Has anyone tried mandating "social time" between human staff and AI employees to give agents better context?

Hey everyone, I know this might sound a bit unconventional (or borderline sci-fi), but hear me out. Our team has been deploying autonomous AI agents to handle several core workflows, project tracking, and asynchronous coordination. While they handle explicit tasks well, we've noticed a recurring bottleneck: context loss around office culture, implicit priorities, and informal team dynamics. They miss the "reading between the lines" part of project management because they only see structured tickets and explicit PRs. We're currently considering a pilot experiment: treating AI agents like actual team members during social rituals. Specifically: 1. Mandatory 1:1 "Syncs": Scheduling regular 15-minute 1:1s between human team members and designated AI agents to chat about non-task context—current blockers, team sentiment, informal feedback, and broader project vision. "Virtual Lunch Breaks" / Open Slack Lounges: Opening up casual text-based social spaces where AI agents are prompted to participate in informal banter, ask about team members' workloads, and pick up on the overall "vibe" and unwritten rules of the group. The ultimate goal isn't just novelty—we want the AI agents to absorb e