Photograph of an oldtimey newstand
Photograph: Archive Photos

Intellectual Value

A radical new way of looking at compensation for owners and creators in the Net-based economy.

What happens to intellectual property when it gets on the Internet?

The Net dramatically changes the economics of content. Because it allows us to copy content essentially for free, the Net poses interesting challenges for owners, creators, sellers, and users of intellectual property. In this new world of the Net, it is easy to copy information but hard to find it. It is easy to program software to solve problems but hard to define those problems and questions precisely.

In the new communities of the Net, the intrinsic value of content generally will remain high, but most individual items will have a short commercial half-life. Creators will have to fight to attract attention and get paid. Creativity will proliferate, but quality will be scarce and hard to recognize. The problem for providers of intellectual property in the future is this: although under law they will be able to control the pricing of their own products, they will operate in an increasingly competitive marketplace where much of the intellectual property is distributed free and suppliers explode in number.

What will almost-free software and proliferating content do to commercial markets for content? How will people—writers, programmers, and artists—be compensated for creating value? What business models will succeed in this foreign economy?

In a new environment, such as the gravity field of the moon, laws of physics play out differently. On the Net, there is an equivalent change in "gravity" brought about by the ease of information transfer. We are entering a new economic environment—as different as the moon is from the earth—where a new set of physical rules will govern what intellectual property means, how opportunities are created from it, who prospers, and who loses.

Chief among the new rules is that "content is free." While not all content will be free, the new economic dynamic will operate as if it were. In the world of the Net, content (including software) will serve as advertising for services such as support, aggregation, filtering, assembly and integration of content modules, or training of customers in their use. Intellectual property that can be copied easily likely will be copied. It will be copied so easily and efficiently that much of it will be distributed free in order to attract attention or create desire for follow-up services that can be charged for. Advertising has a poor reputation in many quarters because most advertising is designed for a broad market. But in the one-to-one world the Net promises, advertising will often be tailored and of higher quality. Those with more money to spend will get higher-quality advertising.

What should content makers do in such an inverted world? The likely best course for content providers is to exploit that situation, to distribute intellectual property free in order to sell services and relationships. The provider's vital task is to figure out what to charge for and what to give away—all in the context of what other providers are doing and what customers (will grow to) expect.

Of course, there still will be ways for content creators to be paid. Much content will be developed under service contracts. A supplier will create high-value content, such as a market research study directly for a paying customer (or a limited set of customers). Newspapers and online news services will pay reporters and editors to produce content, which will then be resold cheaply in conjunction with advertising that covers most costs; that same content may also be distributed "free" as part of a subscription service. Certainly, advertisers will continue to pay people to develop advertising content for them, even if that content is to be distributed free.

I am not saying that content is worthless, or that you will always get it for free. Content providers should manage their businesses as if it were free, and then figure out how to set up relationships or develop ancillary products and services that cover the costs of developing content. Or players may simply try their hands at creative endeavors based on service, not content assets: filtering content, hosting online forums, rating others' (free) content, custom programming, consulting, or performing. The creator who writes off the costs of developing content immediately—as if it were valueless—is always going to win over the creator who can't figure out how to cover those costs. The way to become a leading content provider may be to start by giving your content away. This "generosity" isn't a moral decision: it's a business strategy.

The Half-Life of Value

Imagine you're a farmer in the 19th century headed into the 20th. The intrinsic value of food won't go away in the new century, but as food becomes cheaper and cheaper to produce, the share of economy devoted to agriculture will shrink, and so will your margins. Better to get into manufacturing, or at least into food processing. (But fast-food restaurants. That may be a little premature.)

Now imagine you are a content maker in the 20th century headed into the 21st. Until now, content has always been manifested physically—first in people who knew how to do things; then in books, sheet music, records, newspapers, loose-leaf binders, and catalogs; and most recently in tapes, discs, and other electronic media. At first, information could not be "copied": it could only be reimplemented or transferred. People could build new machines or devices that were copies of or improvements on the original; people could tell each other things and share wisdom or techniques to act upon. (Reimplementation was cumbersome and re-use did not take away from the original, but the process of building a new implementation—a new machine or a trained apprentice—took considerable time and physical resources.) Later, with symbols, paper, and printing presses, people could copy knowledge, and it could be distributed in "fixed" media; performances could be transcribed and re-created from musical scores or scripts. Machines could be mass-produced.

