dbts: Georecursive Auctions (was Re: Friedman (The Younger)Sings...)

[Robert Hettinga <rah@shipwright.com>]

-----BEGIN PGP SIGNED MESSAGE-----


At 4:01 PM -0500 on 12/21/98, Somebody wrote:


> > If it's on my hard drive, it ain't yours anymore, in other
> > words.
>
> What are the assumptions?  Assuming that you were given what is on your hard

> drive, or can you have procured it by any means, like piracy, theft, etc.

:-).

I'm going to do a real rant on this, someday Real Soon Now, and you can see
the germ of the idea in Freidman's talk, and in my old Shipwright rants
<http://www.shipwright.com/>, and also, I think, on the Philodox site
<http://www.philodox.com/>.

More to the point, we all, as far as I can tell, stole it from Eric Hughes,
who even went so far as to present a paper on the subject, long ago at a
DEFCON far away. Who knows where *he* got it from, but that's enough geneology
for the time being...



The point I'm making above comes from Ronald Coase, the father of modern
microeconomics, winner of Nobel for same, circa 1992?.

The most important thing that Coase said, the thing which got him the Nobel,
was that firm size was directly related to transaction cost.

In an information economy, especially in a world of Moore's law (yielding
increasingly geodesic networks, and the markets overlaid on them) and strong
cryptography (digital bearer transactions, perfect pseudonymity, etc.),
tranaction costs get *very* small indeed, and in a considerable hurry.

To some, I believe clueless :-), people, the above means that we still get
economies of scale, because the first copy of a given modern bloatware
application, like Word for Windows, is enormously expensive compared to its
replication cost, which is functionally free by comparison, even if you put a
box around it, much less if you just put it up for download from the web.

However, *I* say that if you can do profitable transactions in, say, the
micropenny range, you get a world where the bloated, industrially-delivered
code-in-boxes that we've all come to know and love starts to surfact, under
its own weight, into smaller and smaller functional bits, each maintained by
different, and much smaller, firms.

I use open source software as my case in point. Notice that, since the cost of
code itself is free, there is a huge, and expanding, diversity of very
specialized code out there now. The writers of that code don't make any money
on that code except by proxy, through enhanced reputation, and so forth.
However, if you could drive the cost of *transactions* down, people *would*
write code just for the money, and simply auction it off their websites to the
highest bidder, and could do so, profitably, for rediculously cheap prices.
More about that in a minute.


Sure, Red Hat grosses $20 million, but that's more a function of the
cost-domination of sneaker-netted CDROM over the still-scrawny lower
capillaries of the internet. More to the point, I claim, it's the result of
the cost of *credit-cards*, even *checks*, as a way to pay for code.
Book-entry settlement, in other words, which is done "out of band", over
private, hierarchical, and proprietary financial transaction networks.

If it were possible to pay to download code, as you needed it, for
instantaneously net-settled *cash*, and for sufficiently small enough bits of
money, then the need, the price-economy, if you will, for large glops of code
would go away.

And, Coase's theorem says (I think) so too would the large firms required to
generate those large glops of code.


So, what we're left with is a world where the legalities of intellectual
property are physically impossible to enforce, thus words like "privacy" cease
to have any real meaning. If you can't enforce such an intellectual "property"
law, it's kind of hard to call it a law, as most law professors and
philosophers of law will tell you.

Ironically, I claim, that's a very strong form of a *second* observation that
Ronald Coase made, that you can't have markets without private property.
Sounds like a motherhood, now, but he was the first person to articulate that
clearly, back in the early part of this century, and it's one of the reasons
he got the Nobel, of course.

*My* variant on Coase says that you don't need *laws* to ensure private
property in digital form. Cryptography will do just fine. Like I said, if it's
encrypted, and it's on my hard drive, it's now *mine* no matter where I got
it, laws or not. I mean, who's to know I have it to begin with, and, more to
the point, who's going to take it away without my permission, especially in a
world of encrypted offsite net.backups, m-of-n archival reconstruction, etc.

Write software, not laws, to quote the my old cypherpunk chestnut.


Yet, I claim, to finally answer your point :-), people who *make* new code in
such abrave new world, *still* make the most money. No, not "publish", or even
distribute new code, like Microsoft and Ingram do, but the actual guy plinking
away in CodeWarrior, or whatever. The mind behind the code, the developer,
let's call him, in a proper use of the now mostly corrupted word.


