Andrew D. Todd Software Piracy and the Crisis of Capitalism (1993) Software piracy has sometimes been talked about as a crime-- or as a non-crime. It has even been accused of being the major source of computer viruses. But on a more profound level, it is a harbinger of the end of almost all the political and economic institutions of present-day society. Let us consider the possible role of the software pirate as a kind of techno-economic revolutionary vanguard. I. The Economic Crisis of Information Back in 1987, Jay Kinney wrote an article for Whole Earth Review ("If Software Countries Ran The Country," Winter 1987, #57:98-100), in which he protested the business practices of major software companies by imagining a situation in which common tools like hedge trimmers were sold on the same terms as commercial software, with limited warranties, penalty clauses, and all. As an alternative, he offered a vision of people sharing software about freely, taking advantage of the fact that it could be easily reproduced, drawing lightheartedly on the cartoonist Al Capp's character, the Schmoo, an amiable beast that would effortlessly produce whatever people wanted. As later transpired, Kinney was mostly talking about the advantages of shareware, and did not actually advocate software piracy, but he threw out an intriguing point: that in the face of software's replicability, property claims in software were obsolete by virtue of being unenforceable, and attempts to maintain them can only end in an escalating cycle of conflict. For, as he put it: "...Software is Schmooware, and it loves to reproduce." In the very next issue, there was an approving comment by Neil Rubenking, the author of Pianoman, a shareware music 'note processor,' (WER, Backscatter, Spring 1988, #58:139), who elaborated Kinney's fantasy with the idea of a 'super duplicator.' Again, it's most likely that he was only referring to software, but he was skating on the edge of something more profound: that such a machine is the logical conclusion of manufacturing technology, a point that Arthur C. Clarke had made in his Profiles of the Future. Clarke had, in effect, taken Al Capp quite seriously, and begun thinking about the technical problem of how to build a Schmoo. Science fiction writers sometimes reach degrees of profundity that mere social critics, tied to reality, cannot plumb. Clarke's discovery implied that since all manner of goods and services are just information and energy, the economic laws of information will ultimately become universal economic laws. The most basic law of traditional economics is TANSTAFL, or 'There Ain't No Such Thing As a Free Lunch.' But this is merely a reformulation of the first law of thermodynamics, that energy is neither created nor destroyed. Naturally, it does not apply to information. So Kinney and Rubenking's pedagogic fantasy is on the verge of becoming reality. However, very few people have even begun to think about the implications of a shift to an economy of information. There has been a certain amount of discussion about issues of property, as applied to information, of course. But this discussion is crippled by its assumption that with certain qualifications business will go on as usual. Even so perceptive critics as Alvin Toefler and John Naisbit are still thinking in terms of an individual, or cottage capitalism. Indeed, there is still a tendency to apply TANSTAFL to software, where it manifestly does not apply. In the same issue of WER that carried Rubenking's letter, there was also an angry denunciation from one Anne Katherine Soleil (ibid. p138), who apparently regarded Kinney as little more than a covert apologist for software piracy. Oddly, enough, Soleil, Rubenking, and Kinney all shared one common assumption: that there were people who produced software (and information in general), and there were people who used it, sometimes without paying for it. What they all failed to ask was one crucial, burning question: what if these two groups of people were one and the same. Alvin Toefler had come close to this point with his concept of 'prosuming,' but he was still thinking in terms of a comparatively uninvolved user doing a less profound kind of designing things. He did not address the appetites of the people at the very center of the creative culture, those producing the most profound kinds of tools. II. The Artist as Thief That is probably the most fundamentally subversive fact of an information economy: creating information is mostly a matter of recombining it in new ways, and therefore information creators themselves use information in mind-boggling quantities, which they acquire with all the restraint and subtlety of a Viking raider or a Mongol horseman. Since computers and software are where this all started, let's take them as our first example. Because of the uniquely intimate relation that computers have with their programmers, certain characteristics, or rather, social implications, of the new information technologies have first become visible here, but they ultimately apply to most arts, sciences, and technologies. In computing, the most productive and valuable type of person is what one might call a 'young computer expert' (terms such as 'hacker' are too imprecise and sometimes too exclusionary to apply completely). A young computer expert is commonly a bit impecunious: This is a normal corollary of his emphasis on expanding his skills in preference to merely trading upon them. Now one of the remarkable thinks about young computer experts is that all the best ones seem to be very serious