Glossary of Key Concepts

Concepts are grouped thematically. Each entry gives the term as used in the thesis, its source (where not original), and a one-paragraph account of what it does analytically.

I. Central Thesis Claim

Proletarianisation (of advisory) Source: Stiegler (1998, 2010), reworking Marx. The selective displacement of savoir-faire - practical know-how, professional judgment, formative engagement - into technical systems whose ownership lies institutionally elsewhere. In the thesis’s deployment the term is taken in Stiegler’s extended sense: not the Marxist loss of the means of production but the capture of the capacity to contribute psychically and collectively to the process of professional becoming. The thesis applies this to European infrastructure consulting by locating the site of capture in the temporal structure of formative engagement rather than in content (see tacit-as-formation). The qualifier selective is critical: proletarianisation does not abolish the practitioner or substitute their outputs wholesale; it displaces the formative and judgmental components that constituted their identity as a particular kind of practitioner.

Selective displacement Original analytical qualifier applied to Stiegler’s mechanism. The mechanism through which proletarianisation operates in knowledge-intensive professional work: not wholesale substitution of practitioners by AI systems but the capture of specific components - routine analytical tasks, expressive output layers, formative work at early career stages - while leaving situated judgment, relational intelligence, and contributory expertise structurally intact (for now). Selective displacement is what makes proletarianisation visible at the level of formation rather than employment.

Temporal structure collapse Original analytical term. The specific form the proletarianisation mechanism takes in consulting under AI mediation. Because consulting content-knowledge has always been substantially externalised into frameworks and playbooks, AI’s distinctive capture is not of content (already externalised) but of the temporal structure of formative engagement: the duration of struggle, error, correction, and witnessed senior judgment through which the capacity to use externalised content wisely develops. AI mediation collapses this temporal structure by delivering competent outputs in hours rather than weeks, eliminating the formative time the thesis argues is constitutive of professional becoming.

II. Stiegler–Simondon Framework

Tertiary retention Source: Stiegler (1998), extending Husserl. Technical memory objects - writing, phonography, digital records - that are external to consciousness but constitutive of its temporal structure. Stiegler’s philosophical danger is when tertiary retentions controlled by capital substitute for rather than support the primary retentions (lived experience) and secondary retentions (memory) through which psychic individuation occurs. In the thesis, AI-generated outputs function as tertiary retentions that bypass formative engagement: the output is retained without the practitioner having undergone the formation that historically produced such outputs.

Individuation / dis-individuation Source: Simondon (2005), extended by Stiegler. Simondon’s account of becoming: individuation is not the instantiation of a pre-formed self but an ongoing resolution of tensions in a pre-individual field. Psychic and collective individuation are coupled - I individuate as a practitioner through collective engagements, and the collective individuates through its members’ individuating activity. Dis-individuation is Stiegler’s term for the impoverishment of this process when technical capture reduces the practitioner’s contribution to operating a system rather than engaging with it. The thesis operates the concept at three scales: individual (practitioner formation), organisational (firm’s distinctive way of seeing), national-cultural (European public-interest tradition).

Savoir-faire / savoir-vivre / savoir-penser Source: Stiegler. Three registers of know-how at stake in proletarianisation. Savoir-faire: practical know-how, the capacity to perform the work. Savoir-vivre: know-how-to-live, the cultural and ethical capacity constituting a way of being. Savoir-penser: know-how-to-think, the cognitive capacity for independent analysis. Consulting proletarianisation operates in all three registers: the capacity to build analytical models (savoir-faire), the European public-interest ethos (savoir-vivre), and the capacity to reason from first principles rather than AI-generated patterns (savoir-penser).

Symbolic misery / noetic proletarianisation Source: Stiegler (2004); Alombert (2024). Stiegler’s terms for the cultural and cognitive extension of proletarianisation beyond the industrial-economic register. Symbolic misery: the impoverishment of the capacity to produce and sustain symbolic meaning. Noetic proletarianisation: the capture of the cognitive faculties through which individuals contribute to collective thought. The thesis situates its consulting argument within this Stieglerian AI literature (Nony 2024; Alombert 2024) before differentiating the consulting-specific temporal-structure-collapse mechanism.

