The 2026 Stanford AI Index Report paints a complex picture of artificial intelligence's current capabilities and limitations, showing that today's models excel in narrow, well-defined domains while stumbling on seemingly simple tasks. According to hosts Dan and Chris on the latest Practical AI episode, the report highlights AI's "jagged frontier" of intelligence—a phenomenon where systems can solve advanced mathematics problems yet struggle to read an analog clock, illustrating the disconnect between benchmark performance and real-world reasoning. Beyond capability gaps, the report documents significant workforce disruption, with junior technology positions disappearing as AI automation accelerates. The episode explores the economic implications of widespread AI adoption alongside growing safety concerns, while also examining the intensifying competition between the U.S. and China for AI dominance. The hosts debate a fundamental question facing technologists and policymakers: should AI be optimized to automate every possible task, or are certain functions better preserved for human judgment and engagement?