Difference between revisions of "Rapid capability gain vs AGI progress"
| Line 3: | Line 3: | ||
these two notions might eventually coincide, if the AI is the one doing AI research (so the state of the art is pushed upward as the AI system itself improves over time) | these two notions might eventually coincide, if the AI is the one doing AI research (so the state of the art is pushed upward as the AI system itself improves over time) | ||
| + | |||
| + | ==See also== | ||
| + | |||
| + | * [[Counterfactual of dropping a seed AI into a world without other capable AI]] | ||
==External links== | ==External links== | ||
* https://www.greaterwrong.com/posts/PzAnWgqvfESgQEvdg/any-rebuttals-of-christiano-and-ai-impacts-on-takeoff-speeds/answer/TCqBaYgoJYhbayqFp | * https://www.greaterwrong.com/posts/PzAnWgqvfESgQEvdg/any-rebuttals-of-christiano-and-ai-impacts-on-takeoff-speeds/answer/TCqBaYgoJYhbayqFp | ||
Revision as of 01:54, 24 February 2020
- rapid capability gain seems to refer to how well a single AI system improves over time, e.g. AlphaGo going from "no knowledge of anything" to "superhuman at Go" in a short span of time as it trained.
- AGI progress refers to progress of AI systems in general, i.e. if you plot the state of the art over time
these two notions might eventually coincide, if the AI is the one doing AI research (so the state of the art is pushed upward as the AI system itself improves over time)
