Agent Foundations for Aligning Machine Intelligence with Human Interests: A Technical Research Agenda
By Nate Soares, Benya Fallenstein. 2017
71 citations recorded by the source corpus.
Subfields: Computational Theory and Mathematics, Information Systems, Safety Research, Safety, Robustness & Interpretability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computational Theory and Mathematics
- Tom Everitt
- Jan Leike
- Shane Legg
- Jessica Taylor
- Yejin Choi
- Prithviraj Ammanabrolu
- Eliezer Yudkowsky
- Hannaneh Hajishirzi
In Information Systems
- David Krueger
- Yejin Choi
- Tom Everitt
- Prithviraj Ammanabrolu
- Hannaneh Hajishirzi
- Maarten Sap
- Shane Legg
- Mohit Bansal
In Safety Research
- Shane Legg
- Jessica Taylor
- Yejin Choi
- Tom Everitt
- Patrick LaVictoire
- Eliezer Yudkowsky
- Andrew Critch
- Miljan Martic
In Safety, Robustness & Interpretability
Citations in the corpus
- Scalable agent alignment via reward modeling: a research direction (2018, background): “Soares & Fallenstein (2017) outline a research agenda of a very different flavor.”
- Alignment for Advanced Machine Learning Systems (2020, background): “Soares and Fallenstein (2014) refer to the former approach as value specification , and the latter as error tolerance .”
- Aligning to Social Norms and Values in Interactive Narratives (2022, background): “ment learning (RL) in sequential decision making environments—are prone to learning behaviors harmful to themselves and their surroundings due to optimal task performance being often misaligned with socially beneficial human values (Moor, 2006; Soares and Fallenstein, 2017; Russell, 2019).”