Inherent calls for reliable AI evaluations, responsible pacing and industry coordination
Inherent says it is building systems for people and machines to work together, aiming to make those systems easier to monitor, increase human agency and enable socially beneficial cooperation.
TLDR
Inherent argues that reliable evaluations and responsible pacing are critical to safe, trustworthy AI, and that success requires industry coordination—including newer labs. The lab says its own work focuses on human-machine teaming, with goals of improving monitorability, increasing human agency and enabling socially beneficial cooperation.
Combined views
—
1 Source, first seen ago
