°°°°~x§x-<@>
Introduction
The application of new AI technologies, such as Large Language Models (LLMs) and machine learning (ML), to strengthen democracy must be carefully scrutinized through the lens of Ezequiel Di Paolo’s enactive philosophy and Barbara Stiegler’s critique of expertocracy, as these thinkers challenge the foundational computational and elitist assumptions often underlying such technologies.
Based on their conceptions, AI technologies should be utilized not primarily as tools for optimizing governance or aggregating opinions, but rather as instruments for enhancing participation, fostering critical consciousness, and providing “intellectual arms” for genuinely autonomous public deliberation.
The important key here is to avoid at all costs the intellectual and cognitive proletarianization (loss of know-how, cf. Bernard Stiegler) engendered by a passive use of these pharmakological tools.
1. Utilizing AI through the Lens of Ezequiel Di Paolo’s Enactive Philosophy
Ezequiel Di Paolo’s work, particularly concerning Linguistic Bodies and Participatory Sense-Making (PSM), emphasizes the embodied, historical, and non-representational nature of mind and language. His perspective fundamentally critiques the idea that the mind operates like a computer processing information, a metaphor that often drives AI and ML development. This metaphor can lead to a sense of alienation, where lived experience is reduced to “neurons firing in the brain” and the individual views themselves as a programmed machine, lacking true agency
Therefore, any AI technology seeking to be compatible with an enactive view must prioritize interaction, participation, and ethical agency:
1.1 Critique and Transformation of the Computational Model
AI researchers and developers must engage in a critical investigation of LLMs and ML systems to reveal how they perpetuate the functionalist/computational metaphor, which conceives of cognition as the manipulation of internal representations. The goal should be to articulate positive, non-representational ideas about action and perception that move beyond mere information processing
1.2 Enhancing Participatory Sense-Making (PSM)
Democracy relies on shared engagement and sense-making, which Di Paolo and his colleagues describe as constantly affecting participants’ sense-making
AI could be developed to support this process by:
◦ Facilitating Co-regulation: AI tools could map or analyze the interactive dynamics of large-scale digital deliberations, focusing on the quality of co-regulation and mutual interpretation between participants
◦ Identifying and Resolving Breakdowns: Since genuine participation involves dealing with conflicts and breakdowns, AI could help identify where and how interactive breakdowns occur in digital discussions, offering pathways for “reconnections” rather than simply shutting down dialogue
1.3 Fostering Critical Linguistic Consciousness
Di Paolo and his colleagues highlight the ethical dimension of linguistic bodies, which are perpetually in a state of “becoming” and must navigate the tensions of “incorporation” (of received utterances) and “incarnation” (of other agencies)
AI could be used to:
◦ Expose Ideological Manipulation: LLMs could be trained to identify patterns of “enunciatorless enunciations” (language lacking clear authorship, often characteristic of ideological narratives) or “blanket references to collectives” (“People think that…”) which enhance systemic asymmetries
◦ Cultivating Resistance: AI tools could provide intellectual support for critical thinking by revealing hidden assumptions, prompting users to attune to “feelings of discomfort” when facing ambiguous orders or institutionalized language, thereby strengthening the “powers of dialogic criticism, counterframing, interpellation, creativity, etc.”. This aligns with the principles of critical pedagogy, which teaches that human beings are unfinished creatures capable of collectively becoming conscious of their potential and limitations.
2. Utilizing AI through the Lens of Barbara Stiegler’s Democratic Conception
Barbara Stiegler critiques prevailing conceptions of democracy—particularly the neoliberal version—which reduces it to competitive elections (minimalist/procedural view) and relies heavily on experts and “leaders” to govern the “inadapted masses”. This view promotes the “manufacture of consent” rather than genuine self-governance (auto-gouvernement du peuple)
To strengthen democracy based on her ideas, AI must actively counter expertocracy and support collective, pluralistic deliberation:
2.1 Empowering Autonomous Opinion Formation (Intellectual Arms)
Barbara Stiegler emphasizes that sociological knowledge and science should provide “intellectual weapons” (armes intellectuelles) to citizens so they can deliberate freely and make informed decisions, rather than being forced to delegate authority to experts
◦ AI for Accessible Knowledge: LLMs could be crucial in democratizing access to complex scientific and intellectual productions by translating specialized research (like climate science or economic theory) into accessible forms, thereby enabling citizens to “forg[e] their own opinions autonomously”
◦ Avoiding Prescriptive Answers: Crucially, AI must ensure that scientific information is presented as contributing to understanding the world (“what is possible/probable”) rather than dictating the specific political choice that “it is necessary to make”. The final political choice must remain the object of deliberation.
2.2 Countering the Aggregation of Opinions (Doxa)
Genuine democratic “truths” emerge from a collective elaboration involving a multiplicity of confronting perspectives. AI should therefore not be used merely to aggregate individual choices or predict political results (like the fictional supercomputer Multivac), which fails to produce collective knowledge
◦ Supporting Multiplicity and Confrontation: ML systems should be designed to highlight the multiplicity of perspectives necessary for deliberation, avoiding the imposition of a “unique interpretation of the social world
2.3 Facilitating Collective Inquiry
Drawing heavily on John Dewey’s pragmatist approach, the use of ML could help communities define their shared problems and evaluate solutions collectively, serving the “inquiry” process
◦ Democratizing Problem Definition: AI must resist the tendency of elite groups (like political partisans or certain intellectual reviews) to elaborate the questions and possible responses before public debate. Instead, AI could help diverse publics articulate and define the problems that affect them.
Conclusion
In summary, for AI to enhance democracy according to Ezequiel Di Paolo and Barbara Stiegler, its role must shift from being an information processor that aids elite governance (the functionalist/expert model) to a dialectical tool that supports radical, informed, and participatory self-governance (the enactive/critical model).
°°°°~x§x-<@>
References
Stiegler, B., & Pébarthe, C. (2023). Démocratie ! Manifeste. Lormont : Le Bord de l’Eau Éditions.
Barbara Stiegler – “Le ressentiment : pulsion populaire ou structure de notre grammaire ?”
Di Paolo, E. A., Cuffari, E., & De Jaegher, H. (2018). Linguistic Bodies: The Continuity Between Life and Language.
Di Paolo, E. A., Buhrmann, T., & Barandiaran, X. (2017). Sensorimotor Life: An Enactive Proposal. Oxford: Oxford University Press.
Originals | Ezequiel Di Paolo: Linguistic Bodies and Sensorimotor Agency
Bringing Forth Worlds with enactive philosopher Ezequiel Di Paolo and Mirko Prokop
°°°°~x§x-<@>