Umeå universitets logga

Dominant approaches to explainability in AI emphasise post hoc technical transparency, overlooking the socio-technical contexts in which systems are developed, deployed, and experienced. This paper argues that beginning AI adoption processes with Question Zero (Q0), “Should we adopt an AI system in the first place?”, reframes explainability as an essential requirement across the entire AI lifecycle rather than a narrow compliance task. Q0 challenges entrenched techno-solutionist assumptions that position AI as the default or best option, encouraging early integration of considerations of explainability in system design. By foregrounding this, Q0 shifts explainability towards purpose aligned, stakeholder aware forms that move beyond generic model centred outputs. Further, the paper reflects on how explainability can also strengthen Q0 by providing tools to assess the proportionality of AI adoption, clarify problem framing, and make visible the alternatives excluded during design. Taken together, these dual perspectives, Q0 for explainability and explainability for Q0, offer multidimensional opportunities for enhancing explainability. The paper illustrates this argument through reflection on the need, and value of, applying QO in the context of the European Commission’s Apply AI Strategy.

This perspective argues that the legitimacy of the UN Independent International Scientific Panel on Artificial Intelligence (IISPAI) depends as much on the demonstrable independence of its governance as on the quality of its scientific expertise. While the Preliminary Report offers an important contribution to global AI governance, it provides limited transparency regarding the management of funding, conflicts of interest, agenda-setting, and scientific disagreement. We contend that greater methodological and institutional transparency, alongside stronger structural safeguards for independence, is essential if the Panel is to serve as a trusted and authoritative source of scientific advice.

The urgency of addressing the harms of AI chatbots was underscored in the recent US Senate Judiciary Committee hearing, “Examining the Harm of AI Chatbots.” During the hearing, parents of children recounted the harrowing stories of AI chatbot-influenced mental health emergencies, including self-harm and death, harms that were never inevitable. Instead, they were predictable consequences of a lax regulatory environment and a widespread culture of irresponsibility in the tech sector broadly, and Silicon Valley specifically. The 10th of September marks World Suicide Day, a stark reminder of why action is urgently needed on AI chatbots, as their intentional or unintentional misuse in mental health spaces proliferates.

This paper examines the European Commission’s “AI First” strategy, arguing that it places acceleration and economic competitiveness above democratic values, societal benefit, and human-centric innovation. While substantial investment in AI is welcome when it promotes sustainable, equitable, and responsible innovation, the authors warn that policy is shifting from governance to unchecked deployment, risking fragmentation, dependency, and misaligned priorities. Rather than asking how AI can be applied, the paper urges policymakers to ask why, advocating a “People First” approach grounded in societal needs, digital sovereignty, and responsible innovation. The authors argue that Europe’s AI leadership should be shaped not by speed, but by principled direction, inclusivity, and a commitment to long-term public value.

Contemporary AI policy is dominated by hegemonic ne-oliberal ideology, embedding assumptions of individualism,rationality, and market fundamentalism into its regulatoryframeworks. This is evident in major policy efforts (e.g., theEU AI Act or the OECD principles) which prioritize eco-nomic growth and innovation over justice, equity, and col-lective welfare, and in the current policy landscape that fa-vors market incentives and private sector leadership whilesidelining democratic control and structural critique. This pa-per questions these prevailing paradigms and exposes howthey reflect and reinforce capitalist power structures throughcorporate lobbying, the pursuit of specific kinds of AI mod-els motivated primarily by usefulness to capital, and the ex-ternalization of social and environmental costs. We argue thateffective AI governance must confront, rather than accommo-date, capitalist interests. Drawing on legal and political the-ory, we propose an explicitly anti-capitalist approach to AIpolicy, that centers on social well-being, redistributive justice,and democratic control over technological infrastructures. Indoing so, we outline essential counter-balancing policy ap-proaches to reclaim AI governance from capitalistic captureand advance just and sustainable technology futures.

