The last decade has been characterised by significant progress, and widespread use, of Artificial Intelligence, thanks to the abundance of data and the possibility of extracting meaning out of them. Of course, high performance computation and storage capabilities at very low cost made this feasible. This has led the big guys to create a variety of services that were just unfeasible in the previous century.
Natural language processing has enabled a new way of interaction with our devices, be it Alexa, the television set or the car navigation system. Chatbots have become more and more pervasive, image recognition and computational photography are becoming pervasive and change the way we look at our environment.
The basic stepping stones for ambient awareness (IoT, seamless interactions, Intelligence/Awareness) have been laid.
Artificial Intelligence is becoming pervasive and now we can tap on AI as we can tap on the Cloud for extra processing and storage capability.
Google announced AutoML Natural Language (ML stands for Machine Learning) and Natural Language API, basically offering anyone of us the possibility to tap on their AI engines to add intelligence to the processing of text in our documents, emails and so on. The beauty of this is that any person can use these services with very little specific skill (you need to know a bit about it but it is at the same level of what you need to know to use a sord processor and a data analytics tool).
This means that the millions of apps developers have now the possibility of extending the features of their apps by embedding artificial intelligence (in the domain of natural language processing and understanding) to enhance the quality of interaction and to better derive meaning. AI is going to become really pervasive. On the other hand, I wonder if this won’t be another stream of data that Google will capture.
There are now many start ups (a subset is shown in the graphic) and quite a few of them are working to provide AI as a Service, AIaaS. In other words they are transforming AI from a service into a tool, like creating a new programming language is not creating a service but it is enabling the creation of services.
I can imagine each one of us, by the end of this decade, to use AIaaS to extend our intelligence. You are stumbling onto a document, a paragraph you don’t understand? Just highlight it, as if you were doing a cut and paste, and invoke AIaaS to work on it and tell you what it means to you… Of course, you can also imagine reading that paragraph and furrowing your brow. That will be enough to invoke AIaaS and get the explanation through Augmented Reality… isn’t that nice?