Tom leaves stealth: Hortus AI

Hortus AI addresses the disconnect between rapid, top-down AI deployment and the practical needs of public sector constituents by prioritizing active, feedback-driven integration. Rather than optimizing solely for technical accuracy or efficiency, this approach treats AI as a tool for rearticulating human goals within legacy organizations like city governments. Current market dynamics often force agencies into opaque vendor solutions that ignore local stakeholder values, mirroring the destructive, non-responsive urban planning of the Robert Moses era. By focusing on observability and community-informed workflows, organizations can move beyond passive consumption of frontier models to create systems that reflect genuine public interest. This shift requires a commitment to specific, human-centric outcomes, rejecting the "AI for AI's sake" mentality in favor of infrastructure that empowers users and maintains accountability in an increasingly automated environment.