Robin Hanson forecasting the next mode of growth.
Mode Doubling Date Began Doubles Doubles Grows Time (DT) To Dominate of DT of WP ---------- --------- ----------- ------ ------- Brain size 34M yrs 550M B.C. ? "~16" Hunters 230K yrs 2000K B.C. 7.2 8.7 Farmers 860 yrs 4700 B.C. 8.1 7.5 ?? 58 yrs 1730 3.9 3.2 Industry 15 yrs 1903 1.9 >6.3
What if was like the industrial transition.
2067-2120 timing estimate
first adopter gains New mode Initial % Final % Brains less than 1% 20 Humans 5 100 Farming 10 40 Industry 20 50 ? ? ?
Knowledge was always key
Five steps to Radical Production
1. Atomic Precision
2. General Plants – like PCs beat signal processor. Scale economy to make, easier design, efficient enough. More differentiation faster evolution of products
3. Local Production
4. Usuallky idle
5. Self reproduction
Lower marginal costs goes then Costco model may be better. Buy into a club to get lower price on a collection of items
Factor Shares of Income
Labor, human capital, other non-human capital, natural resources, stock dividends
My Views on Singularity
Yes General machine intelligence will make huge difference
Not soon – roughly 20 to 200 years
Not Trend – econ growth has been steady
not Local – an integrated economy grows together, not basement takes over world
Not hand coded – probably brain emulation
Not Horizon – we can see past if fuzzier
-new economy doubles weekly to monthly
– natural wages fall below
Economics of Robots
Staple of fiction
if have more of X, do not want Y more (complement) or less (substitute)?
Machine as substitute to human labor
-ricardo 1821, most science fiction
-wages fall to machine cost
Automation as complement to human labor
-wicksell 1923, modern econmomics consensus
-wages have risen as automation cost have fallen
So are robots a substitute or complement?
Robots substitute on task but tasks are complements
complement tasks – if certain tasks become nearly free or lower value then the remaining tasks become more valuable.
The effect of a rising tide. the importance of the shape of the curve of tasks versus human advantage.
A simple robot growth model
complex growth model – many things shrinking or growing
Hanson has worked on AI before and thinks it was hard and is still hard.
Thinks whole brain emulation is hard but doable.
Pivotal : What Ready Last
– other techs fast or fine brain detail key
– broad smooth anticipated transition
2. Scanning (least likely)
– large coalitions, first dominates, diversify
– most in future descend from one human
– may be big surprise so disruptive change.
Cheap for robots
* Immortal (even so most can’t afford)
* travel – transmit to new body (but security)
* Nature – don’t need ecosystems
* Labor – work less tool intensive
– malthusian population explosion, rapid growth
– Wages may fall to fast falling hardware cost
+ depends on mental task landscape shape
+ happens if they slave or if free
+ only draconian population wage
Emulations feel human
*They remember a human life
*retain human tendencies
-love, gossip, argue, sing, violate,play, work, innovate
*more alientated worlds – as were farms, factories
– office work in VR
* Wages well below human subsistence
– some humans may find servant jobs
* but rich if held non-age assets
*copies rent bodies or own on loan
– evicted if can’t pay
* to recoup trainin invetment copy cabal limits copy wage
– security to prevent bootleg copies
* fast growth discourages transport encourages local prodution