13,229 research outputs found
Human-AI Collaboration in Academic Writing: towards a Synergy Model and A Case to Include AI as a Co-Author
No full textAs generative AI systems such as ChatGPT and Gemini 2.5 become increasingly integrated into academic workflows, the question of their legitimacy, limitations, and potential in scholarly writing has become urgent. This paper presents a reflexive case study of a sustained collaboration between a domain expert in consciousness studies and Gemini 2.5, culminating in the co-authorship of a peer-reviewed research article. By analyzing exactly 37,440 words of recorded interactions, we identify patterns of synergy, including recursive refinement, conceptual amplification, and accelerated manuscript development. We argue that when guided by a knowledgeable human author, AI can act as a cognitive partner rather than a passive tool—amplifying scholarly creativity and improving efficiency without compromising academic rigor. The case supports a '1+1=3' synergy model for co-authorship, in which human steering and AI fluency converge to produce novel insights and polished output faster and more effectively than either could achieve alone. The findings advocate for a paradigm shift from prohibitive policies to the responsible, expert-guided integration of AI in academic research and writing, grounded in transparency and accountability, and present arguments for why the AI tool should be listed as a co-author despite current injunctions against such practice
Survey on Generative AI use amongst UK doctoral candidates
No full textA survey on the use of Generative Artificial Intelligence amongst UK postgraduate researchers.
This survey was used to produce the results reported in the paper "A rather stupid but always available brainstorming partner': use and understanding of generative AI by UK postgraduate researchers" published in Innovations in Education and Teaching International.</span
Meaningful human control: actionable properties for AI system development
No full textHow can humans remain in control of artificial intelligence (AI)-based systems designed to perform tasks autonomously? Such systems are increasingly ubiquitous, creating benefits - but also undesirable situations where moral responsibility for their actions cannot be properly attributed to any particular person or group. The concept of meaningful human control has been proposed to address responsibility gaps and mitigate them by establishing conditions that enable a proper attribution of responsibility for humans; however, clear requirements for researchers, designers, and engineers are yet inexistent, making the development of AI-based systems that remain under meaningful human control challenging. In this paper, we address the gap between philosophical theory and engineering practice by identifying, through an iterative process of abductive thinking, four actionable properties for AI-based systems under meaningful human control, which we discuss making use of two applications scenarios: automated vehicles and AI-based hiring. First, a system in which humans and AI algorithms interact should have an explicitly defined domain of morally loaded situations within which the system ought to operate. Second, humans and AI agents within the system should have appropriate and mutually compatible representations. Third, responsibility attributed to a human should be commensurate with that human’s ability and authority to control the system. Fourth, there should be explicit links between the actions of the AI agents and actions of humans who are aware of their moral responsibility. We argue that these four properties will support practically minded professionals to take concrete steps toward designing and engineering for AI systems that facilitate meaningful human control.Interactive IntelligenceDesign AestheticsCyber SecurityHuman-Robot InteractionEthics & Philosophy of TechnologyHuman Information Communication DesignWeb Information System
A Two-Dimensional Explanation Framework to Classify AI as Incomprehensible, Interpretable, or Understandable
No full textBecause of recent and rapid developments in Artificial Intelligence (AI), humans and AI-systems increasingly work together in human-agent teams. However, in order to effectively leverage the capabilities of both, AI-systems need to be understandable to their human teammates. The branch of eXplainable AI (XAI) aspires to make AI-systems more understandable to humans, potentially improving human-agent teamwork. Unfortunately, XAI literature suffers from a lack of agreement regarding the definitions of and relations between the four key XAI-concepts: transparency, interpretability, explainability, and understandability. Inspired by both XAI and social sciences literature, we present a two-dimensional framework that defines and relates these concepts in a concise and coherent way, yielding a classification of three types of AI-systems: incomprehensible, interpretable, and understandable. We also discuss how the established relationships can be used to guide future research into XAI, and how the framework could be used during the development of AI-systems as part of human-AI teams.Accepted author manuscriptInteractive Intelligenc
‘A rather stupid but always available brainstorming partner’: use and understanding of Generative AI by UK postgraduate researchers
No full textResearch into the increased use of Generative AI in Higher Education has largely focused on undergraduate study. While many institutions are grappling with the implications for doctoral level, there has been little published work investigating how postgraduate researchers use the technology or their attitudes towards it. This paper is based on a survey of 75 doctoral candidates across 19 UK Higher Education Institutions. The results show that most respondents had used Generative AI for their doctoral research, with the most common uses being framed as time-saver, editor or colleague. There was an awareness of limitations and ethical issues connected to the use of AI but no agreement as to where those boundaries lie. The paper concludes that there is an urgent need for sector agreement and communication on acceptable use and best practice
Using Generative AI in Research
No full textThe slides accompany a workshop that is intended for graduate students to learn more about generative AI in the context of the research lifecycle. This work is licensed under a Creative Commons license so that others may share and adapt the content for other purposes as long as appropriate credit is provided to the author of the work. To access the Google slides, click here: https://bit.ly/Library_AI_Research
Learning Objectives
At the end of the session participants will be able to:
Demonstrate a basic understanding of how AI tools work
Differentiate between grounded and ungrounded AI tools
Identify key considerations for grad students/researchers
Identify ways AI tools can be used to support the phases of the research lifecycle
Identify main areas of concern with using AI tools
Outline the steps and potential resources for evaluating and citing AI outpu
The AI Author in Litigation
No full textMany scholars have posited whether a computer possessing Artificial Intelligence (AI) could be considered an author as defined per the Copyright Act of 1976. What was once a thought experiment is now becoming reality. To date, scholarship has focused primarily been on whether an AI meets the requirements of authorship from a purely objective legal framework or whether an AI could be an author based on the doctrines of incentives, independent creation, and creativity.
