Difference between revisions of "Continuous takeoff"
| (One intermediate revision by the same user not shown) | |||
| Line 1: | Line 1: | ||
| + | '''Continuous takeoff''' (also called '''slow takeoff''', '''soft takeoff''', or '''gradual takeoff''') is any of a number of views about [[AI takeoff]]. All views present an alternative to [[Eliezer Yudkowsky]]'s [[FOOM]] scenario, arguing that the transition from early AI systems to superintelligent systems will not be so immediate, but these views disagree on the details. | ||
* the word "continuous" in continuous takeoff does not mean the mathematical definition of continuity; even in a [[hard takeoff]] (i.e. "discontinuous takeoff") scenario, the AI system's capability can be modeled as a continuous function. | * the word "continuous" in continuous takeoff does not mean the mathematical definition of continuity; even in a [[hard takeoff]] (i.e. "discontinuous takeoff") scenario, the AI system's capability can be modeled as a continuous function. | ||
| Line 5: | Line 6: | ||
* [[Changing selection pressures argument]] | * [[Changing selection pressures argument]] | ||
| + | * [[Christiano's operationalization of slow takeoff]] | ||
[[Category:AI safety]] | [[Category:AI safety]] | ||
Latest revision as of 22:16, 1 March 2021
Continuous takeoff (also called slow takeoff, soft takeoff, or gradual takeoff) is any of a number of views about AI takeoff. All views present an alternative to Eliezer Yudkowsky's FOOM scenario, arguing that the transition from early AI systems to superintelligent systems will not be so immediate, but these views disagree on the details.
- the word "continuous" in continuous takeoff does not mean the mathematical definition of continuity; even in a hard takeoff (i.e. "discontinuous takeoff") scenario, the AI system's capability can be modeled as a continuous function.
pages about continuous takeoff:
