Research
Field notes from building software with governed, semi-autonomous AI agents: how quality and safety are built in continuously, and what the evidence shows.
Cross-provider AI juries
Across 1,171 independent AI jury rulings, seats on models from different providers disagreed in 42% of pairs, against 23% for seats on the same provider's model. What independent model review finds in governed agent work.
November 8, 2026 · Willem Larsen