A5009631016- Evolutionary Game Theory and Cooperation
- Ethics and Social Impacts of AI
- Reinforcement Learning in Robotics
- Language and cultural evolution
- Multi-Agent Systems and Negotiation
- Embodied and Extended Cognition
- Social Robot Interaction and HRI
- Neural dynamics and brain function
- Neuroethics, Human Enhancement, Biomedical Innovations
- Memory and Neural Mechanisms
- Robot Manipulation and Learning
- Artificial Intelligence in Games
- Child and Animal Learning Development
- Experimental Behavioral Economics Studies
- Cognitive Science and Mapping
- Psychology of Moral and Emotional Judgment
- AI-based Problem Solving and Planning
- Law, AI, and Intellectual Property
- Opinion Dynamics and Social Influence
- Primate Behavior and Ecology
- Evolutionary Algorithms and Applications
- Neural and Behavioral Psychology Studies
- Action Observation and Synchronization
- Culture, Economy, and Development Studies
- Evolutionary Psychology and Human Behavior
Hertie School
2019-2024
University of Bath
2012-2022
Princeton University
2016-2022
Center for Information Technology
2016-2022
University of Edinburgh
1997-2019
Applied Mathematics (United States)
2015
Engineering and Physical Sciences Research Council
2015
Bath Spa University
2014
Konrad Lorenz Institute for Evolution and Cognition Research
2007-2009
Massachusetts Institute of Technology
1995-2006
Artificial intelligence and machine learning are in a period of astounding growth. However, there concerns that these technologies may be used, either with or without intention, to perpetuate the prejudice unfairness unfortunately characterizes many human institutions. Here we show for first time human-like semantic biases result from application standard ordinary language---the same sort language humans exposed every day. We replicate spectrum as by Implicit Association Test other...
This report surveys the landscape of potential security threats from malicious uses AI, and proposes ways to better forecast, prevent, mitigate these threats. After analyzing in which AI may influence threat digital, physical, political domains, we make four high-level recommendations for researchers other stakeholders. We also suggest several promising areas further research that could expand portfolio defenses, or attacks less effective harder execute. Finally, discuss, but do not...
Conferring legal personhood on purely synthetic entities is a very real possibility, one under consideration presently by the European Union. We show here that such legislative action would be morally unnecessary and legally troublesome. While AI may have some emotional or economic appeal, so do many superficially desirable hazards against which law protects us. review utility history of fictions personhood, discussing salient precedents where resulted in abuse incoherence. conclude...
Artificial intelligence and machine learning capabilities are growing at an unprecedented rate. These technologies have many widely beneficial applications, ranging from translation to medical image analysis. Countless more such applications being developed can be expected over the long term. Less attention has historically been paid ways in which artificial used maliciously. This report surveys landscape of potential security threats malicious uses technologies, proposes better forecast,...
AI is here now, available to anyone with access digital technology and the Internet. But its consequences for our social order aren't well understood. How can we guide way impacts society?
The question of whether AI systems such as robots can or should be afforded moral agency patiency is not one amenable either to discovery simple reasoning, because we societies constantly reconstruct our artefacts, including ethical systems. Consequently, the place in society a matter normative, descriptive ethics. Here I start from functionalist assumption, that ethics set behaviour maintains society. This assumption allows me exploit theoretical biology sociality and autonomy explain...
This paper proposes a set of five ethical principles, together with seven high-level messages, as basis for responsible robotics. The Principles Robotics were drafted in 2010 and published online 2011. Since then the principles have influenced, continue to influence, number initiatives robot ethics but not, date, been formally published. remedies that omission.
The EPSRC's Principles of Robotics advises the implementation transparency in robotic systems, however research related to AI is its infancy. This paper introduces reader importance having transparent inspection intelligent agents and provides guidance for good practice when developing such agents. By considering expanding upon other prominent definitions found literature, we provide a robust definition as mechanism expose decision-making robot. continues by addressing potential design...
Social and political polarization is a significant source of conflict poor governance in many societies. Thus, understanding its causes has become priority scholars across disciplines. Here we demonstrate that shifts socialization strategies analogous to identity politics can arise as locally-beneficial response both rising wealth inequality economic decline. Adopting perspective cultural evolution, develop framework study the emergence under shifting environments. In contexts, interacting...
A scientific methodology in general should provide two things: first, a means of explanation and, second, mechanism for improving that explanation. Agent-based modelling (ABM) is method facilitates exploring the collective effects individual action selection. The explanatory force model extent to which an observed meta-level phenomenon can be accounted by behaviour its micro-level actors. This article demonstrates this applied biological sciences; agent-based models, like any other...
Autonomous agent architectures are design methodologies—collections of knowledge and strategies which applied to the problem creating situated intelligence. This article attempts integrate this across several architectural traditions, paying particular attention features have tended be selected under pressure extensive use in real-world systems. We determine that following provide significant assistance autonomous intelligent agents: (i) modularity, simplifies both control; (ii)...
Punishing group members who parasitize their own group's resources is an almost universal human behavior, as evidenced by multiple cross-cultural and theoretical studies.Recently, researchers in social behavioral sciences have identified a puzzling phenomenon called "antisocial punishment": some people are willing to pay cost "punish" those act ways that benefit shared group.Interestingly, the expression of antisocial punishment behavior regionally diverse linked socio-psychological...
OPINION article Front. Comput. Sci., 18 May 2023Sec. Human-Media Interaction Volume 5 - 2023 | https://doi.org/10.3389/fcomp.2023.1210421
Abstract Religion is, at the very least, a highly complex social phenomenon. The theories we use to understand religion – and sociocultural systems more generally are often so that even experts in field may not be able see all their consequences. Social simulations can help us communicate consequences of theory, provided describe theory with sufficient precision comprehensiveness order run it on computer. In this article demonstrate benefits simulating predictions well-known Cognitive...
Deciphering the behaviour of intelligent others is a fundamental characteristic our own intelligence. As we interact with complex artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these are incorrect or inadequate, run risk self deception even harm. Here demonstrate that providing simple, abstracted real-time visualisation robot's AI can radically improve transparency machine cognition. Findings from both an online experiment using video...
Real time planning strategy is crucial for robots working in dynamic environments. In particular, robot grasping tasks require quick reactions many applications such as human-robot interaction. this paper, we propose an approach grasp learning that enables to plan new grasps rapidly according the object's position and orientation. This achieved by taking a three-step approach. first step, compute variety of stable given object. second learns probability distribution based on computed grasps....