LESSWRONGTags
LW

Robust Agents

•

Applied to [Aspiration-based designs] 2. Formal framework, basic algorithm by Jobst Heitzig 1mo ago

•

Applied to [Aspiration-based designs] 1. Informal introduction by Jobst Heitzig 1mo ago

•

Applied to AISC project: SatisfIA – AI that satisfies without overdoing it by Jobst Heitzig 6mo ago

•

Applied to Desiderata for an AI by Nathan Helm-Burger 10mo ago

•

Applied to Even Superhuman Go AIs Have Surprising Failure Modes by AdamGleave 10mo ago

•

Applied to Robustness to Scale by RaemonTest2 1y ago

•

Applied to A multi-disciplinary view on AI safety research by Roman Leventov 1y ago

•

Applied to Temporally Layered Architecture for Adaptive, Distributed and Continuous Control by Roman Leventov 1y ago

•

Applied to Reward is not Necessary: How to Create a Compositional Self-Preserving Agent for Life-Long Learning by Roman Leventov 1y ago

•

Applied to Sets of objectives for a multi-objective RL agent to optimize by Ben Smith 1y ago

•

Applied to Leveraging Legal Informatics to Align AI by John Nay 2y ago

•

Applied to Can we achieve AGI Alignment by balancing multiple human objectives? by Ben Smith 2y ago

•

Applied to Humans are very reliable agents by Alexei 2y ago

•

Applied to Gradations of Agency by Daniel Kokotajlo 2y ago

•

Applied to Thoughts on the 5-10 Problem by Multicore 3y ago

Ruby4y20

some AIs might be programmed with a narrow set of goals.

I don't think this precludes you from being an agent or should preclude you from being a "robust agent".

Ruby v1.6.0Sep 14th 2020 GMT (+14) 2

See also

Agency

Ruby4y20

I actually think this is definition is basically just what "agency" is.

2Ruby4y

I don't think this precludes you from being an agent or should preclude you from being a "robust agent".