Chapter 9 - Conclusion: Recommendations and Residual Questions
9.1 What the thesis has delivered
This chapter closes the thesis. It restates the central claim, names the contributions the thesis has delivered, and develops differentiated recommendations for the four constituencies whose decisions will shape the trajectory European infrastructure advisory follows in the next decade: individual consultants, advisory firms, infrastructure clients, and European policymakers. It acknowledges the limits of what the thesis has done and identifies the residual questions that further work should take up. It closes with a forward-looking statement about what is at stake in the conditions European infrastructure advisory now faces.
The chapter works at a different register from the analytical chapters. The first-person voice is more present here than in Chapter 7 or Chapter 5, deliberately so, because the conclusion is the chapter where the reflexive practitioner-researcher voice that the methodology rests on takes its most direct expression. The recommendations to constituencies are suited to the four cross-scenario findings from Chapter 8: the formation question, the sovereignty pressure, the third-box question, and the requirement of explicit legitimacy narration. These findings hold across the trajectory variation Chapter 8 examined, and the recommendations that follow from them are robust to the question of which trajectory unfolds.
The chapter also acknowledges, more directly than the analytical chapters could, that the framework the thesis develops is not the philosophically complete framework the thesis points toward. Three limits are registered honestly: the single-domain character of the analysis, the reflexive and conceptual rather than systematically empirical methodology, and the deliberate lightness of the philosophical apparatus that Chapter 3 bis would deepen. Five residual questions for further work are identified. These limits and questions do not undermine the contribution the thesis claims; they situate it accurately within the wider conversation it is contributing to.
9.2 The central claim and the contributions
The thesis has advanced a single central claim, which I restate here in the form it has occupied throughout: 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.
Eight chapters have assembled, stress-tested, and deployed the framework’s analytical apparatus - from the philosophical foundations in Chapters 2 and 3, through the domain materialisation and political-economic analysis in Chapters 4 and 5, through the comparative evidence and stage-by-stage core in Chapters 6 and 7, and into the futures analysis of Chapter 8. Together, the contributions produce what the existing literature on AI in consulting does not yet provide: a philosophical account of what is at stake in the profession’s transformation that addresses formation, sovereignty, and the third box simultaneously.
What has been delivered, in terms of the six contributions claimed in the introduction:
Stiegler’s proletarianisation thesis applied to management consulting has been delivered. The thesis has developed the application across the analytical chapters, demonstrating that proletarianisation as selective displacement of savoir-faire operates in the consulting profession, that the dual-scale individuation argument operating across individual, firm, and professional culture provides conceptual unity for the framework, and that the application produces analytical purchase the existing literature on AI in consulting does not provide. (Annex A develops proletarianisation as I deploy it.)
Jonas’s imperative of responsibility applied to AI-mediated advisory accountability in infrastructure has been delivered. The thesis has grounded the normative argument in Jonas’s framework, demonstrating that the third-box constituency I treat as central is institutionally available within the European public-interest tradition, that the answerability of advisory work to publics that bear consequences is the ground of legitimacy in the domain, and that the answerability is structurally at risk under contemporary AI mediation in ways I name.
The ChMC competency framework mapped against AI substitutability across the three stages has been delivered. Chapter 7’s stage-by-stage analysis provides what is, to my knowledge, the first systematic mapping of the ChMC framework against AI substitutability for the three stages, with the four ChMC competencies deployed analytically at each stage. The mapping gives the framework a stage-by-stage diagnostic register that supports the differentiated recommendations developed in this chapter.
Pipeline rupture theorised as a cross-professional philosophical mechanism has been delivered. Chapter 6’s comparative evidence demonstrates that the mechanism operates in software and law, with the AI-washing critique engaged directly to show that the framework’s argument addresses formation rather than headcount. Chapter 7 demonstrates the same mechanism in European infrastructure advisory specifically. The cross-professional pattern is consistent enough to support the theoretical framing.
The dual-scale framing of sovereignty as identity-formation under technical mediation has been delivered. Chapter 5 develops sovereignty at the national-cultural and organisational scales as connected through the same Stieglerian mechanism, and the framework’s analytical contribution at this point is the connection itself: sovereignty operates not across one scale but across multiple, and the same dis-individuating pressure acts at each.
Four scenarios for European infrastructure advisory as philosophical thought-experiments has been delivered. Chapter 8 constructs the scenarios - Commodity Advisory, Digital Twin Economy, Artisanal Advisory, Disintermediated Clients - and identifies the cross-scenario findings that hold regardless of trajectory. The scenarios are explicitly framed as thought-experiments rather than predictions, in the Williamsonian sense, and they generate ground for the recommendations developed in this chapter.