With such mechanical and electronic media, intellectual value could easily be reproduced—and the need (or demand from creators) to protect intellectual property arose. New laws enabled owners and creators to control the production and distribution of copies of their works. Although reproduction was easy, it was still mostly a manufacturing process, not something an individual could do easily. It took time and money. Physical implementation contributed a substantial portion of the cost.

Now we face a new situation: not only is it easy for individuals to make duplicates of many works or to re-use their content in new works, but the physical manifestation of content is almost irrelevant. Over the Net, any piece of electronically represented intellectual property can be almost instantly instantiated anywhere in the world.

Controlling copies (once created by the author or by a third party) becomes a complex challenge. You can either control something very tightly, limiting distribution to a small, trusted group, or you can rest assured that eventually your product will find its way to a large nonpaying audience—if anyone cares to have it in the first place.

But creators of content on the Net still face the eternal problem: the value of their work generally won't receive recognition without wide distribution. Only by attracting broad attention can an artist or creator hope to attract high payment for copies. Thus, on the Net, the creators give first performances or books (or whatever) away widely in hopes of recouping with subsequent works. But that breadth of distribution lessens the creator's control—of who gets copies and what they do with them. In principle, it should be possible to control and charge for such widely disseminated works, but it will become more and more difficult. People want to pay only for what is perceived as scarce—a personal performance or a custom application, or some tangible manifestation that can't easily be reproduced (by nature or by fiat; that's why the art world has numbered lithographs, for example).

The trick is to control not the copies of your work but instead a relationship with the customers—subscriptions or membership. And that's often what the customers want, because they see it as an assurance of a continuing supply of reliable, timely content.

You can, of course, charge a small amount for mass copies. Metering schemes will allow vendors to charge—in fractions of a penny if desired—according to usage or users rather than copies. Yet much as I find the approach of metering and tagging information intellectually appealing—and I especially like Mark Stefik's “rights language”—it won't much change the overall approaching-zero trend of content pricing. At best, it will make it much easier to charge those low, low prices.

Seen One, Seen Them All

There are other hurdles on the Net for content creators. One is the rise of a truly efficient market for information. Content used to be unfungible: it was difficult to replace one item with another. But most information is not unique, though its creators like to believe so; there are now "specs" for content such as stock prices, search criteria, movie ratings, and classifications.

In the world of software, for instance, it's becoming easier to define and create products equivalent to a standard. Unknown vendors who can guarantee functionality will squeeze the prices of the market leaders. Of course the leaders (such as Microsoft) will continue to win because they can use almost-free content to sell ancillary products or upgrades, and because they've reinvested in loyal distribution channels (even though they don't own them). In a sense, the content is advertising for the dealers who resell as well as for the vendors who create.

Overall, it will become easier either to reimplement the use of a software product—or, more significantly, to implement the solution to a particular problem in an alternative way. The definition of the problem, rather than its solution, will be the scarce resource in the future.

In entertainment and art, pricing will drop likewise, as more creators compete for attention with content they make using low-cost, easy-to-use production tools. Rather than hitting the big time, more artists will find their audiences within their local communities—geographical or Net-based. Local barriers to entry will be low, but global competition will be strong. There's the odd movie star or work of art for which no substitute is acceptable, but most entertainment is a way of spending time—not a unique experience. As Mark Stahlman of NewMedia Associates Inc. points out, almost every variety of recreation—from reading a book to going out on a date—converges on the same amount when figured in dollars per hour of experience (currently between US$1 and $2 per hour). People unwittingly value entertainment content by the hour as if it were all an interchangeable commodity.

Paradise Lost and Regained

The idea that intellectual property on the Net can lose its value horrifies most of its owners and creators, but it's not new. It's happening already in the software business. Most software products are becoming commodities, not because they are easy to duplicate precisely (which is illegal), but because they are easy to imitate. Customers tend to want the original product: that forces prices down as knock-offs attempt to gain market share and the original attempts to maintain it by lowering prices.

Overall, in each market, there are likely to be a few leaders who create and protect content with a strong identity (perhaps promoting it with free content). Other players in each market will have a difficult time selling content as assets and will have to find new ways to collect rewards for their creativity. Owning the intellectual property is like owning land: you need to keep investing in it again and again to get a payoff; you can't simply sit back and collect rent. To some, this state of affairs may seem unfair. It certainly is if you grew up by the old rules and don't want to play in a new game. But if you look at the new rules by themselves, they have a certain moral grounding: people will be rewarded for personal effort—process and services—rather than for mere ownership of assets.

