When Intelligence Becomes a Commodity
For more than a century, economic theories of the firm have largely assumed that intelligence is a scarce resource. Whether expressed through the knowledge of specialised workers, the expertise of managers, or the strategic judgement of executives, the capacity to process information, make decisions, and coordinate complex activities has been both expensive and limited. Firms emerged not simply because they owned physical assets, but because they provided an institutional mechanism through which scarce intelligence could be organised and directed towards productive ends.
Artificial intelligence invites us to reconsider this assumption. Much of the current discussion surrounding AI focuses on automation, labour substitution, or productivity gains. While these are undoubtedly important consequences, they may obscure a more fundamental transformation. If increasingly capable AI systems make high-quality reasoning, planning, analysis and execution widely accessible, then intelligence itself may begin to resemble a commodity. In such a world, the question is no longer how firms employ intelligence more efficiently, but whether intelligence continues to define the economic purpose of the firm at all.
This is not simply a technological question. It is a question that strikes at the heart of organisational economics.
Historically, firms have created value by assembling resources that were individually scarce. Financial capital was concentrated to enable investment at scales that individuals could not achieve independently. Labour was organised to exploit specialisation and economies of scale. Equally important, firms accumulated knowledge and managerial capability. They brought together engineers, scientists, accountants, marketers, lawyers and executives whose combined expertise exceeded what any individual could possess. Organisational hierarchy itself can be understood, at least in part, as a response to the scarcity of managerial cognition. Decisions required synthesis, coordination and oversight, all of which depended upon finite human attention.
The emergence of highly capable AI systems challenges this model because they increasingly perform functions that were once regarded as uniquely cognitive. They synthesise information across vast corpora, identify patterns that escape human observation, generate sophisticated analyses, draft legal and technical documents, write software, and increasingly develop and execute complex plans. While these systems remain bounded by human-defined objectives and conceptual frameworks, they nevertheless reduce the scarcity of many forms of intellectual labour that organisations have traditionally accumulated internally.
This distinction is important. Previous waves of technological change primarily reduced the cost of physical production or information exchange. Steam power transformed manufacturing. Electricity transformed industrial organisation. Digital computing transformed information processing. Artificial intelligence is different because it acts directly upon cognition itself. Rather than merely accelerating communication or computation, it increasingly substitutes for activities that have historically been performed by knowledge workers.
If intelligence becomes abundant, the traditional basis upon which firms have competed begins to weaken. Throughout the twentieth century, competitive advantage was frequently associated with superior expertise. Organisations differentiated themselves by recruiting exceptional talent, developing proprietary knowledge, and building internal capabilities that competitors struggled to replicate. Consulting firms sold analytical capability. Law firms sold legal expertise. Investment firms sold superior judgement. Universities sold accumulated knowledge. Increasingly, however, many of these capabilities are becoming widely accessible through general-purpose AI systems.
This does not imply that expertise disappears, nor that all organisations become identical. Rather, it suggests that intelligence itself may cease to be the principal source of competitive differentiation. As with other technologies that have transitioned from strategic assets to essential infrastructure, AI may become something that every organisation possesses, but few organisations derive lasting advantage from possessing alone. Just as access to electricity no longer distinguishes firms, widespread access to advanced intelligence may similarly become a prerequisite rather than a differentiator.
Economic value therefore shifts towards resources that remain genuinely scarce.
One such resource is purpose. Artificial intelligence is capable of optimising almost any objective function presented to it, yet it remains dependent upon externally specified goals. Decisions concerning which objectives should be pursued, how competing values ought to be balanced, and what constitutes success remain fundamentally normative rather than computational. A hospital may optimise patient throughput or patient wellbeing. A corporation may maximise quarterly earnings or long-term resilience. These choices cannot be derived from optimisation alone because they require prior judgements about what ought to be valued. Organisations therefore retain an essential role as institutions that define collective purpose.
A second scarce resource is trust. Markets depend not simply upon capability but upon confidence, accountability and legitimacy. AI systems may generate recommendations or execute actions, yet they do not possess legal personality, bear fiduciary responsibilities, or maintain reputations accumulated over decades. Firms continue to provide the governance structures through which responsibility is allocated, obligations are enforced, and relationships with customers, regulators and society are sustained. Indeed, as autonomous systems assume greater responsibility for operational decisions, institutional trust may become increasingly valuable rather than less.
Third, firms possess context that cannot easily be commoditised. General-purpose AI systems are trained upon broad corpora of publicly available information, but organisations accumulate highly specific knowledge through years of interaction with customers, suppliers, regulators and employees. This institutional context includes tacit knowledge, historical experience, organisational culture, informal practices and relational capital. Such knowledge is often difficult to codify, yet it fundamentally shapes effective decision making. The comparative advantage of future organisations may therefore reside less in possessing superior intelligence than in providing uniquely rich contexts within which abundant intelligence can operate.
Perhaps the most significant resource that remains scarce is judgement. Although intelligence and judgement are frequently treated as synonymous, they are conceptually distinct. Intelligence concerns the capacity to reason, analyse and generate solutions. Judgement concerns the capacity to determine which problems deserve attention, which trade-offs are acceptable, and which risks ought to be undertaken. Judgement necessarily incorporates ethical reasoning, institutional values, political constraints and long-term consequences that cannot be reduced entirely to optimisation problems. While AI may substantially augment judgement by providing richer information and broader analysis, responsibility for those judgements continues to rest with human institutions.
This distinction suggests that the future purpose of the firm may differ substantially from its historical role. Rather than functioning primarily as repositories of scarce expertise, organisations may increasingly become institutions that define objectives, cultivate trust, steward unique contextual knowledge and exercise collective judgement. Their value lies not in owning intelligence, but in directing it towards socially and economically meaningful ends.
This perspective also reframes contemporary debates surrounding artificial intelligence. Much attention has been devoted to the extent to which AI will replace particular occupations or automate existing workflows. These discussions are important, but they arguably remain focused upon first-order effects. The deeper economic question concerns the changing nature of the firm itself. If intelligence becomes abundant in much the same way that information or computing power became abundant, then theories of competitive advantage, organisational design and corporate strategy may require significant revision.
The implication is not that firms become obsolete. Rather, the basis of their existence evolves. Organisations cease to derive advantage primarily from assembling scarce cognitive resources and instead derive advantage from defining purpose, maintaining trust, curating institutional knowledge and exercising judgement. Intelligence becomes an input into production rather than the principal source of differentiation.
Artificial intelligence therefore represents more than another technological innovation. It challenges one of the implicit assumptions that has underpinned modern theories of organisation, namely that intelligence is inherently scarce. If that assumption no longer holds, economists and management scholars may need to reconsider one of the most fundamental questions in organisational theory. The question is no longer simply why firms exist, but what unique function firms continue to perform when the capacity to reason, plan and analyse is no longer their scarcest resource.
