S. Kate Devitt

Humans and artificial intelligences (AI) will increasingly participate digitally and physically in conflicts yet there is a lack of trusted communications across agents and platforms. For example, humans in disasters and conflict already use messaging and social media to share information, however, international humanitarian relief organisations treat this information as unverifiable and untrustworthy. AI may reduce the ‘fog-of-war’ and improve outcomes, however current AI implementations are often brittle, have a narrow scope of application and wide ethical risks. Meanwhile, human error causes significant civilian harms even by combatants committed to complying with international humanitarian law. AI offers an opportunity to help reduce the tragedy of war and better deliver humanitarian aid to those who need it. However, to be successful, these systems must be trusted by humans and their information systems, overcoming flawed information flows in conflict and disaster zones that continue to be marked by intermittent communications, poor situation awareness, mistrust and human errors. In this paper, we consider the integration of a communications protocol (the ‘Whiteflag protocol’), distributed ledger technology, and information fusion with artificial intelligence (AI), to improve conflict communications called “Protected Assurance Understanding Situation & Entities” PAUSE. Such a trusted human-AI communication network could provide accountable information exchange regarding protected entities, critical infrastructure; humanitarian signals and status updates for humans and machines in conflicts. Trust-based information fusion provides resource-efficient use of diverse data sources to increase the reliability of reports. AI can be used to catch human mistakes and complement human decision making, while human judgment can direct and override AI recommendations. We examine several case studies for the integration of these technologies into a trusted human-AI network for humanitarian benefit including mapping a conflict zone with civilians and combatants in real time, preparation to avoid incidents and using the network to manage misinformation.

Australia is a leading AI nation with strong allies and partnerships. Australia has prioritised the development of robotics, AI, and autonomous systems to develop sovereign capability for the military. Australia commits to Article 36 reviews of all new means and method of warfare to ensure weapons and weapons systems are operated within acceptable systems of control. Additionally, Australia has undergone significant reviews of the risks of AI to human rights and within intelligence organisations and has committed to producing ethics guidelines and frameworks in Security and Defence. Australia is committed to OECD’s values-based principles for the responsible stewardship of trustworthy AI as well as adopting a set of National AI ethics principles. While Australia has not adopted an AI governance framework specifically for Defence; Defence Science has published ‘A Method for Ethical AI in Defence’ (MEAID) technical report which includes a framework and pragmatic tools for managing ethical and legal risks for military applications of AI.

Effective robots and autonomous systems must be trustworthy. This chapter examines models of trustworthiness from a philosophical and empirical perspective to inform the design and adoption of autonomous systems. Trustworthiness is a property of trusted agents or organisations that engenders trust in other agent or organisations. Trust is a complex phenomena defined differently depending on the discipline. This chapter aims to bring different approaches under a single framework for investigation with three sorts of questions: Who or what is trustworthy?–metaphysics. How do we know who or what is trustworthy?–epistemology. What factors influence what or who should we trust?–normativity. A two-component model of trust is used that incorporates competence (skills, reliability and experience) and integrity (motives, honesty and character). It is supposed that human levels of competence yield the highest trust whereas trust is reduced at sub-human and super-human levels. The threshold for trustworthiness of an agent or organisation in a particular context is a function of their relationship with the truster and potential impacts of decisions. Building trustworthy autonomous systems requires obeying the norms of logic, rationality and ethics under pragmatic constraints–even though there is disagreement on these principles by experts. Autonomous systems may need sophisticated social identities including empathy and reputational concerns to build human-like trust relationships. Ultimately transdisciplinary research drawing on metaphysical, epistemological and normative human and machine theories of trust are needed to design trustworthy autonomous systems for adoption.

Moving away from the strong body of critique of pervasive ‘bad data’ practices by both governments and private actors in the globalized digital economy, this book aims to paint an alternative, more optimistic but still pragmatic picture of the datafied future. The authors examine and propose ‘good data’ practices, values and principles from an interdisciplinary, international perspective. From ideas of data sovereignty and justice, to manifestos for change and calls for activism, this collection opens a multifaceted conversation on the kinds of futures we want to see, and presents concrete steps on how we can start realizing good data in practice.

Robotic and Autonomous Agricultural Technologies (RAAT) are increasingly available yet may fail to be adopted. This paper focusses specifically on cognitive factors that affect adoption including: inability to generate trust, loss of farming knowledge and reduced social cognition. It is recommended that agriculture develops its own framework for the performance and safety of RAAT drawing on human factors research in aerospace engineering including human inputs (individual variance in knowledge, skills, abilities, preferences, needs and traits), trust, situational awareness and cognitive load. The kinds of cognitive impacts depend on the RAATs level of autonomy, ie whether it has automatic, partial autonomy and autonomous functionality and stage of adoption, ie adoption, initial use or post-adoptive use. The more autonomous a system is, the less a human needs to know to operate it and the less the cognitive load, but it also means farmers have less situational awareness about on farm activities that in turn may affect strategic decision-making about their enterprise. Some cognitive factors may be hidden when RAAT is first adopted but play a greater role during prolonged or intense post-adoptive use. Systems with partial autonomy need intuitive user interfaces, engaging system information, and clear signaling to be trusted with low level tasks; and to compliment and augment high order decision-making on farm.

