Scott Amyx

Amyx+McKinsey

AI Ambient Virtual Assistant Will Coordinate & Negotiate with the Network’s Edge

The Open Compute Project and the Telco Project for a new 5G IoT platform is one of many trends moving the industry from centralized cloud computing to decentralized edge computing. The low-latency, error-correcting 5G IoT networks will be capable of coordinating thousands of autonomous vehicles, robotic systems, drones, equipment and M2M sensors, designed to scale to 500 billion IoT devices.

Full abstract

The Open Compute Project and the Telco Project for a new 5G IoT platform is one of many trends moving the industry from centralized cloud computing to decentralized edge computing. The low-latency, error-correcting 5G IoT networks will be capable of coordinating thousands of autonomous vehicles, robotic systems, drones, equipment and M2M sensors, designed to scale to 500 billion IoT devices.

Moreover, the Moore’s Law is driving the miniaturization of chips used to power the network’s edge. For instance, Intel’s Curie, the size of a shirt button, is equipped with a 32-bit Intel Quark system-on-a-chip (SoC) with 384kB of flash memory and 80kB SRAM. It’s capable of handling computing power for wearable devices.

At the International Solid State Circuits Conference (ISSCC), MIT researchers introduced a 168-core chip dubbed 'Eyeriss' that will enable IoT devices to run artificial intelligence algorithms locally without the need to have Internet connection or access to the cloud.

On privacy and security, blockchain or decentralized ledger systems will support trustless peer-to-peer messaging and secure distributed data sharing with a robust and scalable form of device coordination.

These fundamental shifts in the network, hardware and standards will mean that the future of data management will occur on the edge and we as human beings will be at the epicenter of these concentric circles.

AI/ machine learning is ever becoming pervasive and integrated into our daily lives, from Siri, Google Now, Microsoft Cortana to Amazon Echo’s Alexa. Cognitive science, affective computing, neural networks, natural language processing, computer vision, facial recognition detection, voice analysis, robotics and other AI disciplines and techniques are maturing towards strong AI. No longer will we need to interface one-to-many with apps and devices but rather we will have an intermediary -- an AI ambient virtual assistant that will coordinate and negotiate will apps and devices on the network’s edge on our behalf.