Satoshi and the Dangers of Narrative Momentum
A new documentary claimed to unmask Bitcoin's creator, Satoshi Nakamoto. Does the evidence back it up?
(For regular Shell Game readers/listeners: This one pushes beyond the usual topics here, as I’m gradually expanding a bit into Shell Game’s wider “things that are not what they seem” purview, including this particular obsession of mine. But we’ve got some more Season 1-specific stuff coming in the next couple weeks. -Evan)
On a recent Monday evening in Manhattan, a crowd began to gather at Pubkey, a Bitcoin-themed basement bar in Greenwich Village. There was an electricity in the air, with Bitcoin hitting an all time high of over $84,000, a fact reflected on a digital ticker affixed to the bar’s wall. (The number has since gone higher.) But the fifty-odd people jockeying for seats in the back room had come not just to celebrate the cryptocurrency’s new peak. They were waiting to hear directly from the man recently outed as the inventor of Bitcoin itself.
A month earlier, with the release of the HBO documentary “Money Electric,” filmmaker Cullen Hoback claimed to have finally unmasked Bitcoin’s pseudonymous creator, Satoshi Nakamoto. Satoshi’s identity is a mystery that’s bedeviled journalists and the Bitcoin-curious alike since 2011, when Satoshi sent his, her, or their last verified email—“I’ve moved on to other things,” it read—and disappeared. The mystery endures in part because Satoshi is assumed to have possessed, and could still, a large supply of early Bitcoin—enough to place the inventor among the world’s wealthiest people. It’s potentially enough to move the market, were it ever spent.
I count myself among the frustrated obsessives: Years ago, I investigated the arguments for and against Satoshi being programmer and former criminal cartel boss Paul Le Roux, the subject of my book The Mastermind. I ultimately concluded that Le Roux was among the better of the known candidates—everything matches up nicely, he had the ability and motivation, there are intriguing coincidences, and no hard facts need to be explained away. I also concluded that Satoshi was not likely among the known candidates. I’ve since spent years quietly chasing down evidence for Le Roux and others, amassing spreadsheets of clues, and occasionally debunking bogus claims around Le Roux.
My most conclusive realization, though, has been about the hunt itself. Nearly every attempt to unmask Satoshi begins (as mine had) by ignoring the previous candidates and adding a new one. It’s an investigation-by-addition approach that inevitably evolves into a textbook exercise in confirmation bias. I hoped, based on the pre-release publicity, that “Money Electric” might offer something different. Alas, there’s scant evidence the documentary got it right, and plenty that it guessed wrong.
Before the film’s debut, Hoback’s claim to have solved the mystery surfed the kind of Satoshi hype wave unseen in years. The New York Times, Time, The Washington Post, The New Yorker, and Wired published lightly-hedged reports trumpeting the documentary’s potential investigative triumph. Politico Europe went a step farther, asserting on X that “if its findings are widely accepted, it could send shockwaves through world financial markets and even the U.S. presidential election.”
Spoiler: They weren’t; it didn’t. But the publicity did drive betting markets where millions of dollars were gambled on the usual suspects: known cryptographers and OG Bitcoiners like Adam Back and Hal Finney, digital currency pioneers like Nick Szabo and Wei Dai, outsiders like Len Sassaman and Le Roux, and even a court-established Satoshi faker, Craig Wright.
Hoback’s answer, it turned out, was none of the above. Based on a few bits of circumstantial evidence, the documentary asserted that the real Satoshi was a relatively unexplored candidate: longtime Bitcoin developer Peter Todd. Todd was little known outside the Bitcoin community but highly prominent within it. He’d occasionally been alleged to be Satoshi, but in insular forums, and had sometimes joked cheekily that he was. Hoback asserts that this was all a misdirection. Satoshi had been hiding in plain sight all along.
It’s a fun gambit. Unfortunately, the documentary’s case for Todd rests largely on a 2010 post on the Bitcointalk forum—uncovered by a blogger nearly a decade ago and then supposedly rediscovered by Hoback in a dramatic scene in the film. In the forum post, only Todd’s second on Bitcointalk, he corrects a post made by Satoshi hours earlier. Hoback interprets this correction as Satoshi slipping up, accidentally logging in to his “real” account, and continuing his thought. (Todd has since dismissed the theory, and the documentary obscures the fact that Todd himself was posting under a pseudonym at the time. As for the post, various commenters have observed that it reads like one nerd “well actually-ing” another, a common occurrence on technical forums.)
