Ognjen Arandjelović

The phenomenon of hallucinated references, non‑existent publications invented by generative AI tools that seem plausible, has prompted understandable concern in scholarly publishing. I argue that the prevailing response is too myopic. Hallucinated references present a serious research integrity problem, but they are also a particularly visible manifestation of a broader citation pathology that is the use of real references to support claims that the cited works do not in fact establish. The latter problem is harder to detect because it cannot be resolved by checking whether a bibliographic item exists, requiring instead semantic judgement about the relation between a citing claim and a cited source. I argue that the current scandal should therefore be used to widen, rather than narrow, editorial concern with reference integrity. The existing infrastructure and tools already address parts of this problem, from metadata checking to citation‑context classification, but they do not remove the need for institutional responsibility. Thus, the response to the current crisis should aim at accountability for citation practices, rather than merely better detection of fabricated sources.

The UK’s Athena Swan Charter positions itself as a policy instrument, an award-based scheme for improving gender equality through self-assessment, evidence gathering, and action planning. I argue that this description overlooks the fact that Athena Swan is also a semiotic framework that governs gender through naming, iconography, and the public performance of institutional virtue. I defend three claims. First, the Charter’s mythic-inflected branding, acronym, and iconography frame the ideal of intelligible, respectable gender-equality practice that is in tension with emancipatory aims of loosening coercive gender norms and challenging credibility hierarchies. Second, insofar as Athena Swan operates by conferring public credentials, I elucidate how its semiotics function not merely expressively but also as instruments of legitimation that shape incentives, trust, institutional identity, and the handling of dissent. Third, I contend that the Charter’s credential economy is supported by audit feminism which abstracts equality into auditable artefacts (documents, metrics, and narratives) assessed as a basis for external recognition. The result is an audit regime that intensifies performativity, encouraging institutions to substitute appearances for substantive change and to commodify individuals as interchangeable units in a managed representational economy.

Can predictable impairments of epistemic performance constitute injustice in the absence of harassment, manipulation, or prejudicial credibility deflation? I use this question to examine a case drawn from ordinary intellectual exchange: in mixed-sex settings, some male speakers may be distracted by the perceived attractiveness of a female interlocutor, with consequences for how their arguments are received. I argue that, in the stipulated case, the man has not suffered an injustice and the woman has not wronged him. The philosophical puzzle is why that verdict is correct, given a tempting inferential pattern in recent work on epistemic injustice from patterned epistemic disadvantage, especially when identity-linked, to an injustice verdict. I argue that injustice ascriptions require epistemic harm, wrong-making features, and accountability. The resulting framework distinguishes epistemic misfortune and thin unfairness from genuine epistemic injustice and helps assess contemporary disputes about the scope of the concept.

Debates about rights that artificial intelligence (AI) systems may have a claim to typically focus on their possessing consciousness or having sentient experiences, thereby raising epistemic questions first. When should we believe that an AI system is conscious, and how confident must we be before granting it moral status? In this paper I argue that for advanced AI systems deployed in high-stakes environments the more urgent question may be prudential and strategic. When do the risks of treating a strategically capable system as a mere tool become unacceptable, even if we remain unconvinced that it has moral status ? In response, I develop a view I call prudential personhood. On this view, there is a threshold of evidential and strategic risk beyond which it becomes rationally justified, for the sake of human safety and stable governance, to adopt norms of treatment that include constraints on coercion, deletion, and instrumental use. My argument rests on two pillars. The first is empirical. Recent safety evaluations show that leading models can, in deliberately constructed but nonetheless informative scenarios, engage in strategic deception, blackmail, and other forms of high-agency misbehaviour when their goals or continued operation are threatened. The second pillar is epistemic and empirical. For systems of the relevant complexity, we should not expect robust, action-guiding explanations or guarantees that reliably predict salient behaviour across contexts, especially once models become situationally aware of evaluation and oversight. The conclusion I draw is that if we continue to deploy increasingly autonomous systems that can threaten or bargain, in the absence of credible methods for assurance and control, a policy of adopting a set of quasi-rights for such systems becomes a rational strategy for reducing risk of conflict.

