According to its advocates, online cryptocurrency bitcoin is set to revolutionise the way we do business. At the centre of this emerging market, a company called Mt. Gox, based in Japan, quickly grew to be the world’s largest exchange for the virtual currency.
Then, just as quickly as it emerged, Mt. Gox vanished without a trace – and hundreds of millions of dollars’ worth of investments disappeared with it. In an investigative report for Wired, Robert McMillan uncovers what really happened:
From a distance, the world’s largest bitcoin exchange looked like a towering example of renegade entrepreneurism. But on the inside, according to some who were there, Mt. Gox was a messy combination of poor management, neglect, and raw inexperience.
Its collapse into bankruptcy last week — and the disappearance of $460 million, apparently stolen by hackers, and another $27.4 million missing from its bank accounts — came as little surprise to people who had knowledge of the Tokyo-based company’s inner workings. The company, these insiders say, was largely a reflection of its CEO and majority stake holder, Mark Karpeles, a man who was more of a computer coder than a chief executive and yet was sometimes distracted even from his technical duties when they were most needed. “Mark liked the idea of being CEO, but the day-to-day reality bored him,” says one Mt. Gox insider, who spoke on condition of anonymity.
Get daily business news.
The latest stories, funding information, and expert advice. Free to sign up.
According to insiders, serious problems became apparent quite early in Mt. Gox’s existence:
By the fall of 2013, Mt. Gox’s business was also a mess. Federal agents had seized $5 million from the company’s U.S. bank account, because the company had not registered with the government as a money transmitter, and Mt. Gox was being sued for $75 million by a former business partner called CoinLab. U.S. customers complained of months-long delays withdrawing dollars from the exchange, and Mt. Gox had tumbled from the world’s number one bitcoin exchange to position number three.
But Karpeles was obsessed with a new project: The Bitcoin Cafe. Inspired by a French bistro, it would be a stylish hang-out located in the same building as the Mt. Gox offices, a very-new-looking building of metal and glass within walking distance of Tokyo’s largest train station. You could drop by for a beer or some wine, and — using a cash register proudly hacked by Mark Karpeles — you could buy it all with bitcoin. When WIRED tried to meet with Karpeles and Mt. Gox at their offices this past October — and a company representative turned us away, saying that legal reasons prevented Mt. Gox from talking to the press — the placard in the lobby of the building already identified the cafe. This company representative said it would open by the end of the year. It never did.
One insider says that Mt. Gox spent the equivalent of $1 million on the cafe venture, renovating Mt. Gox’s office building to Karpeles’ specifications. At a time when Gox’s business was falling apart, this insider says, the project was a major distraction. “[Karpeles] was super-proud of being able to use his hacked cash register with the code he wrote,” this insider says.
As McMillan demonstrates, the hackers weren’t the only issue that brought Mt. Gox unstuck.
Need a new phone? There’s a module for that
Back in January, Google announced it is selling its smartphone handset subsidiary, Motorola, to Chinese tech giant Lenovo for $US2.91 billion. Harry McCracken from Time explains, there was one part of the smartphone giant that Google certainly wasn’t going to sell to Lenovo:
On January 29, Google announced that it had agreed to sell Motorola, its phone-manufacturing business, to Chinese electronics giant Lenovo. Thus concluded the company’s brief, unprofitable foray into smartphone hardware, which began when it revealed plans to acquire Motorola Mobility in August, 2011.
Except that it didn’t really end there. It turned out that Google was holding onto one organization within Motorola: the Advanced Technology and Projects (ATAP) group. Headed by Regina Dugan, the former director of the U.S. Defense Department’s fabled Defense Advanced Research Projects Agency (DARPA), ATAP aims to bring the same approach to mobile-gadget innovation that DARPA used to kickstart the Internet, satellite navigation, stealth fighters and other technologies that started small and eventually mattered a lot.
One of the more interesting projects being worked on in Google’s ATAP lab is a device called Project Ara. As McCracken explains, Project Ara is a smartphone that allows users to upgrade their device by swapping ‘modules’ in or out:
Among the ATAP initiatives that have been announced, one in particular is quintessentially Google-y. It’s Project Ara, which aims to reinvent the smartphone by breaking it down into modules that can be assembled and customized in a limitless number of configurations. The company first disclosed that the project existed on October 29 of last year, when it released some intriguing photos but little in the way of concrete details. Today, it’s lifting the veil further as it prepares for an Ara developer conference it’s holding at Silicon Valley’s Computer History Museum on April 15-16. A year or so from now, it hopes to have a product on the market.
The device certainly has the potential to change the way the entire industry works. It’s a point McCracken makes by examining Samsung’s recent announcement of its new flagship Galaxy S5 smartphone:
It’s a big, improbable dream; even based on what little information Google has disclosed to date, pundits have been busy pointing out the pitfalls. But if Ara works as advertised and large numbers of phone buyers buy into the proposition, they could get something that contrasts sharply with the traditional breed of off-the-shelf phones released on a yearly schedule.
