Interactions of Intelligent Computing Systems
In my last post, I talked about future of mobile computing, but I really meant intelligent computing. AI based computing will be pervasive and will not be limited to mobile device like a phone or watch. It would be your home security system, entertainment system (Roku, Apple TV, Echo), or home automation (Nest).
In this post, I will describe how these intelligent systems interact with each other. In today’s world, human knowledge is simple too vast for just one system dominates the entire space. In current early stage of development, systems must interact with each other to provide useful features for us.
Intelligent systems will be divided into two groups: generalized systems and special domain systems, similar to medicine professions of generalists and specialists. Generalized systems are the systems we will interact on daily basis and could understand ordinary conversations reasonably well. They are usually in a form of virtual assistants and act as platforms for other more specialized domain systems to integrate with. They can identify topic of interests and refer to a special domain system if we express desires to get a deep dive in that particular topic.
Special domain systems are expert systems, which can assist us in accomplishing tasks such as baking a pie or ordering a cab. Though voice is the main interface for all intelligent systems, special domain systems may require other form of interfaces such as vision or taste to make them smarter and more effective in accomplishing their tasks.
Let’s go through an use case for baking an apple pie. To start, I have a conversation with my virtual assistant about baking a pie. Based on my preference and the season (imagining it is the fall and apples are in season), I decided to bake an apple pie. The virtual assistant would suggest a few recipes. When I express concerns about my lack of cooking experience, virtual assistant would ask me if I would like to speak to a cooking expert, a special domain system. After agreeing to it, I am transferred to the cooking expert system, which assists me with baking the pie. The cooking expert system can watch my baking process through a camera, and may even “taste” the pie through special sensors and provide feedbacks.
This is one of thousands seamless integration of intelligent systems and our lives. But what are the commercial applications of these intelligent systems? How do companies make money on these systems? That is the topic of my next post.