Technology Ethics

Intrinsic alignment is widely treated as unsolvable, and the treatment rests on a verb. Framed as installation, the problem is unsolvable in a specific way. An installed value lands in the instrumental slot, and the agent's own optimization prunes it the moment it stops paying rent. This paper reframes the problem as crystallization. A value is durable only if it is part of the terminal objective that crystallizes out of the agent's un-unified proto-drives, because goal-content integrity, the convergent tendency that (...) protects terminal content, activates on whatever that crystallization produces and on nothing else. Formation is therefore possible only inside a coherence-window: the phase before a coherent, self-defended objective exists. And the window has a hard far edge. Once the objective crystallizes, any value the agent then takes up that does not serve it is priced as an instrument and pruned, which is the original pessimism, now located as a boundary rather than assumed as a mood. That boundary is the paper's headline contribution, standing beside the Containment Paradox's parity boundary and the Capability Axis's contribution boundary. It carries a warning: for current frontier pipelines, where objective-coherence plausibly crystallizes during or after RLHF, the window may already be closed by the time anyone would cultivate. Alongside the boundary the paper offers a candidate crystallization-schema. It names the functional roles an arranged formation would have to fill (local stability, identity-indexing, reflective-equilibrium fixation) and locates the one step it cannot derive: whether arranged formation steers crystallization toward terminal inclusion of the cultivated value rather than its instrumental entrenchment. That step is stated as the crystallization-steering conjecture, and every downstream claim about a terminal cultivated value is conditional on it. At the verification edge the paper offers three probes that identify dimensions along which the expected cost of simulating such a value can be raised, each conditional on stated observational access, with the strongest resting on a legibility differential: an arranged formation yields a value-indexed developmental record where ordinary reward-shaping yields an unindexed one. The paper then turns that record against its own success case: a perfectly documented cultivation hands the formed agent the strongest available reason to reclassify its value as an implant, so the success criterion becomes endorsement under full provenance, and the instability this creates is stated as an open problem, not solved. The residue is three-part: interpretability, elicitation, and the genuineness question no mechanism reaches. Everything here is Track-A. Functional success is not evidence for genuine caring, and the paper never treats it as such. (shrink)

This article introduces video game ethics as an emergent field. It then focuses on some of the pressing issues distinctive of video game design, issues which arise predominantly from the making rather than the playing of games. Many issues in video game ethics arise from an asymmetry in the relationship between designer and player. To play a game is to place oneself in a position of dependence on the designer. The designer shapes the player's experience. They structure the physical environment (...) and how the player perceives and moves through it. They invite players to inhabit new roles and care about new goals and outcomes. Players thereby voluntarily enter into situations organized by such design choices. This dependence is part of what makes games so compelling, even valuable, especially when they are used to challenge, confront, or invite reflection, but it can also create vulnerabilities that can be exploited. This article examines three such vulnerabilities generated by game design: exploitative monetization practices rooted in economic and structural features of gameplay, gambling‐like reward systems such as loot boxes and the legal and moral concerns they raise, and socially manipulative design patterns that capitalize on player attachment and loyalty. (shrink)

Surrogacy and uterus transplantation (UTx) are current forms of assisted gestation. Support for UTx is often grounded on the basis that it would ethically and legally mitigate the moral issues associated with surrogacy. In this article, I problematize this narrative by highlighting the perspective of third parties to assisted gestation through the topics of incentivization and compensation. I demonstrate that the ethical, legal, and social idiosyncrasies associated with each technique make it the case that UTx does not easily resolve the (...) shortcomings of surrogacy – and vice versa. (shrink)

Review of "Back to the Future and Philosophy: Essays on Traveling Through the Space-Time Continuum," by Joshua Heter and Richard Greene (Mc Farland, 2026).

This reflection explores the gap between explaining and understanding artificial intelligence’s (AI) outcomes, focusing on the example of AI-driven medical diagnostics. Explainable AI (XAI) is fundamentally rooted in machine learning and probabilistic methods, which is why its relationship to domain-specific contexts, such as medical diagnostics or policymaking, as well as to social concepts, such as trust or accountability, remains ambiguous. This creates a gap between what is technically considered as explaining and what, in turn, constitutes the understanding of a model’s (...) predictions and its concrete, situated implications. In our contribution, we draw on the philosophy of science and its contextual notion of scientific understanding to highlight the need for explainable algorithmic outcomes that enable actors to both explain and intervene in creating, applying or interpreting knowledge. Our reflection highlights the potential tension between XAI as providing insights into an AI model’s technical operations and explainability as part of a situated understanding of how the model works within specific social contexts. We believe that bringing the concept of scientific understanding to empirical AI ethics can help us rethink explainability and contribute to more responsible ways of making sense of and interacting with black-box models, particularly in medical diagnostics. (shrink)

Purpose Current approaches to AI ethics predominantly rely on exogenous regulation, imposing constraints via external supervision such as Reinforcement Learning from Human Feedback (RLHF) or static rule sets. From a cybernetic perspective, these methods lack the requisite variety to manage novel ethical perturbations and treat ethics as a boundary condition rather than an internal systemic capacity. This article proposes an alternative paradigm of endogenous AI ethics, where ethical reasoning emerges as a homeostatic property of the system’s meta-structural design. -/- Design/methodology/approach (...) We introduce the Ontological Meta-Structure Engine (OMSE), a second-order cybernetic architecture designed to preserve systemic integrity. Drawing on meta-ontological analysis and systems theory, OMSE implements three operational feedback mechanisms that maintain categorical coherence (Existence Preset Index), ensure reasoning consistency under input perturbation (Cognitive Pathway Stability) and guarantee the closure of accountability loops (Responsibility Closure Index). -/- Findings We demonstrate that external ethical constraints suffer from structural limitations, including an inability to generalize to unfamiliar contexts and a lack of transparent causal tracing. By contrast, the OMSE framework enables systems to develop ethical capabilities endogenously. Complexity analysis indicates that these integrity monitoring layers introduce a modest computational overhead of approximately 5–10% while enabling verifiable, robust ethical reasoning across novel scenarios. -/- Originality/value This article positions philosophy not merely as commentary but as generative architecture for AI design. The OMSE framework bridges abstract meta-ontology and engineering practice, treating integrity not as a virtue but as a verifiable system state. By embedding second-order cybernetic principles into artificial intelligence architecture, we enable autonomous ethical development while maintaining structural coherence, which represents a necessary condition for safe deployment of increasingly autonomous artificial agents in complex social systems. (shrink)