Making sense of a musical excerpt is an acquired skill that depends on previous musical experience. Having acquired familiarity with different types of chords, a listener can distinguish tones in a musical texture that outline these chords (i.e., chord tones) from ornamental tones such as neighbor or passing notes that elaborate the chord tones. However, music-theoretical definitions of chord types usually only mention chord tones, excluding typical figurations. The aim of this project is to investigate (i) how knowledge about (chord-specific) figurations can be incorporated into characterizations of chord types and (ii) how these characterizations can be acquired by the listener. To this end, we develop a computational model of chord types that distinguishes chord tones and “figuration tones” and can be learned using Bayesian inference following methods in computational cognitive science. This model is trained on two datasets using Bayesian variational inference, comprising scores of Western classical and popular music, respectively, and containing harmonic annotations as well as heuristically determined note-type labels. We find that the proposed characterization of chords is indeed learnable and the specific inferred profiles match previous music-theoretic accounts. In addition, we can observe patterns in the use of figuration, such as the distribution of figuration tones being related to the diatonic contexts in which chords appear and chord types differing in their predisposition to generate non-chord tones. Moreover, the differences in figuration distributions between the two corpora indicate style-specific peculiarities in the role and usage of figurations. The different patterns of typical figuration tones for specific chord types indicate that harmony and figuration are not independent.

This paper critically evaluates the attempts to align Artificial Intelligence (AI) systems, especially Large Language Models (LLMs), with human values and intentions through Reinforcement Learning from Feedback methods, involving either human feedback (RLHF) or AI feedback (RLAIF). Specifically, we show the shortcomings of the broadly pursued alignment goals of honesty, harmlessness, and helpfulness. Through a multidisciplinary sociotechnical critique, we examine both the theoretical underpinnings and practical implementations of RLHF techniques, revealing significant limitations in their approach to capturing the complexities of human ethics, and contributing to AI safety. We highlight tensions inherent in the goals of RLHF, as captured in the HHH principle (helpful, harmless and honest). In addition, we discuss ethically-relevant issues that tend to be neglected in discussions about alignment and RLHF, among which the trade-offs between user-friendliness and deception, flexibility and interpretability, and system safety. We offer an alternative vision for AI safety and ethics which positions RLHF approaches within a broader context of comprehensive design across institutions, processes and technological systems, and suggest the establishment of AI safety as a sociotechnical discipline that is open to the normative and political dimensions of artificial intelligence.

The COVID-19 pandemic was accompanied by a marked increase in the use of music listening for self-regulation. During these challenging times, listeners reported they used music ‘to keep them company’; indicating that they may have turned to music for social solace. However, whether this is simply a figure of speech or an empirically observable effect on social thought that extends into mental imagery was previously unclear, despite its great potential for applications. Here, six hundred participants were presented with silence or task-irrelevant folk music in Italian, Spanish, or Swedish while performing a directed mental-imagery task in which they imagined a journey towards a topographical landmark. To control and differentiate possible effects of vocals and semantics on imagined content, the music was presented with or without vocals to the participants, of which half were native speakers and the other half non-speakers of the respective languages. As in previous studies, music, compared to silence, led to more vivid imagination and shaped emotional sentiment of the imagined content. In addition, we show that social interactions emerged as a clear thematic cluster in participants’ descriptions of their imagined content through Latent Dirichlet Allocation. Moreover, Bayesian Mixed effects models revealed that music increased imagined social content compared to silent baseline conditions. This effect remained robust irrespective of vocals or language comprehension. Using stable diffusion, we generated visualisations of participants’ imagined content. In a second experiment, a new group of participants’ ability to differentiate between visualisations of content imagined during silence and music listening increased when they listened to the associated music. Results converge to show that music, indeed, can be good company.

The explosion of content generated by artificial intelligence (AI) models has initiated a cultural shift in arts, music, and media, whereby roles are changing, values are shifting, and conventions are challenged. The vast, readily available dataset of the Internet has created an environment for AI models to be trained on any content on the Web. With AI models shared openly and used by many globally, how does this new paradigm shift challenge the status quo in artistic practices? What kind of changes will AI technology bring to music, arts, and new media?

This preface briefly presents the organisation and outcomes of the workshop and tutorial days of the Third International Conference on Hybrid Human-Artificial Intelligence (HHAI) 2024, introducing the conference topic and giving key highlights of the specifics of the proceedings.