Databricks CEO Ali Ghodsi: AI existential risk near zero, real threat is cyberattacks

Ali Ghodsi argues AI existential risk is near zero, dismissing recursive self-improvement (RSI) with a four-condition test that all fail. Real near-term risk is AI-driven cyberattacks, where CVE-to-weaponization time has shrunk to hours. Enterprise AI adoption bottleneck is missing organizational ontology, not model capability.