How could we make a social robot? A virtual bargaining approach

Chater, Nick

doi:10.1098/rsta.2022.0040

Cited by 6 publications

(6 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Of course, there are many questions beyond only whether LLMs can serve as models of meaning. As [83] argues, 'intelligence is inherently social, and artificial systems, like humans, will be viewed as intelligent, and will be valued as collaborators, to the extent that they can align and coordinate their thoughts and actions with human thoughts and actions'. It may well be the case that showing that LLMs exhibit intelligence is a higher bar than showing they encode meaning, and it may be that communicative intent and other types of grounding will play a crucial role in such debates.…”

Section: (C) Summary and Discussionmentioning

confidence: 99%

Section: (I) Does Meaning Require Communicative Intent?mentioning

confidence: 96%

“…Human intelligence is inherently collective: the creation of language, conventions, norms, systems of religious and scientific belief, technologies and institutions and organisations of all kinds are far beyond the powers of any individual human mind. Rather, we are like termites each making only a small local contribution to a vast edifice whose extent and complexity far exceeds our understanding [83].…”

Section: (I) Does Meaning Require Communicative Intent?mentioning

confidence: 99%

“…Essentially: we know a guy who knows a guy who knows a guy who has directly observed the referent. Chater [83] echoes this argument (ironically, in the context of arguing for a need for communicative intent):Human intelligence is inherently collective: the creation of language, conventions, norms, systems of religious and scientific belief, technologies and institutions and organisations of all kinds are far beyond the powers of any individual human mind. Rather, we are like termites each making only a small local contribution to a vast edifice whose extent and complexity far exceeds our understanding [83].…”

Section: Groundingmentioning

confidence: 99%

“…Chater [83] echoes this argument (ironically, in the context of arguing for a need for communicative intent):Human intelligence is inherently collective: the creation of language, conventions, norms, systems of religious and scientific belief, technologies and institutions and organisations of all kinds are far beyond the powers of any individual human mind. Rather, we are like termites each making only a small local contribution to a vast edifice whose extent and complexity far exceeds our understanding [83]. Adopting this collective view of human intelligence, language models reading text on the internet, which was written by a human, are as good a member of the causal chain as anyone.…”

Section: Groundingmentioning

confidence: 99%

See 4 more Smart Citations

Symbols and grounding in large language models

Pavlick

2023

Phil. Trans. R. Soc. A.

View full text Add to dashboard Cite

Large language models (LLMs) are one of the most impressive achievements of artificial intelligence in recent years. However, their relevance to the study of language more broadly remains unclear. This article considers the potential of LLMs to serve as models of language understanding in humans. While debate on this question typically centres around models’ performance on challenging language understanding tasks, this article argues that the answer depends on models’ underlying competence, and thus that the focus of the debate should be on empirical work which seeks to characterize the representations and processing algorithms that underlie model behaviour. From this perspective, the article offers counterarguments to two commonly cited reasons why LLMs cannot serve as plausible models of language in humans: their lack of symbolic structure and their lack of grounding. For each, a case is made that recent empirical trends undermine the common assumptions about LLMs, and thus that it is premature to draw conclusions about LLMs’ ability (or lack thereof) to offer insights on human language representation and understanding. This article is part of a discussion meeting issue ‘Cognitive artificial intelligence’.

show abstract

Section: (C) Summary and Discussionmentioning

confidence: 99%

Section: (I) Does Meaning Require Communicative Intent?mentioning

confidence: 96%

Section: (I) Does Meaning Require Communicative Intent?mentioning

confidence: 99%

Section: Groundingmentioning

confidence: 99%

Section: Groundingmentioning

confidence: 99%

See 3 more Smart Citations

Symbols and grounding in large language models

Pavlick

2023

Phil. Trans. R. Soc. A.

View full text Add to dashboard Cite

show abstract

Beyond Preferences in AI Alignment

Zhi-Xuan,

Carroll,

Franklin

et al. 2024

Philos Stud

View full text Add to dashboard Cite

The dominant practice of AI alignment assumes (1) that preferences are an adequate representation of human values, (2) that human rationality can be understood in terms of maximizing the satisfaction of preferences, and (3) that AI systems should be aligned with the preferences of one or more humans to ensure that they behave safely and in accordance with our values. Whether implicitly followed or explicitly endorsed, these commitments constitute what we term a preferentist approach to AI alignment. In this paper, we characterize and challenge the preferentist approach, describing conceptual and technical alternatives that are ripe for further research. We first survey the limits of rational choice theory as a descriptive model, explaining how preferences fail to capture the thick semantic content of human values, and how utility representations neglect the possible incommensurability of those values. We then critique the normativity of expected utility theory (EUT) for humans and AI, drawing upon arguments showing how rational agents need not comply with EUT, while highlighting how EUT is silent on which preferences are normatively acceptable. Finally, we argue that these limitations motivate a reframing of the targets of AI alignment: Instead of alignment with the preferences of a human user, developer, or humanity-writ-large, AI systems should be aligned with normative standards appropriate to their social roles, such as the role of a general-purpose assistant. Furthermore, these standards should be negotiated and agreed upon by all relevant stakeholders. On this alternative conception of alignment, a multiplicity of AI systems will be able to serve diverse ends, aligned with normative standards that promote mutual benefit and limit harm despite our plural and divergent values.

show abstract