The documentary presents a few smaller lines of evidence—although in only the loosest sense of the word—including:
Hoback’s discovery that Todd deleted from his resume a line about his coding skills in C++, the language of the original Bitcoin code. (Todd told The New Yorker that he was in high school when he made the claim to C++ proficiency, then later reevaluated it “in the context of professional work.”)
Muddled speculation about another forum poster named John Dillon, who once allegedly paid Todd to push a specific Bitcoin development—and who, according to leaked email correspondence, claimed to be an intelligence agent. Hoback asserts Dillon was actually Todd, pretending to be an intelligence agent paying Todd. Somehow more convoluted than even this sounds, the allegation is meant to show Todd’s propensity for pseudonyms and game-playing. (No evidence is presented to support any of this, and if you read the actual leaked exchanges between Todd and Dillon, the supposed motives for the ploy are difficult to envision.)
An old chat log that shows Todd claiming to have deliberately “sacrificed” a stash of Bitcoins. “I’m probably the world’s leading expert in sacrificing Bitcoins,” he wrote. “I’ve done exactly one such sacrifice and I’ve done it by hand.” This is meant to explain why Todd, being Satoshi, hasn’t spent the mammoth pile of Bitcoin mined back in 2009. (It’s a fair question to ask about any Satoshi candidate. The answer for Len Sassaman and Hal Finney is that they died. For Le Roux, it’s that he’s been in jail since 2012.)
Hoback asserts, based on these two sentences, that Todd may have destroyed his access to the original stash, rendering himself unable to spend the money and thereby keeping the secret safe. (Todd has countered that Hoback misunderstood a technical conversation, and that “sacrificing” requires first moving the Bitcoins to a new address in order to prove that you actually own them. But the coins in Satoshi’s supposed stash have never moved.1)
And then there’s the proof Hoback has referenced repeatedly since the documentary aired, as a kind of journalistic ace-in-the-hole: Todd’s “body language” when Hoback confronted him with these theories on camera. To Hoback, Todd behaves like a kid caught in a lie, jokingly agreeing he is Satoshi, laughing nervously, looking shifty. (Set aside for a moment that these kinds of I-know-it-when-I-see-it lie detector assertions are the stuff of junk science, even among experts. The thesis that awkwardness is proof of guilt—awkwardness in the face of on-camera accusations that you are secretly worth billions of dollars and have deceived everyone you know—refutes itself. Especially when you’re talking about a species of programmer with, let’s just say, a penchant for social awkwardness.)
After the documentary’s debut, Todd explicitly denied being Satoshi repeatedly, and reportedly sent Wired time-stamped photos showing he was offline at times Satoshi is known to have posted. Hoback maintains that it’s all part of Todd’s grand design. “He’s a master of game theory,” he told the magazine.
Perhaps. But there also exists publicly available evidence that casts doubt on the idea that Todd could be Satoshi. It’s easy enough to find with a few hours of research. Some of it requires a mild technical understanding of how software is written and delivered, so stick with me for a minute.
Software engineers are creatures of habit. While anyone can learn and evolve in their profession over time, at a given moment programmers tend to have preferences about every technology they use, from tabs vs spaces in code, to programming languages, to file types. One way to consider the Satoshi question is to ask what Satoshi’s habits were when writing Bitcoin’s code, and whether they match up with any given candidate.
As it happens, in 2008—the year Satoshi wrote Bitcoin—Peter Todd was devoting much of his time to another software project. In April of that year Todd announced on PIClist, a listserv for microcontroller enthusiasts, that he was releasing the first version of a circuit board design program called Tuke. In his announcement, Todd noted the significant hours he’d devoted to the project: “I've been working an average of about 20-30 hours a week on this since the beginning of February, and 40-50 hours since school ended in early April,” he wrote. “However I start a new full-time job in two days, so don't expect me to get very far very fast with any of the above. :)”
This time-consuming development overlapped with the months where Satoshi was toiling away on Bitcoin. That November, Satoshi announced the Bitcoin white paper on the Cryptography mailing list, writing in a follow up email that “I've worked through all those little details over the last year and a half while coding it.” Fast forward to December 2008, just a few weeks before the official launch of Bitcoin, when Satoshi would have been in the most intense final stretch. What was Peter Todd working on? Porting his Tuke software to another programming language, called Haskell.
Of course Todd could be one of the world’s great multitaskers, developing Bitcoin alongside a full-time job and another software project that seemed to be occupying his free time. Or perhaps he was exaggerating the time Tuke was taking up. So it’s worth examining his habits around writing and releasing his software, as compared to Satoshi.