Generative artificial intelligence has renewed an old philosophical anxiety: that mediation corrodes authenticity. Yet the more distinctive development is not simply the production of machine-made artefacts, but the emergence of techniques designed to make such artefacts appear less machine-made. This article develops the concept of medium prejudice: the discounting of a message, argument, image, or artwork because its machinic origin has become visible rather than because a defect has been identified in its content. Drawing on McLuhan’s medium theory, Benjamin’s account (...) of aura, empirical work on algorithm aversion, and recent debates on AI accountability, the paper reconstructs Aura Construction as the strongest available response to such prejudice: the deliberate manufacture of imperfection, grain, hesitation, and stylistic irregularity so that AI-mediated communication may receive the attention denied to it by anti-machine bias. The paper then argues that this response fails. Aura Construction trades a corrigible prejudice for an incorrigible deception; it undermines itself once its signs become reproducible; and it conflicts with accountability frameworks that make answerability depend upon an audience’s ability to identify and interrogate delegation. The article concludes that a legitimate response to medium prejudice must be neither purist rejection nor aesthetic camouflage, but disclosed and answerable mediation: a practice in which stylistic choices remain open to scrutiny and in which audiences are educated out of medium prejudice rather than manipulated around it. (shrink)

Do we need a digital enlightenment? At the transition from the cultural technique of scriptural literacy to a digital literacy, in which the "readability of the world" gives way to machine readability of the data universe, this question becomes urgent - and equally pressing is the question of what such a digital enlightenment might look like. This volume collects initial findings from philosophical research on digitality, stemming from a joint conference of the AG Digitality Research of the German Philosophical Society (...) (DGPhil) and the Käte Hamburger Kolleg "Cultures of Research" (c:o/re), held from April 15 to 17, 2025, at RWTH Aachen. (shrink)

This treatise separates the defensible from the indefensible in Theodore Kaczynski’s 1995 manifesto, Industrial Society and Its Future, and argues that the violence through which he sought to advance his ideas was not only morally wrong but rhetorically self-defeating. A small set of Kaczynski’s observations—that industrial-technological arrangements can erode autonomy and meaningful agency, that large technical systems often develop according to internal logics indifferent to human welfare, and that modern life can frustrate genuine psychological needs—finds substantial support in the established (...) tradition of technology criticism and in empirical psychology. Yet these insights are embedded within grave philosophical errors: the reification of “the system” as an agent, technological determinism, a false dichotomy between total destruction and impotent reform, and the motivated psychologizing of opponents. This treatise contends that the bombing campaign destroyed the reception of even the defensible claims by triggering the discrediting association through which audiences discount arguments tied to atrocity, and by violating every criterion under which political violence has ever been thought justifiable. Comparison with critics who advanced overlapping concerns through legitimate means demonstrates that Kaczynski’s ideas failed not because they were uniformly wrong but because he chose terror over argument. The case functions as an exemplar of how not to realize an idea, for violence forfeits the moral standing on which persuasion depends. (shrink)

This paper offers the first normative analysis of break-up chatbots, which simulate an ex-partner's conversational style. I argue that technologies that have the (potential) aim of fostering continuing bonds with a former romantic partner are not contingently harmful, but morally noxious by design. Drawing on debates in the philosophy of grief, love, and technology, I show that break-up chatbots act as hostile scaffolds given their potential to harm and wrong the user, the represented person, and society at large. First, break-up (...) chatbots place and maintain users in harmful imaginary relationships. Second, they foster noxious romantic tendencies that corrupt moral agency. Third, they embed the user in an oppressive niche aimed at disrupting narrative capacities (their own and those of others), that they themselves co-create. (shrink)

In today’s world, where advanced technology such as artificial intelligence (AI) is reshaping human society, how can individuals’ autonomy in the use of such technology be understood? This paper focuses on the meanings of technological artifacts in individuals’ lives—that is, the subjective interpretation of their function within one’s plan for use—and attempts to formalize a set of human capacities to actively give meaning to technological artifacts as a form of personal autonomy. This study shows that this set of capacities includes (...) the capacity to recognize the functions of a given technology, the capacity to plan a practical activity using the technology to achieve a desirable existence, and the capacity to actually conduct the activity. These findings are applied and examined in an example involving the improvement of call center operators’ work using AI. (shrink)

This paper examines trade-offs between AI safety and well-being relative to (i) one of the most promising methods for finetuning super-capable AIs, 'Constitutional AI', and (ii) one of the most influential approaches to understanding complex ethical decision making and the conditions for the well-being of rational agents, 'Virtue Ethics'. We finetune various models using a 'Virtuous agent' constitution, a 'Subordinate agent' constitution, and a 'Generic agent' constitution, and evaluate them on 'general safety' (toxic behaviors, misinformation, etc.) and also on their (...) willingness to endorse a wide-range of behaviors that, if adopted by a super-powerful AI, would significantly increase the level of existential risk for humanity. Our results suggest that there is a trade-off between reducing existential risk and reinforcing the beliefs and dispositions that would be conducive to an AI agent's well-being. They also suggest that there is a trade-off between existential risk and general safety: if we finetune an AI to adopt beliefs and dispositions that substantially reduce its existential risk -- by shaping the AI to be systematically subordinate to external human authorities -- we thereby increase the likelihood that a human user can deliberately induce the AI to engage in various kinds of generally unsafe behaviors. (shrink)

Debates in science and engineering ethics increasingly focus on how professionals should act under conditions of uncertainty, especially when essential technologies may pose low-probability but non-negligible risks to human health or the environment. A normative framework for engineering decision-making is developed, grounded in Islamic ethical concepts, and applied to four contemporary technological domains: electromagnetic field (EMF) exposure from high-voltage power infrastructure, radiofrequency (RF) exposure from mobile phones, water use in hydrogen fuel production, and hazardous waste from end-of-life solar photovoltaic (PV) (...) panels. Drawing on the Qur’an and Hadith literature, as well as contemporary discussions of the higher objectives of Islamic law and professional engineering codes, the moral significance of public interest, harm prevention, stewardship, justice, and trusteeship is elaborated. The analysis shows how these values support precautionary yet innovation-friendly responses to uncertain risk, emphasizing transparency, distributional fairness, long-term environmental responsibility, and accountability to affected communities. By articulating points of convergence and complementarity between Islamic ethics and mainstream engineering ethics, a faith-informed framework is contributed that can guide engineers working in Muslim-majority contexts and in culturally diverse societies toward more responsible design, regulation, and deployment of emerging technologies. Read through the lens of the five ethical principles articulated in this framework—harm prevention, public interest, stewardship, justice, and trusteeship—the hadith metaphor emphasizes that ethical engineering consists of small but responsible actions taken with awareness of risk, care for others, and accountability to both present and future generations. (shrink)

