Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Amanda Knox Redux: is Satoshi Nakamoto the real Satoshi Nakamoto?

12 Post author: Cosmos 06 March 2014 11:33PM

Many of you here have likely heard of Bitcoin, and maybe know something about it.

Earlier today, a story broke that a reporter had apparently tracked down the real Satoshi Nakamoto, infamous creator of the Bitcoin protocol.

This seems like an excellent opportunity to practice our Bayesian updating!

So, how likely do you think it is that this man is the founder of Bitcoin? What do you believe and why?

Comments (54)

Comment author: gwern 07 March 2014 12:00:46AM 14 points [-]
Comment author: Pfft 07 March 2014 02:46:18AM 3 points [-]

Does the fact that the Satoshi Nakamoto account denied it really add any confidence beyond knowing that Dorian Nakamoto denied it? If they were the same, he could issue a denial through both channels.

Comment author: Kaj_Sotala 07 March 2014 03:42:39PM 6 points [-]

My gut instinct would be that Satoshi denying it actually makes it more probable that it's Dorian, given that we've had other suspects before but Satoshi has never bothered commenting on them. If nothing else, there's something about Dorian that makes him interesting to Satoshi in a way that hasn't been true for any previous suspect.

Comment author: gwern 07 March 2014 05:14:21PM *  27 points [-]

My response to that is never before has a Satoshi proposal led to a scrum of reporters on a target's front lawn, and also none of them have been sick or old (except Finney). If Satoshi was ever going to deny it, of the targets so far, Dorian is the best. And there's an additional factor here: Satoshi may feel directly responsible for Dorian Satoshi Nakamoto's problems, given that he could have chosen a globally unique pseudonym (it's not that hard) but failed to do so and so put people like Dorian at risk. This is untrue of any suggestions about Wei Dai, Hal Finney, Nick Szabo, etc.

(This has, incidentally, been a good reminder of why I don't post my Satoshi research publicly, and generally limit my comments to debunking proposals.)

Comment author: ESRogs 07 March 2014 06:21:43PM 9 points [-]

This has, incidentally, been a good reminder of why I don't post my Satoshi research publicly, and generally limit my comments to debunking proposals.

Despite the fact that I would personally be very interested in this research, I would like to endorse this as a good policy.

Comment author: wwa 11 March 2014 03:02:07AM 2 points [-]

(This has, incidentally, been a good reminder of why I don't post my Satoshi research publicly, and generally limit my comments to debunking proposals.)

If you don't mind sharing, what was the reason for research, besides curiosity?

Comment author: gwern 11 March 2014 03:15:36AM *  5 points [-]

The usual: curiosity; deriving security lessons; research practice; prolonged & severe irritation with the incompetence of other Satoshi hunters; the outside chance of undying renown & deathless fame.

EDIT: Seth Robert furnishes a handy example of the crappy reasoning & rampant confirmation bias in such discussions: http://blog.sethroberts.net/2014/03/11/nick-szabo-is-satoshi-nakamoto-the-inventor-of-bitcoin/

Comment author: wwa 11 March 2014 11:34:04PM 2 points [-]

Thank you. I asked because I don't understand other people's attraction to personal details like this. Nothing specific to Satoshi or bitcoin.

Comment author: ESRogs 25 March 2014 10:45:53PM 1 point [-]

In case you haven't seen it, here's a great piece about Hal Finney, which includes the strongest arguments I've seen so far about why he's not Satoshi (despite the author's initial suspicions):

http://www.forbes.com/sites/andygreenberg/2014/03/25/satoshi-nakamotos-neighbor-the-bitcoin-ghostwriter-who-wasnt/

Comment author: gwern 26 March 2014 12:39:04AM *  1 point [-]

Yes, I too agree Finney is less likely to be Satoshi than is commonly assumed. Mostly based on the early emails to each other - it beggars belief that Satoshi would set up two accounts, sockpuppet them talking to each other, find crashing bugs as one account and report it with debug logs to the other account, etc. Satoshi was trying to be pseudonymous, but nothing suggests he was so paranoid that he would resort to such ruses and stratagems that early on.

Comment author: Kaj_Sotala 07 March 2014 09:06:23PM 1 point [-]

Good points.

Comment author: ESRogs 07 March 2014 02:53:49AM 0 points [-]

Was Dorian back home from his day out with reporters yet?

Comment author: ESRogs 07 March 2014 02:18:24AM 0 points [-]

astounding. Satoshi has had no publicly known activity since November 2013, and no public messages since ~2011

I agree! I wasn't aware of the Nov 2013 activity though, what was that?

Comment author: gwern 07 March 2014 03:51:59AM 0 points [-]

If you can figure it out yourself, I'm willing to discuss the details.