However, a burden inherent in the rights and liabilities of authorship is the ability to be held liable if that author’s expressive work is infringing on another’s. A cause of action is meaningless if a copyright owner cannot enforce it by suing the infringer or if the infringer is judgement-proof. Thus, when contemplating whether an emancipated AI—or any non-human—can be an author under the Copyright Act, part of that examination should be whether the AI which created the work can sue or be sued for infringement.
This article considers issues from the theoretical, like civil procedure and remedies, to the practical, such as legal representation and discovery. How is an AI served with a lawsuit? What would be an adequate, enforceable remedy for an AI’s infringement? Is an AI even bound by our laws? Additional questions—and procedural barriers—are raised when considering other roles an AI might play in an infringement action: as a witness, a co-party, or even a plaintiff seeking to protect its own creative expression.
This morass of legal headaches goes beyond any doctrinal issues regarding authorship, and provide ample reason to keep legal authorship in the hands of humans or entities controlled by humans—at least until legal procedure catches up to technological realities and possibilities for litigation that AI parties present
Recommended from our members
The AI Author in Litigation
No full textMany scholars have posited whether a computer possessing Artificial Intelligence (AI) could be considered an author as defined per the Copyright Act of 1976. What was once a thought experiment is now becoming reality. To date, scholarship has focused primarily been on whether an AI meets the requirements of authorship from a purely objective legal framework or whether an AI could be an author based on the doctrines of incentives, independent creation, and creativity.
However, a burden inherent in the rights and liabilities of authorship is the ability to be held liable if that author’s expressive work is infringing on another’s. A cause of action is meaningless if a copyright owner cannot enforce it by suing the infringer or if the infringer is judgement-proof. Thus, when contemplating whether an emancipated AI—or any non-human—can be an author under the Copyright Act, part of that examination should be whether the AI which created the work can sue or be sued for infringement.
This article considers issues from the theoretical, like civil procedure and remedies, to the practical, such as legal representation and discovery. How is an AI served with a lawsuit? What would be an adequate, enforceable remedy for an AI’s infringement? Is an AI even bound by our laws? Additional questions—and procedural barriers—are raised when considering other roles an AI might play in an infringement action: as a witness, a co-party, or even a plaintiff seeking to protect its own creative expression.
This morass of legal headaches goes beyond any doctrinal issues regarding authorship, and provide ample reason to keep legal authorship in the hands of humans or entities controlled by humans—at least until legal procedure catches up to technological realities and possibilities for litigation that AI parties present
Optimin achieves super-Nash performance
No full textSince the 1990s, AI systems have achieved superhuman performance in major
zero-sum games where "winning" has an unambiguous definition. However, most
social interactions are mixed-motive games, where measuring the performance of
AI systems is a non-trivial task. In this paper, I propose a novel benchmark
called super-Nash performance to assess the performance of AI systems in
mixed-motive settings. I show that a solution concept called optimin achieves
super-Nash performance in every n-person game, i.e., for every Nash equilibrium
there exists an optimin where every player not only receives but also
guarantees super-Nash payoffs even if the others deviate unilaterally and
profitably from the optimin.Comment: arXiv admin note: substantial text overlap with arXiv:1912.0021
- …