III. Ethics and Normativity

Third box Source: Dewey (1927) + Jonas (1984); original thesis coinage for the composite. The constituency that bears the consequences of infrastructure decisions without holding the contracts: the communities, publics, ecosystems, and future generations whose conditions of life are shaped by infrastructure choices. In the standard advisory dyad (client–consultant), the third box has no seat at the table and no mechanism of expression. Thesis’s original coinage: infrastructure advisory is structurally different from management consulting because the third box is substantively present - the scale, irreversibility, and public-interest character of infrastructure decisions give the third box real weight that cannot be discharged through service to the contracting client alone. See also Annex B.

Jonasian responsibility / imperative of responsibility Source: Jonas (1984). Hans Jonas’s normative principle: the capacity to act at technological scale creates obligations extending to the conditions of future capability, not only to present harm. Obligations are asymmetric with power: those who can shape the future conditions of others bear responsibility for those conditions even where the affected parties have no present voice. In the thesis this grounds the advisory profession’s answerability to the third box and to future generations who will bear the consequences of infrastructure built now.

Internal goods / practice (MacIntyre) Source: MacIntyre (1981). MacIntyre’s distinction between goods internal to a practice (excellences achievable only through genuine engagement with the practice’s standards and objects, developed through the virtues the practice requires) and external goods (money, status, power, achievable by any means). In the thesis, MacIntyre grounds the argument that what AI mediation threatens in consulting is specifically the internal goods of the practice - the judgment and sovereign contribution that can only be formed through sustained engagement with consequential work - and that organisations pursuing external goods at the cost of internal goods undermine the practice from within.

IV. Knowledge and Expertise

Tacit knowledge (Polanyi) Source: Polanyi (1958, 1966). Knowledge we have but cannot fully articulate: “we know more than we can tell.” Polanyi distinguishes focal awareness (explicit, articulable) from tacit awareness (the background competence through which focal knowledge is applied). In the thesis’s deployment, the IS-theoretical sense of tacit-as-constraint (tacit = hard to transfer but ultimately transferable by better technology) is the reading that generic AI tools operationalise, and it is the reading the thesis argues is wrong for the consulting case.

Tacit-as-constraint vs. tacit-as-formation Original analytical distinction, central to the thesis. Two readings of tacit knowledge with opposite implications for AI mediation.

• Tacit-as-constraint: tacit knowledge as information that is currently difficult but not impossible to transfer. The IS-theoretical sense. Under this reading, AI tools that make tacit knowledge more explicit are straightforwardly beneficial.
• Tacit-as-formation: tacit knowledge as constitutive of the practitioner’s becoming - not knowledge-as-content but knowledge-as-process, where the process is the formation itself. Under this reading, bypassing the formative process (even while preserving the output) eliminates what was valuable about the process. The distinction rescues the Stiegler application in the consulting domain: even though consulting content-knowledge has always been substantially externalised, there remains a layer of practice that is constitutively tacit-as-formation - the temporal engagement through which a practitioner becomes capable of using externalised content wisely.

Contributory expertise (Collins) Source: Collins & Evans (2007); Collins (2010). Harry Collins’s taxonomy of expertise distinguishes interactional expertise (sufficient to communicate with experts in a field) from contributory expertise (sufficient to contribute original work to the field). The thesis uses contributory expertise to describe what senior consultants deploy and what junior consultants must be formed to develop: the capacity to make genuinely original analytical and judgmental contributions in novel situations, not merely to apply frameworks. This is the expertise whose formation the thesis argues is threatened by AI mediation.

Counsel (Rat) and information (Benjamin) Source: Benjamin, “The Storyteller” (1936), in Arendt (ed.), Illuminations (1968). Benjamin’s distinction between two modes of advisory transmission. Counsel (Rat) is practical wisdom embedded in the adviser’s lived experience, transmissible only from one experience to another - its authority derives from the adviser’s sustained engagement with what they speak of, and its meaning is completed only when it meets the receiver’s experience. Information arrives complete and self-explanatory, with its explanation already attached; it does not require the receiver’s experience to complete it. In the thesis, AI-generated advisory output is information in this precise sense: it does not require the practitioner’s Erfahrung to produce it and does not leave counsel’s interpretive space for the client to engage with. Sophisticated clients who have received genuine counsel can sense the difference even when they cannot articulate it. The counsel/information distinction extends the tacit-as-formation argument to the client’s receiving end: what clients historically received from formed senior advisers was counsel; what AI-mediated advisory delivers is information.