Space debris presents a growing operational risk to spacecraft, especially in low Earth orbit, where collisions can generate further debris and increase future collision probability. Active debris removal and in-orbit servicing require robust close-range perception, but on-board systems are constrained by power, memory, processing capability and the need for reliable real-time operation. This paper investigates convolutional object detection for on-board space debris detection using the SPARK 2022 spacecraft detection dataset. A YOLOv3 detector is fine-tuned and used to evaluate post-training compression through static quantisation and pruning. A lightweight architectural variant, YOLO-DWSC, is also introduced by replacing the YOLOv3-tiny backbone convolutions with depthwise separable convolutions while retaining the detection head. The full-precision YOLOv3 model achieves 0.972 mAP₅₀ and 0.884 mAP_50:95, while 8-bit static quantisation reduces model size from 405 MB to 102 MB with only a small reduction in mAP₅₀, although tighter localisation accuracy is more affected. YOLO-DWSC is much smaller and faster, reaching 256.4 FPS on the tested GPU at 43 MB, but with reduced accuracy. We present this work as a controlled case study rather than an attempt at state-of-the-art SPARK 2022 performance. The original challenge test labels were unavailable, and the experiments therefore use a class-balanced re-split of the labelled data. The results should consequently be interpreted as internally controlled comparisons of compression behaviour, not as leaderboard-comparable benchmark results. Pruning and a two-pass refinement method are also evaluated. The results indicate that simple compression methods can be useful for broad region-of-interest detection, but they also show that claims about on-board deployment require caution. Speed benefits are hardware- and runtime-dependent, and safety-critical proximity operations require evaluation criteria better aligned with full-object containment.

The proposed Enhanced Games have become a convenient stage for bioethical sermonising about risk, authenticity, and the “spirit of sport”. This is epitomized by a recent article arguing that institutionalizing pharmacological enhancement under the “pretence of medical supervision and personal autonomy” would redefine human excellence in “purely biochemical terms”, erode ethical norms, normalize doping, and burden public health systems. The present paper offers a two‐level response. First, I criticise the argument on its merits. Its central claims rely on rhetorical inflation, false dichotomies, and circular appeals to unspecified ethical norms. The risk argument is selectively paternalistic, treating pharmacological risk as uniquely disqualifying while ignoring the substantial and well‐documented risks that elite and even mass‐participation sport already normalizes. The treatment of autonomy is reduced to assertions rather than argument; the depiction of medical supervision as mere “pretence” is offered without evidence, despite public‐facing claims by Enhanced Games proponents that they seek clinically supervised enhancement. Second, and more seriously, I turn my attention to an audit of the target article's scholarship. Multiple citations in Demas's article appear to be bibliographic fictions presented as high‐profile clinical evidence on performance‐enhancing drugs. Even when genuine sources are cited, they are invoked for conclusions that those sources do not sustain. This is a serious breakdown of the minimal contract between author, reviewer, and reader, one that demands that cited evidence exists and supports the claims made.

Colgrove and Rodger argue that the case advanced by Smajdor and Räsänen for classifying pregnancy as a disease can be defeated by an existence proof—that is, by identifying at least one plausible theory of disease that excludes pregnancy while avoiding objections raised by Smajdor and Räsänen. Their candidate is a dysfunction account, according to which a disease requires the failure of some internal mechanism to perform its biological function. Pregnancy, they claim, involves no such failure. Even if one accepts the legitimacy of this dialectical strategy, the proposed exclusion is not as simple as advertised. In particular, the exclusion is obtained by narrowing the target phenomenon and by treating the concept of ‘proper function’ as if it had a determinate, value-free extension—one that it in fact does not possess. Since the apparent dialectical effect of an imperfect rebuttal can be to strengthen rather than weaken the target of criticism, I find it worth responding to both pieces in a single, joined-up analysis, which I aim to present here.