These six contributions are claimed as originality. None is the final word on the questions it engages, and each opens further work that the conclusion’s residual questions identify. What I claim is that the combination of contributions, working together through the framework the thesis has developed, constitutes a philosophical contribution to the analytical conversation about AI in European infrastructure advisory that the existing literature has not yet provided.
Before developing recommendations, I name the failure mode that institutional responses to this argument most readily fall into. Debord described it as récupération: the spectacle’s capacity to neutralise critical challenge by converting it into spectacular form. The critique produces a representation of critique, which satisfies the demand for opposition without delivering it. In the advisory context, this failure mode is precise and recognisable. A recommendation that results in a responsible AI use policy, a human oversight requirement, or an editorial judgment protocol has been recuperated. The governance response produces the representation of a solution: the junior validates the model’s output; the senior approves the validated output; the policy specifies that human judgment must be applied; the firm records the application. Each step is a staged performance of the engagement the critique demanded. None of it restores the temporal structure of formative engagement.
The junior’s validation is spectacular validation. They receive the model’s output and apply their current judgment to it - judgment formed by the same AI-mediated conditions the validation is meant to correct. The senior’s approval is approval of the spectacle of analysis. The policy’s specification of human oversight is the representation of human oversight. What is not specified, and cannot be specified by a governance framework, is the thing the argument requires: the time of genuine formative difficulty, the unassisted encounter with analytical challenge, the sustained engagement with consequence through which practical wisdom develops. The recommendations in this chapter are designed against this failure mode [@debord1967spectacle].
9.3 What this means for consultants
The recommendations developed in this section are suited to the cross-scenario findings: the formation question becomes inescapable across all four scenarios, sovereignty pressure operates across multiple scales, the third-box question persists, and explicit narration of professional legitimacy is required. For individual consultants, the implications are differentiated by ChMC stage but converge on responsibilities I treat as constitutive of legitimate practice in the domain.
Brussels, 2022. A Foundation-stage consultant is assigned to accompany a senior partner to a European Commission technical working group on cross-border energy infrastructure governance. Her formal contribution to the session is minor. The partner considers leaving her behind - the preparation is substantial, the hearing complex, and she will not speak. He brings her anyway, with a specific brief: watch what happens when the Commission official asks a question we have not prepared for, and watch how I handle what I do not know. The session runs three hours. The official asks, forty minutes in, a question about the interaction between the Energy Union framework and a specific bilateral treaty obligation that nobody in the room has anticipated. The partner handles it. In the taxi afterwards, they spend forty minutes talking about it: what he did, why, what he could not say and why, what the question revealed about what the official actually needed. The Foundation consultant describes that conversation five years later, in a session on professional development, as the most formative single afternoon of her career. She is not sure she could have had it via a briefing document. She is certain she could not have had it via an AI tool.
The first recommendation is about deliberate formation. Formation is the ground on which contributory expertise is reproduced across generations, and the framework’s analysis demonstrates that contemporary AI mediation operates on the formation pathway in ways the productivity framings cannot engage. For Foundation-stage consultants, this implies seeking out engagements that develop contributory expertise through sustained engagement with consequence, even where AI-mediated alternatives would be faster. For Applied-stage consultants, this implies recognising the formative role they play for Foundation-stage juniors and the supervisory engagement that role requires beyond AI-mediated review. For Chartered-stage consultants, this implies recognising the formation responsibility their position carries - what they teach the next generation, and what they refuse to delegate to AI mediation in the contexts where formation is occurring.
The second recommendation is about reflexive practice. Reflexive practice is the methodological discipline through which the practitioner’s own engagement with AI mediation is examined as part of the work itself. For practitioners at any stage, this implies a working awareness of when one’s own analytical contribution is being mediated, what the mediation is doing to the work, and what is being preserved or eroded in one’s own contributory grasp. Reflexive practice does not require refusal of AI mediation; it requires consciousness of what mediation is doing and what it is not doing, and the cultivation of judgment that survives mediation rather than being absorbed by it.
The third recommendation is about refusal of unattributed mediation. Where AI mediation operates in advisory work, I treat explicit attribution as a professional obligation. The Foundation-stage consultant whose draft has been substantially produced through AI mediation owes attribution to the consultant who reviews it. The Applied-stage consultant whose deliverable has been substantially produced through AI mediation owes attribution to the client who receives it. The Chartered-stage consultant whose advice has been substantially produced through AI mediation owes attribution to the public who bears the consequences. The attribution is a matter of professional integrity, and I treat it as institutionally important - without explicit attribution, the legitimacy of advisory work is structurally undermined in ways Chapter 6’s fourth finding identified.
