From: avatar (avatar@renegadeclothing.com.au)
Date: Sun Jan 26 2003 - 09:40:07 MST
Robin Hanson:
--------------------------------------------------------------------------------
Thank you Robin for your article on Machine Intelligence and Economic Growth. I read it with interest. I wish I had a better grasp of algebra to understand it fully. [http://hanson.gmu.edu/aigrow.pdf]
When I did economics at the pre-university level I was thinking some of these things. My dad was a political scientist with a background in economics. He didn't think much of my ideas of self-reproducing modular automation in the 1970s or linear accelerators on the Moon! Of course, if technology had "frozen" with the space shuttle and rough tech of the time, it would have taken decades to get things going and a century or two to really get into the swing of things, but it was inevitable.
Some things gave me pause as a student.
a.. The shrinking working class and agricultural sector in numbers.
a.. The weirdly even distribution of unemployment in Western countries, particularly if adjusted. The way this began in the early 70s in overt form. The way western governments couldn't prevent it despite a variety of approaches. Inflation being so strong at the beginning of the 70s (even with Vietnam and the oil crisis).
a.. The growth in the service industries didn't make up for the fall. [Perhaps growth in material goods helped offset the potentially higher unemployment somewhat.]
a.. The shifting of displaced 3rd world labour from "cheapest country" to "newly cheapest country".
a.. The astounding steadiness and strength of growth in Western countries. [And its interaction with inflation rates.]
a.. The complete death of labour-intensive jobs (by 19th century standards). Skilled artisans, etc. Transport for example used up 30% of country labour in the 19th century (wheelwrights, saddlers, stablers, blacksmiths etc.)
a.. The role of automation in daily life especially since the 1910s (transport, irons, kettles, heaters, air conditioning, electrics, lights, washing machines, vacuum cleaners, fridges, toasters, phones etc.).
--------------------------------------------------------------------------------
I guess just some issues of debate are:
Damien Broderick and others: software coding slaves become the new paradigm
Drexler: molecular computing "engineering system" software design comes on line [one of my friends is an architect and I can already see such "software desgin" software is going to make more than half their work redundant]
Tipler and Moravec: AI and IA take over in managing intelligent production.
Me: despite their being many potential tasks (protein design possibilities exceed the number of particles in the universe one scientiest said) tradtional 20th century social existence for traditional humans can become an accessible software core very quickly. Just as a "poor" person in Australia can buy extremely cheap generic versions of everything the software for the "average modular house" will be available for the equivalent o f US$300, with some modularity and ability to modify and automation for construction [my architect said that tilt-slab construction is required because modern cement needs to be shaken in the mold off-site and then transported to site to set properly. When I told him about 3-D printers the first thing he said was "I wish I could print a house": a truly liquid concrete poured into a printed mold would be interesting for example. I guess you could assemble a house-print as a lattice structure.]
When I was in final high school and university no one seemed to be commenting on what was causing the unemployment or inflation. No one commented on growth or wealth distribution or inflation and causes. No one even commented on why the working class was shrinking or why service jobs did not expand beyond fast food etc. to domestic service. Half the things Alvin Toffler had raved about seemed to be coming true (more so since the Internet). I felt it must relate to technology. Doubtless people did comment about these things somewhere, but not in the library books or papers I read or the tv programmes I watched. There used to be a greater degree of truth in the notion that knowledge cost money, in the days before the Internet.
--------------------------------------------------------------------------------
I find things make more sense now I compile figures of global statistics on offshore banking, the total world and Western economices, total pies of expenditure and so on. Also (as you note) knowledge of population growth is important.
Your figure of 2 years for vast expansion of machine-driven growth in the 2020s has a weird correlation with arbitrary extrapolated Malthusian population growth (non-capped) [see the Kapitza pop. article. http://ufn.ioc.ac.ru/ufn96/ufn96_1/ufn61c.ps.gz].
--------------------------------------------------------------------------------
I have often tried to calculate real per person work activity in Victorian Australian society (5 million for all persons total pop.) by grabbing known numbers of professions and occupations and family structures etc.
I have tried to classify them by such boxes as:
agricultural-natural product management and production process
object moving/construction
writing/information management
service essential [e.g. teachers]
service non-essential [e.g. hairdressers]
part-time social work
full-time social work (raising children up to a certain age and percentage of week-time)
studying full-time
studying part-time
non-working
with extra categories and various arrangements etc. Finding figures for some occupations was very interesting, such as power production and engineering management [i.e. plumbing/roadworks/telecommunications etc.].
I don't have access to the exact results but from memory I think I concluded society could lurch along without 7 out of 8 members and still eat and drive cars! Real unemployment as a % of total actual population was well above standard figures.
--------------------------------------------------------------------------------
Your article will be severely modified by the effects of MEMS, santa claus machines, 3D printers, nanoassemblers and self-reproducing nanoassemblers. Also, your article mentions uploading but not augmentation. [As an aside, anyone wanting an hilarious SF read on machine production in a world without organic life - on Titan - should read Code of the Lifemaker/The Immortality Option by James P. Hogan. Or Mack Reynold's Tomorrow Might be Different for a Soviet Union high on automation.]
Thanks for putting the work in. I'll mention it to anyone I think could appreciate it, for what that's worth. I think I've told every single close and moderate friend of mine about Engines of Creation (always prefaced by "As important as Darwin" - a phrase I doubt I will ever retract).
Towards Ascension
Avatar Polymorph
34 After Armstrong
Maximum choice and minimum non-consensual force
This archive was generated by hypermail 2.1.5 : Sun Feb 02 2003 - 21:26:03 MST