Bitcoin's Fatal Flaw Was Nearly Exposed

Panic loomed in Bitcoin land on Thursday after the community’s largest mining operation, GHash.IO, approached half of all computing power in the system, the limit of the cryptocurrency’s well-documented fatal flaw.

If any one mining pool commands the majority of the market—the dreaded so-called "51 percent attack"—the integrity of the system becomes compromised, since the act of mining is how new transactions are processed. An attack like this can therefore essentially hijack the currency by controlling how consensus is reached.

This would theoretically allow the individual or group to spend money that they didn’t have or that wasn’t theirs, known as “double-spending,” among other potential abuses, some of which could go undetected. In such an event, even the mere possibility of abuse would undermine trust, and the resulting panic could cause Bitcoin to collapse.

The online Bitcoin community understandably lost its shit. “LEAVE GHASH,” implored one user in a post on Reddit’s Bitcoin community, home to nearly 100,000 members. Numerous other posts that shared this sentiment littered the subreddit’s front page, including one noting that GHash had successfully solved 42 percent of new blocks in the last 24 hours. “The network health of Bitcoin is awful,” wrote another. “The 2 largest pools comprise 60% of the network.” In other words, the owners of those pools could team up and instantly take over the market.

Users over at BitcoinTalk—the original Internet HQ for crypto-coin discussion—were similarly distressed, with some people pointing out that the system’s most powerful mining pool had been accused of double-spending in the past (though those allegations are far from concrete).

The fiasco “opens a whole new can of worms that people just don’t want,” said Yifu Guo, a long-time Bitcoin insider and one of the creators of the first consumer ASIC-based computer miner. “Inherently, the situation creates possibility of evil-doing.”

Guo declined to go into further detail so as not to advertise potential “attack vectors,” but admitted that if you have control in a 51 percent attack, “the possibilities are kind of endless.”

Some examples of abuse would go virtually undetected despite the embedded transparency of the Bitcoin concept. A majority owner of mining power could delay certain transactions—impossible to prove—and basically hold users to ransom if they wanted to access or spend their stash.

As Guo explains it, mining pools have evolved into a powerful political force within the community. Bitcoin’s open source development is meritocratic in that each new software upgrade only becomes relevant if the majority of miners are willing to update to the new version.

Now that mining has moved towards pools, the owner of each pool has effectively become a politician, a representative for his constituents, who, in this case, are the individual miners who opt in to each pool. The problem with this is that individual miners naturally flock towards the biggest pools, which have the capacity to offer the highest returns with the lowest amount of uncertainty.

This creates a never-ending cycle of threats as each successive mining pool eventually closes in on that majority threshold. Guo compares this to the US political system. You keep having “elections” and new leaders, but, for practical purposes, “the end result is the same even if the names are different.”

“It turns Bitcoin into a currency with a central authority,” Guo told Motherboard, noting that no matter how many bitcoins you own, if you don’t own a similar proportion of mining computing power, “then you don't really control your bitcoins. Mining power is voting power.”

“It’s the same as trusting Paypal,” he said. Of course, Paypal is a regulated US corporation. Though officials could conceivably regulate Bitcoin companies, the decentralized technology itself isn't owned by anyone and arguably can't be controlled (without technically taking over the system).

But if the situation went a long way in exposing one of Bitcoin’s fundamental vulnerabilities, it also exemplified the ecosystem’s ability to self-regulate. As miners heeded the outcry against a looming monopoly, GHash’s share of the market corrected, dropping a few points to 38 percent as users fled the pool fearing a potential takeover.

Still, my Motherboard colleague Victoria Turk, who recently attended the 2013 Bitcoin Expo in London, wondered about the implications of “essentially relying on miners 'doing the right thing.’” (The identities of GHash’s owners aren’t readily available to the public on the mining pool’s website.)

With fiat currencies like the US dollar, we rely on a central authority and the rules and laws the government creates and enforces. With Bitcoin, you can only hope that the incentives for every actor are aligned.

And in fact they are, at least for the most part. By design, it wouldn’t make sense for any individual entity to weaken faith in the system if profit is the end motive, as this would negatively affect the value of bitcoins. The benefit of double-spending quickly diminishes if prices plummet. Logically, an entity would only execute such an attack with the intent to damage the whole system, like some form of protest or Bitcoin terrorism. There is the possibility of potential short-term profits, if the attackers were able to lock-in double-spending gains before the market could react, but this kind of speculative one-off wouldn’t make financial sense relative to the long-term gains of being the most powerful (and benevolent) mining pool in town.

GHash responded by promising to actively prevent "accumulation of 51 percent of all hashing power.”

“GHash.IO does not have any intentions to execute a 51% attack, as it will do serious damage to the Bitcoin community, of which we are part of,” the company said in a press release, channeling the community’s accepted wisdom. “On the contrary, our plans are to expand the Bitcoin community as well as utilise the hashing power to develop a greater Bitcoin economic structure. If something happened to Bitcoin as a whole it could risk our investments in physical hardware, damage those who love Bitcoin and we see no benefit from having 51% stake in mining.”

