I work in logistics for a mid-sized retailer, and we've been testing AI-driven demand forecasting for six months—it cut our overstock by 18%, but the system failed spectacularly during a regional promo event because it couldn't predict human impulsiveness. While automation clearly improves efficiency in stable conditions, do you think AI can ever handle the unpredictable spikes and disruptions that define real-world supply chains, or will human judgment always be essential for the exceptions? I lean toward hybrid models, but I'm curious if others see a future where machines take over even those judgment calls.