Digital replicators

In general I define replicators as information, ideas, or an item of intellectual property that can be copied or can be considered to reproduced. Digital replicators are the same ideas or information but produced in a digital form that can be distributed over the Internet or some other medium. I only make the distinction because this paper discusses issues centred on the digital manifestation of replicators. Examples are well known: CDs DVDs, MP3 files, text files, software.

Digital environment

Although centred on the Internet I extend the idea of the digital environment to include digital products however they are distributed, this includes such items as CDs and DVDs. Human information producers and users are also considered to be part of this extended environment.

Simple analysis of  some digital replicators

Richard Dawkins has established three properties that can be used to define a replicator: Fecundity, Copying-Fidelity, and Longevity. A replicator that is successful in all three of these areas will be a very successful replicator and will spread far and wide. Conversely, a replicator that has weak properties will be a poor replicator.

For this analysis I use these three properties as coefficients in an equation that predicts a replicators success at reproducing and therefore its success at spreading  throughout the environment.

            Replicator success quotient = Fecundity * Fidelity * Longevity

Each coefficient is given a value of zero to one based on the replicators performance in that area and the result will predict the replicators success. (Zero will mean total failure through to one predicting great success).

For example a replicator that always copies itself with so many errors that it is unrecognisable will have a fidelity score near zero and will not get far as a replicator. Equally, a replicator locked in a file with an unknown key will have near zero fecundity and will not spread.

For the Digital replicators mentioned above and a few others for comparison I have estimated values for the three coefficients and produced the table below.

The coefficient values are somewhat arbitrary but I believe give a good and interesting indication of the predicted success, as a replicator, of each of the media.

Considering the vinyl record (replicator value 0.14): Its fecundity is rated at 0.5 because of the specialised equipment required to produce a record. Its fidelity is rated good at 0.7 when you consider the music quality when a new record is played on a high fidelity system but its longevity is rated poorly at 0.4 because vinyl records can wear quickly and can easily be irreparably damaged.

Looking at the coefficients for the raw MP3 file: Its fecundity on the Internet is very high at 0.95 because of its relatively small compressed size and the ease of its practically free distribution. Being digital, its copying fidelity is good (0.9) and its small digital size makes long term storage feasible once it has been copied (0.8). This gives it an overall replicator value of 0.684.

Now look at the coefficients of the new MP3 file with PRD field supported by the distributed rights environment: Its fecundity value (0.7) drops in comparison with the raw MP3 file (0.95) because of higher distribution costs in the rights environment (note that I estimate these costs are still less than obtaining a physical CD (0.45)). Its fidelity is improved (0.95) because many copies are made from a master file and its longevity is ensured (0.95) because any lost files can be replaced from the master through the licence system.

A file that degrades or is inhibited after it has been copied once such as in the SDMI environment will have a much reduced fecundity (0.5) and therefore is a less successful replicator than some of the other formats. (This is obviously the aim of the SDMI and shows that it will reduce the number of file copies but I will argue later that this is not the desired result.)

The following table goes on to expand this analysis over a slightly wider field.

* Assuming sufficient bandwidth is available in the future.

Conclusions from replicator analysis

First off, if you look at the comparison of vinyl records (0.14), cassette tapes (0.189), and CD's (0.385) it shows why CD's have become successful and vinyl records almost obsolete despite the higher costs involved in producing CD's (more complicated technology and new equipment required). Following that it shows why the current rage in MP3 file (0.684) swapping is more successful. It also emphasises the advantage digital replicators have over the analogue equivalents and so demonstrates why the digital environment is going to be 'the future'.

The first table shows that the new MP3 file format with PRD fields added and supported by rights and licence offices has a similar replicator value to raw MP3 files. You might wonder what is the benefit of introducing the new complicated DIRP system. The significant point is that at least 50% of music in the MP3 format would 'prefer' to take the new system route, where the product and owner of the product will be identified. The second point, I would argue, is that as the new system becomes established its fecundity will improve (costs come down) and so will overtake raw MP3 files as the preferred route. In the same way, any environmental pressure that promotes MP3+PRD reproduction over raw MP3 files, either moral, legal or technical, would further swing the balance in favour of new MP3's plus PRD's.

Regarding the second table: Assuming portable reader technology improves, E-books (digital text) will become very successful replicators and more so if producers were protected in the new digital rights system. Digital video on demand has the same promising future and, again, I would argue that if the rights system were in place the identified version would far out perform the raw version partly because the costs of local long term storage will be that much greater.

In the table below I introduce a further coefficient, the desirability coefficient, that includes a cost factor to obtaining the product. I consider the product to be a music file distributed in three formats, CD, SDMI protected product, and MP3+PRD, and for each format I compare its distribution against a 'free' MP3 file containing the same music. The desirability coefficient of 0.1 is arrived at by assuming 10% of the population who liked that particular piece of music would pay a reasonable cost to be able own and play the file and always have a perfect copy available should they loose their copy.

Note the predicted increase of sales of the PRD protected product over the other formats under the same market conditions. Assuming the costs of the electronic MP3+PRD product are less than any physically distributed intellectual products, such as CD's, then the purchase cost could be reduced and you could expect even more sales in the same market. Also, under the Distributed Intellectual Property Rights system, most of the MP3 files shown in the last line of the table would in fact include a PRD and so a least the product would be identified and the creator known even if it was not a purchased product.

If you were to perform this analysis on other products which contain information which might be updated regularly or software which is revised or improved regularly the desirability coefficient could be much higher and the PRD identified product would proportionally gain more of an advantage over a non-identified product.