Organizational Technology Ladders: Remote Work and Generative AI Adoption

TEXT START: This study proposes that firms move along an "organizational technology ladder": adopting one technology transforms hiring and work processes and builds skills and organizational capital that change the cost of adopting subsequent technologies.

The Dissection

The paper identifies a real meso-level mechanism: remote-work adoption builds digital coordination, technical hiring, and managerial routines that make later generative-AI adoption cheaper. Its reported effect is modest but directionally clear: a 10-point increase in remote hiring predicts a 0.4-point increase in firm-level job postings mentioning generative AI and a 0.7-point increase within occupations.

But the paper stops at organizational adoption. It measures whether firms mention AI in job postings, not whether AI replaces workers, reduces headcount, compresses wages, or destroys productive participation. It maps the gearing of the machine while refusing to trace the machine's effect on labor.

The Core Fallacy

The ladder metaphor quietly implies cumulative organizational progress. Under the Discontinuity Thesis, the ladder is not a route to stable human prosperity. It is an installation sequence for labor displacement.

Remote work lowers coordination friction. Generative AI then exploits the digitized workflows, technical capabilities, and managerial systems remote work helped create. The paper's central finding may therefore be evidence of acceleration toward automation, not evidence that workers become safer or more valuable.

It establishes an adoption complementarity, not the survival of the mass employment-to-consumption circuit. It does not test P3: the collapse of economically necessary human labor.

Hidden Assumptions

• Job-posting mentions are treated as a meaningful proxy for actual generative-AI deployment.
• The instrumental-variables strategy isolates remote-work pressure rather than broader differences in firm sophistication, labor scarcity, or technology readiness.
• RTO mandates are interpreted as evidence of low remote productivity, though they may also reflect control preferences, status politics, or sunk real-estate commitments.
• Remote-work and AI adoption are assumed to remain embedded in a labor market whose basic structure survives.
• Organizational capital is treated as an advantage for firms without asking whether that advantage eliminates the workers who created or occupied the routines.
• The 2023–2024 observation window is treated as informative about a process whose decisive effects may arrive later, after deployment moves from experimentation to substitution.
• Firm and occupation averages conceal the discontinuity between frontier firms that automate rapidly and laggards protected only by institutional inertia.

Social Function

Classification: partial truth and transition management.

The paper is not empty. It correctly identifies that technologies arrive through organizational preconditioning rather than as isolated gadgets. But its framing converts a potential labor-market execution mechanism into neutral management vocabulary: “technology ladder,” “capabilities,” “organizational capital,” and “adoption.” The language sanitizes the destination.

For executives, it is a deployment manual. For workers and policymakers, it is an anesthetic unless paired with analysis of displacement, ownership, bargaining power, and distribution. It describes how firms become ready for AI while remaining silent on who owns the resulting productive capacity.

The Verdict

A useful but strategically incomplete paper. It supplies evidence for a technology-ladder channel and indirectly strengthens the case for P1 and P2: digitized organizations become easier to automate. Its fatal omission is P3. The organizational ladder is not rescuing mass employment; it is making the workforce easier to remove. Under DT logic, remote work is not the defense against generative AI. It is one of the preparatory rungs.