Pezzano (2024) departs from two conceptual distinctions in the analysis of properties in both an ontological and an anthropological sense. On the one hand, properties may or may not require a substrate, that is, it must be clarified whether fundamental properties in ontology and anthropology require an entity or subject on which to be grounded. On the other hand, the concept of property is not univocal; on the contrary, we can highlight three typologies: categorical, dispositional and relational. The first derives from the intrinsic nature of the entity; the second from the powers or capacities linked to the development and context of the entity; and the third from the possible relations that the entity can establish with others and itself. In what follows, I will show how the first distinction has been obviated in discussions about moral status in general and why it is central to providing a complete answer. Furthermore, I will argue that the second distinction suffers from a limitation: it fails to incorporate what I will call the overlapping and complementarity dimensions. Once these are considered, I will argue for the need to make the case for a complex anthropology and for a crucial shift in the conception of moral status.

First, some authors have argued that the concept of moral status is founded on an equivocal (Horta, 2017; Rachels, 2004; Sachs, 2011). When we morally consider an entity, we do so not by virtue of the entity that it is, but by certain manifestations of the fundamental property of moral consideration. This thesis is usually associated with utilitarian positions in the following sense: I do not consider an entity morally because it has moral status, but I do so because the fundamental property, sentience, reveals that, at a given time, an entity has interests that deserve to be respected over others. Moral status is understood as the unconditional or objective moral consideration of an entity in virtue of one or more properties (Coeckelbergh, 2012; Kamm, 2008), but moral consideration cannot occur without taking into account the ways in which that property is revealed (conscious states of pleasure, pain, interests, etc.).

In this sense, what matters is not the substrate, the entity, but the property itself and its modes of revelation. This idea is illustrated in Parfit's (1984) thesis on identity: identity is not an additional fact by which I am a person, but the unity of the person is revealed in the degrees of psychological continuity. Similarly, moral status need not be an additional fact associated with an entity, but a property that is revealed regardless of the entity in which it is grounded. For this reason, the debate on the moral status of AI, as with respect to other entities, must address the difficult question of whether moral status is really an adequate concept, since it is intimately linked to the moral value of the substrate through its properties. And it seems that, without the ontological and anthropological notion of substrate, the concept of moral status loses much of its appeal.Footnote 1

Regarding the second distinction, Pezzano offers a brief evaluation in the third section, considering that we have good reasons to reject categorical properties, but this is not the case with dispositional and relational properties. Both constitute interesting alternatives for understanding the kind of properties that are relevant in technological anthropology and moral status.Footnote 2 However, I believe that an assessment that takes each typology in isolation forgets two important dimensions: overlapping and complementarity. This becomes apparent when analyzing each of the typologies separately. Categorical properties do not seem adequate to give an anthropological and moral account of human beings. We are more than intrinsic and essential properties. But the same can be said of the other two. Dispositions do not occur in a vacuum but seem to require categorical properties that drive certain kinds of developments and relations that allow those developments to take place. Relational properties do not seem to work separately either: we need certain intrinsic aspects and dispositions to be able to relate in certain ways (see Latour's speed bump example in my paper, pp. 20–21).

In this sense, it is not only urgent, as Pezzano rightly emphasizes, to undertake a rigorous analysis of technological ontology and anthropology that takes into account the concept of property, but also the ways in which the three types of properties overlap and complement one another. The three typologies require each other, so it is useful to differentiate them well (overlapping) and to study how they combine with others (complementarity). This idea leads to the exploration for models of complex technological anthropology. That is, we should not focus on analyzing which type of property is the most appropriate to explain anthropologically and morally human beings and technologies, but rather what role each property plays and how they interact with each other.

To conclude, I would like to transfer this argument to the debate on the moral status of AI. The literature around this debate has addressed a huge variety of dimensions and positions (Gunkel, 2023; Llorca Albareda, 2023; Llorca Albareda et al., 2024). Nevertheless, I believe that a complex concept of moral status needs to be developed and this is only possible from a complex technological anthropology. And this is due to the fact that, in spite of the various arguments provided, two major alternatives are offered: to seek the fundamental categorical property or properties, or to defend a relational approach. As we have seen, an attentive examination of the concept of property invites us to look for a complex approach, which shows how the different types of properties interplay with each other. To return to the conclusion of my article, there is an urgent need to develop hybrid models of moral status that can deal in a complex manner with the challenge posed by AI.