Ben Recht: Shift from Architectures of Legibility to Participation in AI Public Feedback

beenwrekt · x · 2026-08-19

Ben Recht reflects on a Berkeley microconference about public feedback for AI, advocating for a shift from "architectures of legibility" to "architectures of participation." He argues that policy-minded researchers often fall into a "legibility trap," quantifying complex realities into simple statistics to provide "evidence" for policymakers. While this suits bureaucratic needs, it sacrifices the authenticity and complexity of public engagement. Recht argues for building mechanisms that allow the public to substantively participate in system construction and feedback, rather than just supplying curated data.

Original post →

More from AGI Musings

AGI Musings channel →