tremendous variety of possible transaction types that I designed years ago,” Satoshi wrote. “Escrow transactions, bonded contracts, third party arbitration, multiparty signature, etc. If Bitcoin catches on in a big way, these are things we’ll want to explore in the future, but they all had to be designed at the beginning to make sure they would be possible later.”
Satoshi had advertised Bitcoin as a trustless system that didn’t require its users to rely on any central authority. But like all forms of money, Bitcoin did rely on its users’ trusting the ideas and integrity of the system supporting it—in this case, code and math—and the small elite of cosmopolitan coders was more than willing to do that. These new converts, in turn, were providing not just enthusiasm, but also fresh sets of eyes to examine the code with a level of programming experience that had been scarce up to this point.
In late July Gavin and Satoshi got an e-mail from one such user, a programmer from Germany going by the screen name ArtForz, who had found a previously undiscovered weakness in the code that governed transactions on the network. The flaw made it possible to spend Bitcoins in someone else’s wallet.
Gavin and Satoshi immediately realized this was not just a bug but a fatal flaw that could doom the entire project. If someone else could spend your coins the whole system was all but useless.
Satoshi quickly put together a fix—the flaw was not actually difficult to correct. But in the meantime, Gavin and Satoshi agreed to keep the flaw secret until they got everyone on the network using new, repaired code, for fear that someone would take advantage of it.
“For now, don’t call it the ‘1 RETURN’ bug to anyone who doesn’t already know about it,” Satoshi wrote to Gavin.
Because the patched software “has a dozen changes in it,” Satoshi wrote, “it won’t necessarily be obvious what the worst vulnerability was. That may give people a head start to upgrading if any attackers are looking for the vulnerability in the changes.”
That ArtForz had not taken advantage of the bug himself was a minor miracle. But it was also what the incentives in the Bitcoin system were designed to encourage. ArtForz had been mining coins himself—using the GPU technology that Laszlo had first pioneered—and he knew that if confidence in the system was undercut his coins would be worthless. The market incentives were working as they were supposed to work. This turn of events also confirmed Gavin’s confidence in the power of decentralized systems. ArtForz was a part of the network, and as such, he didn’t just passively use the network. He and Gavin, and all the others, were helping to build this thing.
A FEW MONTHS earlier the big concern plaguing the Bitcoin forum was how to attract new users, but now the problem was how to deal with the influx of new users, their potentially malicious behavior, and their competing interests.
These problems became particularly pronounced after Bitcoin’s next big jump into the spotlight. In November, WikiLeaks, the organization founded by a regular participant in the old Cypherpunk movement, Julian Assange, released a vast trove of confidential American diplomatic documents that revealed previously secret operations around the world. The large credit card companies and PayPal came under immediate political pressure to cut off donations to WikiLeaks, which they did in early December, in what became known as the WikiLeaks blockade.
This move pointed to the potentially troubling nexus between the financial industry and the government. If politicians didn’t like the ideas of a particular group, government officials could ask banks and credit card networks to deny the unpopular group access to the financial system, often without requiring any judicial approval. The financial industry seemed to provide politicians with an extralegal way to crack down on dissent.
The WikiLeaks blockade went to the core of some of the concerns that had motivated the original Cypherpunks. Bitcoin, in turn, seemed to have the potential to counteract the problem. Each person on the network controlled his or her coins with his or her private key. There was no central organization that could freeze a person’s Bitcoin address or stop coins from being sent from a particular address.
A few days after the WikiLeaks blockade began, PCWorld wrote a widely circulated story that noted the obvious utility of Bitcoin in the situation: “Nobody can stop the Bitcoin system or censor it, short of turning off the entire Internet. If WikiLeaks had requested Bitcoins then they would have received their donations without a second thought.”
It wasn’t clear if Bitcoin could actually be used in this particular instance, but whatever the practical possibilities, the blockade was helping elevate the debate around Bitcoin beyond the rather narrow issues of privacy and government money-printing that had been dominant in the early days. Here was a broader philosophical issue that could attract a wider audience, and the forums were full of new members who had been drawn in by the attention. One new user, a young man in England named Amir Taaki, proposed making Bitcoin donations to WikiLeaks. Amir argued this could raise Bitcoin’s profile at the same time that it could help WikiLeaks raise money.
This kicked off a vigorous debate on the forum. A number of programmers worried that the Bitcoin network was not ready for all the traffic—and government scrutiny—that might come if it started to be used for controversial donations.
“It is extraordinarily unwise to make Bitcoin such a highly visible target, at such an early stage in this project. There could be a lot of ‘collateral damage’ in the Bitcoin community while you make your principled stand,” one programmer wrote.
Satoshi eventually ended the debate. When someone on the forum wrote, “Bring it on,” Satoshi responded forcefully:
No, don’t “bring it on.”
The project needs to grow gradually so the software can be strengthened along the way.
I make this appeal to WikiLeaks not to try to use Bitcoin. Bitcoin is a small beta community in its infancy. You would not stand to get more than pocket change, and the heat you would bring would likely destroy us at this stage.
This was enough to convince Amir.
“I’ve done a U-turn on my earlier view and agree. Let’s protect and care for Bitcoin until she leaves her nursery onto the economic killing fields.”