Here's how. You write software for a living. You upgrade your existing code
base, and you announce its availablity to the net. Everyone who has your code
knows how good your work is, and they probably have a standing bid in to buy
revs of it at some specific price, anyway. So, just like the specialist's book
at the New York Stock Exchange, you, or your website's CGI, accepts those
bids, highest to lowest, and, as new bids come in, you sort them into the
queue of outstanding orders, by price, of course, :-), and accept them in that
order.

Notice something important, but obvious, here. The people who bid the highest
get the first copy. We'll worry about Vickery variations on that auction some
other time.

Secondly, there's the fun part, the *recursive* part of "geodesic recursive"
or neologically, and, this very afternoon, "georecursive", auctions.

That is, if I paid a monster pile for the first copy, I must have such a
profitable need for it that it's worth it to me to pay that much. Of course,
one of the most profitable things I can do especially in a world where the net
isn't quite geodesic enough yet, and Freidman hints at this in his Cato talk,
is to *flip* that code.

Frankly, this is nothing really now. It's done identically the same way that
large institutions flip IP0's in the market for a hot stock, or, more
properly, a recording company or software "publisher" puts "talent" under
contract and remarkets their output in classical vertically-integrated, and I
say, now-industrial, markets for those goods.


So, back to the future, I pay through the nose to get to the first copy of
your code, because I have an orderbook of my *own* to fill, and, more
properly, I have sufficient bandwidth and processing power of my own to
remarket it. Even more fun, someone *else* can buy from me, and resell what I
sell them, *recursively*, ad infinitum, and, frankly I don't really *care*
what they do with what I sold them, because I've already bought low and sold
high.

I claim, even with all this rampant "piracy", and, especially as the network
gets more geodesic -- a process which these markets would probably
*accellerate* by the way think what would happen if you could sell bandwidth
this way -- the person who collects the most economic "rent" would be *you*,
the originator of the new code. Just like a rock star, if the code was popular
enough, but without the recording company. Well, actually, with lots of very
*little* record *distributors*, all taking a haircut, and, rarely would any of
them make anything near like what you made selling the first copies.

I mean, new information always costs more than old information, right? And, of
course, you, as a developer, will always have the *newest* information of
all...


Finally, it even behoves huge companies who control large blocks of
intellectual property, record companies, and Microsoft, for instance, :-), to
compete in this new kind of market, precisely *because* the the transaction
costs are so low and the initial profit margins are correspondingly higher.

It's *profitable* to play in this market, even when you're losing market
share, paradoxically, to smaller players. Think of what happened to large
movie studios in the 1960's and '70's, and how they were replaced, in the
*production* of movies, by director-owned, mostly ad-hoc, "virtual" production
companies, like Coppola's Zoetrope, or Speilburg's Amblin, or Lucasfilm/ILM.
Even the new "studios", like Dreamworks/SKG are really director-run
enterprizes (Spielburg and Geffen, remember?).

The "directors", the "auteurs" of software, are pretty much the guys who write
the code, yes?

Everything else is just transaction cost.


Cheers,
Robert Hettinga


-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.5.5

iQEVAwUBNn7OucUCGwxmWcHhAQFidwf+MZ8SLZffWvH7g+QMY6Q2Q3P4flVkEVpr
keXiF7cXXKlv/zL48wX1fjan93MMwN65b9/o7aIe1Wvq1nQfFE1AhPaShYZwciz9
NpInA2+9Mrz30cU4Nj775gXHfCaFl68tUmKd8GwUBI4rwjrHKgFy5hFSWlxNQSFF
yQ8idMnmC7fSIELZJOurejYM90nA4z9QZZibekVuP+u3e6ogkclCS8m3e3QznpWY
yCTYHqFFc5JHHEHc4PCOALWei74/rTRVvaziEU3QQCvytXgsZu2j0OLkpukc4zqb
eJD/+Zkc3d1pfTqxaR4tLatr+QXVEuxozwz3nyoo3mzfd0LbKdBdnA==
=kKXa
-----END PGP SIGNATURE-----
-----------------
Robert A. Hettinga <mailto: rah@philodox.com>
Philodox Financial Technology Evangelism <http://www.philodox.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'