Brookings paper: AI pulls expert work in-house as specialization's advantage vanishes
JMateosGarcia · x · 2026-10-12
Brookings published Luis Garicano's (LSE) paper for the BPEA Fall 2026 conference (presented Sept 25, alongside papers on why AI is contentious and financing the AI buildout). The paper predicts AI will push firms to rely far less on outside experts: as AI lowers the fixed cost of knowledge, frequent problems move in-house, general verifiable knowledge production shifts to model makers, and specialists are left with rare problems. Preliminary evidence from ChatGPT usage, corporate legal departments, and freelance markets supports the shift, with the shares of lawyers, accountants, software developers, and management analysts in their main specialist industries falling at an accelerating rate.
Related event: LSE's Garicano: AI Erodes Specialization and Reshapes Firm Boundaries(3 posts)→
More from AGI Musings
- Gary Marcus: generative AI can't be trusted to follow simple, generalizable instructions — patpat_mit · 2026-10-12
- Tejas Kumar: in the AI era, frontend value is taste and the web belongs to agents — TejasKumar_ · 2026-10-12
- Reading LLMs Through Foucault: Humans Are Stochastic Parrots Too — Merzmensch · 2026-10-12
- Automation doesn't care about your PhD either, commentator notes — VraserX · 2026-10-12
- "SaaS has been dead since 2020": fatigue and slop predate AI, argues investor — michalmalewicz · 2026-10-12
- Pro-AI Redditor argues AGI will make 90% of humanity an economic burden, not abundance — Baroness_Munchausen · 2026-10-12