Tuke was, first of all, written in the coding language Python. Bitcoin, famously, was written in C++. Different languages for different purposes, you might say. But the ways in which the programs were released also differ in telling ways.
In 2008, there were two main varieties of what are called “version control systems”—software that allows multiple programmers to collaborate on the same code, managing their changes. The older and more predominant variety at the time was called subversion, or SVN. SVN was employed on a site called Sourceforge, which served as a code repository for many prominent open source software projects. However, a newer version control system called Git was taking off, particularly with a Sourceforge competitor called Github, which launched in early 2008.
Todd, it turns out, was an enthusiastic early adopter of Github. He launched Tuke there, and an archive of his Github account from the time shows that many of his other projects are hosted there as well. In fact, in 2007, he had described specifically having tried and abandoned Sourceforge’s flavor of version control, and raved about newer more sophisticated forms, like Git and a similar system, Monotone. “I used to be a CVS [another older version control system] guy for years, then subversion for a few months,” he wrote, “but these new revision control systems are just mind-bending in how they make you rethink how you work.”
So we know that Peter Todd is passionate about version control, believed newer options had surpassed SVN, and had moved to Github instead of Sourceforge. What about Satoshi? Well, in January 2009, Satoshi released the first version of Bitcoin as a single compressed file on the bitcoin.org web site and… on Sourceforge. Not only that, but Satoshi didn’t even bother to deploy Sourceforge’s version control option, SVN. (When other developers wanted to collaborate with Satoshi, they simply exchanged code by email.)
Bitcoin’s version control was eventually set up in 2009 by the early Bitcoin developer Martti Malmi (a.k.a. Sirius). When this happened, did Satoshi point out that Git or Monotone could be a more advanced choice, or even mention those or Github as possibilities? No. “Thanks for setting that up,” is all he wrote to Sirius about SVN, in October of 2009. Later, Satoshi actually complained about Sourceforge, as if there is no better option out there. “Sourceforge is just so darn slow,” he wrote to Sirius. “I don't know what else to do though.”2
There are more disconnects that emerge across Todd’s old habits, and within Satoshi’s own posts and emails. Some are more technical, while others are uncomplicated. Satoshi, for instance, was clearly a Windows-first developer. Todd, around the same time, noted his own operating system preference on his website: “I do all my development on Linux,” he wrote.3
Believing Todd is Satoshi requires believing that he adopted an entirely new approach to developing and releasing software, forgoing his feelings on version control, and then later approved a version control format he believed was dated. It requires thinking that Todd (as Satoshi) settled for what he saw as inferior approaches for what he believed to be world-changing software, which he’d spent 18 months on, having finally cracked the biggest challenges in digital currency.
Another line of evidence against Todd being Satoshi actually shows up in “Money Electric” itself, although it passes without notice in the film. Hoback discovers that as a teenager back in 2001, Todd exchanged messages with both Adam Back (inventor of the Bitcoin precursor Hashcash and regular Satoshi candidate who features in the film) and Hal Finney (a key early Bitcoin figure and Satoshi candidate who died in 2014) about digital currencies on a cryptography-related listserv called Bluesky. The documentary leans heavily on these exchanges, as they reveal how early Todd was thinking about cryptocurrency, and even communicating with the field’s major thinkers.
But the documentary also shows, in passing, Satoshi’s first emails with Adam Back, before the white paper is published. One of the most telling clues contained in those emails is the fact that Satoshi seems unaware of a digital currency concept floated years earlier by Wei Dai, called “B-Money.” Back, in these emails, suggests Satoshi look into Dai’s work.4 The exchange suggests that Satoshi invented Bitcoin without studying some of the major cryptocurrency ideas that preceded it.
Peter Todd, however, would have known about Wei Dai and B-Money. Why? Because the Bluesky list was in fact owned and moderated by none other than Wei Dai himself. In fact, B-Money was literally discussed between Back and Dai, on the Bluesky list, in the same weeks Todd was posting to it.
Perhaps Todd, as Satoshi, pretended to be unaware of Wei Dai, knowing that his teenaged emails from years before would look suspicious? But then… why email Adam Back at all, since doing so would also make those emails look suspicious? The game theory starts to run out of moves.
At Pubkey, the crowd had settled into their seats when Todd wandered in, bearded and wearing a backpack. Wired had recently reported that he’d gone into hiding after the documentary. But he wasn’t flanked by security or hustled in a side door, in the manner of a secret multi-billionaire. (“I am in hiding right now,” he later joked to the crowd. “I’m hiding behind all of you.”) He seemed relaxed, and took a seat on stage with Pete Rizzo, a journalist and Bitcoin historian who spent the next half hour quizzing him about the documentary and its conclusions.
