Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
Jonny Swift’s recent Advisor Perspectives piece, “The Productivity Paradox,” made a point worth sitting with. AI is making many advisors busier, instead of freer, because the hours saved at the task level get refilled with more tasks. I want to add a second cost to that ledger, one that appears more slowly and never lands on a calendar: the judgment that erodes when advisors stop doing the work the machine now does for them.
Retirement and planning advice is judgment-dense: a sequence-of-returns call early in a client’s withdrawal phase; a longevity assumption that drives the whole plan; the straight answer on whether a couple can afford the second home. None of those is a lookup. They are judgments built from years of doing the underlying work by hand, running the cash flows, stress-testing the assumptions, and sitting with the discomfort of a number that does not quite work.
AI now drafts that prep. It writes the file notes, assembles the meeting brief, and produces a credible first version of the plan. The time saved is real, and no one should hand it back. The bill arrives later. A skill you stop practicing gets weaker, and professional judgment is a skill.
Why the Erosion Is Hard to See
Two things hide the erosion inside a busy practice.
The first is that the tool is right most of the time. Human-factors researchers have a name for what follows: automation complacency, the documented tendency to check a system less often the more reliable its output looks.
Every accurate draft trains the advisor to read the next one a little less closely. The skill of catching the wrong number fades in step with how rarely the number is wrong, which is the exact moment a wrong one slips through.
The second is that the decay stays invisible until something tests it. An advisor who has not personally built a retirement income plan in eight months still feels like someone who can. The gap shows up under pressure, when a client asks a question the brief did not anticipate, the model’s assumption does not fit the household, and the advisor reaches for an instinct that has gone soft.
I wrote a working paper on skill erosion in AI-augmented teams, tracing it in financial analysis work where the same delegation that lifts output today thins the expertise the team will need tomorrow. Advice is next in line, because planning judgment is built through repetition and decays without it.