Governments must keep agricultural systems free of pests that threaten agricultural production and international trade. Biosecurity surveillance already makes use of a wide range of technologies, such as insect traps and lures, geographic information systems, and diagnostic biochemical tests. The rise of cheap and usable surveillance technologies such as remotely piloted aircraft systems presents value conflicts not addressed in international biosurveillance guidelines. The costs of keeping agriculture pest-free include privacy violations and reduced autonomy for farmers. We argue that physical and digital privacy in the age of ubiquitous aerial and ground surveillance is a natural right to allow people to function freely on their land. Surveillance methods must be co-created and justified through using ethically defensible processes such as discourse theory, value-centred design and responsible innovation to forge a cooperative social contract between diverse stakeholders. We propose an ethical framework for biosurveillance activities that balances the collective benefits for food security with individual privacy: establish the boundaries of a biosurveillance social contract; justify surveillance operations for the farmers, researchers, industry, the public and regulators; give decision makers a reasonable measure of control over their personal and agricultural data; and choose surveillance methodologies that give the appropriate information. The benefits of incorporating an ethical framework for responsible biosurveillance innovation include increased participation and accumulated trust over time. Long term trust and cooperation will support food security, producing higher quality data overall and mitigating against anticipated information gaps that may emerge due to disrespecting landholder rights.

This paper examines the role of coherence as a source of epistemic justification, particularly the argument that all beliefs must cohere within one’s ‘web of belief’, aka confirmational holism. Confirmational holism runs across a potentially devastating argument that a more coherent set of beliefs resulting from the addition of a belief to a less coherent set of beliefs is less likely to be true than the less coherent set of beliefs. I propose confirmational chorism to avoid this troubling outcome. CC posits that coherence adds epistemic justification by limited, logically consistent sets of beliefs exhibiting a satisficing degree of strength, inferential and explanatory connection. Limited coherence may resolve the above argument, but raises the need for another kind of justification: coordination. Belief coordination requires suppressing some beliefs and communicating other beliefs to ensure convergence on the right action for performance success. Thus, a belief in any particular context is justified not just because it is reliably formed and coherent, but also because of how it is coordinated between local and holistic goals.

Coherence between propositions promises to fix the vexing circumstance of prior probabilities for subjective Bayesians. This paper examines the role of coherence as a source of justification for Bayesian agents, particularly the argument that all propositions must cohere within an agent’s ‘web of belief’, aka confirmational holism. Unfortunately, Confirmational holism runs across a potentially devastating argument that a more coherent set of beliefs resulting from the addition of a belief to a less coherent set of beliefs is less likely to be true than the less coherent set of beliefs. In response, I propose confirmational chorism to avoid this troubling outcome. CC posits that coherence adds epistemic justification by limited, logically consistent sets of beliefs exhibiting a satisficing degree of strength, inferential and explanatory connection. Limited coherence may resolve the above argument, but raises the need for another kind of justification: coordination. Belief coordination requires suppressing some beliefs and communicating other beliefs to ensure convergence on the right action for performance success. Thus, a Bayesian formed belief in any particular context is justified not just because it is reliably formed and coherent, but also because of how it is coordinated between local and holistic goals.

This research addresses the justificatory role of conscious reflection within a naturalized, reliabilist epistemology. Reliabilism is the view that implicit, mechanistic processes can justify beliefs, e.g. perceptual beliefs formed after a history of consistent exposure to normal lighting conditions are justified in a given context with normal lighting. A popular variant of reliabilism is virtue epistemology where the cognitive circumstances and abilities of an agent play a justificatory role, e.g. the cooperation of the prefrontal cortex and primary visual cortex of the individual perceiving the Müller-Lyer illusion partly justify the belief that the lines are equi-length. While virtue epistemology is a well-endorsed reliabilism for implicit beliefs, its application to explicit, consciously reflective processes is more controversial. Critics ask: How can iterations of dumb reliabilist processes produce higher order justification? To respond to this concern, I draw on another agent-centred, normative and reliabilist epistemology—Bayesian epistemology. A Bayesian virtue epistemology argues that reflective hypothesis-testing generated by implicit Bayesian mechanisms offers higher order reliabilist justification for beliefs. Iterative Bayesian mechanisms explain the development of higher order beliefs about abstract concepts such as causation, natural laws and theoretical entities traditionally explained by recourse to vague concepts such as ‘the a priori’, ‘intuition’ or ‘the intellect’. A hybrid Bayesian virtue epistemology offers an iterative reliabilist framework to explain how conscious reflection justifies beliefs. However, I acknowledge limitations on Bayesian accounts of justification such as confirmational holism, commutativity, and the frame problem.