Erfahrung and Erlebnis (Benjamin) Source: Benjamin, “The Storyteller” (1936); “On Some Motifs in Baudelaire” (1939), both in Illuminations (1968). Benjamin’s distinction between two forms of experience. Erfahrung is accumulated, reflective, transmissible experience - the practical wisdom that sediments over years of sustained engagement with consequential problems and that makes counsel possible. Erlebnis is isolated, momentary, non-accumulating lived-through-ness: the experience of the modern subject who encounters but does not assimilate. In the thesis, the formation process the tacit-as-formation register describes is precisely the process through which Erlebnis accumulates into Erfahrung: isolated encounters with consequential work - the model built from scratch, the case argued before a regulator - sediment over time into the transmissible practical wisdom that counsel requires. AI-mediated formation substitutes Erlebnis without accumulation: the junior’s engagement is with AI-generated outputs rather than with the direct challenge of the work, and the Erlebnis of those engagements does not accumulate into Erfahrung. The apparatus of capture (D&G) reduces Erfahrung to training data; what it returns is Erlebnis. Connects directly to tacit-as-formation (Erfahrung = what tacit-as-formation produces; Erlebnis = what AI-mediated engagement produces in its place).

V. Professional Formation Framework

ChMC (Chartered Management Consultant) framework Source: CMI Chartered Management Consultant Award. The professional formation framework the thesis uses as its empirical anchor for European infrastructure consulting. Organised around four competency domains (strategic insight, analytical rigour, collaborative delivery, ethical responsibility) and three stages: Foundation (analytical execution), Applied (managed delivery), Chartered (sovereign advisory). See Annex C for full treatment.

Foundation stage The first stage of the ChMC formation pathway. The practitioner develops basic analytical competence through directed engagement with consequential work: building financial models, structuring outputs, executing discrete workstreams under supervision. The thesis argues this stage is most directly threatened by AI mediation, because the formative tasks that historically built Foundation-stage competence are precisely the ones AI tools perform most fluently.

Applied stage The middle stage of the ChMC pathway. The practitioner manages delivery, mediates between senior judgment and junior execution, and develops the relational and contextual intelligence required to translate analytical competence into consequential outputs for real clients. The thesis argues AI mediation threatens this stage by compressing the scope of managed work and by eroding the supervisory function through which senior judgment is transmitted.

Chartered stage The senior stage of the ChMC pathway. The practitioner exercises sovereign advisory: contextually authoritative, normatively grounded, institutionally trusted judgment on complex consequential questions. The thesis argues this is the stage most resistant to AI displacement (the judgment required cannot be captured as a pattern from training data) but that it depends on Foundation and Applied stage formation to reproduce, and that the threat to lower stages therefore eventually threatens the Chartered stage by starving its pipeline.

Pipeline rupture Original analytical term. The thesis’s term for the mechanism through which AI-mediated proletarianisation at Foundation stage propagates to senior levels over time. When Foundation-stage formation is compromised, the supply of practitioners developing toward Chartered-stage sovereign advisory is reduced. The rupture is not immediately visible (current senior practitioners are already formed) but is structurally irreversible if the temporal structure of formative engagement is not restored: the cohort that should be forming now into the senior practitioners of 2035–2040 is not receiving the formation. Ide’s (2025) overlapping-generations model provides economic-theoretic grounding for why this rupture is structurally self-reinforcing.

Upskilling inhibition Source: Natali et al. (2025), adapted from the medical context. The phenomenon in which AI assistance, while improving immediate output quality, inhibits the development of the underlying competence the junior practitioner would have developed through unassisted work. Natali et al. demonstrate this in medical training (PACES-MRCPUK context); the thesis deploys it as the mechanism through which AI mediation at Foundation stage produces practitioners who are fluent at a surface level but lack the deeper judgment the ChMC pathway is designed to develop.