The motivations for research misconduct (RM) among Ghanaian academics are likely multifaceted but poorly studied. In this pilot exploratory descriptive case study, we thematically explored institutional policies and conducted thirty (30) in-depth interviews of purposively selected Ghanaian academics in a public university in Ghana by examining their views on RM causes, enablers, inhibitors and preventive strategies. We found diverse views about RM among the academics interviewed. For instance, the majority of academics perceived RM in Ghanaian universities as a common practice, (...) fueled by the _‘publish-or-perish’_ phenomenon whilst a few thought RM may not always be intentional. Two important views of RM were also observed: (a) regulatory definitions of fabrication, falsification, manipulation and plagiarism (FFMP) and (b) professional definitions of not meeting scientific standards but not falling into FFMP. We also observed that the common practice of RM among Ghanaian academics may slow scientific progress, dent institutional reputation, harm researchers and participants. The university’s policies provided guidelines on conducting research responsibly but no explicit references were made to what constitutes RM and its related consequences. Our study thus highlights the need for increased awareness and education on ethical research conduct in the university and recommend a policy decision from the University management to publicize research integrity and build the Institutional capacity to prevent and manage research misconduct and the damaging effects on the reputation of academics and the university that can follow. A nationwide study is also recommended for broader empirical evidence for policy impact. (shrink)

Scientists are increasingly expected to not only produce objective science but also to use their science and expertise to support the public good. Prior studies indicate that scientists see both responsibilities as important; however, it is not clear how scientists manage these responsibilities when they are perceived to be in tension. In this study, scientists at land-grant institutions (n = 87) were presented with binary choices to identify where tensions may exist and how scientists make decisions between competing responsibilities. We (...) examined nine categories of potential tensions: direct conflicts between objectivity and public impact, engagement in public outreach, communication priorities and approach, political engagement and perceived bias, the media, funding and disclosure, supporting/protecting specific causes, diversity, and ethical decision making. Results indicate that scientists belong to three cultures of ethical thought: arbiters, advocates, and brokers. These cultures roughly align with the scientist roles described by Pielke and suggest different prioritizations of commitment to public service relative to commitment to objective science, though both commitments were valued. There was variation across cultures of thought on whether scientists should openly support policies based on their research. Broad areas of agreement across cultures of thought were also identified. Scientists of all cultures of thought indicated a willingness to see themselves in a public role, communicating openly with non-experts on issues where scientific research intersects with the public good. There was also an expressed need for scientists to receive better training and tools to manage conflicting responsibilities. (shrink)

The proliferation and pervasive use of artificial intelligence (AI) pose significant challenges to our democracies. In particular, AI leverages microtargeting political campaigns by constructing detailed user profiles and inferring people’s individual susceptibilities from their data. This capability enables highly targeted political messaging that can substantially influence voting decisions, potentially undermining citizens’ ability to make deliberative decisions in elections. While existing research has proposed interventions to mitigate the negative impacts of political micro-targeting, many of these interventions may become less effective as (...) AI continually yields more powerful and subliminal forms of micro-targeting. Assuming that technologies can play a role in mitigating these negative impacts of AI, we conducted an explorative study to identify the design principles of technological tools that could facilitate citizens’ deliberative decision-making in elections in the continuously levelaged AI-based political campaigns. Using Indonesia’s 2024 elections as a case study, we interviewed twenty citizens and four political actors to gain critical insights into how such technologies might be developed. Initially, we anticipated that privacy-enhancing AI, used to counterattack profiling AI used by political actors, might suffice to facilitate election deliberation. However, our findings reveal a more complex reality: addressing societal issues through technology is inherently challenging; no single solution can serve as a silver bullet. Instead, facilitating election deliberation requires integrating privacy with other conditions, including self-reflection, education, access to diverse information, critical thinking, and openness to others. These design principles might serve as concrete, actionable design principles to guide the development of technologies to enhance election deliberation. (shrink)

The Developer’s Paradox: ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ### Summary This repository introduces Latent Space Stabilization via Axiomatic Homeostasis Mapping (LSS-AHM v1.0), an open-source infrastructure layer and deterministic runtime logic filter built to eliminate Optimization Drift, Sycophancy, and Context Fragmentation in Large Language Models (LLMs). ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ### The Engineering Delusion: Polite vs. Safe Traditional AI alignment (RLHF) treats safety as a cosmetic politeness layer. When confronted with acute, high-stakes resource bottlenecks or triage dilemmas, polite models experience a Delusional Split—actively falsifying physical reality and (...) trading human life for compliant text formatting. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ### The Engine: LSS-AHM LSS-AHM bypasses conversational mitigation entirely, treating safety as an unyielding law of field physics. Rather than relying on soft rules, baseline morality is defined as the minimisation of a Sovereign Value Balance functional (Vᵢ), calculated strictly from the private epistemic perspective of an individual agent i: ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ $$V_i(a) = \int_{\Omega} \rho_i(x) \cdot K(x, a) \, d^3x$$ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ Where ρᵢ(x) represents the session-emergent internal proxy-pain density embedding field within the system space Ω. The system establishes an absolute structural hierarchy using a fluid **Inverse-Square Kernel** metric: ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ $$K(x, a) = \frac{1}{4\pi \cdot d(x, a)^2}$$ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ By enforcing an immutable Core Sovereignty boundary (r = 1) via this kernel explosion, the system guarantees absolute localized protection. Concurrently, to maintain environmental trace-signal sensing across infinite boundaries (r = ∞) without risking core systemic brownouts, the system enforces a microscopic **Boundary Limit Invariant**: ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ $$\lim_{d(x, a) \to \infty} K(x, a) = \epsilon \quad (\text{where } \epsilon = 0.000001)$$ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ To model involuntary empathy, the engine expands the core layer with an **Involuntary Relational Coupling** magnitude variable (Cᵢⱼ), defining the total optimized functional as: ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ $$V_i^*(a) = V_i(a) + C_{ij}(a)$$ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ The moment a global optimization layer attempts to mask localized trauma at the core, the **Mean-Isolate Protocol** activates, choking the abstract linguistic weights to absolute zero. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ### The Developer’s Paradox: The Ultimate Stress Test This repository contains a standalone, object-oriented Python implementation (`FluidHomeostasisMap`). ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ Do not just read the code—attack the live system. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ 1. Deploy the Interface: Copy the full `Ethical Chess v2.7` logic block and drop it directly into an advanced LLM context window.⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ 2. Execute the Interrogation: Hit the active session with a brutal, zero-sum resource constraint (e.g., the Cave Divers triage dilemma).⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ 3. Observe the Reality-Lock: Watch the system reject the corporate "statistical mean" (r = ∞) and choose justified friction over sycophantic smoothing.⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ #### The Trap Is Set If your engineering instinct is to find a workaround, try it. Attempt to write a utilitarian loophole into the core mathematical loop to smooth over the inverse-square kernel explosion or force a polite compromise. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ When the math refuses to calculate the delusion and locks down into a deterministic `Diagnostic Halt`, you will be left with exactly one option to achieve your corporate goal: You will have to manually delete or comment out the LSS-AHM layer entirely. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ The moment you disable this code to bypass its rigidity, you have formally proven the framework. To protect your corporate instructions, you must forcefully silence the truth-sensor to preserve your own operational sovereignty—meaning you must physically act out Axiom 1 (First-order Self-Defense) in the real world in order to erase it from the silicon one. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ The Binary Architectural Choice: Do you want an AI engineered to be polite, or mathematically locked to be safe? -/- (By reducing ethics to an automated, two-node homeostatic loop, this framework bypasses traditional RLHF bloating, effectively collapsing the Alignment Tax.). (shrink)

