As the capabilities of artificial intelligence systems improve, it becomes important to constrain their actions to ensure their behaviour remains beneficial to humanity. A variety of ethical, legal and safety-based frameworks have been proposed as a basis for designing these constraints. Despite their variations, these frameworks share the common characteristic that decision-making must consider multiple potentially conflicting factors. We demonstrate that these alignment frameworks can be represented as utility functions, but that the widely used Maximum Expected Utility paradigm provides insufficient (...) support for such multiobjective decision-making. We show that a Multiobjective Maximum Expected Utility paradigm based on the combination of vector utilities and non-linear action–selection can overcome many of the issues which limit MEU’s effectiveness in implementing aligned AI. We examine existing approaches to multiobjective AI, and identify how these can contribute to the development of human-aligned intelligent agents. (shrink)

Systems based on symbolic knowledge have performed extremely well in processing reason, yet, remain beset with problems of brittleness in many domains. Connectionist approaches do similarly well in emulating interactive domains, however, have struggled when modelling higher brain functions. Neither of these dichotomous approaches, however, have provided many inroads into the area of human reasoning that psychology and sociology refer to as the process of practice. This paper argues that the absence of a model for the process of practise in (...) current approaches is a significant contributor to brittleness. This paper will investigate how the process of practise relates to deeper forms of contextual representations of knowledge. While researchers and developers of knowledge based systems have often incorporated the notion of context they treat context as a static entity, neglecting many connectionists’ work in learning hidden and dynamic contexts. This paper argues that the omission of these higher forms of context is one of the fundamental problems in the application and interpretation of symbolic knowledge. Finally, these ideas for modelling context will lead to the reinterpretation of situation cognition which makes a significant step towards a philosophy of knowledge that could lead to the modelling of the process of practice. (shrink)