A senior regulatory adviser in a UK infrastructure practice has, over fifteen years, developed a way of framing engagement preparation that does not appear in any methodology document: she consistently asks, for every engagement, what is the regulatory body trying to protect itself from? - not what it wants, not what the legislation mandates, but what institutional fear is shaping its behaviour right now. It is a question that produces different analytical entry points than standard stakeholder interest mapping, and it has been right often enough that her clients have come to expect it as a distinctive feature of the firm’s work. She has never written it as a framework. She does not teach it through a slide. She teaches it by working alongside people. She tested it once, informally, against an LLM’s stakeholder analysis for an engagement she had already completed. The model produced a correct and comprehensive answer. It had not asked what the regulator was afraid of. The question had not occurred to it.
The fourth recommendation is about cultivating judgment that resists generification. The distinctiveness of the practitioner’s analytical voice is part of what makes contributory expertise valuable, and the framework’s analysis demonstrates that generic AI mediation produces dis-individuating pressure on practitioner-distinctive analytical content. The recommendation is to cultivate, deliberately, the analytical voice that distinguishes one’s own work - the framings, vocabularies, and ways of seeing that one brings to engagements and that would not survive uncritical AI mediation. This is a professional commitment as much as a strategic one: the practitioner whose distinctive voice is preserved is making a contribution that the practitioner whose work is fully generified cannot make.
The fifth recommendation, particularly for Chartered-stage consultants, is about senior responsibility for the next generation. The framework’s analysis demonstrates that the formation pipeline is structurally at risk under contemporary AI mediation, and that the long-term substance of European infrastructure advisory depends on the formation pathways that current senior practitioners support or fail to support. Senior responsibility is not only the consultant’s own work; it is also the formation of the practitioners who will succeed them. The recommendation is to take this responsibility seriously - to invest in the formation pathways within one’s firm, to engage with the supervisory role, and to recognise that the firm’s capacity in two decades depends on choices made now.
These five recommendations to consultants are differentiated by stage but cohere on the responsibilities I treat as constitutive of legitimate advisory practice. They are addressed to consultants who recognise the concerns the thesis has named and who choose to take them seriously in their own practice. They do not depend on institutional change at the firm or sector level; they are commitments individual consultants can make within the institutional contexts they currently work in.
9.4 What this means for advisory firms
The recommendations to advisory firms are suited to the same cross-scenario findings, but they engage the institutional choices that firms make about how the formation pipeline is configured, what tooling decisions are made, and how the firm’s distinctive way of seeing is preserved or eroded.
Paris, 2024. An infrastructure advisory practice is evaluating three enterprise AI platforms for integration into its workflow. The evaluation criteria are capability, price, integration with existing systems, and - at the managing partner’s insistence - data governance. Her single non-negotiable: no client analytical content used in vendor training data, contractually enforced, with audit rights written into the agreement. One of the three vendors cannot meet the condition at any price. One meets it at a significant premium. One meets it at standard enterprise pricing but only for structured data, not unstructured client documents. The firm takes the expensive option. Three months later, the managing partner includes the data governance clause as a specific item in her annual presentation to the firm’s client advisory board - evidence that the firm takes the confidentiality of client strategic material seriously in an era when the boundary between workflow tool and data collector is not always clear. Two clients mention it unprompted in contract renewal conversations.
The first recommendation is about formation pathway preservation. The framework’s analysis demonstrates that contemporary AI mediation operates on the formation pipeline through which contributory expertise is reproduced, and the firm’s institutional choices about how AI mediation is integrated into Foundation-stage and Applied-stage work shape whether the pipeline is preserved or eroded. The recommendation is for firms to make explicit choices about formation pathways: which engagements are routed through formation tracks that preserve formative engagement with the work, what supervisory practices are maintained for Foundation-stage consultants, what investments are made in Applied-stage development beyond AI-mediated review. The choices have a content that productivity framings do not engage, and I treat them as institutionally consequential. Ide’s formal result - that protecting the labour-intensive components of task structures is the institutional design lever that prevents entry-level automation from severing tacit knowledge transmission across generations [@ide2025intergenerational] - provides economic-theoretic grounding for why this recommendation is not merely a philosophical preference but a structurally derivable institutional requirement. The absence of an alternative formation mechanism in firms that have already automated entry-level work [@tarki2025consulting] gives urgency to the recommendation: the pathway is being actively dismantled through commercial logic without replacement, and the institutional choices called for here are the only available counterweight to a structural pressure that operates whether or not it is recognised.