Comment author: ESRogs 07 March 2014 04:27:08AM 0 points [-]

I had a feeling I might get that kind of reply :) I'll see what I can find...

Comment author: ESRogs 07 March 2014 02:04:58AM 10 points [-]
Comment author: khafra 10 March 2014 10:59:57AM 2 points [-]

My first-pass pattern recognition parsed this comment as a claim that you, personally, were not Dorian Nakamoto, with a link to a confidence level for that claim on predictionbook.

Comment author: ESRogs 10 March 2014 08:38:00PM 2 points [-]

Haha, I bet I could appear pretty well calibrated if I bet only on claims about myself. :)

Comment author: jimrandomh 07 March 2014 12:43:33AM 9 points [-]

Nope. I say p~=10%.

  • The Newsweek article mentions that there are several people named Satoshi Nakamoto
  • No evidence that Dorian Nakamoto programs or has programmed
  • The reporter tells a straightforwardly false narrative of when and why Bitcoin Satoshi left, which makes me think she's not honest in general
  • Bitcoin Satoshi was privacy conscious from the beginning, as evidenced by using vistomail for his email address and registered bitcoin.org through a proxy
  • The only writing sample of Dorian Nakamoto I've seen (an email about model trains) is a mismatch
  • Dorian Nakamoto is old
  • Dorian Nakamoto isn't wealthy
Comment author: gwern 07 March 2014 01:00:22AM 10 points [-]

Dorian Nakamoto isn't wealthy

That would actually be a point against: since now you need to explain how Satoshi could have cashed out a lot of coins early (at low prices) or any coins recently (at high prices), without anyone noticing, keeping in mind people (such as myself) actively are watching early coin-rewards and especially the ones tagged as the Satoshi miner by Sergio's research.

Comment author: jimrandomh 08 March 2014 07:38:07PM 10 points [-]

Not quite what I meant. The fact that Satoshi didn't sell many of the early coins, and didn't mix them around enough to keep the option open either, is evidence that he felt financially secure without them. That suggests that he had money already, from a source other than Bitcoin - which, for a skilled programmer, has a high prior already.

Comment author: gwern 08 March 2014 08:41:03PM 7 points [-]

Ah, that's a good point. Yes, I agree with you that I expect Satoshi to be well off based on his other activities (or alternately, have a good excuse for being poor, like being a grad student).

Comment author: solipsist 08 March 2014 12:18:16AM 3 points [-]

That would actually be a point against.

Really? If Dorian Nakamoto had recently and mysteriously come into a lot of money, that would lower your estimate that he was Satoshi? Satoshi is less likely to have accumulated an unexplained fortune than the base rate?

Comment author: gwern 08 March 2014 04:42:09AM *  4 points [-]

Yes. The early coins have been scrutinized pretty heavily, and the more time that passes, the harder it is to move a lot of coins (eg. with each day, the coin-days-destroyed increases, irrespective of exchange rate; and blockchain analysis gets better, which because Satoshi did not move the coins early on is going to work against him ever more - normally, the more time that passes and the more transactions that happen, the worse and more unreliable blockchain analysis gets, but since he's done no transactions...).

If Dorian came into a lot of money, I would want an explanation: did he sell an early Bitcoin startup? Did some other project (the things Satoshi moved onto) yield dividends? For example, one problem with Jed McCaleb as a candidate was that he seems to be rich, but that problem was satisfactorily addressed by his MtGox sale & Ripple work.

(That investigation proved pointless, incidentally: I ultimately scrapped McCaleb as a candidate because the password hashing on his MtGox codebase was inexcusably sloppy for Satoshi, and I simply can't see Satoshi deliberately endangering MtGox users just for the off-chance of a hack revealing the sloppiness and inducing researchers to write him off as cryptographically ignorant.)

Comment author: ThisSpaceAvailable 10 March 2014 08:07:34AM 2 points [-]

Your first paragraph seems to be directed more at P(Dorian came into a lot of money | Dorian is Satoshi) than P(Dorian is Satoshi | Dorian came into a lot of money). It explains why you think that being Satoshi does not significantly increase the probability of coming into a lot of money, but it doesn't explain why you think being Satoshi decreases the probability.

Comment author: gwern 10 March 2014 04:58:45PM *  3 points [-]

I think I did explain that. Satoshi is one of the early Bitcoin adopters whose finances we understand best: ie he has cashed out and moved ~$0. Being Satoshi substantially decreases the odds of mysteriously being wealthy as compared to all the other early Bitcoin adopters, whose finances are not so nailed down and so who routinely surprise us watchers.

