Difference between revisions of "Resource overhang"
(Created page with " ==Resource overhang and AI takeoff== Whether we are already in hardware overhang / other "resource bonanza". So whether we are in overhang depends on whether future earl...") |
(→Resource overhang and AI takeoff) |
||
| (3 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
| + | Does overhang just mean "not a bottleneck"? | ||
==Resource overhang and AI takeoff== | ==Resource overhang and AI takeoff== | ||
Whether we are already in [[hardware overhang]] / other "resource bonanza". So whether we are in overhang depends on whether future early AGI systems can make use of existing resources much more efficiently, which means overhang has a dependence on whether few insights are needed to get to AGI? More precisely, there's a feedback loop, where resource overhang amplifies any existing discontinuity. | Whether we are already in [[hardware overhang]] / other "resource bonanza". So whether we are in overhang depends on whether future early AGI systems can make use of existing resources much more efficiently, which means overhang has a dependence on whether few insights are needed to get to AGI? More precisely, there's a feedback loop, where resource overhang amplifies any existing discontinuity. | ||
| + | |||
| + | Andy Jones argues we are already in a resource overhang because GPT-3 could be scaled up to be much more capable but no company has tried to do so.<ref>https://www.greaterwrong.com/posts/N6vZEnCn6A95Xn39p/are-we-in-an-ai-overhang</ref> See [[scaling hypothesis]]. | ||
| + | |||
| + | ==External links== | ||
| + | |||
| + | * https://parallel-forecast.github.io/AI-dict/docs/dictionary.html#overhang | ||
[[Category:AI safety]] | [[Category:AI safety]] | ||
Latest revision as of 03:19, 24 February 2021
Does overhang just mean "not a bottleneck"?
Resource overhang and AI takeoff
Whether we are already in hardware overhang / other "resource bonanza". So whether we are in overhang depends on whether future early AGI systems can make use of existing resources much more efficiently, which means overhang has a dependence on whether few insights are needed to get to AGI? More precisely, there's a feedback loop, where resource overhang amplifies any existing discontinuity.
Andy Jones argues we are already in a resource overhang because GPT-3 could be scaled up to be much more capable but no company has tried to do so.[1] See scaling hypothesis.