The second recommendation is about tooling decisions that protect organisational knowledge from extraction. Chapter 5’s extraction argument demonstrated that the structural pathway through which firm knowledge can flow outward into vendor systems is real, and that the empirical instantiation of the pathway varies by vendor, by contract, and by deployment configuration. The recommendation is for firms to take this pathway seriously in their tooling decisions - to negotiate vendor contracts that explicitly exclude customer data from training where this is achievable, to use deployment configurations that mitigate the structural risk, and to make institutional choices that recognise the firm’s externalised knowledge as an asset whose protection is institutionally important. These decisions are not technical; they are institutional, and I treat them as part of the political-economic landscape Chapter 5 described.
The third recommendation is about cultivation of distinctive firm voice as strategic asset. Chapter 5’s organisational sovereignty argument demonstrated that generic AI mediation produces dis-individuating pressure on firm-distinctive analytical content, and that the firm’s distinctiveness - the way of seeing, the methodological reflexes, the sectoral instincts that make one firm meaningfully different from another - is at risk under cumulative generic mediation. The recommendation is for firms to treat distinctive voice as a strategic asset that requires deliberate cultivation: through the cultivation of senior practitioners whose analytical voice is recognisably theirs, through the institutional practices that reproduce firm-distinctive content across generations, and through the explicit recognition that what makes the firm valuable is the content the framework Chapter 4 described, not the productivity that AI mediation can provide.
The fourth recommendation is about explicit attribution practices. The recommendation to individual consultants to refuse unattributed mediation has an institutional correlate: firms can develop attribution practices that make the use of AI mediation transparent to clients and to the firm’s own practitioners. These practices vary - disclosure on deliverables, attribution conventions in engagement reports, internal practices for Foundation-stage formation - but the institutional commitment is consistent: the legitimacy of the firm’s work depends on explicit narration of how the work has been produced, and the institutional architecture for that narration is a firm-level choice I treat as important.
Brussels, 2025. A European infrastructure advisory firm contributes to a ChMC-convened working group on AI and professional development - one of four firms invited to share practice experience with the standards body. The firm’s written contribution argues that the ChMC’s current CPD guidance on AI literacy is addressing the wrong problem: it trains qualified practitioners in responsible AI use and governance compliance while leaving entirely unaddressed the question of how Foundation-stage formation functions when AI mediates the analytical work through which formation has historically occurred. The contribution does not receive a formal response from the working group’s secretariat. Three months after the submission, the ChMC announces a consultation on its competency framework, with formation pathway review identified as a priority topic. The firm is not sure whether its paper had any effect. It is sure the conversation has started.
The fifth recommendation is about active engagement with the formation question at industry level. The formation pipeline is not a firm-level question alone. It operates across the wider profession, and individual firm choices interact with the institutional architectures of the wider sector - chartering bodies, professional associations, training and accreditation frameworks, regulatory architectures. The recommendation is for firms to engage actively with these institutional architectures on the formation question: contributing to chartering body development of standards that recognise AI-mediated practice, participating in industry-level policy development on disclosure norms, supporting the institutional architectures through which the wider profession reproduces its capacity. The recommendation reflects the framework’s recognition that the formation question cannot be solved firm-by-firm; it requires institutional engagement at multiple levels. The risk any governance response runs here is what Debord calls récupération: the institutional absorption of a critique through conversion into spectacular form. A policy specifying oversight, a disclosure requirement, a framework for responsible AI use - each produces the representation of engagement with the formation argument without restoring the temporal structure of formation. The industry-level engagement this recommendation calls for is formative, not procedural: investment in the formation pathways themselves, not in the institutional architecture of oversight [@debord1967spectacle].
These five recommendations to advisory firms cohere on the institutional choices that shape whether the integrated three-scale system survives the transformation contemporary AI mediation has begun to produce. They are addressed to firms whose institutional decisions recognise the concerns the thesis has named and who choose to take them seriously in their institutional choices.
9.5 What this means for infrastructure clients
The recommendations to infrastructure clients are suited to the same cross-scenario findings, but they engage the procurement decisions, the disclosure norms, and the strategic in-house capability development that clients can institute.
Vienna, 2025. ÖBB, Austrian Federal Railways, has established an internal AI centre of competence for regulatory and strategic analysis. The team of five has spent eighteen months building a library of ÖBB-specific regulatory intelligence: past submissions, regulator correspondence, expert reports, outcome data from price control reviews going back to 1999. They are fine-tuning an internal model on this corpus. They are not attempting to replace external advisory input for major regulatory proceedings. They are attempting to be a more intelligent client: to arrive at advisory engagements having completed the preparatory analysis themselves, so that the advisory conversation can begin at the strategic question, not the background briefing. The advisory firms who work regularly with ÖBB have noticed. The initial sessions run shorter. The questions arrive harder. One senior partner, after an unusually demanding opening session with the ÖBB regulatory affairs director, tells a colleague: they have done our homework. Now they want to discuss the exam.