Cognitive debt Source: Kosmyna et al. (2025). The reduction in neural engagement and deep processing associated with AI-assisted cognitive work, shown to persist even when AI assistance is withdrawn. The thesis uses this as part of the empirical grounding for the claim that formative engagement with AI-mediated work is structurally different from formative engagement with unassisted work - the cognitive work simply is not being done.

VI. Organisational and Political-Economy Concepts

Sovereignty (advisory) Original analytical term. The thesis’s term for the form of advisory authority at stake in Chartered-stage work: contextually authoritative, normatively grounded, institutionally trusted engagement with consequential questions. Advisory sovereignty is not independence from clients but the capacity to maintain an independent interpretive standpoint - one that can disagree with the client’s framing, represent the third box, and exercise judgment not reducible to the patterns of the AI’s training distribution. Operates at three scales, connected by a single mechanism: individual (the practitioner’s sovereign judgment), organisational (the firm’s distinctive way of seeing), and national-cultural (the European public-interest tradition as a collective capability not reducible to any individual firm or practitioner). The thesis’s dual-scale sovereignty framing names the organisational and national-cultural scales as jointly at risk: generic AI mediation erodes the firm’s distinctive analytical identity while simultaneously pulling European advisory practice toward the Anglo-American patterns dominant in model training corpora. The national-cultural scale is the thesis’s own extension of Stiegler’s individuation framework and is not directly present in Simondon.

Organisational identity (firm’s way of seeing) Developed from Stiegler + Nonaka/Takeuchi. The firm’s distinctive interpretive framework: the accumulated, sedimented way of analysing situations, framing problems, and calibrating recommendations that distinguishes one firm’s advisory output from another’s. In the thesis’s analysis, this is at risk of erosion under generic AI mediation because the model’s training distribution carries the discursive patterns of the dominant anglophone consulting culture, and repeated use of the model as an interpretive starting point gradually pulls the firm’s output toward the model’s default framings.

SECI model / dark mirror Source: Nonaka & Takeuchi (1995); dark mirror is original. The SECI model (Socialisation, Externalisation, Combination, Internalisation) describes the knowledge spiral through which organisations create and transmit knowledge. The thesis’s dark mirror is the inversion of this spiral under generic AI mediation: where SECI describes knowledge creation (tacit → explicit → re-internalised), the extraction dynamic describes knowledge capture (tacit externalised into AI training data, then recirculated as an organisational substitute for the original formation process). See Annex F.

Aura and mechanical reproduction (Benjamin) Source: Benjamin, “The Work of Art in the Age of Mechanical Reproduction” (1935), in Illuminations (1968). Benjamin’s concept of the aura - the quality of authentic presence that derives from a work’s embeddedness in a specific tradition, its “here and now” - is deployed in the thesis to name what is at stake in the organisational-scale sovereignty argument. The firm’s distinctive way of seeing is an auratic property: it derives from the firm’s accumulated history, the sedimented judgments of its practitioners, the conventions of its analytical approach built across decades of engagement with specific European regulatory institutions. What makes the firm’s output recognisable to a client across multiple engagements is not a brand or a style but the auratic quality of embeddedness - the traces of a specific tradition expressed through a specific firm’s sustained encounter with it. AI-generated advisory outputs are mechanically reproduced advisory work in Benjamin’s sense: they lack this aura, bearing the generic marks of the training corpus rather than the marks of a specific tradition. Benjamin’s analysis is ambivalent: mechanical reproduction also democratises, stripping cult value and making work widely available. The thesis engages this ambivalence directly - what is democratised by AI mediation is competent output (information), a real gain; what cannot be mechanically reproduced is the auratic quality of counsel. The gain and the loss are real and simultaneous. See counsel and information, Erfahrung/Erlebnis, and §5.3 of the main text.

Apparatus of capture (Deleuze & Guattari) Source: Deleuze & Guattari, A Thousand Plateaus (1980), trans. Massumi (1987). D&G’s term for the state-like mechanism that overcodifies and appropriates what flows outside its borders - transforming independent activity into captured, quantifiable, taxable, and directed labour. In the thesis the concept is deployed to name the market-side capture mechanism that operates alongside and beneath Stiegler’s temporal-structure-collapse: AI mediation is not merely a technical displacement of formative tasks but a component of an apparatus that redirects the flows of professional knowledge (tacit, formative, judgmental) into proprietary model architectures controlled by actors outside the profession and outside European public-interest governance. Where Stiegler describes what is lost from the practitioner (the temporal structure of formation), D&G describe what captures it and why the capture is structurally incentivised. The apparatus of capture is relevant both to the political-economy analysis in Ch05 and to the sovereignty argument across Ch07 and Ch09.

