As firms deploy large numbers of AI agents built on the same underlying models, the unique judgment and productiveness that companies paid humans for converges toward a single, model‑driven style and set of decisions. That homogenization compresses product and messaging differences, making competition cheaper on inputs but harder on proprietary insight, and shifts rents toward model providers and platform integrators.
— If true at scale, this trend shifts where economic value and political power lie — from firms and workers to the small set of model and agent providers — with consequences for antitrust, labor policy, and national industrial strategy.
Aaron Horwath
2026.09.24
100% relevant
The article’s vignette of Lukas Weber checking and managing multiple agents that produce identical-seeming outputs across competing firms, and the observation that everyone 'hired Mr. Claude' for their thinking, exemplifies this homogenization.
← Back to all ideas