The first recommendation is about procurement norms that distinguish judgment from output. The framework’s analysis demonstrates that AI mediation produces interactional fluency more readily than contributory expertise (the distinction Annex E develops ), and that the value of advisory work depends on contributory expertise that is structurally harder to demonstrate through deliverables alone. The recommendation is for clients to develop procurement norms that recognise this distinction - to procure for the judgment that advisory work brings, not only for the deliverable that demonstrates the judgment, and to develop procurement frameworks that distinguish between these two registers. These frameworks vary by subdomain and by client, but the institutional commitment is consistent: clients who procure for judgment are supporting the institutional architectures through which judgment is reproduced, and clients who procure only for output are accelerating the dis-individuating pressure I name.
Benjamin’s counsel/information distinction makes this recommendation precise: what the client is procuring is counsel - practical wisdom transmissible only from one Erfahrung to another - not information, which any adequately calibrated model can produce. The test is not whether the deliverable is technically correct but whether it bears the marks of the adviser’s sustained engagement with the specific regulatory tradition the project answers to: whether what arrives carries the authority of Erfahrung or only the adequacy of generated information [@benjamin1936storyteller].
Procurement for auratic advisory work means procuring the adviser’s embeddedness in the tradition the work answers to - not the delivery of accurate outputs, but the marks left by sustained formation within a specific institutional context. The practical question is whether the firm’s practitioners carry the Erfahrung of the relevant regulatory tradition, or whether the advisory work is reproduced from a training corpus in which that tradition is one input among many. This is not a question of prestige or firm size; it is a question of whether the work bears the marks of genuine formation in the tradition the client’s problem belongs to [@benjamin1935artwork].
The second recommendation is about disclosure norms that recognise the third box. Chapter 5’s sovereignty argument and Chapter 7’s stage-by-stage analysis both demonstrate that the third box - the publics who bear the consequences of infrastructure decisions (developed in Annex B) - is structurally less visible under contemporary AI mediation, and that the institutional architectures that engage the third box require explicit disclosure to function. The recommendation is for clients to develop disclosure norms that make the use of AI mediation visible to the publics whose interests are engaged: in public consultation processes, in regulatory submissions, in the institutional documentation that infrastructure decisions are recorded in. The disclosure norms are not a procedural compliance matter; they are institutional architectures through which the third box’s standing is recognised, and I treat them as important.
The third recommendation is about strategic engagement with in-house capability development. The disintermediation dynamic Chapter 7 examined is partly driven by client decisions to build in-house AI capability, and those decisions matter for what the wider professional ecology becomes. The recommendation is for clients to engage strategically with their own in-house capability development - to recognise that in-house capability development affects the wider professional ecology that the client depends on, that advisory expertise reproduced through that ecology has value the client cannot reproduce internally, and that the client’s procurement decisions and capability decisions together shape whether the ecology continues to produce the expertise the client itself benefits from.
The fourth recommendation is about recognition of the client’s role in supporting the wider professional ecology. This is a more diffuse recommendation than the first three, but it is institutionally consequential. The framework treats the wider professional ecology as an institutional architecture whose continuation depends on choices made by multiple actors, and clients are one of the actors whose choices matter. The recommendation is for clients to recognise their role - not as charitable support for the consulting industry, but as engagement with the institutional architectures through which European infrastructure advisory is reproduced, and on which the client’s own interests depend.
These four recommendations to infrastructure clients cohere on the institutional choices that shape the trajectory of European infrastructure advisory from the demand side. They are addressed to clients whose institutional decisions recognise the concerns the thesis has named and who choose to take them seriously in their procurement, disclosure, and capability development decisions.
9.6 What this means for European policymakers
The recommendations to European policymakers are the most institutionally consequential of the four sets, because they address the conditions under which the political-economic landscape Chapter 5 described could be different. They are suited to the same cross-scenario findings but engage the policy architectures through which the wider conditions of European infrastructure advisory are shaped.
A note on the analytical context is worth registering before the recommendations are developed. The dominant operative framework for European AI governance - visible in the EU AI Act, in adjacent policy frameworks, and in the institutional architectures of contemporary European AI policy more broadly - has crystallised around principlist approaches synthesised most influentially by Floridi [@floridi2023unified]: the five principles of beneficence, non-maleficence, autonomy, justice, and explicability. The recommendations I develop in this section sit alongside this framework, not displacing it. The principlist apparatus addresses the procedural level - what AI systems ought to do - and produces real normative work in the policy contexts where it is deployed. The recommendations I make engage the level the framework works at: what European public-interest infrastructure advisory is constitutively for, and what institutional architectures protect that content under contemporary AI mediation. The two levels are complementary; the recommendations developed here are intended to operate within the institutional space the principlist tradition has helped construct, not to replace it.
