Cheaper than printing it out: buy the paperback book.

Out of Control
Chapter 12: E-MONEY

We manage the disconnection of domestic utilities, such as water or electricity, through metering. But metering is neither obvious nor easy. Thomas Edison's dazzling electrical gizmos were of little use to anyone until people had easy access to electricity in their factories and homes. So at the peak of his career Edison diverted his attention away from designing electrical devices to focus on the electrical delivery network itself. At first, very little was settled about how electricity should be created (DC or AC?), carried, or billed. For billing, Edison favored the approach that most information providers today favor: charge a flat fee. Readers pay the same for a newspaper no matter how much of it they read. Ditto for cable TV, books and computer software. All are priced flat for all you can use.

Edison pushed a flat fee for electricity -- a fixed amount if you are connected, nothing if you aren't -- because he felt that the costs of accounting for differential usage would exceed the cost of variances in electricity usage. But mostly Edison was stymied about how to meter electricity. For the first six months of his General Electric Lighting Company in New York City, customers paid a flat fee. To Edison's chagrin, that didn't work out economically. Edison was forced to come up with a stop-gap solution. His remedy, an electrolytic meter, was erratic and impractical. It froze in winter, it sometimes ran backwards, and customers couldn't read it (nor did they trust the company's meter readers). It wasn't until a decade after municipal electrical networks were up and running that another inventor came up with a reliable watt-hour meter. Now we can hardly imagine buying electricity any other way.

A hundred years later the information industry still lacks an information meter. George Gilder, hi-tech gadfly, puts the problem this way: "Rather than having to pay for the whole reservoir every time you are thirsty, what you want is to only pay for a glass of water."

Indeed, why buy an ocean of information when all you want is a drink? No reason at all, if you have an information meter. Entrepreneur Peter Sprague believes he has just invented one. "We use encryption to force the metering of information," says Sprague. His spigot is a microchip that doles out small bits of information from a huge pile of encrypted data. Instead of selling a CD-ROM crammed with a hundred thousand pages of legal documents for $2,000, Sprague invented a ciphering device that would dispense the documents off the CD-ROM at $1 per page. A user only pays for what she uses and can use only what she pays for.

Sprague's way of selling information per page is to make each page unreadable until decrypted. Working from a catalog of contents, a user selects a range of information to browse. She reads the abstracts or summaries and is charged a minuscule amount. Then she selects a full text, which is decrypted by her dispenser. Each act of decryption rings up a small charge (maybe 50 cents). The charge is tallied by a metering chip in her dispenser that deducts the amount from a prepaid account (also stored on the metering chip), much as a postage meter deducts credit while dispensing postage tapes. When the CD-ROM credit runs out, she calls a central office, which replenishes her account via an encrypted message sent on a modem line running into her computer's metering chip. Her dispenser now has $300 credit to spend on information by the page, by the paragraph, or by the stock price, depending on how fine the vendor is cutting it.

What Sprague's encryption metering device does is decouple information's fabulous ease in being copied from its owner's need to have it selectively disconnected. It lets information flow freely and ubiquitously -- like water through a town's plumbing -- by metering it out in usable chunks. Metering converts information into a utility.

The cypherpunks note, quite correctly, that this will not stop hackers from siphoning off free information. The Videocipher encryption system, used to meter satellite-delivered TV programs such as HBO and Showtime, was compromised within weeks of its introduction. Despite claims by the meter's manufacturer that the encrypto-metering chip was unhackable, big moneymaking scams capitalized on hacks around the codes. (The scams were set up on Indian reservations -- but that's a whole 'nother story). Pirates would find a descrambler box with a valid subscription -- in a hotel room, for instance -- and then clone the identity into other chips. A consumer would send their box to the reservation for "repairs" and it would come back with a new chip cloned with the identity of the hotel box. The broadcasting system couldn't perceive clones in the audience. In short, the system was hacked not by cracking the code but by subverting places where the code tied into the other parts of the system.

No system is hack-proof. But disruptions of an encrypted system require deliberate creative energy. Information meters can't stop thievery or hacking, but meters can counteract the effects of lazy mooching and the natural human desire to share. The Videocipher satellite TV system eliminates user piracy on a mass scale -- the type of piracy that plagued the satellite TV outback before scrambling and that still plagues the lands of software and photocopying. Encryption makes pirating a chore and not something that any slouch with a blank disk can do. Satellite encryption works overall because encryption always wins.

Peter Sprague's crypto-meter permits Alice to make as many copies of the encrypted CD-ROMs as she likes, since she pays for only what she uses. Crypto-metering, in essence, disengages the process of payment from the process of duplication.