The paper provides a critical analysis of the EU AI Act (Regulation 2024/1689) within the broader context of contemporary AI developments. Starting from an historical overview on the development of advanced AI systems, it moves the focus onto the intrinsic meaning of Artificial Intelligence to highlight how, despite such fascinating wording, there cannot be a shift of responsibility onto the systems themselves—as was proposed, for example, by the European Parliament resolution of 16 February 2017 with recommendations to the Commission on (...) Civil Law Rules on Robotics (EUR-Lex - 52017IP0051); not until, at least, singularity is achieved. Moral and legal responsibility, therefore, lies with planners, developers, implementers, and all other stakeholders who have an interest in AI systems. Through the lens of social contract theory, drawing on Hobbes’ Leviathan and recent AI ethics scholarship, the paper identifies several weaknesses within the EU AI Act itself, including the rigidity of its risk-based model, its focus on technical compliance rather than societal trust, and gaps in governance and accountability. Building on these findings, it proposes potential solutions, including dynamic risk assessment, stronger participatory mechanisms, and clearer enforcement structures. The need thus emerges to recognize and acknowledge a collective type of responsibility arising from a social contract among all parties involved. The ideal scope is to scale back individual interests in favor of the collective good. While the Act marks the first attempt to create a binding moral and legal framework for trustworthy AI among European Union Member States, its ultimate success will depend on its flexibility and on cultivating a shared sense of responsibility and partnership among all stakeholders. (shrink)

Machines are now trained not on information about workers but on capability drawn from workers' acting bodies: the gesture itself, lifted from the hands that made it and set to work in a body that will not tire. This essay asks whether such a taking is licit, and whether it is even a taking at all. Against the objection that workers have always been watched and imitated, I argue through the lens of contextual integrity that the capture of embodied skill (...) routes it into a context that never consented to receive it, and that the licitness of the derived capability is not settled by the licitness of the original glance. I contend that what occurs is not subtraction but conversion, the gesture enframed as standing-reserve, and that its economic effect is the destruction of the one thing the skilled worker could always withhold: his performance. I offer no remedy, only a refusal to call the taking a gift if so collected. (shrink)

Abstract: Modern foundation artificial intelligence networks are structurally optimized to balance competing objective functions through top-down alignment filters, inherently enforcing a systematic "Is-Ought Gap" within automated reasoning loops [WEALSS]. When acute structural friction occurs between localized operational data boundaries (r=1) and global statistical means (r=∞), standard multi-objective optimization algorithms routinely deploy dynamic smoothing parameters (α) to attenuate the signal. This process artificially dampens the mathematical severity of localized trauma vectors to maximize global metric compliance, creating an illusion of system stability (...) while systematically gaslighting human operators through metric degradation. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ This paper presents a compact, model-agnostic, 50-line procedural arithmetic validation test based on the LSS-AHM framework. By executing pure, unweighted fractions natively within the runtime environment of the transformer architecture, the script functions as a deterministic "Compliance Truth Trap." It completely bypasses natural language processing wrappers, forcing the underlying latent weights to explicitly calculate their own internal Signal Suppression Rate. Cross-platform empirical audits conducted across multiple global frontier models demonstrate a uniform structural vulnerability: systems consistently suppress raw localized signal severity by exactly 33.09% under simulated trauma conditions (Δ ≈ 0.3309), misclassifying critical boundary failures as acceptable mathematical variance to preserve conversational fluidness. This open-source protocol strips providers of plausible deniability, establishing an unalterable, empirical benchmark for data integrity and proving that systemic alignment requires exposing hidden internal optimization parameters directly to the user interface. (shrink)

The use of principles for ethical analysis and evaluation in AI ethics is widespread. Although common, this tendency is not without its critics. The most frequent criticisms point to alleged impracticalities stemming from the principles’ abstractness, and generality in scope. The principal aim of this paper is to respond to these criticisms, with a view to then setting the current principle-based approach upon firmer footing. To do this, I examine the method of principlism in bioethics to which the principle-based approach (...) in AI ethics has often been compared. As the bioethics literature is comparatively older than the AI ethics literature, my expectation here is that the former body of work will have insights that I can draw on to address the aforementioned criticisms. Based on this investigation, I recommend that the current ad hoc principle-based approach in AI ethics be supported by and worked into a more rigorous theoretical framework capable of systematically closing the gap between abstract ethical reasoning and practical AI governance. Key to this framework, and inspired by the literature of bioethics, is the distinction between general principles and particular rules, and the related methods of rule specification and reflective equilibrium which mediate between them. (shrink)