Paris, 2025. The French Secrétariat général pour l’investissement publishes revised procurement criteria for AI-assisted advisory tools used in engagements that inform public infrastructure investment decisions above a defined threshold. The criteria require that such tools: use training data that does not include confidential French public-sector content without explicit consent; provide explainability of outputs on request from the commissioning public authority; and be subject to French data protection law for all data processed in the engagement. The criteria do not specify European ownership, French manufacture, or public-sector operation. They specify epistemic and accountability conditions: what data trained the model that informed the recommendation, and whether the reasoning behind the recommendation can be interrogated. A legal adviser reviewing the criteria describes them to a client as “the EU AI Act applied to advisory procurement, before the AI Act applies to it.” They are, in the vocabulary the thesis develops, a partial engagement with genuine sovereignty rather than procedural sovereignty alone - not asking whose tool it is, but asking what it knows and whether it can be held to account.
The first recommendation is about extension of strategic-autonomy framing from defence to other infrastructure subdomains. Chapter 5’s defence reference case demonstrated that European policymakers already treat sovereignty as a strategic concern in defence, and that the logic - the recognition that the political community has standing in decisions that affect its autonomy - is institutionally available within European policymaking. The recommendation is for policymakers to extend this logic to other infrastructure subdomains where the European public-interest tradition is constitutive of the domain’s content: energy, water, transport, urban planning. The extension does not require defence-equivalent institutional architectures; it requires modulated institutional architectures that recognise the sovereignty logic as it applies in each subdomain.
The recommendations developed in this section can be situated within a broader analytical context worth registering explicitly. Acemoglu, Kong and Ozdaglar’s recent formal model of agentic AI under collective learning [@acemoglu2026collapse] develops, through formal economic modelling, an information-design policy proposal that complements the recommendations I am developing here. Their analysis identifies a two-phase optimal policy: a moratorium phase in which agentic AI is suppressed (through additive Gaussian noise added to AI recommendations, formally equivalent to capping AI precision at zero) to rebuild the depleted stock of general knowledge, followed by a permanent cap on AI precision at a welfare-maximising level that they characterise analytically. The two-phase structure has a specific economic-theoretical justification: in the multiple-steady-states regime, where the collapse trap and the high-knowledge equilibrium coexist, the moratorium phase moves the system into the basin of attraction of the high-knowledge equilibrium, and the subsequent cap maintains the system in that basin under conditions of partial AI integration. This is structurally what the thesis is recommending in different vocabulary: institutional architectures that protect the European public-interest tradition (the moratorium-equivalent function of recognising sovereignty), combined with procurement frameworks that cap the operational substitution of AI mediation for the formative engagements through which contributory expertise is reproduced (the cap-equivalent function of distinguishing judgment from output). The convergence between formal economic theory and the philosophical political-economic framework deployed in the thesis is analytically significant: it suggests that the recommendations the thesis makes from a philosophical commitment to answerability are also, on independent economic-theoretical grounds, what would maximise long-run welfare under the conditions their model treats as central.
The convergence is supportive but should not be overstated. Acemoglu et al. operate within a welfare-economic framework that defines the policy problem as maximising long-run welfare under specific assumptions about human preferences, AI capability, and aggregation technology. The thesis operates within a framework that treats European public-interest infrastructure advisory as constitutively answerable to publics who bear the consequences of decisions, including future-generation publics whose conditions of life are being shaped now. The two frameworks converge on what the operational policy should be - institutional architectures that protect formation, procurement frameworks that recognise sovereignty, governance instruments that cap substitution of AI for formative engagement - but they justify the policy from different normative grounds. Where the thesis adds is in articulating what the policy is protecting: not merely a stock of general knowledge in the formal model’s information-aggregation sense, but the answerability of the European public-interest tradition to publics whose interests it has historically served. The economic-theoretical convergence is welcome support; the philosophical-normative argument remains the thesis’s distinctive contribution.
The second recommendation is about procurement architectures that recognise sovereignty. The existing literature on AI sovereignty operates predominantly at the procedural level - data residency, cloud infrastructure, regulatory compliance - and where sovereignty is engaged, it tends to be at the national or supranational scale rather than at the level of how procurement is institutionally configured. The recommendation is for European policymakers to develop procurement architectures that recognise sovereignty: procurement frameworks for advisory services in infrastructure that recognise the European public-interest tradition, that engage the third box institutionally, and that protect the formation pathway through which contributory expertise is reproduced. These frameworks would vary by subdomain, but the institutional commitment is consistent: procurement is not a neutral procedural matter; it is an institutional architecture through which European public-interest practice is reproduced or eroded.
