[cryptography] (off-topic) Bitcoin is a repeated lesson in cryptography applications - was "endgame"

ianG iang at iang.org
Fri Feb 24 21:53:28 EST 2012


I changed the title so those only interested in pure crypto can delete 
and move on.


On 25/02/12 04:24 AM, lodewijk andré de la porte wrote:
>     This was an offtopic discussion from the start. The original paper does
>     not include anything about crypto.


Yeah, except Bitcoin is a cryptographic construct from the get-go.

It is a singular lesson in what happens when cryptographers dabble 
outside their particular expertise - in this case, monetary 
architecture.  I.e., some very smart cryptographer thought he had solved 
the 'gold' problem with a cryptographic construct.  But alchemy has a 
longer history than he realises.

It is also a singular lesson in the emotive power of cryptography to 
encourage large numbers of people to hash their intelligent thought 
processes.  What we are seeing is otherwise rational people invest much 
time & effort into what amounts to a ponzi or bubble or pyramid scheme.

Many people are being hurt by this, and more to come.  Same or similar 
thing happened with PKI, digital signatures, Digicash, SSL, e-gold, etc...

I suggest it is professionally important for serious cryptography people 
to follow this story and try to understand what goes wrong.  Or right. 
One day, you might be asked to design something like this.  Then you'll 
be involved.

Then you'll find out about Santayana's curse - those that don't study 
history are doomed to repeat it.  For reference, start with read John 
MacKay, _Extraordinary Popular Delusions and the Madness of Crowds_.

Published in 1841.



>     Anyway, the problem you mention is exactly the one described in the
>     paper.
>
>     "Using Mancur Olsen's rationale that a prince is a bandit that stops
>     roving, the notion of the mining franchise being captured by the botnets
>     might have been an acceptable compromise to the economy growing up
>     around bitcoin mining, if it went no further [Olsen].  However,
>     criminals are rarely satiated.  Several things happen: (a) incentives
>     for easy money naturally cause an increase in criminal participation at
>     all levels, such as direct theft of bitcoins.  This increase across the
>     board encourages (b) honest users to pack up and leave.  Both of these
>     effects combine to create rising criminality, and (c) at some stage the
>     Feds get involved.  Finally, (d) the system collapses."
>
> So "criminals" exist and they want to make money (which they already
> could but now they want more).  Now something happens that summons an
> unbeatable* nemesis/third party and everything goes to hell.
> Nice line or reasoning. Very certain, unbiased, etc.
> Funny thing is that everyone believes them because they can use LaTeX,
> put references (to websites, most of which are bullocks themselves) and
> call it a paper. It's just another rambling about something that could
> but really won't happen.
>
> Don't forget to put things into perspective.


Perspective ... there's a funny word.  Here's mine, from the e-gold 
experience and others [*].

What happened to e-gold was this:  it worked, it boomed.  It was 
structured fairly well to avoid the obvious attacks.  It was good.

It was very cheap for users - around 0.5% transaction fees.  And 
payments were final.  Which happened to work for an "arbitrage" class of 
customer variously known as real-time games, ponzis and pyramids.   This 
is e-gold's "porn video" story, if you remember the 1980s.

The problem with these games was that (a) they were making a lot of 
money and doing a lot of transactions, (b) they were taking a lot of 
money from people who wanted to play, (c) which latter would then turn 
around and complain to "authorities" and (d) the games were more or less 
illegal in some sense or other.

Each of those points by themselves was challengable and acceptable and 
mitigatable at some level.  Together however the result was as laid out 
in that paper.  At some point something happened and /the feds/ or the 
nemesis got interested.  Endgame.

The weaknesses that brought down the e-gold story were two-fold [$]. 
Firstly the management had a certain belief set that led them to charge 
on and not take pause - to not respond to emerging developments (in a 
way that Paypal were able to do for example).  They did not respond so 
well to the games.

Secondly, the exchange network.  Although highly diversified, 
independent and robust as a system, each exchanger had to be somewhere 
and had to have relationships with the primary banking system.  This 
made them vulnerable to a single phone call.  And so it happened - phone 
calls were made, exchange relationships were dropped.  Most of the 
exchangers had perpetual trouble opening and losing bank accounts. 
Eventually most dropped out or were raided or had their funds seized or, 
etc etc.



This second weakness exists with Bitcoin.  The bigger the exchangers get 
the more vulnerable they get, hence the exchange network has scalability 
problems at the nodal level.  To get a taste of what this is about, see 
Jon Matonis' blog:

http://themonetaryfuture.blogspot.com.au/2012/02/foreign-located-money-services.html
http://themonetaryfuture.blogspot.com.au/2012/02/paxum-exits-from-bitcoin-business.html

You can almost hear the phone call....

So yes - I'm as certain as history can make it.  What is predicted in 
that paper is certainly based on history.  The pattern was repeated.  It 
was even known in advance of e-gold:   someone I knew laid out the 
e-gold story for me before it happened.  He said it would take 6 years. 
  He was right.


> *can't really beat anything, they can only make it crime-exclusive. (you
> make it illegal and only those that don't care about the law can use it.)


That's it!  Now, leave aside the libertarian hopes and the politics and 
the freedom bias and right to code and the "this time it's different" 
and all that crap -- and ask yourself.

Where do you want to invest your future?



iang



[*]   I wasn't e-gold, I was running a company called Systemics that 
made hard crypto payment systems, and was preparing the next generation 
of software, in a sense.  But over time I became a sort of trusted 
advisor to e-gold, one of the inner circle, and I saw most of everything.

As well as being part of the e-gold experiment, I came from the same 
town as the DigiCash experiment.  I was influential in the goldmoney 
experiment, and involved in countless other small scale issuances of 
money.  About a decade++ building payment systems.

[$] ok, there were other weaknesses.  I am handwaving them away to get 
the point across.



More information about the cryptography mailing list