Todd claimed to have been unaware that “Money Electric” would actually out him as Satoshi, up until journalists with early copies of the film started calling. “You’ve got to remember, Cullen asked a lot of questions,” he said, noting that he’d done hours of interviews over multiple sessions. “I didn’t leave with the impression he was really serious about this.” The film’s evidence “was bullshit,” he said, offering up his explanations about the forum posts and “bitcoin sacrifice” chat.
But on stage he also repeatedly joked about being Satoshi, “accidentally” slipping into first person when discussing him. It was the same kind of awkward meta-humor that Hoback found suspicious in the first place. One person’s game theory is another’s trolling.
The journalistic work of investigating Satoshi’s mysteries can be technical and tedious—I know, having lost countless hours to it. Code comparisons and spreadsheets of time stamps don’t make for riveting TV like a guy squirming on camera. But the trap of the Satoshi hunt, as ever, is that a how-about-this-guy approach tends to dampen investigative rigor. Locked onto a suspect, the investigator piles up circumstantial detail that, as Hoback says in the film, “checks a lot of boxes.” Coincidence becomes proof.
Meanwhile, contrary evidence is held just out of frame, along with other suspects for whom larger piles of evidence have already been gathered. “Money Electric” spends barely a scene dispensing with candidates like Hal Finney and Nick Szabo. The former is dead and the latter declined to speak with Hoback, ruling out a cinematic confrontation. The film only name-checks Len Sassaman—the case for whom is by many measures more compelling than anything provided for Todd (although also not without counter-evidence). And “Money Electric” briefly feints toward Adam Back as a candidate. But to arrive at the end of the documentary is to see that it’s been building a case for Todd all along.
One wonders whether, once the narrative train got rolling, Hoback ever slowed it down to consider the logic around Satoshi’s motivations. Does it seem likely that someone who developed Bitcoin and then left the community, saying they’d “moved on to other things,” returned anew under their own identity to wade back into its developer wars? And if they did, that they battled in these conflicts for over a decade, in forums and on panels, never leveraging the one fact that would give them ultimate authority to win Bitcoin’s endless arguments? (Whatever one might say about Peter Todd, he seems like a guy who likes to get credit for being right.5)
Does it make sense that one of the modern world’s most meticulous practitioners of operational security, whose secrets have resisted cracking for 15 years by journalists, hackers, and obsessives, would voluntarily sit for hours of interviews for a documentary? And that they’d then be shocked into bodily confession of something they’d jokingly “admitted to” many times?
It’s all possible, of course. Peter Todd could be Satoshi. Real proof may yet emerge. He’ll now be enshrined in the ever-growing list of viable candidates, no matter what evidence is presented to the contrary. The beauty and vexation of Satoshi’s mystery is the way nearly any objection can be explained away as 4D chess or game theory mastery. He knew that’s what we’d think, the story goes. He did that to throw us off the trail. But as each new box-checking subject is added to the pile, the chance of ever convincingly solving the riddle slips further out of reach.
“I can’t prove to you I don’t have Satoshi’s coins,” Todd told the assembled crowd with a smile. “I can persuade you that it’s unlikely. But I cannot prove to you that I do not have Satoshi’s coins.”
Hoback’s assertion that Todd destroyed the stash, if correct, would also mean that Satoshi is not one of the richest people on earth. The contradiction goes unremarked in the film, which hangs its premise in seeking Satoshi on that very fact.
In nearly two years of emails between Satoshi and Sirius, Github is never even mentioned. Bitcoin was eventually moved over to Github, but not until 2011, as Satoshi stepped away. There’s no record of Satoshi having an account on Github, or contributing to the code there.
Other divergences followed from this one. For example: Satoshi released the original, Bitcoin code as what’s called a RAR (.rar) file, a form of file compression popular with Windows developers at the time. Todd habitually released his software using a file packaging format called tarball (.tar)—more common among Linux developers.
“Thanks, I wasn’t aware of the B-Money page,” Satoshi replies. He then proceeds to email Wei Dai, asking how to cite B-Money, which he thinks dates to 2006. Dai responds, correcting him that he released it in 1998, on the Cypherpunks mailing list. Satoshi follows up with Adam Back afterward: “Thanks for the pointers you gave me to Wei Dai’s B-money and others.”