## Notes from 21 January 2026 [[2026-01-20|← Previous note]] ┃ [[2026-01-22|Next note →]] I saw this [in a post](https://www.linkedin.com/feed/update/urn:li:activity:7418932490221973505/) from Joe Hill at [[Re.State Think Tank]] about the UK government reopening a school of government for senior civil servants. I'm skeptical of creating institutional structures before clearly diagnosing the problem. What exactly needs to change in how senior officials are trained? What's the gap that justifies this institutional response? The risk is falling into the trap of "fixing form before content" - building a nice facility with generic curriculum on AI, data, finance and leadership without confronting the structural issues that actually prevent innovation in the civil service. There's a real danger this becomes civil servants teaching civil servants to be civil servants, reproducing the same practices that need transformation. The underlying challenge is finding ways to teach creative problem-solving about implementation and delivery. But there's a deeper operational dilemma here about how to actually manage training provision. If you outsource everything without strong internal oversight, you end up with a large generic training company that throws together a curriculum on the spot - the path of least resistance in procurement. But if you try to structure everything in-house and contract separately for specific needs, you can go after boutique providers who actually excel in particular domains. The problem is internal capacity: government doesn't always have people who truly understand these subjects well, and even when they do, those experts rarely have time to teach regularly while doing their actual jobs. So outsourcing becomes inevitable, but how do you do it well? How do you evaluate good providers versus mediocre ones? How do you even find the specialized boutique firms instead of just the ones who know how to win government tenders? The UK's previous approach failed not because of outsourcing itself, but because procurement was structured to favor large framework providers who acted as gatekeepers, rather than allowing departments to seek out genuine expertise. A new school of government needs to solve this operational challenge - building internal capability to curate and quality-control external provision, rather than either doing everything in-house (impossible) or outsourcing blindly (the previous disaster). It's a procurement and quality assurance problem as much as a pedagogical one.