An ensemble model to sustain variability power in the theory of developmental pathways?

The theory of developmental pathways is that technology develops not due to some spontaneous event which is often referred to in the media as a “discovery”, but because technology and learning has reached a stage where it is inevitable that it will be discovered. So, for example, whilst someone today claimed that CERN invented the internet … the fact was that other technology required a form of communication, and whilst the exact form of that communication may be down to one particular group or individual, whether or not they made their contribution it is certain that something very similar and arguably also quite possibly better would have been “invented”.
The importance of this theory is that it makes “invention” or “discovery” subject to scientific investigation. This is because we do not attribute discovery to chance, but instead to inevitability. Therefore, the theory states that in the same conditions with enough people working on it, given a reasonable time, the same invention would always have come about.
Thus if we know what the technological environment was like before a change in technology, we can determine the likely “developmental pathway” – as the most likely route by which the technology changed step by step from one state to another.
This allows us to make a science out of understanding technological development. Because rather than accepting the vague and often unsupportable claims of “inventors” (like Al Gore invented the internet), we can create hypothesis as to what conditions led to changes and use experiment to determine the most enerconic pathway

An ensemble model to sustain variability power.

Today, I discussed the fiddler dilemma in which I discussed the issues for the data fiddler as they constantly cherry pick random data leaving little randomness left. I then explained how a similar problem besets ensemble forecasting of climate. Because climate models inject variation as initial conditions – and the “variability power” diminishes over time meaning that much of the “variability space” remains unexplored by their ensemble forecasts.
However, I now realise I may have been a bit of a hypocrite. Because in the theory of developmental pathways, I have constantly suggested that there is an optimum developmental route. In part this is the reverse of the fiddler’s dilemma – physics/engineering and practicality of the technological environment means that only certain pathways are feasible. And since the physics/engineering does not change, if we can reproduce the technological environment we should be able to find an optimum route.
However, what about “natural variation”. What about (using the example I gave in the fiddler’s dilemma of navigating a city) … what about the individual who by chance goes down a technological back alley?
Hence this note … which is really saying that rather than trying to find the one single “developmental pathway”, I should consider development as a group of possibilities from an ensemble of pathways … and that I shouldn’t just consider optimums but as I suggested re climate models, I should continually allow for the continuous vibrant natural variation in the way technology develops.
In other words – yes the optimum might be one way – but sometimes the suboptimum fluke variations like micro$oft wins out.

This entry was posted in Climate. Bookmark the permalink.