This was one of an ever-diminishing number of communications from Satoshi during the fall of 2010. Messages from both Satoshi and Martti had been increasingly rare. In Martti’s case, after a year of working on Bitcoin free, he needed a regular source of income. In September, two months after the Slashdot story, he took a full-time job with a firm that analyzed social-media data. On top of having a full schedule, Martti also saw that he was no longer needed. Gavin and a few others were taking over many of the day-to-day tasks that Martti had previously handled. And the chat channels were crawling with people ready to help out.
Satoshi’s gradual fading was less explicit. He still posted occasionally to the forums when there were specific questions, but he never appeared on the chat channel and increasingly shifted to infrequent private communications with Gavin and just a few other developers. In December, Satoshi asked Gavin if he would mind having his e-mail address posted on the Bitcoin website, as a point of contact. After his own name went up, Gavin noticed that Satoshi’s e-mail came down.
When the last public forum post came from Satoshi, on December 12, 2010, there was nothing marking it as such. Announcing the latest version of the software, version 0.3.19, the post was markedly different in tone from those early messages, selling the world-beating potential of Bitcoin. The main sentiment now was a warning that Bitcoin was still extremely susceptible to denial-of-service attacks, which overwhelm a system with message traffic.
“There are still more ways to attack than I can count,” Satoshi wrote in the brief note.
This came just days after Hal Finney checked back in for the first time since early 2009. His disease, ALS, had progressed quickly and he was now largely confined to the family living room, in a special setup his family had concocted so that he could continue working on a computer.
Hal made an unassuming return to the community with some relatively dry comments about patterns in the price of Bitcoin and the possibility of using Bitcoin’s blockchain as a new kind of database. He was as enthusiastic as ever about the network.
“I’d like to hear some specific criticisms of the code. To me it looks like an impressive job, although I’d wish for more comments,” he wrote on the forum. “This is some powerful machinery.”
This provoked Satoshi’s second-to-last post: “That means a lot coming from you, Hal. Thanks.”
This exchange set off a discussion among people who had never heard Hal’s name before.
“Who is Hal on the forum?” one user wrote. “Satoshi seemed to know of him.”
The question quickly gave way to the bigger mystery: Who is Satoshi?
“Is he a real person? ;-)” a forum user asked.
“Hmm, there are almost no results for Satoshi unrelated to Bitcoin,” another user wrote after some quick research.
This set off the first stages of a hunt for Satoshi that would continue for years. People on the chat channel began debating the available details about Satoshi and their significance. It was noted that Satoshi occasionally used British spellings and words like “bloody.” There was also a fragment from a British news story written into the first block of Bitcoins created by Satoshi’s computer.
A Bitcoin user in Japan noted that Satoshi was a common name in Japan, but he argued that Satoshi was unlikely to be Japanese given that Satoshi had never used Japanese words and had always written his name with the family name last, contrary to Japanese tradition.
“Maybe this is a gambit to trick us to think he’s not Japanese,” another user wrote.
“I like the pseudonym theory the best. It’s so much cooler for someone to have a secret identity than just a boring name,” someone wrote.
“Jesus, this is a great story. I’m amazed the NY Times hasn’t picked up on it yet,” another poster chimed in.
In the early days, Martti had never asked Satoshi any personal questions but had assumed that Satoshi Nakamoto was probably not a real name. Martti’s access to the Bitcoin websites allowed him to see that Satoshi was joining the sites through a Tor network that obscured his geographic location and IP address.
Gavin had asked Satoshi some personal details in his first e-mail, but Satoshi ignored the questions and Gavin never pressed for more.
One regular forum user asked Satoshi: “Suppose, god forbid, you were no longer able to program or were unavailable due to unknown circumstances. Do you have a procedure in mind to continue Bitcoin in your absence?”
Satoshi didn’t answer, but others on the forum noted that because Bitcoin’s software was open source, available to all the users, Satoshi’s involvement shouldn’t matter: “As long as the source code remains open, that is sufficient. If there is a need, and enough interest, the community will provide. Trust in the community :)” one developer wrote.
Satoshi was, in many ways, just as powerless, or powerful, as every other user on the network. All the coins were on the communal blockchain, but only the person with the private key corresponding to each address on the blockchain could use the coins in that address. Satoshi could try to change the software in some way that would give him more control, but doing so wouldn’t gain traction unless a majority of the network adopted the changes.
Still, Gavin, who was now perhaps the most central figure in Bitcoin, knew that the platonic ideal of open source software was somewhat more complicated underneath the surface. While anyone could propose changes to the protocol, he and Satoshi were still essentially the only people who could sign off on changes—and this gave them an unusual amount of power in the system. What’s more, while Satoshi had written a program designed to eliminate the need for trust, users of the technology still had to have faith that it would work as intended. On the forum, Gavin wrote: “Trust is Bitcoin’s biggest barrier to success. I don’t think there is anything we can do to speed up the process of getting people to trust that Bitcoin is solid; it takes time to build trust.”
At this point, though, the primary cause for distrust was not the lack of information about Satoshi. Satoshi’s anonymity, if anything, seemed to increase the level of faith in the system. The anonymity suggested that Bitcoin was not created by a person seeking personal fame or success. What’s more, Satoshi’s absence allowed people to project their own vision onto Bitcoin.
Those who could cause problems, though, were the very people who were making Bitcoin grow. The network was expanding, but the people among its growing ranks would also pose the greatest threat to Bitcoin and the trust it needed.