From: "Vladimir Z. Nuri" <vznuri@netcom.com>
Date: Sat, 28 Oct 1995 06:02:35 +0800
To: Steven Levy <steven@echonyc.com>
Subject: Re: newsweek oct 30 Levy digital cash article
In-Reply-To: <Pine.SOL.3.91.951026180246.16411E-100000@echonyc>
Message-ID: <199510272049.NAA29753@netcom7.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain



S.L.:
>>also, my question about whether one loses the downloaded Chaumian bucks
>if one's hard drive crashes was answered in the affirmative by Levy.
>
>But as I said, if the drive is backed up you don't lose it. It's my 
>understanding this is the case with the current Digicash/Twain bucks.  
>But not necessarily with all implementations.

you apparently didn't notice but there were some other posts on this
list about writing down the random seed that a person used to generate
the cash, and then being able to restore the cash somehow based on knowing
the random seed, however privacy is sacrified in this case apparently.
this with the Chaumian implementation, from what I understood.

if this is true, I wanted to point out in my post that this is a property
that real cash does not have. you cannot ever say to your bank from which
you withdrew cash, "oops, I lost a piece, could you replace it? also, if
someone spends the piece that I lost, tell them they can't?"

how would the bank handle a situation in which the consumer says, "oops,
my hard drive was compromised. all of the following cash was NOT spent
by me." notice this is precisely what happens with stolen credit cards today
all the time. the bank, if it was real cash, would have to say, "sorry,
you're out of luck. it's your responsibility to keep your cash secret".

or, the bank might say, "oh, well, that cash was not spent yet. we can issue
a replacement and bar the cash from being spent". this idea of "replacing"
cash is of course unique to the electronic realm. if the bank does it,
the consumer would be required to notify the bank before the cash is 
spent, again another unusual property unique to e-cash.

I would suggest to all companies developing e-cash that they print up
a carefully designed FAQ in which their cash properties 
are compared directly with parallel
properties of credit cards such as like the item above, because the
consumer is of course very familiar with credit cards. (this will be very
crucial for longterm customer education and acceptance imho. otherwise
the customer may make assumptions like, "since the cash comes on a card
it behaves like a credit card", etc.)

I would like to see someone elucidate what happens with lost Chaumian cash
in a detailed post (the oneliners I saw so far don't address the complexity 
and seriousness of this issue. this is a really significant issue with digital 
cash IMHO that will become very obvious once people who were used to the 
"mushiness" of credit card transactions switch to cash).
there seem to be many caveats with this "restoration".

one immediate problem I can imagine: the recent netscape bug showed
very well that a short random number generation process can be hacked.
so in other words, if the random numbers are short enough to write down,
it makes me wonder about the security of the cash generated from this
seed.

>This "lose your money" with digital cash really does seem to get to 
>people. I always point out that when you take $ out of an ATM machine, 
>you don't expect to get it back if you lose it.  For those forms of 
>e-money that are irretrevable when lost users are warned not to download 
>huge amounts. 

yes, it is clearly an emotional issue. the problem with a lot of technology
is that people are very emotional and irrational, and hyperactively so when it
comes to spending money. all marketing campaigns are designed to 
capitalize on this. various competitors can be tainted merely by
cheap psychological ploys that have no basis in fact. (the back and
forth between AT&T, MCI and Sprint and other competitors is an example
of the endless mudslinging, marginally based on reality, that can ensue.)

as I wrote in my earlier post, the key is to let people know of the weaknesses
in the product ahead of time so they are prepared. and the analogies that
you offer are precisely the kind the company can give to the customer to
make them feel more comfortable that "this stuff is more convenient than
real cash and not really any more dangerous". it's critical that they do
so ahead of time however, before the customer gets zapped by his own naivete.

>Question: will people's worries about losing their e-money lead them to 
>accept a higher degree of tracibility as a tradeoff?  

that's exactly the issue I tried to raise in my post, unfortunately perhaps in
not the crisp way you do here. (of course, you're the one that makes a living
off your prose, heh.)

one can see advertising as a kind of yin/yang game. you have to imagine
that if your competitor says, "gosh, you could lose cash with so-and-so's
system", you can always reply "yes, but you retain privacy-- can you do
that with so-and-so's system"?  digital cash is not a panacea. it is
in fact a design compromise that simply applies more importance to
various aspects of transaction capabilities: ease of use, privacy, etc.

checks and cash are at two ends of the economic transaction spectrum.
in contrast to other cypherpunks I don't really see either as intrinsically
superior to the other. many businesses may *never* end up using digital
cash in certain transactions (i.e. with other companies, not individual
buyers) for traceability reasons. it seems to me
there are many agreements in which hiding the identity of one or both
parties is simply intrinsically unacceptable to both, and no amount of
handwaving smoke-and-mirrors appeals to magic "reputation agencies" will
completely alleviate this.

so my bottom line is that I think digital cash is going to become a
very, very important part of the future economic ecology, but it will
coexist with other methods. it may very well become the dominant method,
but I suspect there will be significant areas of commerce that actively
resist digital cash technology out of the preferences of all parties.

something the hardcore cypherpunks in general fail to consider, that was 
brought out in a great article in the WSJ about the counterfeit 
$100 superbills circulating:
the cash aspects of the global economy are negligible in comparison to
"identified" transactions such as through credit and checks. (hence the
treasury's lukewarm approach to stopping counterfeiting, which scares
the willies out of me personally). there is absolutely an *enormous*
dedication to "identified" transactions in the world today, and cypherpunks
seem to me to be somewhat misguided in the ways they exalt cash.

it is true that cash is the medium of choice for the individual, but the
vast majority of economic transactions are *not* made by the individual
in virtually any civilized society. digital cash may change this slightly,
but not significantly, I suspect. businesses exchange money as part of
complex agreements and contracts in which liability is a key ingredient.
liability is an exceedingly problematic issue with digital cash. if anyone
is going to create a widespread cash economy based on total anonymity
they have an uphill battle easily far more difficult than Chaum has
encountered so far. (remember Chaum only implements "semi-anonymity", 
anonymity on payer side but not payee side).

--

anyway, congrats on the fine article SL. you get my "honorary 
cypherpunk of the week" award for the article <g>