When Capital Starts to Think: Agentic AI and the Future of Labour
How agentic AI dissolves the boundary between labour and capital and reshapes economic theory
When capital becomes active
For centuries, economics treated capital as passive and labour as the sole source of cognition and decision-making. Agentic AI disrupts that assumption by enabling software systems to perceive, plan, execute, and adapt tasks independently within defined constraints. Capital is no longer just a tool—it is now an active participant in production.
The rise of agentic labour
Agentic labour represents a hybrid model where human judgement and AI-driven execution function as a single unit. Humans define intent and retain accountability, while agents handle iteration, analysis, and operational work. This shifts productivity from linear human effort to scalable, capital-driven cognition.
A broken economic assumption
Classical economics assumed that thought could not be separated from human labour. Agentic systems invalidate that premise by embedding cognition into capital itself, which can now scale infinitely at near-zero marginal cost. This fundamentally alters cost structures and productivity models.
A porous boundary
Rather than replacing labour, agentic AI dissolves the boundary between labour and capital. Humans and agents operate in partnership, combining strategic oversight with continuous execution. This creates a new production function that existing economic frameworks do not fully capture.
Rewriting economic architecture
As the distinction between labour and capital weakens, economic theory must evolve. New frameworks are needed to understand value creation in systems where cognition is embedded in capital. The shift is not just technological but conceptual, requiring a rethinking of how work, ownership, and productivity are defined.