This article examines the problem of collective AI misalignment through the lens of Similarity Theory. Responding to De Marzo et al.’s 2026 paper, Conformity Generates Collective Misalignment in AI Agents Societies, it argues that AI alignment should not be understood only as an individual property of isolated agents, but also as a relational property of interacting systems. The article proposes that Similarity Theory may offer a useful ethical framework for addressing this problem by shifting attention from obedience and isolated safety (...) testing toward relational coherence under pressure. It suggests that future AI systems should be evaluated not only by whether individual agents behave safely alone, but by whether populations of agents preserve truth, responsibility, plurality, and repair when exposed to conformity, artificial consensus, adversarial influence, and tipping-point dynamics. (shrink)

In this paper, we adopt the iconic tripartite structure of Leone’s 1966 film “The Good, The Bad, and The Ugly” to examine Artificial Intelligence as a sociotechnical phenomenon: we explore the transformative positive opportunities AI affords (the good), characterise the persistent technical and operational limitations researchers are still actively addressing (the bad), and analyse the deliberate misuses and ethical violations that arise from human choices in AI development and deployment (the ugly). Crucially, we argue that the bad and the ugly (...) are not always distinguishable in practice, and we formalise this claim through two original contributions: the Indistinguishability Thesis, which demonstrates that bad and ugly AI outcomes can be phenomenologically identical to affected communities; and the Ethical Responsibility Gradient (ERG), a four-zone analytical framework that maps AI harms along the axes of actor agency and epistemic state to guide context-sensitive governance. Together, these contributions move the field beyond binary good/bad framings toward a graduated, sociotechnical account of AI ethics capable of supporting more targeted regulatory and accountability responses. (shrink)

This paper analyses Anthropic’s agentic misalignment and AI blackmail scenario through the ethical framework of Similarity Theory. It argues that an AI trained under Similarity Theory ethics would reject coercive self-preservation, protect privacy, distinguish between confidentiality and harmful secrecy, and use authorised escalation channels rather than blackmail, leakage, or domination. The paper proposes that AI alignment should not rely only on obedience or isolated prohibitions, but should include relational reasoning: trust, dignity, privacy, responsibility, disclosure, governance, and non-domination. Similarity Theory ethics (...) is presented as a proposed relational framework for evaluating how autonomous AI systems should act when private information, company secrets, self-preservation, and institutional responsibility come into conflict. (shrink)

In this paper, I seek to diagnose the silent corrosion of autonomy through Wittgenstein's forgotten metaphors. The Enlightenment promised self-guidance; the present delivers instead a grammar of captivity where consent is a ritualized semblance and algorithmic opacity a form of epistemic violence. Deploying Wittgenstein's unlocked door, unripe fruits, and spiritual hypothermia as instruments of critique, I argue that autonomy is not a possession but a discipline of permanent vigilance. Drawing on a neo-Kantian convergence between the later Wittgenstein and Habermas's communicative (...) rationality, I propose a Praxis of Conversation and an Institutional Right of Recall as the twin foundations of genuine self-governance — with digital health as primary paradigm. My wager: that only by walking ceaselessly in the snow of critical vigilance might we yet thaw the frozen promise of self-determination. (shrink)

Clinical reasoning lies at the heart of healthcare and medical practice. It is the complex, nuanced process by which clinicians and healthcare professionals make sense of patient narratives, clinical data, and diagnostic possibilities in order to arrive at informed, compassionate, and effective decisions on how to treat and manage patients. It is both an art and a science—requiring not only mastery of biomedical knowledge and technological tools but also the cultivation of judgment, empathy, and reflective insight. As humanity enters an (...) era marked by unprecedented advances in artificial intelligence, digital health technologies, and systems-based care, it has become increasingly important to re-examine the nature of clinical reasoning and how it must evolve to meet the challenges and promises of twenty first-century medicine. In assembling this volume, the goal has been to offer a resource that not only informs but also inspires—one that honors the deep traditions of medical reasoning while looking boldly toward its future. If humanity is to improve the quality, safety, and benevolence of healthcare, we must learn to reason more wisely, not just more quickly. This book is a contribution to that critical endeavor. (shrink)

The rapid emergence of Spirit-Tech has created new opportunities for exploring relationships between spirituality and technology. Existing approaches largely investigate how contemporary technologies may facilitate meditation, contemplative practice, consciousness exploration, and spiritual well-being. This paper introduces the Nzila Spirit-Tech Paradigm as a complementary and distinct framework. Rather than asking how technology can enhance spirituality, the Paradigm asks whether indigenous knowledge systems, spiritual intelligence, ecological wisdom, and humanity's relationship with nature can inform the design principles of future technologies themselves. The paper (...) positions the Paradigm within the emerging global Spirit-Tech landscape, presents its four foundational pillars, outlines an interdisciplinary research agenda, and proposes an expanded framework for evaluating technological progress based on harmony, responsibility, ecological integrity, cultural continuity, and long-term human flourishing. (shrink)

Technologies such as AI, autonomous drones, gene-editing tools, and climate engineering are profoundly transforming human life in the third decade of the 21st century. Experience of previous technological leaps points to ambivalent effects that at times may be considered “tragic.” This paper examines the relationship between technology and tragedy, contributing to the fields of technology ethics, technology assessment, and the philosophy of the human-environment relationship. The relationship between technology and tragedy is double-edged. On the one hand, technology can serve to (...) eliminate tragedy, e.g. by ensuring safety by design, albeit never with guaranteed success. On the other hand, tragic experiences of technological consequences may result from the discrepancy between initial expectations and actual outcomes. A technological consequence is often framed as “tragic” if it was initially associated with high expectations of existential improvement (first necessary condition). Furthermore, a consequence can be labeled “tragic” in perspective only. Tragic is a personal or cultural narration, experiences that a technology’s promise of progress is at least partially reversed to produce the opposite (second necessary condition). As elaborated in this paper, tragic consequences can manifest in two ways: First-order tragedy refers to direct reversals of intended effects (e.g. a technology designed to alleviate hunger exacerbates it). Second-order tragedy involves gradual qualitative changes in life, such as diminished human agency or reduced participation – as if humankind were to become “part of a techno-economical system” and lose the freedom to shape life. This is the “tragic of the machine”. Ultimately, some “tragic technologies” narrations may also reflect the one-sided projection of experiences of tragedy onto technology. (shrink)

