Empowering generative AI through mobile edge computing

Abstract

Generative artificial intelligence (GenAI) has brought about profound transformations across the diverse domains of the Internet of Things such as manufacturing, marketing, medicine, education and work assistance. However, the proliferation of computationally intensive and highly complex GenAI models poses substantial challenges to servers and central network capacities. To effectively permeate various facets of our lives, GenAI heavily relies on mobile edge computing. In this Perspective article, we first introduce GenAI applications on edge devices highlighting its potential capacity to revolutionize our everyday life. We then outline the challenges associated with deploying GenAI on edge devices and present possible solutions to effectively address these obstacles. Finally, we introduce an intelligent mobile edge computing paradigm able to reduce response latency, improve efficiency, strengthen security and privacy preservation and conserve energy, opening the way to a sustainable and efficient application of the different GenAI models. The application of generative artificial intelligence to mobile devices has the potential to enable integrated, personalized, contextually aware experiences. However, the computational energy demand is challenging. This Perspective article introduces an intelligent mobile edge computing paradigm for the implementation of generative artificial intelligence on the Internet of Things system.