The Academic Aristocracy Rebels Against Its Executioner

1

The starting point for this reflection is Pierre Rimbert’s article, “Everyone Hates AI,” published in the latest Spanish edition of Le Monde diplomatique. Rimbert describes a growing resistance to a technological transformation that threatens skilled and administrative jobs, concentrates strategic decision-making in large corporations, and takes material form in data centers with high energy, water, and territorial costs. The opposition, therefore, is directed not only against automation, but against a model of power that combines labor substitution, the privatization of public decision-making, and the deterioration of the material conditions of life (Rimbert, 2026).

That resistance is fully justified. A technological transformation of this magnitude should not be imposed without democratic deliberation, especially when it concentrates benefits in large firms while shifting labor, social, and ecological costs onto workers, communities, and territories. There is also a distributive issue at stake. As Acemoglu warns, AI may widen the gap between returns to capital and returns to labor, even if its aggregate effects on productivity prove far more modest than corporate rhetoric suggests (Acemoglu, 2024).

But being against AI does not, by itself, imply an emancipatory position, or even a progressive one in the usual sense of our political map. Such opposition may express a defense of the commons, of the material conditions of life, or of democratic autonomy. But it may also defend professional privileges, monopolies over knowledge, institutional hierarchies, or threatened positions of authority. It may even present itself in progressive language while protecting class interests tied to the preservation of the bureaucratic-administrative apparatuses through which capital organizes and reproduces its power. Rimbert rightly notes that this opposition is politically heterogeneous and can be appropriated by nationalist and reactionary forces.

The question, then, is not whether one is for or against AI. The question is what social order is being defended or transformed when AI is rejected.

2

AI may destroy professional positions, institutional hierarchies, and monopolies of intellectual legitimacy that do not deserve to be preserved unconditionally. But the destruction of a hierarchy does not automatically amount to an expansion of equality. On the contrary, it may become the means through which a new form of subordination is established.

This is the key to a Marxian reading of capital. In the Communist Manifesto, Marx and Engels show that the bourgeoisie destroys estate-based ties, inherited authorities, and traditional privileges. But it does not do so in order to liberate those subjected to such hierarchies. It destroys them in order to impose a new regime of social relations organized around the market, competition, and capital accumulation (Marx & Engels, 2011).

Machinery forms part of that process. In Capital, Marx shows that technology can reduce necessary labor time, increase productivity, and transform production. But under capitalist relations, these advances are not immediately directed toward freeing time or expanding the capacities of those who work. They become instruments for cheapening labor power, intensifying subordination, and reinforcing the power of those who control the means of production (Marx, 2021b, p. 604).

The ambivalence does not lie in technology as such, but in the social relations that organize it. In the Grundrisse, Marx observes that the development of the productive forces could reduce necessary labor time and expand the time available for human life. Under capital, however, that possibility is inverted: the potential reduction of labor is translated into unemployment, precarity, discipline, and the concentration of wealth (Marx, 2021a).

3

The university is one of the places where the ambivalence of AI becomes most visible. What is at stake is not merely the automation of particular academic tasks, but the transformation of the very conditions of teaching, assessment, research, credentialing, and intellectual authority. The question is not only what AI can do within the university, but what kind of university helps produce the conditions for its implementation.

AI may erode an academic aristocracy founded on the monopoly of certain competencies, credentials, and forms of authority. But that erosion does not necessarily amount to a democratization of knowledge. On the contrary, it may serve to transfer university functions to private platforms, cheapen teaching labor, precarize faculty, and turn education into a market for automated answers.

For that reason, it is not enough to lament the destruction of academic authority. We must ask what kind of authority is being destroyed, who benefits from its destruction, and what new social relation is being imposed in its place.

Under present conditions, AI functions as a new force of accumulation. It lowers costs, accelerates processes, concentrates information, reorganizes labor, and expands the capacity of firms to replace, monitor, and discipline workers. But it also produces a crisis of legitimacy in certain professional strata, including part of the university faculty, especially in the humanities and social sciences.

For decades, these disciplines sought to secure their institutional place by presenting themselves as scientific, methodological, measurable, and functional forms of knowledge. They adopted the language of indicators, accreditation, rankings, research assessment cycles, competencies, rubrics, productivity, and impact. Knowledge was translated into processable information; teaching into the management of outcomes; research into the production of evidence; and the academic career into the accumulation of verifiable merits.

AI therefore does not intervene in order to destroy a university that was previously free. It intervenes in an institution that had already reduced a significant part of its intelligence to standardized procedures.

Faculty selection illustrates the problem. The primary criterion is not who teaches best, reads best, or thinks with the greatest independence. More often, selection favors those who fit most effectively into the bureaucratic apparatus: those who obtain credentials, publish, index their work, manage, apply for grants, produce measurable outputs, and adapt to assessment mechanisms. Intelligence is not necessarily rewarded; institutional conformity is.

4

The degradation of the institutional conditions of thought should not be attributed simply to individual shortcomings. It is a structural effect. An institution that rewards adaptation ends up producing adapted subjects. Slow reading, intellectual risk-taking, demanding teaching, and serious discussion are subordinated to the administration of dossiers, the strategic production of publications, and the preservation of positions.

AI then acts as a mirror. It reveals that many academic tasks had already been transformed into repeatable operations: summarizing, classifying, translating, grading, writing reports, organizing bibliographies, applying rubrics, producing teaching materials, or responding according to predictable formats. When a machine can carry out a significant portion of these tasks, it becomes clear to what extent they had been institutionally reduced to predictable procedures. This does not mean that such tasks involve no judgment, interpretation, or experience; rather, those dimensions had been subordinated to repeatable, measurable, and administrable formats.

The academic aristocracy thus rebels against its executioner. But that executioner is not the machine in the abstract. The machine does not invent from the outside the reduction of intelligence to an administrable function: it finds terrain already prepared by decades of bureaucratization, quantification, credentialing, and institutional competition. Under these conditions, AI can take over and expand tasks that the university itself had previously turned into standardized procedures.

The crisis of that aristocracy may contain an egalitarian possibility. Part of its authority did not rest exclusively on knowledge itself, but on the differential administration of access to it: the monopoly over legitimate language, conventions of writing, certification, access to bibliographies, networks of recognition, and tasks whose difficulty had been amplified by institutional opacity. The crisis of that monopoly should not be lamented without qualification.

5

But that crisis should not be celebrated naively. Under present conditions, democratization may take the form of a degrading leveling: precarized faculty, students left to platforms, universities turned into providers of automated content, and knowledge reduced to a digital service governed by the logic of profitability, scale, and data extraction.

The decisive question is not how to protect the intellectual aristocracy from the machine. It is how to prevent capital from using the destruction of that aristocracy to destroy the university as a space for formation, judgment, reading, discussion, and responsibility.

References

Acemoglu, D. (2024). The simple macroeconomics of AI (NBER Working Paper No. 32487). National Bureau of Economic Research.

Marx, K. (2021a). El capital: crítica de la economía política. Libro primero: El proceso de producción del capital (P. Scaron, Trad.). Siglo XXI Editores.

Marx, K. (2021b). Elementos fundamentales para la crítica de la economía política (borrador) 1857–1858 (P. Scaron, Trad.). Siglo XXI Editores.

Marx, K., & Engels, F. (2003). Manifiesto comunista (P. Ribas, Trad. e introd.). Alianza Editorial.

Rimbert, P. (2026, julio). Todo el mundo odia la IA. Le Monde diplomatique en español.