[ExI] [mta] Re: Bitcion Moore's Law?

[Anders Sandberg]

On 2013-07-16 06:41, Brent Allsop wrote:
> Lots of Lauphs!  You expect to see something other than noise in 10 year
> graphs?  Did you see the graph on this page that at least covers 100 years?
>
> http://www.numbersleuth.org/worlds-gold/

That graph shows that production has increased by a factor of ~5 over a 
century. If that is exponential, the doubling time is many decades. Gold 
price might have rushed over the past ~15 years with a doubling time of 
5 years, but the past curve is fairly bumpy and with big plateaus - 
definitely no constant doubling time.

Now, given that prices are subjective evaluations rather than objective 
value, there is no physical law that precludes eternal exponential price 
growth even of a finite supply of something. Or constant, declining or 
oscillating prices. But the industrial side of gold production seems to 
be improving very slowly: my guess is that high capital costs and the 
benefits of restricted supply are disincentives to improve it. We do not 
get much of Wrightean learning leading to a Sahal's law exponential 
improvement in productivity.

The same might be true for bitcoin: if looking for coins requires bigger 
and bigger investments the productivity per unit of effort will not 
increase (unless some mathematician comes up with something clever and 
blows up the whole market...) But then again, the mining is just a 
clever part to motivate people and maintain a fairly fixed supply 
without central coordination: the eventual viability of the currency is 
pretty independent of it.



-- 
Anders Sandberg
Future of Humanity Institute
Oxford University