Using encryption to force the metering of information works because it does not constrain information's desire to reproduce. All things being equal, a bit of information will replicate through an available network until it fills that network. With an animate drive, every fact naturally proliferates as many times as possible. The more fit -- the more interesting or useful -- a fact is, the wider it spreads. A pretty metaphor compares the spread of genes through a population with the similar spread of ideas, or memes, in a population. Both genes and memes depend on a network of replicating machines -- cells or brains or computer terminals. A network in this general sense is a swarm of flexibly interconnected nodes each of which can copy (either exactly or with variation) a message taken from another node. A population of butterflies and a flurry of e-mail messages have the same mandate: replicate or die. Information wants to be copied.

Our digital society has built a supernetwork of copiers out of hundreds of millions of personal faxes, library photocopiers, and desktop hard disks. It is as if our information society is one huge aggregate copying machine. But we won't let this supermachine copy. Much to everyone's surprise, information created in one corner finds its way into all the other corners rather quickly. Because our previous economy was built upon scarcity of goods, we have so far fought the natural fecundity of information by trying to control every act of replication as it occurs. We take a massively parallel copy machine and try to stifle most acts of reproduction. As in other puritanical regimes, this doesn't work. Information wants to be copied.

"Free the bits!" shouts Tim May. This sense of the word "free" shifts Stewart Brand's oft-quoted maxim, "Information wants to be free" -- as in "without cost" -- to the more subtle "without chains or imprisonment." Information wants to be free to wander and reproduce. Success, in a networked world of decentralized nodes, belongs to those plans that do not resist either the replication or roaming urges of information.

Sprague's encrypted meter capitalizes on the distinction between pay and copy. "It is easy to make software count how many times it has been invoked, but hard to make it count how many times it has been copied," says software architect Brad Cox. In a message broadcast on the Internet, Cox writes:

Software objects differ from tangible objects in being fundamentally unable to monitor their copying but trivially able to monitor their use....So why not build an information age market economy around this difference between manufacturing-age and information-age goods? If revenue collection were based on monitoring the use of software inside a computer, vendors could dispense with copy protection altogether.

Cox is a software developer specializing in object-oriented programming. In addition to the previously mentioned virtue of reduced bugs which OOP delivers, it offers two other magnificent improvements over conventional software. First, OOP provides the user with applications that are more fluid, more interoperable with various tasks -- sort of like a house with movable "object" furniture instead of house saddled with built-in furniture. Second, OOP provides software developers the ability to "reuse" modules of software, whether they wrote the modules themselves or purchased them from someone else. To build a database, an OOP designer like Cox takes a sort routine, a field manager, a form generator, an icon handler, etc., and assembles the program instead of rewriting a working whole from scratch. Cox developed a set of cool OOP objects that he sold to Steve Jobs to use in his Next machine, but selling small bits of modular code as a regular business has been slow. It is similar to trying to peddle limericks one by one. To recoup the great cost of writing an individual object by selling it outright would garner too few sales, but selling it by copy is too hard to monitor or control. But if objects could generate revenue each time a user activated one, then an author could make a living creating them.

While contemplating the possible market for OOP objects that were sold on a "per use" plan, Cox uncovered the natural grain in networked intelligence: Let the copies flow, and pay per use. He says, "The premise is that copy protection is exactly the wrong idea for intangible, easily copied goods such as software. You want information-age goods to be freely distributed and freely acquired via whatever distribution means you want. You are positively encouraged to download software from networks, give copies to your friends, or send it as junk mail to people you've never met. Broadcast my software from satellites. Please!"

Cox adds (in echo of Peter Sprague, although surprisingly the two are unfamiliar with each other's work), "This generosity is possible because the software is actually 'meterware.' It has strings attached that make revenue collection independent of how the software was distributed."

"The approach is called superdistribution," Cox says, using a term given by Japanese researchers to a similar method they devised to track the flow of software through a network. Cox: "Like superconductivity, it lets information flow freely, without resistance from copy protection or piracy."

The model is the successful balance of copyright and use rights worked out by the music and radio industries. Musicians earn money not only by selling customers a copy of their work but by selling broadcast stations a "use" of their music. The copies are supplied free, sent to radio stations in a great unmonitored flood by the musicians' agents. The stations sort through this tide of free music, paying royalties only for the music they broadcast, as metered (statistically) by two agencies representing musicians, ASCAP and BMI.

