Block CFO: AI agents could appear more on org charts as companies work to justify AI spending

TEXT START: Earlier this year, Block CEO Jack Dorsey laid off 40% of his entire staff, part of a structural overhaul within the company as it ballooned its adoption of AI to automate workflows.

THE DISSECTION

The article documents the early conversion of the organization chart from a map of human labor to a map of mixed human-machine production. “AI agents” are being assigned organizational status because companies need to make massive AI expenditures operationally visible, measurable, and defensible. The language of development, context, goals, builders, adoption, engagement, efficiency, and revenue disguises the central transaction: fewer humans supervise increasingly autonomous productive systems.

The most revealing fact is not that workers will train and deploy agents. It is that Block already eliminated 40% of its workforce while expanding AI. The agents are not an extension of a stable workforce. They are the mechanism by which the workforce is compressed, with remaining employees converted into temporary supervisors, integrators, and maintainers of the replacement system.

THE CORE FALLACY

The article treats AI as an organizational augmentation problem rather than a labor-substitution regime. It assumes that because humans are needed to configure, train, deploy, and govern agents today, humans retain durable economic necessity tomorrow.

That assumption fails under the Discontinuity Thesis. Once agents achieve durable cost and performance superiority, every human task involved in building, supervising, and improving them becomes another target for automation. The worker who trains the agent is not securing a permanent role; the worker is transferring operational knowledge into a system that can eventually train, coordinate, and audit itself or require far fewer supervisors.

The article also confuses corporate accounting with social viability. Measuring adoption, engagement, time saved, and revenue may prove that AI creates returns for the company. It does not prove that displaced workers retain access to wages, ownership, or productive participation. The stronger the ROI, the faster competitive pressure forces deployment and headcount reduction.

HIDDEN ASSUMPTIONS

• That new AI-related responsibilities will be created at anything close to the scale of the jobs automated.
• That “builders” will remain scarce after AI makes software creation and workflow design cheaper and easier.
• That human context, judgment, and goal-setting remain permanently superior rather than becoming additional capabilities agents acquire.
• That companies can preserve broad human employment while competitors exploit lower-cost machine labor.
• That training and deploying agents gives ordinary workers ownership or control over the AI capital they create.
• That revenue growth will translate into wages rather than accrue primarily to owners, executives, and infrastructure holders.
• That an agent’s presence on an org chart represents a new job category instead of a formal acknowledgment that software is replacing labor.
• That institutional and cultural adaptation can outrun the competitive mechanics of automation.

SOCIAL FUNCTION

Primary classification: transition management and elite self-exoneration, with a substantial layer of prestige signaling and partial truth.

The partial truth is real: AI deployment requires integration, governance, data preparation, workflow redesign, evaluation, and maintenance. These functions create transition niches and may temporarily reward adaptable workers.

The anesthetic is the framing. Workers are told they will engage with intelligent systems, become builders, and discover hidden talent. Executives can then describe displacement as empowerment and restructuring as innovation. The article converts a labor-demand collapse into a story about organizational evolution, allowing leadership to present mass substitution as a neutral technical upgrade rather than a transfer of productive power.

The “agents on the org chart” concept is especially revealing. It gives machine labor symbolic recognition while leaving human workers with responsibility but no ownership. Humans supply context, accountability, and repair; capital captures the resulting productivity. The org chart becomes a ledger of human subordination to software capital.

THE VERDICT

This is not evidence that the post-WWII employment circuit is adapting successfully. It is evidence that the circuit is being dismantled inside the firm.

The article describes the opening phase of P1 and P3: cognitive work is converted into deployable agents, human headcount is cut, and the survivors are assigned to accelerate the replacement process. “Purposeful” AI spending does not restrain the transition; it makes the displacement legible to finance departments and therefore easier to scale.

The final KPI is indeed revenue growth. Under the existing ownership structure, that means the gains flow upward while wage dependence becomes less defensible. The organizational chart is beginning to show the truth that corporate rhetoric still avoids: the company of the future will contain fewer workers, more machine labor, and a narrow human layer whose viability depends on owning the systems, controlling their deployment, or remaining indispensable to those who do.