By Ido Ramati (The Hebrew University of Jerusalem)

What is a voice in the age of artificial intelligence, and how does AI vocality reshape the technē of the human voice? This paper argues that AI-generated voices should not be reduced to either human expressions or machine outputs. Instead, they constitute a hybrid formation resulting from a human–machine operative ontology that unsettles the traditional assumption that a voice belongs to a singular speaking human subject.

AI systems that generate speech operate by extracting, segmenting, and recombining recorded human voices, producing utterances that are at once familiar and dislocated. Behind every AI vocality, there is at least one human whose voice is algorithmically manipulated. AI voices emerge from a distributed network of human performances, datasets, and algorithmic processes.

To account for this condition, the paper introduces the concept of the “thirdness” of the voice: a co-constituted formation in which human and machine agencies are inseparably entangled.

Drawing on contemporary media theory, the study of cultural techniques, and current philosophies of the voice, this framework extends the idea of the voice as inherently relational, showing how AI systems reconfigure this relationality at the level of technological mediation.

Building the analysis on both contemporary and historical cases, the presentation first analyzes the creativity of the voice on social media, focusing on TikTok videos that engage with synthetic voices. It then situates these practices within a longer history of vocal mediation through phonography, radio announcing, and cinematic voice-over. I argue that AI algorithms transform what a voice is by changing how it is made: the techniques, practices, and forms of skilled production through which voices are trained, performed, and recognized. By situating contemporary AI vocality within this broader genealogy, the paper reconceptualizes the voice as a distributed, relational, and programmable phenomenon, contributing to debates in the philosophy of technology on mediation, hybridity, and the reconfiguration of human expressive capacities.

By Sebastian Lederle (Bauhaus-University Weimar)

This paper proposes Hans Blumenberg’s concept of technicisation as a resource for rethinking artificial intelligence within the philosophy of technology. Against the widespread framing of AI as an external instrument, partner, or quasi-agent added to an otherwise intact human subject, I argue that Blumenberg allows us to understand AI as the late visibility of a more basic technical condition: the operational constitution of perception, behaviour, thinking, and self-description.

The point of departure is Blumenberg’s essay “Life-World and Technicisation under Aspects of Phenomenology,” read together with his wider writings on technology. Blumenberg’s decisive thesis is that technicisation cannot be adequately derived from conscious intentionality, as if technology were merely the result of human plans and purposes. Rather, technicisation becomes effective in advance of reflective consciousness, as an infrastructure that organizes what can appear as obvious, natural, practicable, and true. In this sense, the life-world is not the untouched sphere opposed to technology. It is the sphere of self-evidence in which technicisation becomes habitual, while at the same time depriving the life-world of its character as simply given nature.

This reversal is crucial for AI. Contemporary AI systems present themselves as machines of fitting: they translate, predict, recommend, classify, write, and respond in ways that appear adapted to human needs. Yet from a Blumenbergian perspective, this apparent fit is also a suggestion. AI does not simply conform to human capacities; it makes visible the extent to which those capacities have already been rendered operationalizable, automatable, and detachable from specifically human performances. The transition from automatability to operationalizability, already diagnosed by Blumenberg in the 1960s, thus becomes central for interpreting generative AI. Evidence itself is no longer presupposed as a pretechnical given, but appears as technically produced, stabilized, and made usable.

The paper therefore treats AI less as a rupture than as a late “humiliation” of human exceptionalism. What is unsettled is not only whether machines can imitate intelligence, language, or creativity, but whether the human ever possessed these as a pure, technology-free domain. AI confronts us with the fact that our self-understandings are made possible not despite technical operativity, but through it. Its autonaturalizing tendency lies in everyday use: the more smoothly AI operates, the more it recasts adaptation as the obviousness of the human-machine relation. There is no zero point before AI from which one could judge it without already relying on histories of technicisation that shape judgment itself.

By rereading AI through Blumenberg, the paper contributes to debates on traditions, transitions, and tensions in philosophy of technology. It reframes AI as infrastructure rather than mere tool, as a problem of human–technics relations rather than only agency, and as an epistemological challenge to the self-evidence of the human.

By Floriana Ferro (University of Trieste)

Humanoid robots, especially in their most recent AI-driven developments (Cao, 2025), challenge established categories in the philosophy of technology by blurring the boundaries between object and subject, tool and partner, human and non-human. As their presence becomes increasingly pervasive in social and interactive contexts, the question of agency can no longer be adequately addressed within traditional frameworks. Instrumental approaches reduce robots to passive artifacts (Flusser, 1999; Turkle, 2011), whereas dystopian thinking, often enhanced by science fiction (Meinecke & Voss, 2018), projects the fear that robots might replace humans and become powerful subjectivities. Both positions fail to capture the ontological ambiguity and relational dynamics that characterize human–robot interaction.

This paper proposes a relational and distributed account of agency between in human–robot interaction by bringing into dialogue phenomenology and posthumanism, while grounding the analysis in recent developments in social robotics. My recent work on humanoid robots shows how they can elicit specific forms of empathy through embodied interaction, particularly via perceptual configurations such as gaze and face-like structures (Ferro, 2023; 2026). In this way, they are experienced as “as-if subjects,” whose ambiguity can be understood in terms of embodiment and perceptual expressivity, both grounded in Maurice Merleau-Ponty’s (1968) ontology of the flesh. Unlike more traditional phenomenological accounts of agency, which ascribe it exclusively to self- aware subjectivities (Gallagher & Zahavi, 2020), I argue for a broader concept of agency. Agency is understood as practical engagement with the world, emerging within the lived experience of interaction as a dynamic process, in line with the ontology of the flesh, which destabilizes rigid boundaries between subject and object, human and non-human.

This account is further expanded through non-phenomenological traditions, namely posthuman and relational ontologies. Karen Barad’s (2007) notion of intra-action allows us to conceptualize agency as co-constituted within relational processes, while Bruno Latour’s (2005) actor-network theory reframes it as distributed across networks of human and non-human actants. These perspectives converge in rejecting anthropocentrism and emphasizing relational configurations.

At this point, my work introduces a further theoretical step by engaging with the notion of flat ontology (Bryant, 2011). In this framework, flatness does not imply a lack of depth, but rather the rejection of hierarchical structuring among entities, affirming instead an ontological parity between human and non-human agents. This perspective enables a posthuman reconceptualization of agency as distributed, without privileging human subjectivity.

My thesis is that humanoid robots should be understood not as isolated agents, but as participants within fields of extended corporeality, where agency is distributed, embodied, and situated. From a phenomenological perspective, this implies that humans and robots are not simply related as external entities, but are intertwined within a shared ontological field grounded in an expanded notion of embodiment. Agency is therefore neither located in individual agents nor reducible to relational networks alone, but emerges from the chiasmatic structure of this extended corporeality, where perception, affectivity, and materiality co-constitute the dynamics of interaction. Within this framework, agency is enacted and negotiated in situated encounters and only provisionally stabilized in specific contexts.