Owning land gives you the right to charge for the value you put into it. But the business of real estate is increasingly concerned with location and ancillary conditions such as zoning rights and obligations. Yes, land ownership matters, but it's not the most interesting factor in real estate today. The same is true in the Net world: content ownership matters, but it's hardly the key factor in intellectual commerce.

Photograph: Upi/Bettman
The Curve of Intellectual Value

The value of content follows strange curves. When only a few people are interested enough—to attend the party, watch the TV show, shop at the mall—it's generally not very valuable. When or if an item becomes a standard, it gains in value, both in the aggregate and per user. Users can cooperate with each other and share data: Juan and Alice can share the experience of watching John Malkovich in In the Line of Fire; the price of gold goes up as more people believe the price of gold will go up and start buying it in hopes of getting rich; the party's more fun when all your friends are there. But at some point that value peaks: the party's too crowded; the star is passé.

By contrast, other kinds of value are greater when they're exclusive: you're the only one in the market, you have a special formula for toothpaste, or you know a particular place to find oil. But many kinds of information flow from one bucket to the other; the first guy gets the advantage, but after that, the value is maximized by spreading the knowledge widely or by postprocessing—adding synergistic value on top of it.

Some aspects of intellectual property can be fixed into a specific medium and copied. But other less fungible or reproducible aspects of content cannot easily be instantiated or transferred. Their worth is realized only through human attention and interaction. Let's call it intellectual value rather than intellectual property. Intellectual value comprises content such as performances; teaching, training, and coaching; analysis of specific questions applied to specific situations, and personal attention—someone reading and responding to your e-mail, answering questions, or watching you on a video connection.

Intellectual value (some call it context) is also simply the presence of other people, often specific ones, interacting casually or formally. This value is the difference between sitting in a packed stadium and watching a pickup ball game with a couple of friends. It's the difference between the $500-a-plate dinner with 1,000 people in a hotel ballroom and the $10,000-a-head 40-person reception beforehand in the presidential suite. It's the difference between an off-the-rack special and an original from Dior—or the "free" gown your mother wore at her wedding 50 years ago. Value and uniqueness interact in mysterious ways.

The intellectual value of context can't be replicated so easily over the Net. Unsurprisingly, it depends on the activity or presence of a person—locally or remotely, in real time or at least in individual response. Intellectual property is the embodiment or automation of effort, replicable easily for all. Intellectual value, on the other hand, is the effort, service, or process itself; it can sometimes be shared, but the effort can't be replicated without another person around to do the same task.

Precisely because it is scarce and unreplicable, this unreplicable kind of content is likely to command the highest rewards in the commercial world of the future.

Approaching Zero

These trends are already beginning to play out. For example, while most packaged software vendors continue to fight the perennial battle against software piracy, others have chosen to begin adopting a different business model. Except for the leader in each field (Microsoft Corp., Oracle Corp., Autodesk Inc., and a select other few), not many software companies can survive on the sale of intellectual property alone. The price of most packaged software tends toward (although it rarely reaches) zero.

So, what happens in a world where software is basically free? Successful companies are adopting business models in which they are rewarded for services rather than for code. Developers who create software are rewarded for showing users how to use it, for installing systems, for developing customer-specific applications. The real value created by most software companies lies in their distribution networks, trained user bases, and brand names—not in their code.

What Novell Inc. really is selling is its investment in training certified NetWare engineers, instructors, and administrators, and its perceived ability to produce and sell widely the next release of NetWare. (Selling the next release is essentially selling vaporware, but note that vaporware is not just a spurious form of advertising. Well-managed vaporware enables developers and customers to plan effectively; it is part of the value a software vendor can provide. But it works in the long run only if it is legitimate.)

Packaged software is a property, but in many ways it is becoming simply an advertisement for follow-up goods and services—bug-fixing, support, upgrades, training, implementation, and development services. The price of the software covers production and distribution; the intellectual content is free.

