AI ALIGNMENT FORUM
AF

Wikitags

Robust Agents

Edited by Raymond Arnold, Ruben Bloom last updated 14th Sep 2020

Robust Agents are decision-makers who can perform well in a variety of situations. Whereas some humans rely on folk wisdom or instinct, and some AIs might be designed to achieve a narrow set of goals, a Robust Agent has a coherent set of values and decision-procedures. This enables them to adapt to new circumstances (such as succeeding in a new environment, or responding to a new strategy by a competitor).

See also

Subscribe
Subscribe
Agency
Discussion0
Discussion0
Posts tagged Robust Agents
77Humans are very reliable agents
Alyssa Vance
3y
0
54Embedded Agency (full-text version)
Scott Garrabrant, Abram Demski
7y
4
33Robust Delegation
Abram Demski, Scott Garrabrant
7y
2
23Gradations of Agency
Daniel Kokotajlo
3y
6
18What if memes are common in highly capable minds?
Q
Daniel Kokotajlo
5y
Q
6
45Robustness to Scale
Scott Garrabrant
7y
7
34Even Superhuman Go AIs Have Surprising Failure Modes
AdamGleave, EuanMcLean, Tony Wang, Kellin Pelrine, Tom Tseng, Yawen Duan, Joseph Miller, Michael Dennis
2y
9
412-D Robustness
Vladimir Mikulik
6y
1
16A multi-disciplinary view on AI safety research
Roman Leventov
2y
0
23Beyond the Board: Exploring AI Robustness Through Go
AdamGleave
1y
1
8[Aspiration-based designs] 2. Formal framework, basic algorithm
Jobst Heitzig, Simon Dima, Simon Fischer
1y
0
5Reward is not Necessary: How to Create a Compositional Self-Preserving Agent for Life-Long Learning
Roman Leventov
2y
0
3Temporally Layered Architecture for Adaptive, Distributed and Continuous Control
Roman Leventov
2y
0
Add Posts