This questions has been copied from a duplicate on the main Metaculus site.
Paul Christiano suggests operationalising AI takeoff speeds in terms of economic growth. A slow takeoff is one where the economy doubles in four years before the first time it doubles in one year, and a fast takeoff is one where it does not. (See the same article for Christiano's list of arguments regarding fast vs. slow takeoff.)
Will a 4-year GDP doubling complete (slow takeoff) before a 1-year doubling (fast takeoff) does?
Resolution is by credible assessment of world GDP in constant US dollars (or other appropriately widely-accepted units).
There will be an ambiguous resolution if there is no 4 year (nor 1 year) doubling interval by 2050, to isolate specifically the takeoff speed from other things like the chances of no takeoff occurring at all or human extinction.
As of 2018, it's taken about 16 years for the world economic output to double.