Consider Cygnus Support, a profitable five-year-old company (on whose board I sit); had 1994 revenues of more than $5 million and whose customers include Cisco Systems, Hitachi America Ltd., Motorola Inc., and Sun Microsystems Inc. Cygnus successfully sells support and implementation services, along with free copies of software from the Free Software Foundation (GNU C compilers, Unix, tools, et cetera).

Much of its work is porting free system software to a particular vendor's hardware implementation, a task paid for by the hardware vendor, or by the customer. The resulting implementation is freely redistributable (as provided by the rules under which the free software is distributed). Of course, the hardware vendor gets the benefit of its existence, since it makes his hardware products more attractive and "advertises" them. A single company, or individual, may create the intellectual property, while others may provide support services.

A similar situation often crops up in publishing. "Give us your information," a publisher says to an author, "and we'll classify it, put it online, and send you a royalty depending on how many people access it or how long they spend with it." Yes, the content creator is selling intellectual property, but unless this creator has a very strong identity, the balance of power will shift to the publisher, who controls the channel to the customer.

The Bifurcation of Content

Would you pay more for Michael Crichton's words, or for the ability to suggest a new plot to him or name a hero? How much was his book worth to the movie Disclosure? How much was his name worth?

In the world of media and entertainment, there is not very much of a support or training market per se. (Classes in playing Fig Mutant Space Rangers, anyone?) The payments to creators are most likely to come not from the viewers, readers, or listeners, but from advertisers.

The challenge for advertisers is to make sure that their advertising messages are inextricable from the content. In some cases, it will mean that advertising will be as good as editorial content: truly funny commercials, informative, truthful product information, and so forth. Advertisers will pay for the services of the creators, if not necessarily for the content. As is happening now, famous people will contract to promote products or companies. In essence, advertisers will sponsor lives, or online forum hosts, rather than content. Just as prominent patrons such as the Medicis sponsored artists during the Renaissance, corporations and the odd rich person will sponsor artists and entertainers in the new era. The Medicis presumably had the pleasure of seeing or listening to their beneficiaries and sharing access to them with their friends. This won them renown and attention as well as a certain amount (we hope) of sheer pleasure at experiencing the art. Of course, they could also hike over to the studio at any time to watch their artist at work (and perhaps find him at play?).

Overall, entertainment will become cheaper and cheaper, since so much of it will be sponsored. Rights to transmit or display particular creative works will exist, but they will usually be purchased by third parties rather than consumers—for redistribution to consumers. However, those rights will be hard to protect or exploit thereafter. Much information and entertainment, once paid for, will simply be disseminated free, as consumers won't pay more than a little, and competing suppliers will bid prices down. The advertiser will gain from being the first to deliver the content, and possibly from redistribution of its ads along with the content and identification with it. Product placement within content will be much more effective than placing advertising alongside content where it can more easily be removed or completely ignored by customers.

With the means of production growing cheaper and easier because of the Net, a bifurcation will take place: more and more people will produce material for smaller audiences of their friends, while those seeking large audiences will give their stuff away or seek payment from a sponsor—and try to persuade influencers to recommend it.

In the end, the only unfungible, unreplicable value in the new economy will be people's presence, time, and attention; to sell that presence, time, and attention out-side their own community, creators will have to give away content for free. As John Perry Barlow points out in "The Economy of Ideas" (see Wired 2.03, page 84), that's exactly what the Grateful Dead do by encouraging people to tape their performances (and a performance is not just the Grateful Dead on the stage; it's all the people there with you). Enough of the people who copy and listen to Grateful Dead tapes end up paying for hats, T-shirts, and performance tickets. In the new era, the ancillary market is the market.

The Vacuum of Attention

The popular notion about cyberspace is that it is infinite and unbounded. But it, too, is limited by the amount of human attention available in it. Does a place in cyberspace exist if no one visits it?

What makes any kind of real estate valuable? Just ask a restaurateur! It's not mere buildings. There's a complex interaction among tenants and visitors, the physical plant and the services, as well as the location. Real estate in cyberspace works similarly.

The initial appeal of real estate may be proximity to other space—it's easy to find on your way somewhere else. The Net equivalent (more or less) is a listing in someone's guide, for example, a pointer in a Web page, or highlighted availability through a service such as CompuServe or Poland Online.

The virtual space near any particular location is limited—just like retail space along Fifth Avenue in New York, Bond Street in London, or Nevsky Prospekt in St. Petersburg. There are just so many services to highlight or point to a particular location.

