Very glad to have taken part in the Society for Philosophy and Technology (SPT) 2025 Conference at the Technical University of Eindhoven.

On the 26th of June, I presented a paper titled “Intimacy as a Techno-Human Symbiosis: Reframing the LLM-User Experience from a Phenomenological Perspective.” In the paper, “intimacy” is understood not much as a relational contiguity, but as an onto-epistemological affair, that is, a condition by default marking human and technology’s co-essentiality and co-development.

The paper contains three major takeaways:

1) starting from the idea that today we live in and experience an onto-epistemological crisis of sense, due to the (partial) incommensurability between the language and digital ecologies (this idea is contained in a milder form in my book Quantum Ecology) I look at Large Language Models (LLMs) as a perfect instantiation at the intersection of these two ecologies. Specifically, I propose to reconsider the LLM-User imbrication as a complex (maybe also complicated) affair, that is, one that requires departing from today’s common “essentialist standpoint” which regards LLMs and Users as two essentially distinct poles. By contrast, I suggest adopting a phenomenological (and autopoietic) standpoint that does not pretend to benchmark human vs. AI and vice versa, but rather looks at the LLM-User pair as a whole, thus being concerned with the effects that the symbiotic experience generates on sensing, i.e., knowing and knowledge as embodied practices (works in this direction already exist, cf. Harnad 2024; Ashery 2025; Delacroix 2025)

2) From here, I suggest a research agenda consisting of three major axes:

a) the exploration of the entanglement between prompts and generated outputs, through longitudinal and comparative studies involving different users, LLMs, and writing systems. Such studies would enable to detect either converging and/or diverging meaning-making patterns at the LLM-User nexus, or to provide evidence for the randomization of the LLM-User experience. I would probably put my bet on the first option.

b) longitudinal and comparative approaches to hallucinations aimed at tracing and mapping their emergence across LLMs and contextual uses. Again, such approaches would help either provide evidence for the rhapsodic nature of hallucinations, or reframe hallucinations as traces of broader techno-human onto-epistemologies in the making. On this, it is harder to bet, but the point would remain to focus on the effects that such hallucinations produce on knowing and knowledge as embodied practices.

c) the perceived (by socio-culturally diverse users) affects concerning the fitness and finesse of LLMs outputs. Regardless of a supposed objective quality of the outputs generated, the goal here is to put the LLM-User symbiotic experience into perspective and explore its contextual aptness and perceived robustness, including also an ethical-aesthetic perspective.

3) As one example concerning this last point, I discuss the literary case of 1 the Road (reminiscent of On the Road by Jack Kerouac) the first ever travel novel written entirely by an AI. I argue that this travelogue, dating back to 2018 when LLMs were not there yet, can be regarded as a techno-agent hypomnemata, not only realising Roland Barthes’s idea of intertextuality, but also instituting a synthetic site of techno-human negotiation.