bootleggers of software. Such a person will have what amounts to his own software store, with copies of practically everything under the sun. What he doesn't actually have, he can get within a few days from his friends. This sort of arrangement is a necessary result of what the young computer expert is. He has little or no money (especially after he's bought rather more computer than he can really afford), and yet he must develop and maintain an encyclopedic range of familiarity with different software to maintain his expert standing. That means he really does need twenty or forty different word processors or whatever. And so on for every category of program. Even if he has a job, he could not possibly afford to buy them all, even at student discount prices. He couldn't even rent them (never mind that as Kinney points out, vendors commonly regard this as a cover for piracy, and attempt to stamp it out). What the young computer expert can afford is the price of blank disks. He is too obscure to be regularly and consistently given presents of review copies (including beta copies). Naturally, he will gratefully take them when offered, and will devote quite a lot of time to testing and critiquing them, but it's a fortuitous bonus, rather than a regular and reliable means of supply. So young computer experts bootleg. They bootleg everything they can get their paws on. Along the same lines, many of them engage in unauthorized entry to computers, and phone phreaking. Leaving aside security penetrations proper for the moment, much of this is 'theft of service,' and much of the same logic applies to as for bootlegging: the illicit user could not possibly pay for the range and variety of remote system access he needs to qualify as an expert. Of course, phreaking is much more dangerous than bootlegging, and therefore tends to define a more exclusive club. Security penetrations proper are themselves intimately connected with the display of a certain kind of virtuosity, of perfect understanding of the operation of very complex systems. So, to sum up, being a computer expert more or less goes with being an outlaw in terms of the system of property rights in computers. A very high proportion of conspicuously successful innovators (e.g. John Draper, of Apple and Easywriter), are known to have a background of phreaking. And that is generally because they got caught. The proportion who have merely bootlegged (which is nearly undetectable) must be much higher. So, when considered as a class, computer experts bootleg what they create and create what they bootleg. Perhaps computer people are a special case? Very well then, let's take another group: Hard Core Liberal Arts Graduate Students. They bootleg software too, but they do it less, and it is less of a necessary concomitant of their position. What they mainly bootleg is reading material. Naturally, they copy journal articles, and anything else that is not readily purchased and does not circulate outside the library. But on occasion, they will set out to copy a whole book from cover to cover. With skillful use of reasonably advanced models of copier, this can be done in about half an hour. Even allowing for their labor at the highest hourly wage rate that a grad student could realistically obtain, the cost is generally no higher than a new scholarly paperback, and often less. Due to the sheer tedious difficulty of Xeroxing, the threshold of actual copying is somewhat higher. Usually there is an element of it being impossible to get the material through legitimate channels in a timely fashion. Cost generally enters into the situation when a professor decides that the grad students can't afford a given book, and announces that instead of it being in the bookstore, it will be on reserve in the library. The grad students rapidly decide, either singly or in concert, that lining up to do their reading in a public place is intolerable, and set about doing something about it. Yet another proximate cause of bootlegging is when a grad student has checked out a book, only to have it recalled. He isn't done with it yet, and he may well be committed by his page citations to that particular edition, so he does the sensible thing: he makes himself a copy immediately, and then turns the original back in. Some degree of restraint is provided by the fact that the most advanced copiers, those with collating ability are still attendant-operated rather than self-service. So the copy shop management is in the position of taking responsibility for what is copied, and having an incentive to regulate it. But this may not last much longer. Quite apart from the possibility of new and more automatic copiers, the copier is on the verge of ceasing to be a separate and specialized machine, with definite limits on its operating modes. Soon, when full page scanners and laser printers become cheap enough, it will be merely a capability of a typical personal computer system. Something similar may very likely happen to video and microfilm apparatus in due course. At that point, all possibility of control will end, and the Stack Rats will copy with the same abandon as the Hackers. But again, these are the very people who write books. Any serious interference with 'piracy' would jeopardize the production of knowledge. Even the "Kinko's" decision, banning the unauthorized production of course packets, has proved highly disruptive (of course the advance of hardware will soon put the situation largely to rights again). One could supply other examples, of course, such as the collaging habits of the more innovative