AI and the Rise of Just-In-Time Knowledge Work | Operating Notes

AI and the Rise of Just-In-Time Knowledge Work

24 May, 2026

One of the most interesting effects of AI in the workplace is not automation itself, but the compression of cognitive lead times. For years, organizations applied “Just-In-Time” principles to physical supply chains: reduce inventory, eliminate slack, respond faster to demand, and minimize idle capacity. AI appears to be bringing a similar logic into knowledge work.

Before modern AI tooling, many forms of white-collar output required maintaining large amounts of “cognitive inventory.” This included prepared drafts, research notes, background reading, presentation scaffolding, and partially formed ideas sitting in reserve. That inventory acted as a critical buffer. A sales presentation might require a week of preparation; a technical proposal might require multiple review cycles. Strategy work depended heavily on accumulated context and long periods of deliberate synthesis.

AI dramatically compresses that process. Recently, I was asked to present an account plan with only a few hours of notice. Historically, I would have pushed back and asked for at least several days. Instead, AI generated a workable first draft in under an hour, leaving me the remaining time to refine, restructure, and pressure-test it. The result was not lower quality—in some ways, it was better than what I might have produced before.

But something important changed organizationally: once this becomes possible, expectations recalibrate permanently. Managers ask later. Deadlines compress. Preparation buffers disappear, and immediate responsiveness becomes the norm.

This is where the analogy to Just-In-Time systems becomes highly useful. JIT manufacturing increased efficiency enormously, but it also stripped out resilience. The removal of inventory buffers made systems incredibly brittle during unexpected disruptions, a lesson many supply chains learned the hard way during the pandemic.

AI may create a similar dynamic in knowledge work. The same systems that increase immediate responsiveness can also optimize away the cognitive slack that previously allowed for reflection, error detection, deeper understanding, and independent judgment. In many AI-assisted workflows, the human role shifts from generating original output to reviewing AI-generated output under intense time pressure. That sounds efficient on paper, but it fundamentally changes the cognitive structure of the work itself.

The risk is not simply “AI replacing jobs.” The deeper, more systemic shift may be that organizations increasingly treat human cognition as an on-demand resource: generated just in time, with minimal inventory, minimal slack, and increasingly compressed cycles.

That creates real leverage, but it also creates new forms of structural fragility. We may eventually discover that what looked like “waste” in knowledge work—over-preparation, redundancy, idle thinking time, and accumulated context—was not waste at all. It was resilience.