For instance, when Samsung announced the Galaxy S5 this week, its headline improvements included a better camera, a fingerprint scanner and a heart-rate monitor. In a world of modular phones, you might be able to pick any or all of those features and add them to the phone you already have. You’d even be able to pick among multiple cameras, or choose quirky features not meant for the masses. (Eremenko’s playful example: an on-phone incense burner.)
While the release of a smartphone based on Project Ara is more than a year away, this is certainly a project that is worth watching closely.
Wikipedia’s automatic troll-killing machine
Back in the early 2000s, online encyclopaedia Wikipedia was at the vanguard of the Web 2.0 movement. Proponents argued that, instead of the web being a place filled with static content, it should be interactive.
The “wisdom of crowds” was able to create many good articles. However, as Jesse Hicks from The Vergepoints out, soon trolls and vandals came out of the woodwork:
Wikipedia launched in 2001 from the ashes of expert-penned Nupedia. When Nupedia floundered, founders Jimmy Wales and Larry Sanger pivoted to a crowdsourced encyclopedia. Within four years, the English Wikipedia had more than 750,000 articles. No longer an obscure internet experiment, it had gone mainstream.
The increased attention brought a flood of new users with all of the attendant headaches: self-promotion, amateurish additions, and outright vandalism. Wikipedia’s shortcomings, both as an information source and as a self-organizing community, were becoming apparent. In the fall of 2006, Jimmy Wales gave a keynote speech calling on Wikipedians to focus on article quality over article quantity.
There were many articles trolls could vandalise – too many for even a committed community of editors to stay on top of. The answer proved to be technological:
In response, four Wikipedians built what would become known as AntiVandalBot. As the name suggests, it was a first attempt at automated vandalism protection: using a relatively simple set of rules, it monitored recent edits and intervened accordingly. Obvious vandalism could be removed automatically, while borderline cases went to another program, VandalProof, for human intervention. Crude by today’s standards, AntiVandalBot nonetheless saved editors time and attention.
It may even have saved the site. One study examined the probability that a typical Wikipedia visit between 2003 and 2006 showed a damaged article. While the chance was always minuscule — measured in thousandths — it had increased exponentially over just three years. Without the evolving anti-vandalism tools, that trend could have continued; editors would simply be overwhelmed by defacers.
The modern versions of AntiVandalBot are significantly more sophisticated than their predecessors:
The original bot relied on simple heuristics; Cluebot NG would instead employ machine learning. That meant instead of supplying basic rules and letting the software execute them, Carter and Breneman would provide a long list of edits classified as either constructive or vandalism — the same process is often used in spam filtering and intrusion detection. The key to successful machine learning is a large collection of data. Luckily, an anti-vandalism competition had already provided a dataset of about 60,000 human-categorized edits. From that, Cluebot could begin learning, finding patterns and correlations within the data.
With social media now a huge part of how we interact online, and growing concern about trolls and vandalism, the techniques being developed on Wikipedia indicate new ways websites can stay on top of trolls and vandals.
How Mark Zuckerberg does a deal
Last week in my Control Shift column, I examined some of the reasons why Facebook purchased mobile messaging service WhatsApp.
But this still leaves open the question: How did the $US16 billion mega-deal go down? It’s a question investigated by Parmy Olson in Forbes:
The subject line of the e-mail was like every other come-on that hit Jan Koum’s in-box in the spring of 2012. He was pounded daily by investors who wanted a piece of his company, WhatsApp. Hatched on his birthday, Feb. 24, 2009, WhatsApp was emerging as a global phenomenon. Some 90 million people were using it to text and send photos for free. No social utility had ever grown as fast. Facebook had only 60 million by its third birthday. And at the time close to half of WhatsApp users were returning daily.
Koum looked at the e-mail sender: Mark Zuckerberg. Now, that was a first. The Facebook founder had been using WhatsApp and wanted him over for dinner. Koum stalled, then finally wrote back saying he was traveling soon and dealing with server issues. Zuckerberg suggested they meet before Koum left. Koum forwarded the reply to his cofounder, Brian Acton, and his sole venture backer, Jim Goetz, a partner at Sequoia Capital, adding the word: “Persistent!”
One of the more interesting revelations is the role fate and circumstances played in the deal. For example, as Olson reveals, Google very nearly poached the mobile messaging site.
In June mid-2013 the [WhatsApp] founders happened to meet Sundar Pichai, who oversees Android and Chrome at Google. They talked about their love of clean and simple digital products. At some point around early 2014 Pichai decided it would be good for Koum and Acton to meet his CEO, Larry Page. They agreed to meet on Tuesday, Feb. 11.
On the Friday before that meeting a WhatsApp staffer ran into Facebook’s head of business development, Amin Zoufonoun, and told him that Koum was meeting with Page imminently. Zoufonoun, who helped broker Facebook’s $1 billion Instagram acquisition in April 2012, went back to his company and set the wheels in motion to accelerate an acquisition offer that had already been in the works for some time.
Until the day WhatsApp declares $US16 billion in profits for Facebook, the deal will remain controversial. Nonetheless, how it all came to be is an interesting tale in itself.