Meaningful contestation of algorithms requires new machine-grade regulatory power.

Evidence

This recalls our discussion of Facebook’s “prediction engine” FBLearner Flow, where the machines are fed tens of thousands of data points derived from behavioral surplus, diminishing the very notion of the right to contest “automatic decision making.” If the algorithms are to be contestable in any meaningful way, it will require new countervailing authority and power, including machine resources and expertise to reach into the core disciplines of machine intelligence and construct new approaches that are available for inspection, debate, and combat. Indeed, one expert has already proposed the creation of a government agency—an “FDA for algorithms”—to oversee the development, distribution, sale, and use of complex algorithms, arguing that existing laws “will prove no match for the difficult regulatory puzzles algorithms pose.” — Shoshana Zuboff, The Age of Surveillance Capitalism, 1547

Categorization

• SEED: The ‘FDA for algorithms’ proposal is a concrete institutional-design idea an essay could explore: what would an algorithm-oversight agency actually credential, test, and recall?
• XREF: connects to: Decisions made by legal algorithms need contestation rights comparable to human appeals — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law: A Definitive Look at…, 1203; US law requires governmental decisions be explainable, constraining opaque AI use — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law: A Definitive Look at…, 1199; Algorithmic management strips workers of any human authority to contest discipline — Clara E. Mattei, Escape from Capitalism, 119

Broader Connections

• Decisions made by legal algorithms need contestation rights comparable to human appeals #expands (0.90) — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1203
• Algorithmic management strips workers of any human authority to contest discipline #complicates (0.82) — Clara E. Mattei, Escape from Capitalism, 119
• US law requires governmental decisions be explainable, constraining opaque AI use #qualifies (0.78) — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1199