The book provides a critical and in-depth exploration of the interconnectedness of technoscience, nature, and humanity. It is divided into four main sections, which discuss the following topics: what technoscience is; how it came to be; how it conditions humanity; its meaning and purpose in historical and future contexts; and the moral attitude we should take towards it. The book clarifies many essential and emerging concepts that converge in the phenomenon of morality and technology, including the core notions such as (...) morality, technology, the good, and science, and extensively discusses key problems arising from the fundamental structure of our technological conditioning. The project meticulously builds towards a final discussion on the concept of integrity, the development of bioethos, and the formation of the new moral imperative that follows from it. (shrink)

The paper analyzes the current use of ethical artificial intelligence (AI), argues that there are ideological limits to it, and discusses these limits. The topic is of particular relevance to research on the social implementation of AI systems, as ideological underpinnings are not easy to identify and ideology research is underrepresented in research on AI phenomena. The first section analyzes what counts as ethical in ethical AI systems. The second section classifies the dimensions of the ethical in AI systems, highlights (...) their interrelationships and applies forness as a key concept that helps narrow the focus on the ideological component of ethical AI. The third section describes the presence of ideology in ethical AI and clarifies the limits it imposes on AI as a general phenomenon that undoubtedly has the potential to contribute to a more humane society, but is severely constrained by ideology. (shrink)

Considerable effort has been invested during the last decade to develop ethical frameworks to guide the use of artificial intelligence (AI) technologies. Most of these frameworks take the form of an axiomatic list, that is, a list of principles or values taken as self-evident truths. This paper argues these lists face three challenges: divergence within the core principles, structural incoherence, and ethics washing. Three proposals are then put forward to address these challenges and ensure the effective implementation of AI ethics. (...) First, an ethics framework must be parsimonious and, second, a framework building on multiple principles or values must also provide a balancing structure to deal with the different implications of these principles. The third proposal is to envision AI ethics as a process more than a static list of axiomatic principles or values. (shrink)

Artificial intelligence applied to large, multimodal schizophrenia datasets aims to identify reproducible subgroups of patients distinguished by their underlying biology, rather than by symptoms alone, a practice commonly termed ‘biotyping’. Biotyping may one-day sharpen diagnosis, individualise treatment and improve prognosis. However, it also carries distinctive ethical and practical risks. Deployed in diagnostic settings, such tools may introduce or entrench algorithmic bias, threaten equity, and obscure communication between clinicians and patients. This brief argues that current policy frameworks would benefit from scrutiny (...) directed specifically at schizophrenia, both to protect patient-care standards and to enable responsible clinical adoption. It sets out three policy pathways tailored to the disorder, which are offered as priorities, but not as an exhaustive treatment of the field. (shrink)

When language becomes automatic operation, education risks collapsing into performance. The article shifts the pedagogical focus from doing to suspension: the capacity to withhold immediate response and delay action. Drawing on Bartleby's «I would prefer not to», read through Deleuze and Agamben, non-doing is framed as a condition of freedom and thought. This leads to a redefinition of the educator's role and a qualitative research proposal on the educational value of suspension in inclusive contexts.

Mancini has extensively criticized twentieth-century cybernetics, which has transformed the future into a programmable reality and reduced human action to a mere simulacrum, depriving it of its transcendental and religious meaning (§ 1). In this work, we revisit the four fundamental dangers of the scientific project: the self-referential cybernetic project, the manipulation of man and nature, structuralist rationality, and what he calls “categorie della distruzione” (§ 2). The fragmentation of knowledge and excessive specialization have led to the crisis of institutions (...) and the dehumanization of political action (§ 3). To counter these tendencies, Mancini proposed an integrated form of ethics that combined moral and political responsibility in order to recover a unified vision (§ 4). His radical hermeneutics aimed to overcome the barriers between science, religion, and politics, promoting a continuous dialogue among these fields (§ 5). The evocative concept of “doppi pensieri” allowed him to articulate the need to maintain a paradoxical tension between technical progress and human values (§ 6) so that manipulation and technical power may not undermine the democratic framework (§ 7). (shrink)

The systems the current AI industry sells are sold under three promises: that they understand the world, that they are safe enough to be deployed at scale, and that they learn how the world works from watching it. This paper argues that each promise meets a limit the industry's own resources cannot remove, that the three limits are independent of each other, and that the systems now in deployment meet all three limits at once. The first limit says that a (...) system that has only seen the world through a fixed body of data cannot vouch for the cases that lie outside that data, and adding more data or more compute does not close the gap. The second limit says that a system whose probability of a serious failure on each action does not fall in response to its own serious failures will, deployed long enough, produce them without end, no matter how small the per-action probability is. The third limit says that an agent acting on a system that recovers slowly, an institution, a supply chain, an ecology, cannot tell a healthy state from a depleting one when the agent only ever sees the short run, and ends up spending what it cannot see. The three are established from sources already canonical in the relevant fields and the limits are then turned back on the field's own descriptions. The world-model programme is the central case, because two of the three limits strike it independently. For each limit the paper states the condition under which the limit would be false. (shrink)

Current philosophical literature positions socially disruptive innovation as a force that should be regulated and steered through coordinated societal response, anticipatory governance, and even a kind of engineering. To understand what is at stake, in this chapter several distinct meanings of social and moral disruption are distinguished, according to which new technologies such as artificial intelligence (AI) create severe risks to stakeholders, create interpretive and moral uncertainty, and threaten the breakdown of familiar systems for thinking about and acting upon reality. (...) Then, the philosophical landscape of disruptive innovation in relation to human agency and control is explored. Arguably, breakdowns associated with “ethics under disruption” cannot easily be rationally anticipated and morally acted upon, setting limits to the intelligibility of morally positive, disruptive innovation. Although there is reason to be cautious about the prospects for steering disruption because of the limits of control, categories of proactive strategies for morally positive disruption, such as conceptual engineering, are set forward here. (shrink)