graphic arts and design people, who cut up innumerable picture magazines for raw material. But the point has been sufficiently made: the act of 'theft' is integral to the act of creation. To the extent that information is defined as property, creative people will be outlaws. This means that a society dependent on abundant creativity in connection with comparatively pure information (all goods and services can be defined as mixtures in varying proportions of information and energy, remember) cannot afford to allow property rights to extend to this new information sphere. However, the maintenance of free information will not depend upon society being enlightened about it. Artists, writers, programmers, scholars, scientists, technologists, and other intellectuals who simply take whatever they need will have that much more material to draw upon in creating their own work, which will be that much better in consequence. The natural result will be that it will drive out the comparatively anemic stuff produced with only the sources obtainable through 'legitimate channels.' (Incidentally, I am not talking about anything so crude as merely recycling someone else's code. What I have in mind is more of the order of this: A hacker filches copies of many, many word processors, plays with them all extensively, disassembles parts of them, and obtains from all this certain insights about how to build a better word processor, one better than any of the others, which insights he then acts on. In coding up this new program, he may or may not scavenge bits and pieces of logic from the previous programs, but the borrowing is dwarfed by and subordinated to his own invention. Conversely, even if he writes his own code, the fact remains that he didn't pay for the software by critiquing which he developed his own ideas. There is no release or easy escape from the moral ambiguity of the situation. At all events, our hacker now releases this new program, whereupon some other hacker swipes a copy... And so the cycle repeats endlessly, leading to a result none of the programmers involved could have accomplished on his own.) III. The True Nature of Information So far, it has been assumed that there is such a thing as 'Information,' and that ownership in it can be assigned, whether or not that is a socially prudent thing to do. But could it be that information does not actually exist? Since it shows such a propensity to transfer itself in defiance of all restraints, shouldn't we consider whether that is its essential property? Let us assume that true Information exists only in motion, and that the idea of 'Information' is really just a reification of the concept of 'Communication.' It is a commonplace to speak of Information as something that actually exists, and can potentially be owned, in the same way as moveables or real property. But Communication has a somewhat different nature than Information. It is not essentially a pattern of bits as a state of being connected. Communication is always with someone, even if the other party is now deceased, or if the other party is oneself. This last case would be Auto-Communication, or 'Feedback.' The economics of Communication are different, as well. The value of being hooked up to a communication system derives from being able to get a message to someone (assuming they're in a condition to receive it), and from being able to receive one from them. So the enrollment of additional members increases rather than decreases the benefit to the previous members. That means that the idea of scarcity is knocked on its head. Each individual already in the system finds that the way to maximize his own benefit from it is to get everyone else hooked up. The only real limiting factor is the rapidly declining cost of the necessary equipment. As the cost of hooking up a new member becomes less that the marginal utility of that member to those already enrolled, it becomes rational for them to insure that he is hooked up. For Feedback, incidentally, it does not even matter whether anyone else may benefit by your well-being. You are at once the source and the destination. If you have the sensors, you can get all the feedback you want. Of course, not all information behaves like this. There is some 'non-communicative information.' Non-communicative information can be divided into Congested Communication and Energy-driven Information. Congested Communication is the residual seepage in a situation where the normal flows of advantage-seeking communication do not operate. A classic example is market research. Marketeers go to great lengths to ascertain popular preferences about tastes, scents, textures, etc. Then, the results are treated as a valuable and scarce commodity, and hence as secrets of the highest order. But if you think about it, this is a result of a failure in communication. Such information is not readily available, because the consumer does not sufficiently articulate what he wants. When he can say what he wants and why, it is in his interest to shout it aloud, so that the greatest number of possible vendors might respond. When better communication links are in place, he can do just that. Again, computer people have been remarkably successful in decongesting their own communication. They convert their feelings about software into letters to the manufacturer and letters to the editors of computer magazines. The editors print a selection of it, which becomes public information. An aspiring software developer with no connections whatever can obtain an fairly exhaustive list of what not to do by reading though a few years worth of letters sections. By contrast, advice about how to make more satisfactory canned soup is far more elusive. The consumers do not normally express themselves for the public record: extreme displeasure is almost always expressed in the vigor with which the offending stuff is lobbed into the garbage can. Even if the consumer should be minded to make public complaint, there are simply not many of the necessary channels to do so. 