Avoiding Unintended AI Behaviors

32 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Research Priorities for Robust and Beneficial Artificial Intelligence (2015, background): “For example, it may be possible to design utility functions or decision processes so that a system will not try to avoid being shut down or repurposed [80], and theoretical frameworks could be developed to better understand the space of potential systems that avoid undesirable behaviors [81–83].”
Ethical Decision Making during Automated Vehicle Crashes (2014, methodology): “Hibbard, in the formulation of a consequentialist approach to machine ethics, proposed a similar method by which an independent artificial intelligence agent calculated the moral weights assigned by humans after study participants were polled across a wide range of hypothetical situations (46).”
Self-Modeling Agents and Reward Generator Corruption (2015, methodology): “An alternative, in our finite universe (Lloyd 2002), is to model environments using finite stochastic loop programs (Hibbard 2012a; 2014b).”
Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2009, methodology): “…to develop from its current form (for discussion of various issues regarding the ethics of advanced AGI see (Goertzel and Pitt, 2012; Bostrom, 2014; Hibbard, 2012; Yudkowsky, 2008)) Anyone familiar with the current state of AGI research will ﬁnd it hard to suppress a smile at this ambitious…”