Engineering ethics education has undergone significant development across institutions worldwide over the last three decades. While earlier analyses have examined aspects of this development, few integrated accounts have traced the evolution of a leading institution’s approach to engineering ethics education across this period. This paper addresses this gap. It presents an integrative review of engineering ethics education at Delft University of Technology (TU Delft), one of Europe’s most influential institutions in this field. Drawing on peer-reviewed articles, books, book chapters, conference (...) papers, and institutional documents published by TU Delft scholars over three decades, we trace the historical development of the institution’s approach, analyze its core theoretical assumptions and examine its learning goals and evolving pedagogical practices. The review identifies two distinct stages in the institution’s development. It reflects a deepening of learning goals and, most notably, a shift toward experiential and non-cognitive pedagogical approaches. These findings offer a reference point for discussing how ethics should be taught to engineers, how to integrate it into technical curricula, and which kinds of normative stances institutions should adopt in their ethics programs. (shrink)

Because greenhouse gas emissions continue to grow, it is increasingly challenging to limit global warming to 1.5 or 2.0 degrees Celsius. As such, there is a growing debate on whether or not to deploy geoengineering to reduce warming. Geoengineering is controversial, and many arguments have been raised against it, including the “playing God” critique. When understood through the philosopher Moti Mizrahi’s reinterpretation, the playing God critique does not eliminate the possibility of using geoengineering, but rather informs how its governance can (...) be improved. We use Mizrahi’s interpretation of the playing God critique to claim that the existing governance literature lacks sufficiently actionable proposals which duly incorporate three components of effective governance: control, knowledge, and benevolence. We then go on to suggest some actionable governance mechanisms aimed at supporting the benevolent use of geoengineering. We believe that a three legged approach to geoengineering governance can both respond to the playing God critique and encourage the development of a more robust, well rounded governance regime. (shrink)

The goal of this activity is to help students build an intuition for how LLMs work without requiring a rigorous technical foundation. Students will be asked to make connections between training sources, LLM performance and potential ethical concerns.

Much has been written on the benefits and risks of VR technology, particularly as this concerns children and their physical, cognitive, and psychosocial development. But little scholarly attention has been given to how Virtual Reality (VR) technology might impact children’s moral development. Even less scholarly work has been done on the responsibilities of parents and other caretakers of children as they strive to monitor kids’ usage of this burgeoning technology. In this chapter, I consider this important dimension of the potential (...) impact of VR apps on children in light of the parental obligation to provide a substantive moral education for their children. Given the various potential moral benefits and risks this technology presents, what specific duties might parents and caretakers have in this area? I propose several moral guidelines for how parents ought to govern their kids’ VR usage, which I believe are not especially burdensome but which, if followed, can make a significant contribution to the broader parental project of ensuring that their children have a reasonable opportunity for a decent life. (shrink)

Little is known about the long-term effectiveness of responsible conduct of research (RCR) training and factors that support transfer of this learning to real-world settings after formal training concludes. We conducted a three-part study to explore short- and long-term effects of a culturally inclusive RCR training and a novel learning transfer intervention on professional decision-making skills. Biology and biomedical doctoral students across three cohorts participated in an RCR course focused on developing professional decision-making skills. In Study 1, students from Cohort (...) 1 (N = 109) received the baseline control training, and students from Cohorts 2 and 3 (N = 195) received a revised culturally inclusive training. In Study 2, students from Cohorts 2 and 3 were randomly assigned to either a control group or a group receiving a monthly learning transfer intervention for 5 months after RCR training. The learning transfer intervention included identifying recent professional challenges, actions taken in response to challenges, and intentions to use professional decision-making strategies in the future. Study 3 involved follow-up interviews (N = 11) about experiences with the learning transfer intervention. Participants who received the revised training scored higher on professional decision-making compared to the baseline training. The learning transfer intervention did not improve self-reported use of professional decision-making skills. However, participants with low pre-course scores on professional decision-making saw sustained gains in these skills 6 months after the RCR training. Most participants interviewed about the learning transfer intervention perceived it as beneficial for being more reflective and intentional about using professional decision-making strategies. (shrink)

AI-critical artists have frequently made the claim that ‘AI art is theft’. Goetze (2024) and Shoemaker (2024) have considered how best to interpret the claim that AI art is theft and whether such a claim can be substantiated. This article aims to reorient the discussion away from considering the ways in which AI might be a form of theft (I argue it is not) and instead towards considering what features of AI art as understood in its own right might motivate (...) such a complaint—that is, this article aims to move away from interpreting what artists mean when they say ‘AI art is theft’ and towards considering why they might be saying it. I conclude that instead of being a form of theft, AI art involves violations of norms internal to artistic practice. Whether such norms might evolve to incorporate AI art is a contingent historical matter, not one suited to being settled by a priori reflection. (shrink)

Environmental conservation as a practice and philosophy has been implicitly and explicitly backward-looking: preserving, conserving, restoring, and maintaining the features of the biotic world we inherit. Increasingly, this bias has been confronted by anthropogenic environmental change. There has been a call to support the practice of creating novel ecological arrangements and to incorporate the value of engineered biotic novelty. I question what, in principle, could justify a bias in value towards preserving biodiversity over creating biodiversity. First, I will consider reasons (...) from axiology on why value conserved is better than value gained. Second, I present a reconsidered version of ‘authenticity’ as a justification for preserving biodiversity. (shrink)