'Consumer Reports,' and similar magazines serve a useful function, I admit, but they are not very practical to someone who does not wish to make a profession of being a consumer. Imagine if your comments on the vile rations could be more or less automatically transmitted to anyone who was standing in the supermarket aisle, debating whether to get it. And of course to all of the competing vendors, who in turn would be able to indicate that they had taken cognizance of the complaint. But such is not yet technically feasible. So market researchers have to go to a good deal of trouble to set up tasting sessions and analyze the results in order to find out that So and So's chicken noodle is a nauseating mess. However, this information has the potential of becoming communicative, merely by the removal of the barriers which obstruct its flow. The other and more fundamental kind of Non-Communicative Information is Energy-driven Information. Energy-driven information is information leading to the discovery of a finite energy source. The blueprints for perfectly efficient machinery, for example, would not qualify, because there would be no real limit to how much of it could be used. What make information energy-driven is the fact that it practically conveys the use of the energy, provided it is acted on preemptively. Follow your copy of the treasure map quickly, and you get the chest full of gold, while the next fellow with a copy of the same map only gets a long nose. But if you delay, it's you who gets the long nose! Indeed, the purest real-world case of non-communicative information I can think of would be oil-prospecting data. But the total value of this information is only some fraction of that of the energy it pertains to. As the economy becomes more and more information-driven, Energy-driven information should become less and less important along with energy. IV. The Last Days of Information as Property The dominant activity in an economy tends to impose its own coloration upon other activities, enforcing compromises upon their practitioners. For example, an energy economy compels the producers of information to 'energize' their information, converting it into a form that can be traded for energy. Philosophers must eat, and therefore the philosophy must be made into a salable form. Hence the producers of information have, in sheer self-defense, maintained the illusion that it has price, scarcity, etc. But what happens when Information, or Communication, becomes dominant over Energy? At present certain minimum appearance of unfree information is maintained by the needs of information producers for goods which are economically unlike information. That is, a hacker, be he ever so good, still needs thing which he is not in a position to simply make a copy of. This means he has to have some kind of involvement in the mundane economy, that of energy and TANSTAFL. There are two ways he can do it. The first way is to find a way of selling information. Either he can sell copies of information, or he can teach. In practice, the most lucrative forms of teaching tend be those of large lectures, etc., which approach the quality of a recording. But the problem is that in so doing, he becomes an economic antagonist to bootlegging. The second way is to produce mundanely in proportion to his mundane requirements. This can mean self-sufficiency in mundane requirements, eg. hippie-style back-to-the-earth dropping out, or it can mean producing mundane goods and services to exchange for other mundane goods and services. There are two problems with this. The first problem is that it diverts his time and energy away from the production of information. The second problem is that, given the present state of the art in mundane production methods (assembly lines and such like), an individual cannot compete as an individual, and is therefore compelled to do his mundane production as part of an organization dedicated to mundane production, which will present him with various problems of alienation, ongoing culture shock, and even persecution at work for not being like the others. Together, these tend to marginalize him in the culture of information producers. The result is that if an information producer can humanly manage it, he will try to sell information rather than producing mundanely. This, I think accounts for the exaggerated prestige of certain programmer-moguls, e.g. Steven Jobs, Bill Gates, Mitch Kapor, etc. They are held up as archetypes precisely because they have succeeded in compelling the mundane economy to refrain from impinging on them, and they have done so in a very absolute way, infinitely far beyond mere subsistence requirements. There are analogs in other fields, for example [-----------], whose writing has become less exciting as he has gradually emerged as a major lobbyist for copyright extension. In contrast to all this, someone like Richard M. Stallman or Bjarne Stroustrup, while commanding considerable creative respect, though less uncontrolled worship, does not have the same kind of mystique, because he cannot offer an example of immediate escape from mundane-ness, with