For example, even as I type this, we are being surprised by an early Bitcoin adopter: the 1933x addresses (with 111k btc or ~$70 million's worth), generally believed to have been Ross Ulbricht's Silk Road commissions from the first year of operation, abruptly started moving yesterday after being completely inactive since ~Nov 2011; Ulbricht is still in jail, has not pled, and the FBI presumably long ago finished its forensics, so the inescapable conclusion is that we were completely wrong about it being Ulbricht and it's an unknown actor. If you find someone involved with Bitcoin early on who matches Satoshi's profile but mysteriously has millions, you haven't found Satoshi - you've found someone like 1933x!

Comment author: Eugine_Nier 11 March 2014 01:02:54AM -2 points [-]

It's possible he hid some bitcoins somewhere the FBI didn't get to them, and is now moving them.

Comment author: gwern 11 March 2014 03:18:53AM 3 points [-]

He's in jail; he has no access to the hardware the FBI seized (with all his encrypted files on them, and it's hard to unencrypt an encrypted file when you don't have a copy, even if you know the passphrase); and there is no particular reason to move the coins at this time, with Mtgox shattered & the database leaked. It's extremely implausible that it's Ross, now.

It's much more likely it's either Karpeles or a Mtgox insider moving around gains for some reason, or the insider realizes their Mtgox account information is now public and is scrambling to disperse their fortune to less-easily-tracked addresses.

Comment author: Eugine_Nier 12 March 2014 12:16:58AM -2 points [-]

Maybe he hid a backup on some public data hosting site.

Comment author: ESRogs 07 March 2014 01:20:49AM 3 points [-]

Dorian Nakamoto is old

That part matches though doesn't it, given that SN's code used hungarian notation and otherwise seemed to be a bit old school?

Comment author: Jayson_Virissimo 07 March 2014 01:46:19AM 3 points [-]

That part matches though doesn't it, given that SN's code used hungarian notation and otherwise seemed to be a bit old school?

You don't mean reverse polish notation? There are other signs of age too, like using two spaces after a full stop.

Comment author: jkaufman 14 March 2014 08:33:58PM *  4 points [-]

Hungarian notation in a random bitcoin file I opened: coins.cpp

if (!fZero) {
nLastUsedByte = b + 1;
nNonzeroBytes++;
}

See the letters preceding each variable? Those tell you what type of variable they are. Here I suspect "f" is for float and "n" is for "integer". This is not a common style for programming, and especially this kind of it ("systems hungarian") is associated with Microsoft.

Comment author: Jayson_Virissimo 14 March 2014 08:38:35PM 1 point [-]

I see, thanks.

Comment author: JTHM 07 March 2014 02:41:22AM *  4 points [-]

I use two spaces after every sentence, and I'm 23. It's not a personal quirk either, it was just normal formatting in the American public schools I attended. (By the way, anyone who points out that this very post uses single spaces after a full stop should know that LessWrong messes with formatting. I typed double spaces; it's just not displaying as written.)

Comment author: CellBioGuy 07 March 2014 05:12:25AM 3 points [-]

I use two spaces after every sentence, and I'm 23.

Same. Well, 24.

Comment author: CronoDAS 07 March 2014 08:28:58AM 1 point [-]

By the way, anyone who points out that this very post uses single spaces after a full stop should know that LessWrong messes with formatting. I typed double spaces; it's just not displaying as written.

This isn't a specifically LessWrong thing. Web browsers do that to text - they ignore spaces at the beginning of a paragraph and also between words.

Comment author: Douglas_Knight 07 March 2014 04:40:04PM 3 points [-]

Actually, there is something specific to LW: view source and you'll see that it is the server and not just the browser that is doing it.

Comment author: ESRogs 07 March 2014 02:12:02AM *  2 points [-]

I got the part about Hungarian notation from Gavin's quote here. Hungarian notation is a thing in programming.

Sounds like both systems acquired their names in similar ways (founder / proponent from a particular national origin).

Comment author: jamesf 07 March 2014 01:40:05AM 0 points [-]

The only writing sample of Dorian Nakamoto I've seen (an email about model trains) is a mismatch

The writing may have been done by another person. The original story was that Satoshi Nakamoto is an unknown, positive number of people; is that a worse hypothesis now?

Comment author: CellBioGuy 07 March 2014 07:12:11PM *  7 points [-]

The author of the Newsweek article had no idea what she was talking about. She claimed that a reference to the obsolete concept of "disc space" proved that the author of the white paper had to be old. Its a shameful hack job.

Comment author: ThisSpaceAvailable 10 March 2014 08:30:23AM 0 points [-]

I take it you are ascribing both the idea that it proves that the paper is old, and the idea that "disk space" is obsolete, to her, and are not intending the word "obsolete" to be "your" word? Also, some people draw a distinction between disc and disk, so I think it is at least bordering on acceptable pedantry for me to point out that she used the latter spelling.

