Great point, and we completely agree—consistency is crucial for a good user experience. Ideally, all AI features should run locally for the best speed and privacy, with the cloud reserved strictly for storage.
The reality, however, is that local hardware computing power and DDR resources vary significantly across models. For example, while the G5 Pro and G350 can handle multiple local AI functions, a model like the G100 only has the resources for human detection on-device. Moving certain algorithms to the cloud wasn’t about pushing subscriptions; it was about giving users the option to use advanced features on hardware that couldn’t otherwise handle them locally.
Moving forward, our strategy is to prioritize sufficient local processing power in our product design, ensuring that matching AI features run on the device itself rather than as a paid service.
8 Likes