all its constraints. But this selling of information brings the information producer into a moral contradiction: he must bootleg in order to produce information, and he must avoid being bootlegged himself in order to be free from material necessity so that he can produce information. In short, he must do unto others that which he would not have done to himself. As information becomes more fluid, this moral contradiction becomes more desperate. It becomes harder to convince a 500 person lecture audience that they are getting more value than they would from a videotape. It becomes harder to control the distribution and reproduction of the videotapes and any other recording media for that matter. At the ultimate extreme, we find the hysterical, unwise, and indiscriminate lashing-out of Apple or Lotus. Of course, there is an alternative of sorts, the production of customized information for individuals. This can mean tutoring, teaching in small classes, or custom programming. The catch is that there will always be definite limits on productivity, and it is unclear just what effect computerized information tools will have. In teaching, for example, the problem is that the kind of teaching involved, when done well, consists in teaching the pupil how to work on his own, and even to teach himself. In short it has the effect of disseminating the teacher's skills rapidly outwards. In short, it's a kind of kind of ponzi scheme. As for programming, Jeff Dunteman has remarked that the decreasing price of computers, and the inevitable comparison between the price of the machine and the price of the software that runs on it, is in the process of driving custom programmers into the new 'visual' programming systems (e.g. Visual Basic). In short, professional information producers are increasingly using tools designed for amateurs, and thus forcing the creation of tools designed for the still more amateur. Custom information production is essentially a response to the customer's inability to use the information system. For example, the customer may lack the knowledge of how to go about finding what he needs to know. Or it may be a matter of not knowing how to translate what he knows into an alien language which is meaningful to the information system. But if the information system is reaching out to him in more and more pervasive ways, it becomes increasingly difficult for a class of custom information producers to find a place between them. One can illustrate this by an example-- learning to read and write. Back in the early 1960's, the sociologist Omar Khayam Moore developed a machine for teaching 3-year-olds how to read and write. He started from a psychological fact-- that learning to talk is overwhelmingly the greatest feat of learning that any of us accomplish. And furthermore, almost everyone succeeds in it-- cases of failure, or autism, are quite rare. When a substantial fraction of children subsequently fail to learn to read, and a majority fail to learn the higher forms of reading, such as reading in foreign languages, the causes are purely social, rather than psychological or neurological. Various people, starting with the Italian psychologist Maria Montessori, had demonstrated that properly instructed, toddlers could learn such skills as reading. But apart from the Montessori method's need for early kindergartens (Dr. Montessori started by running a day-care center, and went from there into early childhood education), it required especially skilled teachers. In a characteristically american gesture, Moore decided to mechanize the process. Essentially, his machine, the Edison Response Environment (ERE), was a personal computer with a voice synthesizer (in short, considerably less than a macintosh, but it cost $35,000 at the time, which must have been more than a schoolteacher made in a decade). The ERE ran an educational program of the kind which has since become familiar. When the child pushed a key, the machine responded, both visually and aurally. Moore was very careful to isolate the ERE program from the child's social environment and pressures, so as not to corrupt his experimental results. The child could leave the ERE room at will, was left severely alone while there, and parents and regular teachers were not informed about the child's progress. Under these conditions, it was found that a child would soon learn to read and to touch-type from the macine alone. The thing is much more absorbing than any schoolbook could be, since it feeds the child's own vocabulary back at it, with results that any modern personal computer user could predict. In practice, Moore could not get enough machines for his experimental school program (the one machine was paid for by a grant), so he had to round out his teaching lab with human instructors sitting on the other side of a partition, who were instructed to exactly imitate the machine (he was equally satisfied with a reproducible 'cookbook' method). The result, in economic terms was that Moore's instructional system cost approximately as much as conventional classroom teaching. While he had managed to reach children appreciably younger than ordinary classroom instruction could reach, that was a more problematic achievement. The teachers in upper grades were unwilling to accommodate themselves to Moore's early readers, and as a result, the experimental school withdrew its participation. That ended the matter. Moore did not as yet have a package versatile enough that he could take it on an end run around the educational establishment (see Maya Pines, Revolution in Learning, 