Artificial Intelligence Explained – Computer Business Review

Add to favorites

In five questions or less, an industry expert defines and explains a technology, term or trend with this installment seeing Heather Richards, CEO of Transversal, tackle Artificial Intelligence.

HR: AI is a broad term describing machines that operate with some degree of intelligence. It can refer to a machine that mimics human thought processes or to a machine that achieves a level of creative autonomy, for example by being able to learn things beyond its original programming. AI encompasses many methods and applications, including natural language processing, problem solving, playing games and even recognizing emotions. In our business, self-service knowledge management, it helps make user interactions more intuitive.

HR:Most AI applications are designed for specific tasks. At a basic level, the designer maps out how an intelligent entity might solve a problem, and breaks down that process into steps that can be expressed as computer calculations. AI often uses nonlinear processing techniques, such as neural networks, to approximate more closely how a living mind works. It requires significant processing power and sometimes large volumes of background data to enable the computer to form judgements. Thats why AI has leapt forward during the past decade alongside the upsurge in data storage capacity.

HR:Deep learning and cognitive computing are both divisions of AI. Deep learning is a technique that enables a machine to learn more like a person does, by using a neural network of multiple layers through which calculations pass in succession with a cumulatively sophisticated result. Cognitive computing describes AI that mimics human thought processes in order to facilitate interactions between a person and a computer. For example, a cognitive application might understand input in natural language, or deduce what a user wants by interpreting disparate clues.

HR:Luckily, robots are still some way from overthrowing humanity! The effects of more benign automation are going to vary between sectors. In our business, we want AI to help employees rather than replace them. If routine, repetitive requests are handled by automated knowledge management, users can solve problems faster and employees can give more time to issues that need personal attention. AI can thus reduce effort and make work more interesting, but the human factor is still very much required.

HR:Many AI functions are small things, invisibly embedded in larger applications, that people might not even notice much less find a threat. Regarding the news-worthy manifestations of AI, like robots and self-learning machines, there is of course concern about how such machines might behave were they to become more powerful than their builders. Industry thought leaders are already developing guidelines about the ethics of AI and how AI should be steered for the benefit of humanity. From our perspective, though, AI is simply a helper. It doesnt supplant anyone; it makes for a more effortless experience.