Spectacular pseudo-cyclical time (Debord) Source: Debord, The Society of the Spectacle (1967), §§149–163. Debord’s term for the commodity’s colonisation of lived time: cyclical agricultural time displaced by the linear time of production, which is then projected back as circular through the spectacle - a pseudo-cyclical rotation of novelty (seasons, model releases, product generations) that structurally forecloses accumulated Erfahrung. In the thesis the concept amplifies the temporal-structure-collapse argument by naming the systemic time-logic within which that collapse operates: AI mediation does not merely compress the duration of formative engagement but occurs within a spectacularised temporality of capability releases and tool updates that colonises the practitioner’s sense of what a professional generation looks like. Each model release performs the work of a new season - appearing as renewal while structurally preventing the kind of sustained, non-spectacular engagement through which Erfahrung accumulates. Connects to Erfahrung/Erlebnis and temporal structure collapse.

Récupération (Debord) Source: Debord, The Society of the Spectacle (1967); Situationist International writings. Debord’s term for the mechanism by which radical or subversive elements are absorbed, neutralised, and repackaged by the dominant spectacle as commodifiable content. What began as critique becomes a feature; what began as resistance becomes a style. In the thesis récupération is deployed to name the risk that professional formation responses to AI mediation - upskilling programmes, formation-aware procurement, professional body guidance - are absorbed back into the efficiency narrative they are designed to resist. The danger is that the third-box argument, the sovereignty argument, and the formation-centred response are récupéré: praised, endorsed, marketed, and emptied. The concept is deployed primarily in Ch09 §9.1 and §9.5 as a caution against naïve optimism about professional self-regulation.

TGAIF (Task-GenAI Fit Framework) Source: Tuczek et al. (2026). An IS-theoretical framework for assessing which tasks are well-matched to generative AI assistance based on task characteristics (decomposability, verifiability, knowledge-domain specificity). The thesis treats TGAIF as its primary analytical foil: it represents the dominant IS-theoretical approach to AI mediation in professional work, operationalises the tacit-as-constraint reading of knowledge, and produces recommendations (deploy AI where task-technology fit is high) that the thesis argues are analytically correct but normatively incomplete - they do not capture the formation consequences of optimising for task-technology fit at the junior practitioner’s stage. See Annex G.

VII. Context and Domain

European infrastructure advisory The domain the thesis analyses: advisory work in the family of European infrastructure subdomains - transport, energy (including nuclear), water, defence, environmental, social, built-environment, and urban. Defined by three distinguishing features: (1) a public-purpose engineering tradition with roots in post-war European reconstruction; (2) a regulatory and governance architecture oriented to public-interest outcomes; (3) a pronounced third-box exposure - the scale and irreversibility of infrastructure decisions means the affected constituencies are substantive, not notional. Treated as a coherent domain despite subdomain variation.

Pyramid collapse Source: Tarki & Raczynski (2025). The structural change in consulting firm economics under AI mediation: the traditional leverage model (many junior hours billed at lower rates supporting fewer senior hours at higher rates) is compressed as AI substitutes for junior analytical work. The thesis uses pyramid collapse as the financial-structural argument alongside the formation argument: the economic logic of the traditional pyramid was aligned with the formation logic (juniors needed to do junior work to develop); AI disrupts both simultaneously.

Symbolic domination (AI mediation) Source: Stiegler + Alombert (2024), developed in Ch5. The thesis’s term for the cultural dimension of the sovereignty threat: when the interpretive frameworks, rhetorical conventions, and analytical defaults that AI tools carry - drawn from their anglophone training corpora - are absorbed into European advisory practice, the result is not neutral efficiency gain but the gradual displacement of the European public-interest tradition by the dominant cultural-discursive patterns encoded in the model. The European public-interest tradition is specific: it has distinctive ways of weighing long-term systemic risk, engaging regulatory complexity, and representing public constituencies that differ from the patterns the model carries.