From Autonomy Economics, the reference entry for the discipline
Autonomous capability remains early and uneven. Technical possibility does not automatically produce commercial impact in how activity is carried forward, how interaction takes place, or how access to capability is structured. Several counterarguments therefore deserve to be taken seriously.
The most direct challenge is visible in current data. Firms across professional services, financial institutions, and technology are reducing progression-heavy roles or interaction-dependent costs, holding revenue, and in some cases improving margins. If cost reduction alone preserves the model, the paradox may appear less structural.
The distinction lies in timing. Cost reduction across a sector can become self-cancelling. When firms reduce similar costs using comparable capability, any margin advantage may narrow and the question of value justification may return. Current stability may therefore reflect an early phase, and the resolution may not be durable. The more decisive test is whether clients continue to accept commercial structures built on HBP, HBI, or CSD once the underlying activity, interaction, or capability access no longer depends on them in the same way. Early evidence across professional services, consulting, software pricing, and other progression- or interaction-dependent models suggests how this pressure can begin. A firm that reduces progression-heavy headcount while leaving its billable or access structure unchanged has lowered its cost base, but has not necessarily resolved the structural question. Once clients can see that activity is being carried forward differently, or that capability is becoming easier to access elsewhere, the basis on which the fee was justified may become harder to defend even if the firm remains more profitable in the short term.
Value may concentrate in the platforms and model providers that supply underlying autonomous capability. If that consolidation becomes substantial, firms positioned between the customer and that infrastructure may face narrower margins regardless of how well they redesign their commercial structures. This is a serious objection and deserves a direct answer.
The response is that Autonomy Economics addresses a different level of analysis from the question of where AI surplus concentrates in aggregate. Even if model providers and infrastructure platforms capture significant value from the wider change, an incumbent firm facing client fee pressure, commercial model exposure, or erosion in access-based value capture cannot resolve that pressure by pointing to where surplus accumulates elsewhere in the system. A firm facing weaker pricing power or erosion in the commercial basis of value capture is not protected by the fact that the capability enabling that pressure is profitable for its supplier. The two questions are distinct. One concerns aggregate value distribution at the infrastructure layer. The other concerns firm-level and sector-level durability once autonomous capability changes how activity is carried forward, how interaction takes place, or how access to capability is structured. A firm can therefore face serious structural exposure to the mechanism described in this paper even in an environment where infrastructure providers capture substantial surplus. The two dynamics can coexist, and often will.
Agentic systems can reduce the cost of routine activity, interaction, or capability access while enabling new forms of personalisation and outcome-linked models. Business models may reorganise around new anchors for value as earlier structures lose relevance. The pattern is likely to vary by sector and use case.
This objection is consistent with the framework and does not sit in tension with it. Autonomy Economics predicts that value reorganises and stays in the system. Business-model evolution of the kind described here is the redesign response the paper advocates. The objection becomes a challenge to the framework only if it is taken to mean that this evolution happens automatically and without the structural pressure described in this paper. The historical record in Appendix A suggests that it often does not.
Agentic systems can increase demand for software. They require coordination, monitoring, and integration layers that did not previously exist. The stack can widen as autonomy develops. Expansion in these layers can occur alongside compression in units tied to HBP, HBI, or CSD. The key question is where a firm sits within that structure.
This is correct, and the paper does not dispute it. The pressure described in Autonomy Economics falls on specific revenue lines calibrated to HBP, HBI, or CSD. Software demand in aggregate is a separate matter. A firm whose revenue is anchored in the coordination, monitoring, and integration layers that agentic systems require may face more limited exposure through this mechanism. A firm whose revenue depends on seat-based or cycle-based structures calibrated to human-bounded progression, on interaction-dependent monetisation, or on access to scarce capability faces a different position. Expansion of the stack does not neutralise the exposure of the units within it that depend on HBP, HBI, or CSD remaining the basis for value justification.
Reductions in certain human-dependent inputs can appear as higher margins or increased purchasing power. These effects can support demand and fund reinvestment. Previous technology cycles show that input reduction is often absorbed through reallocation and expansion.
This is historically accurate, and the paper does not claim that autonomous capability produces broad economic contraction. The mechanism described here operates at the level of specific commercial structures and leaves aggregate demand aside. Absorption and reallocation at the macroeconomic level are therefore consistent with significant structural pressure on individual revenue lines and business models during the adjustment period.
Current capability remains narrow in many settings. Even so, once activity can be carried forward, interaction can take place, or capability can be accessed with less dependence on HBP, HBI, or CSD, the economics around it may begin to change. Scale and timing remain uncertain. The possibility that the effect proves larger than current expectations remains material.
This is a boundary condition on the timing assumptions in the paper more than a counterargument. It is included here because it reflects a genuine uncertainty. The framework describes a direction of change with more confidence than it describes a pace. Where that pace proves faster than current adoption patterns suggest, the pressure described in this paper may arrive sooner and with less time for firms that have not begun to redesign.
From section 12 of the foundation paper. HBP, HBI and CSD stand for human-bounded progression, human-bounded interaction and capability scarcity dependence, the three axes of exposure.