AI – our daily companion
Every day, clever voice assistants lip-read all our wishes, suggest suitable products to us from web shops thanks to sophisticated algorithms, manoeuvre our cars into the smallest parking spaces by themselves and our inquiries are answered on various websites by courteous chatbots. The series of such examples, which can be continued almost indefinitely, impressively shows that Artificial Intelligence (AI) has long since become an integral part of our everyday lives. However, we usually associate the popular buzzword with complex machines or human-like robots and overlook the fact that AI is also found in small applications and tools. This means that we use a variety of AI-based applications and services without being aware of it.
This raises the question of what is actually behind the trend topic of Artificial Intelligence and what opportunities or risks arise from it. Despite the current hype, the topic is certainly not new. Experts have been discussing AI for more than 60 years. And in the science fiction genre, relevant visions of the future have an even longer tradition. For a long time, however, no significant progress could be made in this field – mainly due to technical limitations. In the recent past, however, the computing power of our computers has increased considerably. Not least because of this, a decisive step has been taken in the development of AI and related applications in recent years.
The human brain as role model
Currently, Artificial Intelligence is – once again – on everyone’s lips and the road map seems clear: AI will change our daily routine, our lives, our working world and society. But not all contemporaries welcome this development. Diffuse ideas about what AI can and cannot do, or whether it is an opportunity or a threat, lead to uncertainties and ethical concerns. Thus, the first question to be clarified is how Artificial Intelligence is defined at all. Until today, there is no generally valid definition. In the broadest sense, AI is understood as the automation of intelligent behaviour and machine learning. AI is supposed to enable machines to work on tasks that require human-like intelligence to solve them. The basic idea is therefore to imitate the way the human brain functions.
In principle, a distinction can be made between “strong” and “weak” AI. “Weak” AI is what we often encounter – as in the examples mentioned at the beginning – in a supportive way in everyday life. It is limited to a clearly defined area of application, for example the recognition and processing of speech. “Strong” AI, in contrast, has intellectual abilities that are at least equal to those of the human brain. Its conclusions should not be limited to one area but should be transferable to any number of other areas. Until today, it has not been possible to develop a “strong” AI according to this understanding.
AI – but how?
A large part of the (“weak”) artificially intelligent systems, assistants and tools we use every day are based on the principle of machine learning. The aim of this extremely successful sub-area of AI is to identify connections or patterns through the intelligent linking or statistical analysis of data in order to draw conclusions or make forecasts. For this purpose, the learning machine needs a sufficiently large database. In addition, it must be enabled to process this data by programming special algorithms. Thus, human action is necessary before the machine can develop its artificial intelligence.
Machine learning has become particularly popular in application scenarios in which huge amounts of data are to be statistically evaluated and searched for patterns. AI-based assistants, spam filters, language translations, weather forecasts, autonomous driving systems, medical diagnostics and investment advice, for example, work according to this principle.
The so-called Deep Learning, conversely, is a subfield or a further development of machine learning and could become the basis of “strong” AI in the future. It is based on artificial neural networks whose structure and function is modelled on the neurons of the human brain. With a large information base and the structure of the neuronal networks, a system based on deep learning should be able to link learned content with new content independently and in the long term also across departments. In this way, new learning processes would be constantly initiated. The aim is that the machine does not simply learn, but that it learns to learn. The more meaningful data such systems process, the more reliable they become. In the ideal case, the “learning effect” therefore leads to continuous self-optimization.
Machines becoming a threat?
Conventional machine learning requires human intervention to classify conclusions as “right” or “wrong” and make appropriate adjustments. Deep learning algorithms, on the other hand, are supposed to perform this step themselves. In contrast to classical machine learning, it is hardly possible to understand how deep learning systems achieve their results. Their higher precision in pattern recognition is therefore at the expense of methodological transparency. There is a risk that deep learning systems draw wrong conclusions without recognizing or correcting their own errors.
At the latest since the Hollywood classic “Terminator” (1984) showed in apocalyptic images what could happen when machines were given too much power, many contemporaries may have approached the subject of AI with a certain scepticism. In the film, the machines of a technology corporation become independent and have nothing less in mind than to seize world domination. In reality, however, such a threat scenario hardly seems conceivable. Although machines are supposed to be able to perform tasks through AI for which human-like intelligence performance is assumed, this does not lead to their developing their own consciousness. It is also debatable whether machines can ever develop something like intuition or spontaneous creativity. A decisive limit would thus remain.
How can AI be used sensibly?
On the basis of the examples mentioned at the beginning, a few practical fields of application for artificial intelligence have already been outlined. The spread of AI-based applications affects numerous areas of life. While private users are probably mainly happy to receive gadgets that make their everyday lives easier, doctors can use AI to make more reliable diagnoses and politicians can obtain comprehensive analyses of complex relationships in no time at all. Extensive statistics and data volumes can be evaluated in a very short time. For example, AI could be used to make more precise forecasts of key economic figures, such as gross domestic product, unemployment figures or the inflation rate. In addition, intelligent algorithms using microtargeting enable targeted communication with specific groups of voters.
Companies can also benefit from the targeted use of AI. Although AI-based technologies have so far only been used sporadically in the economy, current studies indicate that they are set to become massively more important in the coming years. In the dialogue with customers, the advantages of implementing AI for companies can already be seen today. For example, the intelligent routing of processes in the Contact Center ensures a fair distribution of the workload. Or through chatbots, which currently represent one of the largest areas of application for AI in companies and – just like voice portals – lead to a significant improvement in the service offering with manageable investments. However, it is important to ensure that processes are well thought out and customers are not frightened off by immature applications. It is therefore recommended to implement AI solutions in cooperation with experienced experts, for example a specialized service provider.
The concern that people will be replaced by a bot in the course of automating the customer dialogue can be quickly dispelled. In the customer dialogue, the technology aims to provide the best possible support for service staff. For example, by allowing bots to take over time-consuming routine work, employees can focus on value-adding activities. Artificial Intelligence in customer service, for example, only unfolds its full effect through the combination of state-of-the-art technology and human skills. Instead of a clash, the path to the future leads through the interaction of man and machine.