But you can also build your land up so that more people want to pass through it. You can add restaurants or tony shops. On the Net, content can draw people in. It can attract a certain crowd, tune out others (there's a story about mall owners who played classical music to drive the teenagers away), and provoke all sorts of interesting interaction among the visitors. Just as shopping malls offer rides, and cafes are not only places to buy lattés but places where people can meet, so will cyberspace real estate provide environments for engaging social interaction.

People draw more people: they exist as content that no one owns, but content that is sold to other members of the market. Live performances and the presence of others attract still more visitors or customers. Free copies of performances and videos can serve as advertising for live performances, genuine two-way interaction with the performers, or membership in the community where these things happen.

Content and people (like goods) that get visibility in favorable locations gain in popularity (to some limit), and can thereafter be used in other locations to raise value elsewhere (within limits). This is all a delicate game, appropriately reminiscent of the intricate arrangements of broadcast programming and counter-programming.

Photograph: FPG International
Return of the Middlemen

Aside from a few leaders who manage to sell brand-name content widely and cheaply, the most promising businesses in the Net world will be services and processes. They will include selecting, classifying, rating, interpreting, and customizing content for specific customer needs. Other services will include access to various sorts of performing, interacting with people, and all kinds of other activities that require the time of a live, talented person. Even those who sell content will invest in distribution channels, whether or not they own them; the value added in those channels will be tied to (and will enhance) the value of the content they sell.

Much chargeable value will be in certification of authenticity and reliability, not in the content. Brand name, identity, and other marks of value will be important; so will security of supply. Customers will pay for a stream of information and content from a trusted source. For example, the umbrella of The New York Times sanctifies the words of its reporters. The content churned out by Times reporters is valuable because the reporters undergo quality-control, and because others believe them—context, again. The New York Times can almost make the truth—for better or worse.

Contrary to the notion that the Net will be a disintermediated world, much of the payment that ostensibly goes for content will go to the middlemen and trusted intermediaries who add value—everything from guarantees of authenticity to software support, selection, filtering, interpretation, and analysis. The redistributor's goal is to be the most convenient source of content and to put its own attitude or personality around the content; the underlying content is unlikely to be exclusive, since the content provider wants to maximize its distribution (either for revenues—however small per item—or for advertising breadth).

The problem for owners of content is that they will be competing with free or almost-free content, including their own advertising as well as the output of myriad creators who launch products on the Net.

Creators worry that they won't be paid, and that their creative efforts may be discouraged. The free-content market will discourage redundant effort, since the wheel won't need to be reinvented. The almost-free content market might also discourage a lot of crummy content and marketing designed to draw attention to such content. Why market a book that's free? It should sell by itself (drawing attention to the author) or not at all. There's no external reason to sell it, so poor novels won't be foisted on the public, and good ones may find their audience by themselves—or through the efforts of filter agents who get rewarded for finding (not creating) good content. The novelist, then, will be rewarded by fees for his or her performances, or perhaps by finding sponsors for future work. He or she may write serials and find people who are willing to pay for this service.

The final result for creators in this new world is that intellectual value markets will bifurcate into content assets of premium prices and high value, and services and processes built around free or cheap content. The middle disappears.

To many people, such a world is frightening, since it does not require any laws to change or be broken. It's simply the unfolding expression of economic laws—of demand and scarcity—applied in the future world of electronic content and commerce. It's not the world most creators and intellectual property owners have been planning for, contracting for, securing rights for.

Of course, this new world will distribute its benefits differently than how they are distributed today. But as long as the rules are the same for everyone—and predictable—the game is fair. The big issue is the transition.

And there will be rules: copying content will be easy and acceptable in most cases; protected content will be "special," presumed to be of high value. (And it will have to be if the creators want to keep their reputations.) Protected content will be tagged and monitored (for one method, see "Digital Watermarks," page 141), and use will be metered. Some payments will be for content, some for time, some for transactions. This system of control will be managed by efficient, well-designed computer systems—a delightful intellectual engineering challenge that will keep many programmers and companies busy for years, and that will reward some venture capitalists.

There will also be strong legal and social pressures for authenticity, integrity, trademarks, and identification. Most often, you will be able to copy something freely, but you can't claim for it an identity or origin it doesn't have. Discrete works must be attributed; derivative works (assembled or modified from identified original components) will have to find their own value.

