Analytics has taken significant leaps forward in recent years: utilization of big data, analytical databases, smart algorithms, and increasingly easy-to-use tools have enhanced the business value of data. However, there are many areas where humans have traditionally been superior to computers due to certain qualities that we have but computers don’t. Our observational skills are far better, in addition to which we are better at seeing connections and at creative thinking. However, the situation is changing: for example, IBM has been developing cognitive data processing for years. This development work has produced Watson, an artificial intelligence with cognitive intelligence, i.e. the ability to think. Watson learns, deducts, understands normal language, and communicates with humans in a manner that is far more natural than what traditionally programmed systems are capable of. In addition, Watson is able to absorb limitless amounts of research data.
Cognitive systems can make the connection between the contents and the correct context and find what is essential in increasingly diverse data entities. They are also able to identify and create new deduction models based on available data, and can find answers to complex questions. In this way, cognitive artificial intelligence can speed up and enhance decision-making, bring up new viewpoints, and assist people extensively.
At present, people mainly communicate with cognitive systems via user devices. The difference with traditional systems is the way the system is used. While traditional systems are commanded either through menus or code language, cognitive systems can be controlled using normal language, either by writing or speaking. Increasingly, mobile applications are continuously being developed for business operations. These provide inventive ways of utilizing cognitive technology e.g. through smart phones, spectacles, and clothes.
The second significant developmental step that will be taken in the near future is changing the processes of artificial intelligence. The first examples of this have already been seen. At present, most of the learning processes with a cognitive aspect have been designed and are run by humans, i.e. humans teach the machine. In analytics, teaching models is nothing new. Now a machine can learn more extensively by monitoring human activities, e.g. the work of a surgeon, and perform the same operation based on what it has learned. IBM’s Watson is successfully used all over the world not only to assist medical doctors but also to support business decisions relating to pricing and investment decisions, for example.
Various attempts have also been made to facilitate the independent learning of systems. The attempts have come up against several challenges: for example, Microsoft had to remove from use a chatbot based on artificial intelligence once Twitter users tricked it into making racist and otherwise insulting comments. Sophia, a robot developed by Hanson Robotics, also received international attention after it painted nightmare scenarios for the future of the mankind in an interview by CNBC International.
The volume of increasingly diverse written, spoken, and visual data is growing exponentially in the digital economy with increasing mobilization and use of social media. Editing this kind of data to produce information increases the need for cognitive information processing. Once these trends create a constant need for increasingly personalized customer dialogue, cognitive processes will also take over this area. On the other hand, development challenges political systems to establish clearer regulation for the transparency of the data economy based on the right of individuals to information that concerns them. Therefore, earning and maintaining the trust of consumers/citizens is a requirement for the development of companies and the public sector.
Cognitive data processing will offer huge opportunities in the coming years. The border between work tasks that have traditionally been performed by humans and mechanical data processing will shift. The opportunities for added value generated by deduction based on unforeseen amounts of information and up-to-date learning are limitless. This will mean a breakthrough for the IoT in homes, for example in the form of speech-recognition robots that will assist families with children and senior citizens. We are facing the most significant developmental leap of the computer era, and we cannot afford to miss the opportunities that it provides.