When, in early 1980's, I told everybody that the old bulk production system is to be soon changed in favor of the more progressive mode of on-demand production, people just laughed and called me utopist. I used to say that producing things in advance, in anticipation of request, was bound to waste resources due to statistical demand variations, warehousing costs and logistics losses. It would be much more effective to produce each thing in response to a specific request, immediately shipping it to the requester, who would directly put that thing in its place, never needing to keep it in stock. Of course, for underdeveloped economies, this is not an option, since on-demand production requires flexible technologies, fast communication and reliable transportation means. When it takes weeks to contact the producer, then months to replicate an old product, then weeks to bring it back to the consumer, one can hardly expect any economic stability. Bulk production and stock came up as a natural solution, creating a kind of a buffer between the producer and the consumer, to smoothen any productivity or demand variations. This is exactly like in computers, when data sent to a relatively slow device are first stored in a special buffer to avoid device underrun or congestion. However, with the devices becoming faster, the need in such buffers is being pushed to the periphery, and data can be sent to their destination immediately as they are produced. Multiplexing and framing can cope with communication efficiency problem, combining different data flows to minimize the idle time on the line. Similarly, any article at all could be produced on demand and delivered to the consumer through a "multiplexed" transportation system, utilizing fast communication means.
The idea is not new. It logically follows from the Marxist economic theory, and it was discussed in the Soviet Union since 1930's. Some Soviet science fiction writers (for instance, Zinovy Yuriev) pictured that kind of economy in their books, and many people got contaminated with the idea through belles-lettres. I can only cherish a slight hope that my own discussions with many people have served to the propagation of this idea in the world, especially after the ruining of the USSR, when former Soviet people went worldwide for a piece of bread, unintentionally bringing the unsorted fragments of the progressive ideology whenever they came.
Of course, one can hardly expect the domination of on-demand production under capitalism. The market is a rather wasteful distribution mechanism, and one will always need certain buffer structures to stabilize the capitalist economy. However, the new is always born within the old, and I can see a prototype of the economic order to come in the business of on-demand publishing, that is rapidly developing today. Surprisingly, it is the example of book publishing that I often used to illustrate my theory in 1980's. There were problems with getting books that one wanted to have, while many books filled the shelves of the bookstores without a slightest chance to find a reader. Well, said I, why not to keep anything necessary for printing a book for a while and let people send their requests to the publisher, so that the number of printed copies would match the readers' interest? Probably this was not that easy in that time, with rather cumbrous printing technologies; however, it's becoming quite possible today, when everything one needs to keep the book ready for print is just a few hundred megabytes on the hard disk. With the modern multi-terabyte storage facilities, there is no problem in thus maintaining millions of books, and virtually all of them. The practical development of this system is limited by the market economy; in its full, it would result in a comprehensive library, whence everybody could get the required text, or portion of text, for free.
The same system is bound to penetrate in any business at all, as soon as the production means get controlled by computers and the reproduction of any particular thing can be reduced to a simple program change. The industry of raw materials is the bottleneck of on-demand production, but this difficulty can certainly be overcome with more technological development finding efficient ways of refining and recycling.