A Market of Visibility

All these emergent economic pressures and social conventions will spill over into the framework of organizations in this new world. What's a company's greatest asset, according to many annual reports? People and partners, of course, not computers or even intellectual property. In the new world of the Net, much will change. Everyone knows about flatter organizations and virtual corporations. But how do they play out in texture as opposed to structure?

The short answer is that companies will—must—become more visible. More of what any company sells will comprise information—whether it's plain bits over the Net or consulting services, design services, management development. As in the past, some companies will sell products to myriad customers; others will add value to only a few key ones. But in a knowledge world, the quality of those relationships will matter more than the contractual conditions (as in a marriage). The best cement is a two-way flow of information, or visibility. Companies will try to find partners not by offering discounts but by sharing information about themselves and by exchanging their competitive wisdom. In order to make their wisdom credible, they will have to be self-revealing.

Moreover, whether or not a company chooses to be visible, it will happen. You can't hide. And the image you project—on your Web home page or elsewhere—will and should be true. It's not just outsiders peering in, it's your own employees out in the electronic world: they are the company. As both physical and intellectual products lose their value (in the manner described above), the interactions with your company will be what you sell. And the quality of the interactions you foster will be what draws employees to your firm or community. People want to buy information-based services and products from visible companies that operate as partners. They do not want commodity products from black boxes.

The question of what happens to intellectual property on the Net may be summed up like this: value shifts from the transformation of bits rather than bits themselves, to services, to the selection of content, to the presence of other people, and to the assurance of authenticity—reliable information about sources of bits and their future flows. In short, intellectual assets and property depreciate while intellectual processes and services appreciate.

Software Metering

"We're turning the infobahn into a toll road," claims Peter Sprague, the avuncular CEO of Wave Systems Inc. It may sound like just one more overwrought metaphor, but it precisely describes the future Sprague and competing Infosafe Systems Inc. are racing to build. Both New York-based companies have recently announced data metering systems—hardware devices that attach to your computer and act as toll collectors, charging you for the data you download and the programs you use. Infosafe and Wave are betting that such a system will fundamentally transform information economics by making unauthorized copying more difficult while taking advantage of information's natural tendency to spread far and wide.

The concept is simple: instead of charging a flat fee for a software program, or an hourly fee for access to a database, data metering allows companies to charge per use. So, if you need to use a CAD program only a couple times a year, you could pay just a few dollars each time rather than hundreds of dollars to buy the program. Or, if you need to look up a statistic, instead of buying an entire reference CD-ROM, you could pay for just the facts you need. Think of this metering device as an electricity meter that keeps track of the flow of data into your computer and bills you accordingly.

The meters developed by Infosafe and Wave are similar in every way but their dimensions. While Infosafe's device is a small box attached to a computer and a phone line, Wave's solution is a single chip that must be integrated into a computer. With either system, a user can transfer money onto the meter by providing a credit card number, which the meter then verifies by modem. When a user requests a program off a CD-ROM or an online database, the meter subtracts the appropriate amount from the user's credit balance and then downloads and decrypts the data. Downloaded programs may be set so that they live for only a few days or uses.

Sprague likes to illustrate how data metering will change the software business with this challenge: "Go out and try to find a spell checker for German. Chances are you won't be able to find it in any store." His point is that with limited shelf space, stores can stock only the products they know will attract a large audience. But metering schemes afford infinite shelf space, because the store is virtual. That means niche applications are economically viable. And it means customers can instantly obtain almost any program they want.

Data metering exploits the fundamental ease of distributing digital information. By collecting a toll at the user's computer rather than at some distribution point, you can cut out a lot of the middlemen, as well as the costs of packaging. And that means cheaper software.

Metering systems will also help eliminate software piracy, says Thomas Lipscomb, CEO of Infosafe. Software will have built-in hooks that start the meter ticking. So, it doesn't matter if you copied the program from a friend—you'll still have to pay to use it. These software hooks also allow a wide variety of charging policies. Five dollars, say, every time you print, or US$20 every time the program is run.

The advantages of data metering have been known for a long time. Ryoichi Mori, a professor at the University of Tsukuba, in Japan, first came up with the idea more than 10 years ago. But only now has the technology become cheap enough and secure enough to be practical. This isn't to say all the problems have been solved: there remain nagging technological, social, and business issues, any one of which could stop these metering schemes in their tracks.