1967, for fuller details). But we are approaching a new threshold. It has been established that ERE's work, if they can be gotten to the child. All that is required is that they be sturdy enough to be child- proof, and cheap enough that they can be distributed gratis. At present, what are known as 'pre-computers' are not quite good enough. Those that are really cheap are not real personal computers, and those which are real personal computers are not cheap enough, apart from being not really powerful enough for an audio-visual application. But suppose there was something like a notebook computer, costing about five dollars. At that price, the government could simply arrange that every child in the appropriate age range got a new one every year or so. The birthrate runs about 4 million per year, and assuming an age range of 3-4 years (infancy to kindergarten, at which time, the children are brought together and more efficient equipment utilization becomes possible), that means 12-16 million computers per year, which at 5 dollars each would be 60-80 million dollars. That is a small figure by government standards, and is probably even within the reach of private charities. Which is another way of saying that if one can produce a five dollar computer, one can take it pretty well as read that all children, slum dwellers included, will arrive in their first year of formal schooling knowing how to read, write, and presumably everything else in the Sesame Street curriculum, only with infinitely better results than can be secured by a non- interactive television program. But the subjects of the Sesame Street curriculum represent a high proportion of the elementary and middle school curriculum and almost all of the arduous components thereof (with the exception of handwriting). The justification for the second and third cycles of elementary- middle school (grades 4-6, and 7-8) is primarily remedial, an indication that grades K-3 do not consistently succeed in imparting the Curriculum of Sesame Street. That is to say that a child who enters kindergarten having had a couple of years of continuous exposure to a personal computer (in the sense of owning it and being allowed to use it continuously, unsupervised, and at will, just like any other toy), is going to be able to coast, getting high grades and reading omnivorously, until he or she starts the secondary school curriculum. But here the stage is again set for something similar. Assuming that the student does not have some residual illiteracy problem, subjects like History, Biology and English are not going to be overly demanding. The real enemy in these subjects is likely to be boredom if the students are kept too passive. Assuming the Biology teacher is sufficiently intelligent to be broad-minded about aquaria and rats, there should be no problem. The real meat of secondary school is and always has been languages, and more recently mathematics, which is a kind of language. One might add the smattering of Linguistics and composition found in English, and much of Physics and Chemistry, which are really more of the same. These languages and pseudo- languages are really a continuation of the Sesame Street Curriculum at a higher level of proficiency and abstraction. Again all that is required to make them easy is a suitable addictive mechanism. There are suitable programs, designed for the use of adult professionals of one kind or another, which fill the bill. For languages proper, there are machine translation programs, which if correctly employed, can lower the threshold of difficulty in a foreign language to the point where a normal desire to find out how the story ends is sufficient to keep the student going. It is only very recently that such programs have not all been priced on the assumption that the customer was not a professional translator, or rather a large firm of translators. Such an assumption constituted a self-fulfiling prophecy by making the software much too expensive for anyone else, thus concealing for as long as possible its revolutionary potential. Similarly, something like Mathematica could probably produce the same effect in mathematics instruction-- if it were appropriately priced. For some years, various study groups of academic mathematicians have developed recommendations for curricula involving the beginning of serious mathematical training in the second grade. Such a program generally leads, by age 18, to a proficiency in mathematics sufficient for a higher secondary diploma with that specialty (e.g. the French Baccalaureat or German Abitur, equivalent to junior standing in an american college. Higher secondary diplomas are given in math and science, classics, or modern languages). The key flaw to such a program as the mathematicians proposed, of course, has been the number of teachers who could only teach math at all by rote, and the number of students who were still shaky on basic arithmetic. The program is theoretically possible, and actually possible with bright students, but makes no realistic allowance for failure. That is where the cheap personal computer comes in. Even in secondary schools, their impact has still been blunted by the fact that few schools have enough for the students to really treat them as personal computers. When they actually do become personal in the sense of everyone having one, all sorts of things are likely to happen. So one way or another, cheap computers are going to force the schools into reexamination. It is unclear whether the schools will wind up being dismantled, but they will certainly have to be run