Epistechnology operates at the workbench where physical tolerances dictate scientific possibility. It dismisses the scholastic appetite for disembodied concepts alongside the sociology that treats technical systems as arbitrary cultural fashions. Scientific knowledge and vital agency emerge through the stubborn resistance of material objects: the suspension of mirrors in an interferometer, the thermal drift in an amplifier, and the mechanical wear of a cutting tool.
To prevent both speculative drift and technocratic boosterism, our investigations follow three material heuristics, a dialectical counterfactual protocol, and an epistemic demarcation of authorship centered on second-order curatorial acts.
1. The Methodological Triad
Every inquiry is tested against three operational vectors:
A. The Historical A Priori
Transcendental conditions of knowledge do not reside in immutable mental faculties. They materialize in historical apparatuses (phénoménotechnique). A pendulum clock, an optical grating, or a high-vacuum chamber establishes the empirical boundary of what can be perceived and verified. Theoretical concepts do not descend from contemplative mathematics; physical thermodynamics grew out of the condenser cylinder and indicator diagram of the steam engine before Rudolf Clausius formalized entropy as a universal law.
B. The Heuristic of Breakdown
Machines conceal their underlying compromises under frictionless execution. Smooth automation naturalizes the system, seducing the observer into treating the interface as transparent reality.
Breakdown (la panne) shatters that complacency. Operational crisis reveals architecture. When an oil film fails, a relay welds shut, or an automated pipeline throws an unhandled exception, the concealed value judgments of its human designers surface instantly. Breakdown is our primary epistemic probe: failure strips the artifact of its neutral facade, exposing the economic trade-offs, metric tolerances, and discarded edge cases that structured its design.
C. The Immanent Critique of Objectivity
Every measurement rests on an antecedent pact: quantifier, c'est convenir puis mesurer (Alain Desrosieres).
There is no virgin empirical datum. To produce an objective signal, an apparatus executes an incision: it isolates specific observables, groups disparate inputs into equivalence classes, and discards unquantifiable experience as background noise. Objectivity is legitimate strictly within its explicitly operationalized predicates, as Evandro Agazzi demonstrated. Danger erupts the moment this localized operational validity is hypostasized into an ontology, reducing human suffering and social deliberation to computational vectors. Epistechnology audits the incision, excavating what was excised to manufacture the illusion of neutral fact.
2. Authorship, "Distant Writing", and Second-Order Curatorial Acts
The emergence of statistical language models forces a methodological reckoning with the act of composition. Epistechnology treats computational generation not as an autonomous authorial consciousness, but as an externalized, grammatized archive. Following Luciano Floridi's diagnosis of "distant writing" (2023) and Tiegue Vieira Rodrigues's formulation of the Apt Curation Model (2024), human epistemic agency undergoes a decisive structural shift under generative automation: from the first-order generation of text to second-order curatorial acts.
First-order generation operates at the level of token sequence prediction: minimizing cross-entropy loss across an ungrounded textual archive. That mechanical fluency produces syntax without semantic responsibility. The human agent cannot claim epistemic authorship through passive prompting or by surrendering to statistical drift. Rather, following the virtue epistemology of Ernest Sosa, authorship resides in the virtuous execution of four second-order curatorial responsibilities:
- Architectural Design: Instituting the conceptual scaffolding, epistemic boundaries, and material constraints within which technical systems operate.
- Dialogical Refinement: Subjecting automated outputs to iterative dialectical friction, cross-examining machine suggestions against primary archives and physical reality.
- Evaluative Verification: Auditing claims against empirical evidence, identifying stochastic confabulations, and measuring output against the physical limits of the apparatus.
- Integrative Synthesis: Forging disparate fragments into an authentic, coherent thesis anchored in an embodied form of life (Lebensform).
Second-order curation is not clerical post-processing. It is the locus where ethical accountability, semantic intention, and critical judgment are instituted. The machine calculates variance across an archive; the human curator bears the sole responsibility for truth, error, and normative consequence.
