Re: Economic Growth Assuming Machine Intelligence

J. Maxwell Legg (
Mon, 05 Oct 1998 18:43:33 +1300

Doug Bailey wrote:
> One last point is the idea of efficiency. I read a paper a year or so ago
> in the Journal of Economic Perspectives (I'll try and locate it but can't at
> the moment) that stated some poor countries would do well not to worry about
> acquiring more of the right resources but instead work towards wasting less
> of their existing resources. What impact might a drastically increased
> efficiency in capital utilization have on growth models? How would this
> potential effect of AI-development be factored in to growth models?

I found one site that said the opposite condition applied to countries facing scarcity :

"Resource depletion and degradation in poor economies may have their most inimical
effect not by directly constraining growth but by indirectly affecting the potential of these economies to innovate."

For low and middle income countries, Deacon finds that

"the political variables associated with deforestation tend also to be negatively
associated with ordinary investment." In particular, the strongest (negative)
associations were between investment and guerilla warfare, revolutions, constitutional changes, military executives (i.e. dictatorships or juntas) and
circumstances in which the senior executive of government was not chosen by
elected representatives.

Nonetheless, it is my considered view that the simplistic measurement system of double entry bookkeeping is highly suspect as to it being the cause of this so-called affected potential. A public AI would not and could not make head nor tail of such information in determining how best to structure the public data needed to foster innovation. Thus capitalism, seen as a degradable economic lubricant, will be as foreign to a public AI as oil will be to a Drexlerian assembler. In the YAK fiasco monetary mechanisms will be seen as a hindrance and in the same way that NZ shelved its pounds shillings and pence in 1967 when they were seen as a machine problem and likewise when gold met a similar fate in 1971, in part because of Hewlett & Packard's advice to the government, public AI researchers with a keen eye on the problems of financial statistics are now calling for the next informational change of transactional integrity.