If prices are governed by the ratio of demand to supply, and most things in the knowledge economy are available in infinite quantity, should everything be free? Not quite, but almost, for the basis of value and trade must change.

Information, unlike anything else known to humans, can be replicated indefinitely. This has always been possible, regardless of the medium of transmission - whether a grunt or gesture, a manuscript or several megabytes through optical fibre. There is no inherent cost in such reproduction; it is only the medium for which you may need to pay. While it is part of Internet principles to say that information wants to be free, in fact it has often been so. Although looked upon as valuable, knowledge has in past civilizations been treated much as air - from the oral tradition of ancient India to the early scientific exchange between Arabs and crusaders in the Middle Ages.

Later, it became something to be kept secret, within trade guilds for instance. And at the threshold of the 21th century it has reached the extreme where a sequence of zeroes and ones may not be repeated without paying someone - information has become not free, but property. But technology has advanced simultaneously, so that as the volume of intellectual property increased, so did the ease of replicating it at a small, and decreasing, cost.

Any one morsel of information can now be reproduced infinitely for transmission, in theory, to all those who want it (as well as those who don't). That morsel now has only a notional value - what its 'owner', or originator, could have received if the information was available in some arbitrary, but finite, quantity. It is from the loss of this notional amount that the fuss over intellectual property rights arises - from the problem of infinity.

The key to its solution lies not in ignoring a real problem that is intrinsic to information - because the market will not, whatever the laws - but in understanding its implications. Reproduction, replication, redistribution may be in infinite quantities; still, creation occurs but once. No matter what the demand, the supply of the creation of knowledge, rather than of knowledge itself, is quite limited. Its value, it follows, is just as high as that of raw information is low.

This can lead to some conclusions that are not immediately apparent. The pursuit of original creation is a manifestation of the knowledge economy's hunger for change - mass production simply will not do. But continuous creation can be divided into categories, leading to coexisting bases of value, and sometimes-conflicting models for trade.

First, there is plausibly some value in creation with consistence, which leads to the relatively well understood model where knowledge labour rules, of expertise and the client-consultant relationship.

Then there is creation with diversity, where the value is actually in the difference between ideas, in the variety of knowledge. This is evident on the Internet - which would lose much of its appeal but for its reputation of having something for everyone - but otherwise discounted without much thought. As such diversity is rarely a trait of the expert, this basis for value needs a different model of trade, which just happens to be non-monetary.

This model is the 'cooking-pot market' found on the Internet: the common barter exchange into which everything created is dumped, as in a vast tribal cooking pot (which is bottomless, thanks to the infinity of information) and from which what is wanted is taken, reproduced as it is at no cost, at will. Everyone has something - to put in, for variety is the basis for value here; and to take out, for the cooking-pot is where you go to find something different.

So the problem of infinity does have solutions. These solutions are not particularly easy, and may be especially hard for those who have decided the knowledge economy is where they can create once and then await infinite returns. They may find infinite problems instead.