[Cryptography] What has Bitcoin achieved?

ianG iang at iang.org
Wed Jun 25 01:57:54 EDT 2014


On 25/06/2014 00:50 am, John Levine wrote:
>>> I wouldn't say that mPesa is issued by telco (Safaricom). Its based on
>>> Kenya shillings.  One just open an account and put the Kenya currency on
>>> the account. Its this currency that Kenya use for trading.
>>
>> Well, the term 'issue' is generally to imply that one stands behind the
>> contract, in the sense of 'issuer.'  In this case it is Safaricom that
>> is the issuer of the unit, and they stand behind their contract (to the
>> extent they do, never really looked at those aspects).
> 
> They're acting as a bank.

You mean, a cheque payment system.  As a matter of monopoly marketing,
banks would like you to believe that only a bank can accept cheques
(checks).  This is really a marketing distinction, and it is not
recognised in all countries all the time.  For example, in Europe there
is a Banking Directive and a Payment Systems Directive.

The point of distinction for banks is deposit taking -- which means
taking your money, placing it into long term assets (mortgages) but
promising it back to you on demand.  A contractual fraud for others,
permitted by the 'charter' for banks.

However, to attract deposits, they offer checks, which they will clear.
 It's just a marketing product.  Many non-banks such as S&Ls or credit
unions will also offer checks, in different countries.

There's no reason why your employer couldn't offer checks against your
salary... except his bank would get upset.


> Doing an mPesa transfer is conceptually the
> same as writing a check that directs your bank to pay money from your
> account to someone else.

Right.  But the check (cheque) technique is just a way to send an
instruction to the money issuer.  How would you do it in a digital
sense, otherwise?

> The only way to turn it into cash is to take
> it to a branch of the bank.  They probably don't have the equivalent
> of deposit insurance, but lots of banks don't, either.

Deposit insurance is a central banking invention that can be applied to
anyone in the net.  It's not really a hallmark of banking but of central
banking.  As it happens, post-cyprus, every central bank is scurring
around trying to figure out how to drop their insurance.

But, as far as mPesa goes, they put their aggregate float account with a
bank.  That probably doesn't earn deposit insurance via central bank
rules, but likely they have some other form of accomodation.


> (I realize
> nobody uses checks in Australia any more, but substitute whatever you
> use to tell your bank to pay bills.)

So, checks.  Tech.  How would you instruct your money issuer to move
money from Alice to Bob?  In Bitcoin?  In some hypothetical 3 party
money system with Alice, Bob and Ivan?

You'd write a check as a digital layout specifying Alice, Bob, the
contract and the amount.  Sign it with your public key.  Send it in a
packet to the issuer.  Who would 'settle' it.


> What we seem to be discovering here is that no matter what kind of
> electronic currency you use, you have to trust someone.  The only
> question is who it is, and what if any recourse you have if they are
> crooked or incompetent.


Well, yes.  Pre-blockchain cryptocurrencies, we just called that person
Ivan and wrapped some governance about it.  NS said no, we need to not
have a person, so he wrote a blockchain settlement system, and
eliminated the contract's issuer, and wrote the contract into hard code.

But this shoved the onus onto the dev team.  Which is fine, in some
sense it is just moving the pieces around, the head of engineering at
the the dev team is now the one who can decide to change the formula.
But he can be watched easily, and it turns out that techies are good at
reading code, if not contracts.

Conceptually, one can see who we have to trust, same as before.  In
market results, the difference is radical, and inspiring to a lot.

Your point about recourse is a good one tho, blockchain technologies
don't have much of an answer to that, to their cost.



iang


More information about the cryptography mailing list