What it takes to scale agents in the enterprise: context, control and choice
Databricks CEO says AI is smart enough; enterprises need context, control, and choice to scale agents.
“We think that the AI is plenty smart already. We don't need more intelligence for the AI, we need more context.”— Ali Ghodsi
Databricks CEO Ali Ghodsi argued at the company's conference that frontier AI is already smart enough and the real bottleneck is infusing enterprise context, framed around four pillars: choice (avoiding lock-in), control (security and cost), and context (via Genie ontology, likened to PageRank for the enterprise). He warned that AI costs will become prohibitive for most organizations in the next 6 to 12 months. This signals a strategic reframing of enterprise AI value away from raw intelligence toward data context and governance.