Comment author: Caerbannog 07 March 2014 06:39:02PM 15 points [-]

I'm going to be more arrogant and say p << 1%:

  • Prior probability that a super-intelligent, reclusive, crypto-guy that values anonymity would use his given name as his pseudonym: very very low.

  • Subsequent evidence to believe Dorian is Satoshi: He's an apparently intelligent guy that knows how to program + a quote wherein he appears to admit that he was once involved with Bitcoin, but he later claimed was misinterpreted and out of context.

Comment author: Nic_Smith 08 March 2014 03:52:29AM *  5 points [-]

The reporter began with the very questionable assumption that Satoshi Nakamoto was his real name*. Because of this, it is less surprising that they found someone with that name and some resemblance to "The" Satoshi Nakamoto. 0.05.

* "It was only while scouring a database that contained the registration cards of naturalized U.S. citizens that a Satoshi Nakamoto turned up whose profile and background offered a potential match." -- It's possible from this sentence they were doing an unrelated search and just saw the name, but I doubt it.

Comment author: metatroll 07 March 2014 05:52:08AM 24 points [-]

It's obvious that Amanda Knox is the real Satoshi. Just check the timelines.

Comment author: ThisSpaceAvailable 10 March 2014 08:51:27AM 4 points [-]

"I don't think he's in any trouble," I say. "I would like to ask him about Bitcoin. This man is Satoshi Nakamoto." "What?" The police officer balks. "This is the guy who created Bitcoin? It looks like he's living a pretty humble life."

Anyone else find it odd that a random cop knows who Satoshi is?

Comment author: Nic_Smith 23 March 2014 07:21:44AM *  1 point [-]

Only slightly; Bitcoin had awareness on Less Wrong when it was still fairly obscure, IIRC, but I have heard it discussed recently by people offline and far removed from LW (a coworker, "random" people on the street).

Comment author: fubarobfusco 07 March 2014 02:10:04AM 6 points [-]

When I read this story, my response was something like, "I don't believe this very much, but it sure makes a good story — Mysterious Bitcoin founder turns out to be hiding in plain sight under his real name. He's some unassuming old guy, not the anarcho-capitalist cypherpunk everyone was expecting. The fact that it's a good story means a lot of people are going to believe it, since believing things like that is short-term rewarding."

Comment author: Lumifer 07 March 2014 01:23:00AM 2 points [-]

"The man Newsweek claims is the founder of Bitcoin denies he had anything to do with the digital currency.

In an exclusive two-hour interview with The Associated Press Dorian S. Nakamoto, 64, said he had never heard of Bitcoin until his son told him he had been contacted by a reporter three weeks ago."

says AP

Comment author: maia 07 March 2014 01:44:02AM 1 point [-]

That's good to know. It's not much evidence either way, though. The real Satoshi would probably say something very similar to that.

Comment author: ThisSpaceAvailable 10 March 2014 08:46:39AM 2 points [-]

The real Satoshi would likely be in a situation where claiming complete ignorance would be a poor strategic choice. If the real Satoshi were to make that claim, he (I'm assuming it's safe to use male pronouns) would be making a gamble that nothing would come to light that would contradict that claim. It's possible that Satoshi would feel comfortable making that gamble, but there is a significant probability that the real Satoshi cannot credibly claim complete ignorance. The fact that Dorian made that claim is weak evidence that this is a safe gamble for him to make, and it being a safe gamble is weak evidence that he is not Satoshi.

Comment author: niceguyanon 09 March 2014 12:21:19PM 0 points [-]

I was previously 70% sure that Dorian is Satoshi but this interview and other reports that he was taken out of context by the Newsweek reporter has now made me think this is just a coin flip.

Comment author: Transfuturist 06 March 2014 11:49:20PM 2 points [-]

I think it is correct (>70%), but also unimportant (>95%). Even with a revealed identity, Satoshi remains reclusive and private, and I believe he will continue to avoid public recognition (>80%). This is purely from a few handfuls of background information beforehand, and then the article itself.

I am not sure where updating will come into this.

Comment author: Dagon 07 March 2014 07:53:04AM 0 points [-]

What predictions are we making? How will this bet (with ourselves) be resolved?

I mostly expect that I'll never know for sure. I think it's somewhat unlikely that this guy is (or was) a billionaire without showing it or anyone (including his family) knowing it.

Comment author: CronoDAS 07 March 2014 08:35:08AM *  3 points [-]

Well, let's see you try selling an amount of bitcoins currently valued at a billion dollars without tanking the market price. :P