AI: Converting science fiction to science fact
- Posted on January 22, 2018
- Estimated reading time 3 minutes
It seems that in the current debate about Artificial Intelligence (AI), the desire for science fiction is greater than the need to solve real business problems. But disappointment remains around unfulfilled promises, loss of control and data insecurity.
Automation as an innovation driver
Automation itself is a phenomenon that has driven people and business for centuries. Until the recent past, however, it was all about training physical machines or software so that they always do exactly the same thing, and that they react to a given impulse the same way at any given time. Today, technologies such as AI with machine learning make it possible to automate processes in a way that they learn to do exactly what is "right" at the moment. And without having been specifically been trained to this situation before.
So the difference between conventional and intelligent automation lies in the way a system is able to cope with the unknown. For a robotic lawnmower, trapped in an induction loop pulling its tracks across the lawn, the environment must be adjusted by laying wires. A system becomes intelligent when, like an autonomously driving car in an unknown environment, it can react or even act on unknown events or complex event chains.
A field of application in which this development is currently making great progress is chatbot technology, which has long been used by many companies for first-level communication with customers in the service center. A simple chatbot, with a largely static question-answer list, can be created within 15 minutes. However, anyone who has ever had to deal with such a system knows how frustrating it can be if the digital conversation partner does not understand the actual question or answers in phrases that unerringly bypass the actual problem.
Avanade is therefore working on new concepts that enable man-machine dialogues based on AI and machine learning. The result is, for example, the Avanade Curious Bot that our team in Germany has developed - an intelligent chatbot that is able to capture context, learn from previous dialogues with the same or other interlocutors, with the desire to even recognize the mood of its interlocutor. The goal is to use this technology to automatically serve standard requests and initiate standard processes without human intervention. For employees, this creates new freedom for tasks that a bot cannot (yet) do today, and there are still many of them.
Interaction between man and machine
There is no doubt that intelligent automation is here. Avanade recently commissioned a study that uncovered the need for a significant shift in leadership capabilities for the AI-first world. In it, we confirmed that a clear majority of executives interviewed are convinced that they must deploy intelligent automation to be a leader in their field. Crucial in all of this is how we shape this developing technology trend – we need to define and organize the interaction of people and machines; intelligent automation also means intelligent division of labor. The machines will not outstrip us, but there are some things they can do better and faster.