Blog · essay
When AI can write the deck, what are you being paid for?
A response to Melissa Bailey of Baringa Partners, whose essay poses the question.
Her answer is right, and her closing line deserves to be quoted in full: "Your moat is depth, judgement, and your willingness to take a position. Nobody is paying a partner rate for a literature review." She argues that two things become more valuable as expertise compression gets cheap: genuine depth, the kind that takes a decade inside a domain to build, and judgement at the edges, knowing which fifth of the analysis matters and being willing to say so. I agree with both. I want to add a third, because I have spent this year testing it.
The deck is cheap. The defensible deck is not.
Any model can now produce a credible draft in minutes. What it cannot produce on its own is a draft that survives scrutiny: every figure sourced, every claim graded, every named person and date checked, the unverified parts flagged rather than smoothed over. That is not a writing problem, it is an operating problem. It needs verification gates, provenance back to source, decision records and an audit trail, and those have to be built, not prompted. The gap between an AI artefact and a defensible artefact is exactly where senior accountability lives, because the senior person signs it.
Judgement scales when the rest is systematised
Bailey's advice is to audit your week and move the hours AI can do onto the things it cannot. The audit is right, but reallocating hours inside an unchanged way of working only gets you the first saving. The compounding gain comes from operationalising it: a system that holds your context so you stop re-deriving it, that runs the compression and the checking to a standard you set once, and that is built to challenge your reasoning rather than agree with it. I built one, and ran it daily while leading a live investment-platform replacement through to go-live. The hours it returned went where she says they should: the question, the room, the position taken.
One addition to her moat
Depth, judgement, and the willingness to take a position. I would add: the willingness to show your working. Anyone can claim judgement. Publishing the method, the decision records and the honest gaps is what lets a sceptical reader check the claim, and a reader who can check you is a reader who can trust you. That is what this site is for.