Databricks CEO: AI extinction risk near zero, real threat is cyberattacks

Databricks CEO Ali Ghodsi argued AI existential risk is near zero, calling the public fear-mongering harmful. He proposed a four-condition test for recursive self-improvement (RSI) that all fail in reality. The real near-term risk is AI-driven cyberattacks, with vulnerability weaponization time shrinking to hours. Enterprise adoption bottlenecks are missing organizational context, not model intelligence.