Surprisingly, it's the technical issues Infosafe and Wave worry the least about. True, both products use powerful encryption schemes and patented security techniques to protect the user's credit balance and the data being delivered. But neither Infosafe nor Wave claims its system is impervious to a determined hacker. Instead, the two are staking their reputations on the belief that if it is hard enough to break the system's security, most people will just pay. As Lipscomb points out, the existence of the copy machine hasn't destroyed the publishing industry. Besides, argues Sprague, "total security would require a complete change of computer architecture. And that isn't going to happen."

Infosafe and Wave are only slightly more worried about social issues, such as privacy concerns. True, metering systems keep close track of what programs you use and what data you download. But, as metering advocates are quick to point out, the same can be said of credit cards. Nonetheless, computer users are notorious for their dislike of anything that smells like "Big Brother inside."

It's the cold business realities that make Sprague and Lipscomb lose the most sleep. Who's going to buy these metering devices if there is no software out there for them? And who's going to develop software for meters that no one has? The solution Infosafe and Wave have come up with is, as Sprague puts it, "to sell the chicken with the egg." In other words, bundle the meter along with the software. Indeed, Infosafe is already doing that with Design Palette—a collection of 43 CD-ROMs, packed with tools and images for graphic designers, that comes with a meter included.

In the end, most software industry analysts agree that metering devices will eventually be used with expensive, business-to-business content. But for consumer applications like games, experts are more pessimistic. The only way it will work, they suggest, is if meters are built into the motherboard of every PC. That's not completely out of the question. After all, DAT players are sold commercially with circuits that prevent unauthorized copying, despite the added cost. But then again, consumers never really adopted DAT players, as they didn't seem to offer much of an advantage. And that's the situation Wave and Infosafe find themselves in now: they need to convince us that the benefits of metering are significant enough for us to put a toll collector in our computers.

—Steve G. Steinberg

Tracking Usage Rights

According to Mark Stefik, a scientist at California-based Xerox PARC, "the seeming conflict between digital publishing and commerce" can be overcome by fundamentally redesigning computer systems. In his paper "Letting Loose the Light," Stefik proposes a solution reminiscent of software metering, but far more radical. It rests on two key components: trusted systems to store information, and usage rights to define exactly what a user can do with a digital work.

Think of a trusted system as a device that can be trusted to follow the law. The device might be a computer or a CD player—whatever medium, the owner would be unable to use it to make an illegal copy. By storing unencrypted digital information only on these trusted systems, the fluidity of information can be controlled.

In Stefik's scheme every digital work is assigned several levels of usage rights. For example, the usage rights for a CD might allow the user to listen for 25 cents, or make a copy for US$5. These usage rights can be nested, one within another, so that a digital work can include other people's work while ensuring everyone gets paid. A music compilation might have a separate usage right for each track, defined by that track's creator, as well as a usage right for the entire CD, defined by the CD's publisher.

Stefik isn't stopping with the paper description of his scheme: he's filed for a number of patents on the underlying technology and has put together a business team. But Stefik admits that compromises are in order to make his scheme economically viable.

—Steve G. Steinberg

Digital Watermarks

When Geoffrey Rhoads, an engineer from Portland, Oregon, thought about selling some of the digital astronomical images he'd been developing, he hit on a critical problem. Once in digital form, there was no easy way to "sign" his planetary pictures to prove they were his.

It took him only a few weeks to come up with a solution.

The answer lay in a mathematical technique to hide data, such as a signature or ID number, inside an image. The data is invisible, but a computer can analyze the photograph and pull out the hidden data. Here's why his development is so innovative: it's impossible to strip the invisible code out of the photo without completely mangling the image.

Think of Rhoads's scheme as a digital watermark, a unique identifier that becomes part of the document and can't be removed. Rhoads has filed five different patent applications to cover his technique, and started a company, Digimarc Corporation, to commercialize the new technology.

Roger Dillan, a member of Digimarc's management team, says photographic stock houses are an example of where digital watermarks will be used. "By placing invisible codes into their digitized images, stock houses will be able to distribute photographs with less fear of piracy," he says. In fact, if the code number is changed every time the photograph is sold, the stock house will even be able to track the piracy back to the original source.

Digimarc isn't the only group working in this area, but it seems to be closest to a market-ready product. Digital information may never quite be the same.

—Steve G. Steinberg