The third recommendation is about engagement with the formation question at policy level. The formation pipeline is not only a firm-level or sector-level question; it has policy correlates that European policymakers can engage. The recommendation is for policymakers to engage these : through support for the chartering and accreditation architectures that reproduce professional capacity, through educational policy that supports European public-interest professional formation, through regulatory architectures that recognise the formation question as a policy concern, not a private firm matter. The engagement is more diffuse than the procurement architecture recommendation, but it is institutionally important.
The fourth recommendation is about support for the European public-interest tradition. This is the most general of the recommendations to policymakers, and it is the recommendation I treat as most consequential. The European public-interest tradition that Chapter 4 described is not a given; it is an institutional achievement that has been reproduced over decades through the institutional architectures of European public policy and professional practice. The recommendation is for policymakers to recognise this achievement and to support its continued reproduction - through cultural policy, through educational policy, through regulatory policy, and through the architectures of European public-interest practice that the tradition operates within.
The fifth recommendation is about explicit institutional recognition of the third box. The third box - the publics who bear the consequences of infrastructure decisions, including the future generations whose conditions of life are being shaped now - is the constituency I treat as central to the answerability of advisory work. The recommendation is for policymakers to develop explicit institutional architectures through which the third box’s standing is recognised: through public consultation requirements, through the institutional architectures of regulatory practice, through the explicit recognition of intergenerational responsibility in policy frameworks, and through the engagement with publics that the European public-interest tradition has historically attempted but that contemporary AI mediation makes structurally more difficult.
These five recommendations to European policymakers are the most institutionally consequential of the four sets because they address the conditions under which the political-economic landscape could differ. They are addressed to policymakers whose institutional decisions recognise the concerns the thesis has named and who choose to engage them through the policy architectures available to European public-interest practice.
9.7 The limits of what the thesis has done
The framework the thesis has developed is not the philosophically complete framework the thesis points toward. Three limits need to be acknowledged honestly.
The first limit is the single-domain character of the analysis. The thesis is about European infrastructure advisory specifically, and the framework’s content is suited to that domain. The analysis does not claim to be a general framework for AI mediation in consulting, in professional work, or in knowledge work. The cross-professional comparative evidence in Chapter 6 supports the claim that the proletarianisation mechanism operates beyond infrastructure advisory, but the framework - the third-box constituency, the public-interest tradition, the dual-scale sovereignty argument - is suited to the European infrastructure context. Whether and how the framework extends to other domains is a question the residual questions section identifies.
The second limit is methodological: the thesis is reflexive and conceptual, not systematically empirical. The thesis is reflexive practitioner research, with conceptual and normative claims rather than empirical claims as its contribution. There is no interview programme, no survey of consulting practitioners, no longitudinal study of AI adoption in firms. The empirical evidence the thesis draws on is the comparative evidence from adjacent professions in Chapter 6, the observational and reflexive material the analytical chapters draw on, and the structural argument I develop. The empirical validation of the framework’s predictions in the European infrastructure context specifically is a residual question. The thesis’s contribution is conceptual and normative, and it should be read as such.
The third limit is the deliberate lightness of the philosophical apparatus. Chapter 3 developed the framework at a working register adequate to the analytical chapters, but the philosophical depth of the framework - Stiegler’s wider apparatus including tertiary retention, grammatisation, the pharmakon, and épokhé; Simondon on metastable individuation; Jonas’s relationship to Heidegger, Ellul, and Feenberg; the phenomenological grounding of methodology in Merleau-Ponty and Dreyfus; the philosophical ontology of LLMs grounded in Floridi and postphenomenology - has been deferred to a Chapter 3 bis that this thesis does not include. The lightness is deliberate but it is also a limit: the philosophical apparatus the thesis deploys is sufficient for the analytical work the thesis undertakes, but it is not the philosophically complete framework the thesis points toward. A subsequent doctoral thesis would develop Chapter 3 bis and would extend the framework into the philosophical depth the analytical chapters indicate would be required.
These three limits are acknowledged honestly. They do not undermine the contribution the thesis claims. They situate the contribution accurately within the wider conversation it is contributing to.
9.8 Residual questions
Five residual questions emerge from the thesis’s work that further research should take up.
The first is empirical validation of the framework’s predictions in the European infrastructure context specifically. The thesis’s central claims are conceptual and normative, but they include empirical predictions: that the proletarianisation mechanism operates in European infrastructure advisory in ways analogous to its operation in software and law, that the formation pipeline is being altered, that the firm’s distinctive way of seeing is at structural risk under cumulative generic mediation. These predictions could be tested through systematic empirical work - interview programmes with practitioners across the ChMC stages, longitudinal studies of AI adoption in consulting firms, comparative studies across European infrastructure subdomains. The empirical validation would either confirm the framework’s predictions, refine them, or identify where they require revision. The thesis treats this as the most consequential of the residual questions.
