In the rapidly evolving landscape of video surveillance, the integration of artificial intelligence (AI) is becoming increasingly prevalent. This article explores the three primary methods of deploying AI analytics in video surveillance systems: cloud-based processing, edge computing, and on-camera deployment. While cloud and edge solutions offer distinct advantages, deploying AI directly on cameras emerges as the optimal choice for providing real-time, efficient, and flexible video analysis. Concluding with an introduction to our innovative AI camera motherboard.