_Free to read on publishers website_ In normal child development, both individual and group pretense first emerges at approximately two years of age. The metarepresentational account of pretense holds that children already have the concept PRETEND when they first engage in early group pretense. A behavioristic account suggests that early group pretense is analogous to early beliefs or desires and thus require no mental state concepts. I argue that a behavioral account does not explain the actual behavior observed in children and it cannot explain how children come to understand that a specific action is one of pretense versus one of belief. I conclude that a mentalistic explanation of pretense best explains the behavior under consideration.

My research investigates why nouns are learned disproportionately more frequently than other kinds of words during early language acquisition. This question must be considered in the context of cognitive development in general. Infants have two major streams of environmental information to make meaningful: perceptual and linguistic. Perceptual information flows in from the senses and is processed into symbolic representations by the primitive language of thought. These symbolic representations are then linked to linguistic input to enable language comprehension and ultimately production. Yet, how exactly does perceptual information become conceptualized? Although this question is difficult, there has been progress. One way that children might have an easier job is if they have structures that simplify the data. Thus, if particular sorts of perceptual information could be separated from the mass of input, then it would be easier for children to refer to those specific things when learning words. It would be easier still, if linguistic input was segmented in predictable ways Unfortunately the frequency of patterns in lexical or grammatical input cannot explain the cross-cultural and cross-linguistic tendency to favor nouns over verbs and predicates. There are three examples of this failure: 1) a wide variety of nouns are uttered less frequently than a smaller number of verbs and yet are learnt far more easily ; 2) word order and morphological transparency offer no insight when you contrast the sentence structures and word inflections of different languages and 3) particular language teaching behaviors have little impact on children's tendency to prefer concrete nouns in their first fifty words. Although the linguistic solution appears problematic, there has been increasing evidence that the early visual system does indeed segment perceptual information in specific ways before the conscious mind begins to intervene. I argue that nouns are easier to learn because their referents directly connect with innate features of the perceptual faculty. This hypothesis stems from work done on visual indexes by Zenon Pylyshyn. Pylyshyn argues that the early visual system segments perceptual data into pre-conceptual proto-objects called FINSTs. FINSTs typically correspond to physical things such as Spelke objects. Hence, before conceptualization, visual objects are picked out by the perceptual system demonstratively, like a finger pointing indicating ‘this’ or ‘that’. I suggest that this primitive system of demonstration elaborates on Gareth Evan's theory of nonconceptual content. Nouns are learnt first because their referents attract demonstrative visual indexes. This theory also explains why infants less often name stationary objects such as plate or table, but do name things that attract the focal attention of the early visual system, i.e., small objects that move, such as ‘dog’ or ‘ball’. This view leaves open the question how blind children learn words for visible objects and why children learn category nouns, rather than proper nouns or higher taxonomic distinctions.

Optimal decision-making requires us to accurately pinpoint the basis of our thoughts, e.g. whether they originate from our memory or our imagination. This paper argues that the phenomenal qualities of our subjective experience provide permissible evidence to revise beliefs, particularly as it pertains to memory. I look to the source monitoring literature to reconcile circumstances where mnemic beliefs and mnemic qualia conflict. By separating the experience of remembering from biological facts of memory, unusual cases make sense, such as memory qualia without memory or a failure to have memory qualia with memory. I argue that a pragmatic, probabilistic approach to belief revision is a way to rationally incorporate information from conscious experience, whilst acknowledging its inherent difficulties as an epistemic source. I conclude with a Bayesian defense of source monitoring based on C.I. Lewis’ coherence argument for memorial knowledge.

How do agents with limited cognitive capacities flourish in informationally impoverished or unexpected circumstances? Aristotle argued that human flourishing emerged from knowing about the world and our place within it. If he is right, then the virtuous processes that produce knowledge, best explain flourishing. Influenced by Aristotle, virtue epistemology defends an analysis of knowledge where beliefs are evaluated for their truth and the intellectual virtue or competences relied on in their creation. However, human flourishing may emerge from how degrees of ignorance are managed in an uncertain world. Perhaps decision-making in the shadow of knowledge best explains human wellbeing—a Bayesian approach? In this dissertation I argue that a hybrid of virtue and Bayesian epistemologies explains human flourishing—what I term homeostatic epistemology. Homeostatic epistemology supposes that an agent has a rational credence p when p is the product of reliable processes aligned with the norms of probability theory; whereas an agent knows that p when a rational credence p is the product of reliable processes such that: 1) p meets some relevant threshold for belief, 2) p coheres with a satisficing set of relevant beliefs and, 3) the relevant set of beliefs is coordinated appropriately to meet the integrated aims of the agent. Homeostatic epistemology recognizes that justificatory relationships between beliefs are constantly changing to combat uncertainties and to take advantage of predictable circumstances. Contrary to holism, justification is built up and broken down across limited sets like the anabolic and catabolic processes that maintain homeostasis in the cells, organs and systems of the body. It is the coordination of choristic sets of reliably produced beliefs that create the greatest flourishing given the limitations inherent in the situated agent.