The Ladder-Rebuild Model

The Ladder-Rebuild Model is this chapter’s principal framework, and it rebuilds the apprenticeship as a deliberate design, since the osmosis that once did the work has ended. It works in four movements, applied to each entry-level role whose work is automating. Name the judgment the old work built; design a deliberate substitute that builds the same judgment without the automated repetition; sequence and assess the substitute so that it actually forms the capability; and keep the learner in proximity to the judgment being exercised.

The first movement names the specific judgment the automated work used to build. Diagnosis: name precisely what judgment the now-automated work actually built. The document review built pattern recognition for dangerous clauses; the model-building built a feel for when a number is wrong; the reconciliation built an instinct for where errors hide. This diagnosis is harder than it looks, because the judgment was never made explicit while the work built it incidentally, and naming it requires a senior person to articulate what they know that the junior grind taught them, which they have often never had to put into words.

The second movement designs deliberate practice that builds the same judgment differently. Substitution: design a practice that builds the same judgment without the volume of automatable repetition. If the old learning came from reviewing 10,000 contracts, the substitute might be critiquing the machine’s review of curated hard cases, being handed the exceptions the machine flags as uncertain, or working through a designed sequence of instructive examples chosen for what they teach, with the week’s business needs set aside. The substitute deliberately supplies the judgment-building exposure that the automated workflow no longer provides as a byproduct.

The third movement sequences and assesses the substitutes so judgment actually forms. Sequencing and assessment is what turns a good intention into a working ladder. The substitutes have to be staged so that judgment builds in a sensible order, and assessed so that the organization can see whether the capability is actually forming, because a training substitute that no one checks degrades into a ritual as empty as the hazing the chapter warned against. The model describes a standing ladder, and treating it as a course to be delivered once forfeits it, because the judgment it builds is the organization’s future capability and requires the same standing attention as any other asset the organization depends on.

A fourth movement, easy to omit and fatal to omit. Proximity: keep the learner in the room where the judgment is exercised. The first three movements can all be satisfied on paper by a training program, and the old apprenticeship was never a training program. What transmitted was overheard reasoning, the partner thinking aloud about why this clause was the dangerous one, and none of the substitutes reproduce it if the junior is not present when it happens. Designing the practice and then delivering it at a distance rebuilds the syllabus and loses the thing the syllabus was standing in for.

Where the model misleads. Its diagnosis step assumes the automated work built one identifiable judgment, and much entry-level work built several at once, unevenly, in ways the people who did it cannot articulate on request. A leader who runs the diagnosis honestly should expect to end up with a partial list, and should design for the gap it leaves. The model also cannot tell a leader whether the judgment in question is still worth building. Some of what the old ladder transmitted was genuinely specific to work that no longer exists, and rebuilding it faithfully would be an expensive way to train people for a firm that is gone.