JEIDA, a Japanese consortium of computer manufacturers, developed a chip and a protocol that allows each Macintosh on a network to freely replicate software while metering use rights. According to Ryoichi Mori, the head of JEIDA, "Each computer is thought of as a station that broadcasts, not the software itself, but the use of the software, to an audience of a single 'listener.'" Each time your Mac "plays" a piece of software or a software component from among thousands freely available, it triggers a royalty. Commercial radio and TV provide an "existence proof" of a working superdistribution system in which the copies are disseminated free and the stations only pay for what they use. Musicians would be quite happy if one radio station made copies of their tapes and distributed them to other stations ("Free the bits!") because it increases the likelihood of some station using their music.

JEIDA envisions software percolating through large computer networks unencumbered by restrictions on copying or mobility. Like Cox, Sprague, and the cypherpunks, JEIDA counts on public-key encryption to keep these counts private and untampered as they are transmitted to the credit center. Peter Sprague says plainly, "Encrypted metering is an ASCAP for intellectual property."

Cox's electronically disseminated pamphlet on superdistribution sums up the virtues very nicely:

Whereas software's ease of replication is a liability today, superdistribution makes it an asset. Whereas software vendors must spend heavily to overcome software's invisibility, superdistribution thrusts software out into the world to serve as its own advertisement.

A hoary ogre known as the Pay-Per-View Problem haunts the information economy. In the past this monster ate billions of dollars in failed corporate attempts to sell movies, databases, or music recordings on a per view or per use basis. The ogre still lives. The problem is, people are reluctant to pay in advance for information they haven't seen because of their hunch that they might not find it useful. They are equally unwilling to pay after they have seen it because their hunch usually proves correct: they could have lived without it. Can you imagine being asked to pay after you've seen a movie? Medical knowledge is the only type of information that can be easily sold sight unseen because the buyers believe they can't live without it.

The ogre is usually slain with sampling. Moviegoers are persuaded to pay beforehand by lapel-grabbing trailers. Software is loaned among friends for trial; books and magazines are browsed in the bookstore.

The other way to slay the problem is by lowering the price of admission. Newspapers are cheap; we pay before looking. The ingenious thing about information metering is that it delivers two solutions: it provides a spigot to record how much data is used, and it provides a spigot that can be turned down to a cheap trickle. Encryption-metering chops big expensive data hunks into small inexpensive doses of data. People will readily pay for bits of cheap information before viewing, particularly if the payment invisibly deducts itself from an account.

The fine granularity of information-metering gets Peter Sprague excited. When asked for an example of how fine it could get, he volunteers one so fast it's obvious that he has been giving it some thought: "Say you want to write obscene limericks from your house in Telluride, Colorado. If you could write one obscene limerick a day, we can probably find 10,000 people in the world who want to pay 10 cents a day to get it. We'll collect $365,000 per year and pay you $120,000, and then you can ski for the rest of your life." In no other kind of marketplace would one measly limerick, no matter how bawdy and clever, be worth selling on its own. Maybe a book of them -- an ocean of limericks -- but not one. Yet in an electronic marketplace, a single limerick -- the information equivalent of a stick of gum -- is worth producing and offering for sale.

Sprague ticks off a list of other fine-grained items that might be traded in such a marketplace. He catalogs what he'd pay for right now: "I want the weather in Prague for 25 cents per month, I want my stocks updated for 50 cents a stock, I want the Dines Letter for $12 a week, I want the congestion report from O'Hare Airport updated continuously because I'm always getting stuck in Chicago, so I'll pay a buck per month for that, and I want 'Hagar the Horrible' cartoon for a nickel a day." Each of these products is currently either given away scattershot or peddled in the aggregate very expensively. Sprague's electronically mediated marketplace would "unbundle" the data and deliver a narrowly selected piece of information to your desktop or mobile palmtop for a reasonable price. Encryption would meter it out, preventing you from filching other tiny bits of data that would hardly be worth protecting (or selling) in other ways. In essence, the ocean of information flows through you, but you only pay for what you drink.

At the moment, this particular technology of disconnection exists as a $95 circuit board that can slide into a personal computer and plug into a phone line. To encourage established computer manufacturers such as Hewlett-Packard to hardwire a similar board into units coming off their assembly line, Sprague's company, Waves, Inc., offers manufacturers a percentage of the revenue the encryption system generates. Their first market is lawyers, "because," he says, "lawyers spend $400 a month on information searches." Sprague's next step is to compress the encrypto-metering circuits and the modem down into a single $20 microchip that can be tucked into beepers, video recorders, phones, radios, and anything else that dispenses information. Ordinarily, this vision might be dismissed as the pipe dream of a starry-eyed junior inventor, but Peter Sprague is chairman and founder of National Semiconductor, one of the major semiconducter manufacturers in the world. He is sort of a Henry Ford of silicon chips. A cypherpunk, not. If anyone knows how to squeeze a revolutionary economy onto the head of a pin, it might be him.