Descriptions:
Databricks CEO and co-founder Ali Ghodsi made a provocative claim in this Bloomberg Technology interview: that AGI has already arrived. His evidence is behavioral rather than technical—when asked, roughly 90% of any audience agrees that frontier AI models are smarter than most colleagues they work with most of the time. The bottleneck to AI productivity, Ghodsi argues, is not intelligence but context: agents today have tapped perhaps 1% of their potential because they lack access to the organizational data needed to act effectively.
This thesis drives Databricks’ current product focus on Genie, a data-context layer designed to feed enterprise information to AI agents. Ghodsi cited Novo Nordisk as a concrete example: by infusing clinical trial data into Genie, the pharmaceutical company reduced the time required to analyze ongoing studies from weeks to minutes. Databricks is also betting heavily on LakeBase, a new database product purpose-built for agentic workloads, premised on the observation that every piece of software—including the unprecedented volumes of AI-generated software anticipated over the next two years—requires an underlying database.
On capital markets, Ghodsi dismissed fears that an AI IPO wave will crowd out private funding, arguing that as large companies absorb public-market attention, private capital will flow more freely to companies like Databricks. He endorsed Jensen Huang’s thesis that AI agents will become the dominant users of enterprise software, and predicted that more software will be written in the next one to two years than in all of prior computing history.
📺 Source: Bloomberg Technology · Published June 04, 2026
🏷️ Format: Interview