3. The Counterfactual Protocol
No claim stands on dogmatic assertion. Every thesis must survive a three-phase counterfactual protocol:
- Steelmanning the Counter-Thesis: The inquiry confronts the strongest articulation of the opposing view. We test vital biological extension against Jacques Ellul's fatalist determinism; we test embodied semantics against Alan Turing's formal computationalism.
- Exposing Antinomic Illusions: Dogmatic extrapolation produces antinomies. When an engineer equates statistical token fluency with cognitive understanding, they commit the Kantian error of applying an operational calculus beyond the sensory horizons of practice.
- Resolution through Material Grounding: We reject romantic retreats and utopian syntheses. Antinomies dissolve by locating the concrete physical boundaries of the machine and reconnecting formal manipulation to the living organism.
4. Epistemic Demarcations
Five negative boundaries guard the research from recurring traps:
I. Neither Technophobia nor Technolatry
We reject romantic nostalgia for pristine nature alongside accelerationist techno-worship. Technical objects are neither demonic threats nor autonomous deities. They are material artifacts embodying human labor, cultural conflict, and physical ingenuity. They demand rigorous material dissection, not fear or adulation.
II. Against Disembodied Computationalism
Syntactic manipulation is not thought. Algorithms cannot establish the conditions of their own application without circularity. As Ludwig Wittgenstein demonstrated, no formal rule contains the instructions for its own execution. Pure syntax remains semantically blind. Meaning exists exclusively within shared human practice.
III. Rejection of Ethics as Downstream Compliance
Ethics cannot exist as a regulatory afterthought applied by corporate compliance boards to already frozen code. Black-box technical systems cannot be rehabilitated through administrative checklists. Ethics belongs inside the apparatus: in the selection of variables, the design of loss functions, and the distribution of power across the technical ensemble.
IV. Intelligence Augmentation (IA) over Artificial Intelligence (AI)
Douglas Engelbart's historical bifurcation is binding: - Artificial Intelligence (AI): The alienating fantasy of human substitution. It automates decision-making and hides moral accountability behind an algorithmic curtain. - Intelligence Augmentation (IA): The development of open-ended cognitive prostheses that expand human exploratory power, perception, and normative invention.
V. Research as an Open Dossier
Our work refuses the false closure of academic manuals. We maintain working dossiers, archive anomalies, and track theoretical dead ends in the permanent research compendium, exposing the friction of inquiry rather than smoothing difficulties for rhetorical elegance.
References
- Agazzi, Evandro. Scientific Objectivity and Its Contexts. Dordrecht: Springer, 2014.
- Bachelard, Gaston. Le nouvel esprit scientifique. Paris: Presses Universitaires de France, 1934.
- Canguilhem, Georges. La connaissance de la vie. Paris: Hachette, 1952.
- Desrosieres, Alain. La politique des grands nombres: Histoire de la raison statistique. Paris: La Decouverte, 1993.
- Ellul, Jacques. Le systeme technicien. Paris: Calmann-Levy, 1977.
- Engelbart, Douglas C. Augmenting Human Intellect: A Conceptual Framework. Summary Report AFOSR-3223. Menlo Park, CA: Stanford Research Institute, 1962.
- Floridi, Luciano. "Distant Writing: The Cognitive Ecology of Generative Artificial Intelligence." Philosophy & Technology 36, no. 4 (2023): 83.
- Rheinberger, Hans-Jorg. Toward a History of Epistemic Things: Synthesizing Proteins in the Test Tube. Stanford, CA: Stanford University Press, 1997.
- Rodrigues, Tiegue Vieira. "Distant Writing, Generative AI, and Authorship: An Apt Curation Model." Philosophy & Technology 37, no. 1 (2024): 15.
- Simondon, Gilbert. Du mode d'existence des objets techniques. Paris: Aubier, 1958.
- Sosa, Ernest. A Virtue Epistemology: Apt Belief and Reflective Knowledge, Volume I. Oxford: Oxford University Press, 2007.
- Turing, Alan M. "Computing Machinery and Intelligence." Mind 59, no. 236 (1950): 433-460.
- Wittgenstein, Ludwig. Philosophical Investigations. Oxford: Blackwell, 1953.