Previously I tried to explain how our economic system in combination with easy to condition humans creates an aware AI that does not have human wellbeing as its primary objective. In order to prevent total destruction of our habitat by the unbridled utilization of fossil fuels we need to alter the ‘software’ in a subtle way, namely by making our survival a factor, by demanding that priority is given to projects and activities that maximize the variety and abundance of life, not just maximize cashflow.
The implementation ofthe ML criterium has to be done in a competitive environment, so that projects can be viable but at the same time have better or worse performance in terms of ML. The way the resulting variety and abundance of life is determined requires specification for the system to work. First of all, what is a life?
A life is an independend biological process that replicates itself and has arrived in existence today through natural selection. A life has to have a natural habitat that predates ad 1800.
Second of all what life do we expect?
Any location on Earth will have suited lifeforms which evloved to inhabit it. However, life should be introduced where variety and abundance is low as long as it does not reduce the chance of life succes elsewhere f.i. due to climate effects.
On the surface we can easily understand that a project that clearcuts a piece of rainforest to build a refinery scores very low on the ML scale, and that the extraeconomic dedesertification of an arid region scores high. So perhaps several components can be separated out of the ML ‘index’.