in a far less authoritarian way. But more to the point, these changes can be expected to supercharge the collective intellect, to the point that intellectual autonomy, the ability to grow out of inexpensive books instead of expensive teachers, will be a far more common trait than it is today. In short, there are going to be far fewer people who need an intellectual to mediate between them and the various higher learnings. In anything like the long run, a posture of information selling and energy buying, the economic role of an intelligentsia, is unsustainable. This is still more true if the intelligentsia desires to maintain a superior position. In the long run, each individual will have to participate in each sphere, according to the rules of that sphere, to the extent required to satisfy such of his needs as come from that sphere. This can be seen on an economic level along the lines of the impossibility of maintaining a favorable price for information. But it can also be treated at a moral level, by saying that there is no possible fair price for information, expressed in energy, simply because energy is conserved, and information isn't. If an energy producer buys information with energy, the information producer hasn't been deprived of anything, as the energy producer has, and consequently has gotten something for nothing. But that means that the only price fair to the energy producer is zero. But that defines the information as worthless, which is unfair to the information producer. If there is no equitable basis of exchanging information for energy, then the information producer must provide his own energy. But this is not as dire as it seems. The energy sphere will be contracting. This is somewhat obscured because it does not contract as fast as the information sphere expands and fluidizes itself. But, nonetheless, the energy sphere is contracting. We are conditioned to think of information as what can presently be produced by comparatively versatile machines which were are familiar with, or reasonable developments of them. That is, we think of information as text (in a book or on a screen), programs, pictures, audio-visual, etc., because we have computers, video displays, printers, copiers, scanners, sound systems, VCR's, etc., all fairly general purpose tools in their own right, and we can even seen from the fact that they are built out of substantially the same underlying components that they might well converge in the near future. We can even make the leap of seeing how an advanced printer might have a built-in binding machine, and thus be able to deposit a bound book in the output tray. But we are not conditioned to see manufactured objects in general as potential information because we have no experience of them or anything like them being information. But for all that, the raw materials rarely cost more than a small fraction of what the finished product does, leaving aside that those raw materials are themselves the product of manufacture or cultivation (art applied to nature). As wealth increases, due to the more extensive operation of information, there is a shift in the mix of things the information creator uses. Once he reaches the point of eating all that's good for him to eat without making himself sick, foodstuffs per se become a lesser and lesser proportion of his eating expenses. Either his expenditure in this direction decreases, or it is maintained by a continuing quest for novelty of cuisine, which is to say, for information, and communicative information at that: communication proper in the form of the recipe, and feedback that sees to it that everything is cooked to the proper degree. The same principle applies to all of his other possessions. It becomes possible to enjoy every possible comfort, to have access to every useful tool, without the expenditure of a great deal of energy. Machines will be converging towards the Schmoo. Fixed gears, tracks, ratchets, and so on will be replaced by servomechanisms driven by a microprocessor, and appropriate feedback sensors added. Photographic processes, such as photo-etching, possibly holographically refined into three dimensions, will spread. Instead of many new kinds of machines, there will be mostly new programs, and only a comparative handful of new machines, which may themselves be economically more like programs. With increasing automation in their production, these machines will cost less and less without any compensating increase in demand, so there will be less and less room for capitalism. V. The First Days of Free Energy. As the energy component of one's livelihood decreases, the means by which that component is obtained matter less and less. At some point, energy will cease to be treated according to an economics of energy. When one's energy-labor is down to half an hour a week or so, it no longer really matters if one is being cheated, and it may not be worth keeping accounts of how much one has done. Instead, an economics of information may be imposed on energy. This might take the form of making gifts of old hardware, rather than trying to recover its residual value. It is not actually rational in the strictest sense of the word to give away energy freely to all comers, but if there are very definite limits on how much they might even want, and meeting those limits is very cheap, doing so might be preferable to having to deal with the complexities of juggling two kinds of economic system. Thus will begin the 'Withering of the State,' as a Marxist would put it. Once property and scarcity go, there's no need for government.