The owners of GHash did not immediately respond to Motherboard’s request for comment.

The price of BTC has stayed steady during this period at over $900 per coin on Mt. Gox, likely because GHash never actually reached 51 percent, and it shows the market’s confidence that things would eventually work themselves out. (Price levels also probably got a boost when Overstock.com officially started accepting bitcoins half a year ahead of schedule. Business Insider reported that the major US online retailer made $10,000 worth of BTC sales in the first hour.)

It’s also possible that news of the incident never reached “mainstream” investors before the threat had been mitigated. Or worse, the strength of Bitcoin’s mining network simply isn’t on their radar, says Guo.

The primary source of outrage this time around resulted from Blockchain.info’s chart of mining pool market share, which was passed around as evidence of the impending threat. “What if that chart didn’t exist?” wondered Guo. “Then people wouldn’t even know. They might not even care to know.” On a more sinister note, what if the chart obscured the truth?

“They could easily just split the pool,” said Guo, speculating on the possibility of a single owner controlling apparently separate mining pools. And multiple pools could always collude, “like OPEC.” No one can say whether or not the owners of GHash and BTC Guild, the second largest pool, are having high level discussions behind closed doors.

Guo is keenly aware of the problems that holding such potential power poses. When he first began manufacturing specialized computer chips for mining—which provided exponentially more power and efficiency than the video card powered systems they replaced—some wondered whether or not he might abuse his hardware arsenal. Given the number of chips he controlled at the time, he could have quietly taken over the network. As far as anyone knows, that never happened.

This puts into question not only the strength of the Bitcoin network, but also the efficacy of mining pools themselves, which weren’t coded into the initial release, but added in a later update. However, Guo notes that Satoshi Nakamoto, Bitcoin’s enigmatic creator, was still around to make that call.

Pools have their advantages, too, mainly by reducing randomness. If miners couldn't team up and share the spoils of each successful block, the act of mining would begin to resemble a lottery rather than scientific method. As more miners joined the system, the odds of winning a reward would approach zero for everyone, reducing the incentive to participate. By allowing miners to team up and share the spoils, pools reduce the variance of successful blocks for their participants. But Guo wonders if the current situation was part of Satoshi’s original vision.

“Is this what Satoshi wanted?” wondered Guo. “Or was he unable to come up with a solution?”

Much of Bitcoin’s elegance is attributed to the system’s ability to sidestep human intervention, which, for ardent believers, translates to human weakness and the opportunity for corruption. As the old crypto-saying goes, "In Numbers We Trust."

Seeing as how the current scenario is a consequence of the human element, I wondered aloud whether or not you can ever conceivably avoid it. After all, no matter how any system is initially designed, the situation inevitably changes once people are involved. Is it really possible to eliminate the human part of the equation from any system if we're the ones ultimately using it? (Wasn't this the takeaway from The Matrix?)

The problem for the cryptocurrency movement, at least ideologically, is that over time Bitcoin begins to exude many of the characteristics and downsides the model promises to address.

Guo believes in the possibility of a technological solution, however, even if we haven’t quite nailed it. He points to P2Pool, a decentralized mining pool that would theoretically eliminate these kinds of problems. But he readily admits that the system isn’t totally fleshed out and isn’t yet a perfect answer. He also suggests that if a decentralized peer-to-peer mining model could address these concerns, Satoshi would have included it in the original design, although we can't rule out that the man (or woman, or group of people, or government agency) wasn't infallible.

And neither is his creation, it appears. Part of this then comes down to growing pains as Bitcoin continues to develop and new problems emerge. For Guo, more than anything, the central issue is a lack of awareness by users at large.

“I think [the growing power of mining pools] could be okay,” he said. “I don’t think the current situation is necessarily a bad thing. But people are not informed about it, which is a problem. It’s not a conscious choice.”

Since Bitcoin's development is ongoing, better education of relevant risks will lead to smarter adjustments and decisions concerning the system's overall health.

But being conscious of the situation might not always translate into the opportunity for choice as mining power steadily consolidates in the hands of the financially well-equipped.

In their Bitcoin cover story, Bloomberg Businessweek pointed out Silicon Valley's arrival in the ongoing mining arms race, a secretive startup called 21e6 (which is the scientific notation for 21 million, a reference to the total number of bitcoins that will ever be mined).

The company has raised $5 million to "build what’s believed to be one of the fastest mining chips in the world" with an army of insidery investors that allegedly include the "Winklevoss twins; Marc Andreessen and his venture capital firm, Andreessen Horowitz; early Tesla Motors (TSLA) backer Bill Lee; PayPal (EBAY) mafia member David Sacks; and Naval Ravikant, founder of AngelList, a social network for investors and entrepreneurs."

Such developments raise pointed questions about the undercurrents of power within the burgeoning system. Who really controls Bitcoin? Going forward, it may become increasingly difficult to tell.

@sfnuop