To properly understand how to integrate with agentic tools, we need to rethink collaboration models between humans and machines and continually validate via experiments.
Current agents use natural language as the primary interface with humans and humans have established, machine-readable processes for collaboration.
The question then is: how do we onboard our agentic colleagues? I think by
- adapting human collaboration models to include agents and by
- enabling agents to self-improve and learn from their interactions with humans.
We need to think about how to re-use and adapt our well established human interaction models to include agents: AI Engagement Patterns