The second is comparative analysis across non-infrastructure consulting and non-European contexts. The thesis is suited to European infrastructure advisory specifically, but the framework’s analytical apparatus could be extended to other consulting domains and to non-European contexts. Comparative work could identify which content of the framework is general (the proletarianisation mechanism, the sovereignty argument, the formation question) and which is suited to the specific domain (the third-box constituency in its European public-interest form, the European tradition). The comparative work would refine the framework’s claims about generality and specificity in productive ways.
The third is the philosophical depth that Chapter 3 bis would develop. The deferred Chapter 3 bis would engage Stiegler’s wider apparatus, Simondon’s metastable individuation, Jonas’s relationship to Heidegger and Ellul, the phenomenological grounding of methodology, and the philosophical ontology of LLMs. The development would be a contribution to philosophy of technology in its own right, and it would deepen the framework the thesis has developed at working register.
The fourth is the operational specifics of the recommendations developed in this chapter. The recommendations to the four constituencies are suited to the cross-scenario findings, but they are not operationally specific. The procurement architectures that recognise sovereignty, the chartering body standards that recognise formation under AI mediation, the disclosure norms that engage the third box - these are recommendations whose operational specifics would require further work, ideally in collaboration with the institutional actors whose decisions would shape them. The thesis registers the recommendations but does not develop them operationally; this is residual work for further engagement with the institutional architectures the recommendations address.
The fifth is longitudinal validation of the long-horizon predictions about Chartered-stage erosion. The framework’s most consequential prediction - that contemporary alteration of the formation pipeline propagates forward into the senior practitioners of two decades from now - cannot be validated within the timeframe of any single research programme. It requires longitudinal research across decades, tracking how cohorts of consultants formed under different conditions develop the capacities I name. This is the residual question with the longest time horizon, and it is the most institutionally consequential because the framework’s normative argument depends on what the long-horizon prediction is correct about.
These five residual questions identify the territory that further work should engage. None of them undermines the contribution the thesis claims; together they situate the contribution within the wider research programme it is part of.
I write as a practising consultant in European infrastructure advisory. The puzzle that drove this thesis is one I have lived. The framework I have developed is offered as a contribution to making contemporary advisory transformation visible in terms that the productivity framings cannot capture. The recommendations I have drawn are addressed to the four constituencies whose decisions will shape the trajectory European infrastructure advisory follows in the next decade.
I do not know whether the trajectory will follow Commodity Advisory, Digital Twin Economy, Artisanal Advisory, or Disintermediated Clients. What I now believe, having developed the framework and worked through the analytical chapters, is that the trajectory is not given. The conditions under which it resolves can be engaged by the constituencies I identify. The engagement requires analytical apparatus of the kind the thesis has developed.
The European public-interest tradition in infrastructure has been one of the achievements of the post-war institutional order. It has produced infrastructure decisions that have shaped how people in Europe live across decades - grids that connect, water systems that protect, transport networks that enable, urban architectures that house. The decisions have not been perfect; the tradition has its failures and its contradictions. But the tradition - the public-purpose orientation, the regulated utility architectures, the recognition of intergenerational responsibility, the institutional commitment to publics whose interests the work serves - has been an achievement that the advisory profession has been one of the institutional architectures supporting.
The third box - the publics who bear the consequences of infrastructure decisions, including the future generations whose conditions of life are being shaped now - is the constituency this tradition has historically answered to. The advisory profession that serves this tradition has been one of the institutional architectures through which the answerability has been exercised. The answerability depends on the expertise of the practitioners who exercise it, on the institutional architectures within which they work, and on the integrated three-scale system through which their expertise has historically been reproduced.
Whether the answerability survives the transformation that AI mediation has begun to produce is the question this thesis has tried to make visible. The framework has provided analytical apparatus for engaging the question. The recommendations have identified what the four constituencies can do. The residual questions have identified what further work is required.
I close with a recognition that the work the thesis has done is one contribution to a wider conversation that is already underway and will continue beyond the thesis’s submission. The analytical apparatus the framework provides is offered to that conversation as one resource among others. The recommendations are addressed to constituencies whose decisions matter, in the recognition that the trajectory European infrastructure advisory follows is not given but is being shaped now, by the decisions practitioners and institutions are making, and by the engagement those decisions reflect with what is at stake.
The engagement is itself the European public-interest tradition I treat as constitutive of legitimate advisory practice. The tradition has been reproduced through such engagement across decades. Whether the engagement continues, in the form I name, is the question the next decade will answer.
That is the question the thesis closes with. The work that follows will answer it.