User count = humans + agents
We said enterprises would hire agents before they could manage them. What we underweighted is that everyone else would acquire agents as users, and would have no way to see them.
LLM evaluation
Future #2: Trust Is the Missing Layer Between Foundation Models and Proprietary Data
Bee Partners on Future #2: the evaluation, enrichment, integration, and deployment stack that turns foundation models against proprietary data from a pilot loop into production.