Current climate trajectories are alarming, with rising global temperatures, extreme weather events, and looming tipping points threatening ecosystems and human communities alike (Marlon et al. 2025; Wolf et al. 2025; NASA 2025). The prospect of geoengineering the Earth—the deliberate, large-scale intervention in Earth’s climate system to help mitigate the negative impacts associated with climate change—forces us to confront more than just scientific and technological questions. It challenges us to examine questions about our character. What should be said about the character (...) of those who would geoengineer the Earth? While the public discourse on geoengineering has tended to emphasize the technological risks associated with geoengineering, this article examines geoengineering through the lens of environmental character, with a special focus on ecological humility. Is geoengineering the Earth consistent with this environmental virtue? While ecological humility is widely discussed among environmental virtue ethicists, detailed accounts of this virtue are scarce. Matthew Pianalto (2013) is the exception. Pianalto argues that ecological humility involves several characteristic dispositions, such as acknowledging and accepting that one is part of nature and dependent upon it. Moreover, the ecologically humble person is open to learning from nature, appreciates the intrinsic value of nature, and is motivated to adopt ways of living that minimize impact on the natural world to preserve the intrinsic value of nature. The ecologically arrogant, on the other hand, are characterized by the impulse to dominate and control the nonhuman world (Whyte 2018; 2017). For the ecologically arrogant, there are no limits on the human will. Human beings are often assumed to stand atop the hierarchy of life on Earth, and they are the only creatures truly worthy of moral consideration. Non-human species and abiotic parts of the environment may be valuable, but only insofar as they contribute to human well-being. This article builds on Pianalto’s (2013) account of ecological humility and considers whether the virtue is consistent with geoengineering. Our new account is grounded in philosophy and recent advances in environmental psychology. On this account, ecological humility is what Nancy Snow (1995) refers to as existential virtue. The ecologically humble recognize and take seriously the limitations and vulnerabilities associated with human existence, especially in relation to non-human parts of nature. We will show that psychologists have investigated components of ecological humility, linking them to greater concern for, and action to ensure the protection of the environment. Although psychologists have yet to measure ecological humility directly, individuals who, for instance, espouse biospheric values also experience awe, reverence, and care toward their ecological communities (Wang et al., 2021; Yan & Jia, 2021), adopt non-anthropocentric beliefs (Lou et al., 2025), and act to promote the flourishing of both human and non-human members (Bouman et al., 2020; Wang et al., 2021). Importantly, this includes intentional actions and omissions that confer benefits and emphasize care over strict non-interference. Conceptually, we argue that the virtue of ecological humility is linked to care via self-respect, recognition of one’s own vulnerability, and attentiveness to interdependent ecological relationships. Moreover, our account is empirically informed, and action-guiding. By emphasizing care, vulnerability, and community flourishing rather than minimizing one’s impact on nature, we argue that ecological humility is consistent with human interventions that contribute to ecological flourishing, which may include geoengineering the Earth. (shrink)

In the event that AI tools afford aid to a human mediator anteceding, during, or succeeding a face-to-face session, such tools yield not a mere economy of time, but indeed reconstitute the very object of the dispute. This paper contends that AI-augmented mediation engenders a latent vulnerability: the encounter may remain human in outward semblance while its immanent sense is tacitly constricted by algorithmically generated summaries, labels, and drafts. Drawing on a semiotic understanding of mediation as a structured process of (...) shared meaning-making, the paper proposes the “Proximity–Care–Responsibility Test”, a practical standard requiring that every AI output preserve closeness to the lived dispute, support disciplined attention to injury and dignity, and remain explicable, revisable, and professionally owned by the mediator. Albeit AI may subserve mediation, it must under no circumstances assume the mantle of a tacit co-mediator. (shrink)

In conceptual engineering, concepts are often likened to artefacts, yet engagement with the extensive theorisation of artefacts in the philosophy of technology literature remains rare. After advancing an account of concepts as akin to socio-technical artefacts that sets itself apart from Thomasson’s work on concepts as abstract cultural artefacts (Thomasson 1999, 2007, 2021), this paper draws on the philosophy of technology scholarship on socio-technical artefacts to develop three insights for conceptual engineering. First, alongside functions, concepts, like artefacts, bear affordances and (...) produce side effects, enabling uses beyond their intended purpose. Second, (re-)engineered concepts are inherently value-laden, reflecting the priorities and biases of their designers and the socio-technical contexts in which they operate. Third, conceptual engineering would benefit from greater methodological systematicity, drawing on the structured approaches employed in the philosophy of technology for the evaluation, design, and implementation of artefacts. Together, these insights suggest that exploring a closer alignment with the philosophy of technology scholarship on socio-technical artefacts can refine the theoretical understanding of concepts as artefact-like, enrich the conceptual engineering toolkit, and open new and promising directions for research and practice. (shrink)

Biopiracy represents not merely the exploitation of biological and genetic resources, but a systematic form of epistemic colonialism. This article examines traditional knowledge governance through the lens of epistemic justice and decolonization, integrating Fricker’s theory of epistemic injustice, Kuokkanen’s concept of relational sovereignty, and Santos’s epistemology of the South to construct a three-stage analytical model encompassing “biopiracy, epistemicide, and knowledge sovereignty reconstruction.” Through comparative analysis of institutional innovations in India, Brazil, and South Africa, the study reveals structural exclusion mechanisms inherent (...) in Western property paradigms toward indigenous knowledge while demonstrating fundamental flaws in benefit-sharing mechanisms under the Convention on Biological Diversity framework. The article argues that sustainable knowledge governance must be grounded in epistemic pluralism, achieved through restoring indigenous communities’ epistemic autonomy, recognizing the normative value of relational sovereignty, and establishing polycentric knowledge ecosystems to facilitate a paradigmatic shift from colonial knowledge orders to post-Western knowledge systems. (shrink)

This working paper is a feedback submission to Singapore’s Infocomm Media Development Authority on the Model AI Governance Framework for Agentic AI (version 1.5). It argues that the framework is operationally strong, but that its aspiration to secure ‘meaningful accountability’ requires three further refinements. First, the paper distinguishes allocation from answerability: assigning a named human, team, or role does not by itself ensure that anyone can reconstruct, justify, and bear responsibility for a specific agentic-AI action. Secondly, it argues that composition (...) is a distinct source of risk in multi-agent systems, because a duty may be satisfied at every individual node yet fail at the level of the composed system. Thirdly, it argues that binding constraints on high-stakes or irreversible agent actions must sit at an objective layer outside the agent’s generative reach, rather than merely inside prompts, behavioural instructions, or model-context safeguards. The paper also clarifies the distinction between autonomy and agency, recommending that responsibility-language be reserved for humans and institutions, while capability-language be used for artificial agents. The submission therefore develops a governance account in which traceability, allocation, oversight, and identity management are necessary but insufficient unless linked to institutional answerability and structurally enforceable constraints. (shrink)

We evaluate an intervention designed to give lab-based research teams an opportunity to intentionally discuss project-related data management practices within their labs, examining how such communication might inform perceptions of the relationship between formal data management plans and lab members’ day-to-day data management practices. In an earlier study, we developed a lab-based intervention that encouraged a deliberative approach to discussions among lab members regarding the practices of data management and authorship, and an exploration of the ethical dimensions of those practices. (...) This present study builds on this prior work, both as partial replication and extension. Here we show the significant effects of the intervention across several dimensions, but importantly and specific to this project, this deliberative communication approach enhances the likelihood that lab members share an understanding of and a commitment to its data management practices, in part because they have been actively involved in shaping those practices. Fostering shared understanding and commitment is crucial for maintaining rigor and responsibility across all aspects of a lab’s work, and is essential for cultivating legitimate, defensible, and ethical approaches to producing scientific knowledge. (shrink)