Executive Summary

The Proletarianisation of Advisory: Judgment, Sovereignty, and Responsibility in European Infrastructure Consulting under AI Mediation

Author: LJ (practitioner-researcher, European infrastructure advisory) Award: MPhil Word count: ~62,000 words (nine chapters, seven annexes)

The problem and the central claim

European infrastructure advisory — advising on transport, energy, water, urban, and built-environment projects with long time horizons and direct consequences for publics — depends on a form of professional judgment that is not reducible to the retrieval and synthesis of information. Senior infrastructure advisers bring situated, consequential, answerable judgment: formed through years of supervised practice, exercised in specific regulatory and political contexts, and answerable to the publics whose conditions of life are being shaped by the decisions they advise. That formation pathway is now under structural pressure from the integration of generic AI mediation into the work at all career stages.

The existing discourse addresses this situation primarily in a productivity register: workflow efficiency, task-technology fit, appropriate governance of tool use. This thesis argues that the productivity register misses what is most consequential. What is at stake is not whether AI tools make consultants faster, but whether the formation pipeline through which advisory judgment has historically been reproduced — and through which the profession’s legitimacy has been sustained — is being structurally altered in ways the productivity framings cannot detect.

At the general level:

Generic AI mediation, under current conditions, proletarianises the situated, answerable judgment on which the legitimacy of professional advisory work rests.

At the specific level:

Generic AI mediation, under current conditions, risks proletarianising the situated, answerable judgment on which the legitimacy of European infrastructure advisory rests — and on which its obligations to publics and futures depend.

Proletarianisation is taken from Bernard Stiegler (reworking Marx): the selective displacement of savoir-faire — practical know-how, professional judgment, the capacity to contribute — into technical systems the worker does not own. The displacement is not total; it is formative. It does not eliminate the advisory workforce but progressively absorbs the expertise that constituted its professional identity, replacing contributory expertise with interactional fluency — the capacity to work with AI outputs without the independent cognitive ground required to evaluate them.

The stage-by-stage analysis

Chapter 7 applies the framework to each ChMC stage and identifies three distinct transformations:

Foundation stage - the apprenticeship blow. The formative tasks at this stage (drafting, modelling, structuring, presenting) are precisely the tasks AI mediation absorbs most efficiently. Junior consultants are still hired and assigned; what changes is the formative content of what they do. AI-assisted production of competent outputs is not formation. The medical education literature calls this never-skilling — the failure to develop foundational competency, distinct from the deskilling of experienced practitioners. False proficiency names why the deficit is invisible to productivity metrics: the performance looks adequate until the support is withdrawn.

Applied stage - moat erosion. The polished deliverable has historically been the visible expression of the firm’s distinctive analytical contribution. Generic AI mediation produces that expressive layer rapidly and competently, compressing the advantage the deliverable once demonstrated. The Applied consultant’s contribution now has to be demonstrated through what is behind the deliverable — analytical judgment, relational grasp — rather than through the deliverable itself.

Chartered stage - disintermediation and relational erosion. Clients building in-house AI capability compress the scope of engagements commissioned from advisory firms. The immediate pressure is on revenue; the longer-term pressure is on the relational fabric through which senior judgment has historically been transmitted. The pipeline rupture at Foundation propagates forward: the senior practitioners of two decades hence are the Foundation-stage consultants of today.

Four scenarios

Chapter 8 constructs four scenarios as philosophical thought-experiments:

Scenario Expertise held by AI ownership
Commodity Advisory AI systems Commercial Anglo-American platforms
Digital Twin Economy AI systems European/firm-governed systems
Artisanal Advisory Human practitioners Used but bounded
Disintermediated Clients Human practitioners (in-house) Client-side

Futures findings

The chapter’s contribution is seven cross-scenario findings — concerns that hold regardless of which trajectory unfolds, and which provide the robust ground for the recommendations that follow:

  1. The contributory/interactional distinction is the primary determinant of professional value. What AI mediation reproduces at scale is interactional fluency; what it cannot reproduce is contributory expertise formed through sustained engagement with consequence. This is the load-bearing mechanism beneath every other finding.
  2. The formation question becomes inescapable. How the profession reproduces contributory expertise is at stake in all four scenarios — pathways collapse, are technically reconfigured, are institutionally protected, or compress, but the question cannot be avoided.
  3. Sovereignty pressure operates across multiple scales. Both national-cultural sovereignty (the distinctiveness of European public-interest practice) and organisational sovereignty (firm-level analytical voice) are under pressure in every trajectory.
  4. The third-box question persists. The publics who bear the consequences of infrastructure decisions without holding the contracts — including future generations — remain a live normative concern across all four scenarios.
  5. Explicit legitimacy narration becomes inescapable. The implicit legitimacy that historically operated through institutional architectures is insufficient; legitimacy now requires explicit narration of what is being done, by whom, and with what AI mediation.
  6. Extraction operates structurally in all scenarios. Knowledge flows outward from European advisory practice into the generic mediation apparatus in every trajectory; the variable is not whether extraction happens but under whose ownership.
  7. The Applied stage has no safe trajectory. The intermediary analytical function connecting formation to authority is commoditised, technically reconfigured, defended only by deliberate design, or disintermediated outright — and requires its own institutional design logic.

Contributions

  1. A philosophical account of AI-induced proletarianisation of advisory work — the first sustained application of Stieglerian framework to professional consulting
  2. The tacit-as-formation distinction (tacit knowledge as a process of formation, not merely as hard-to-transfer content)
  3. The temporal structure collapse thesis (AI collapses the time of formation, not merely its content)
  4. The first systematic mapping of ChMC competencies against AI substitutability across three career stages
  5. Dual-scale sovereignty analysis (national-cultural and organisational, connected through the same dis-individuating mechanism)
  6. Four scenarios for European infrastructure advisory as philosophical thought-experiments, with seven cross-scenario findings

Recommendations

Differentiated recommendations are developed for four constituencies:

Read the full thesis · Download the ebook · Plain-language versions