Man and machine must collaborate

DR. LARS SCHWABE: The great potential of artificial intelligence for companies is in automation. Self-driving cars, inconceivable a few years ago, are a very visible example of this. The intersection of companies to their end customers will also be strongly marked by AI in the future, in the form of speech-based interfaces like chatbots, for instance. A major topic will also be the archiving and IT-related activation of knowledge that has accrued in a company over the years and usually exists only implicitly in the minds of a few employees – who often are also about to retirement. 

JOACHIM WOLF: This latter point shows that artificial intelligence is no longer a purely technical matter. Instead, we in the industry speak of a triangle: man, machine and data. If AI is to enrich the interaction between people and technology, and thus enhance value added, it is only sensible to make use of it. One example is predictive maintenance, where the machine uses intelligent data analysis to signal the need for maintenance in advance, enabling humans to make the decision before damage occurs.

JOACHIM WOLF: As a matter of fact, it all depends on the level of maturity. For instance, the level of automation in the process industry is necessarily high, since it is required by its very nature to collect and administer data. Things look very different, however, in manufacturing. Here unit quantities of one are practically the norm. In addition, there are also the individual product components from various suppliers. Automation – that is, increasing efficiency with AI – is more difficult to achieve here. The consumer goods market, in contrast, can make excellent use of AI, to engage in scouting, for instance, deduce consumer behavior and respond with appropriate individualized advertising.

DR. LARS SCHWABE: Apart from the differences between industries, the level of maturity also depends on the specific company. There is often a lack of IT infrastructure that AI needs to achieve the desired value added. And this begins with companies neatly managing their data and paying attention to data protection. This is where many companies are, in fact, only just getting started. 

DR. LARS SCHWABE: At present, 95 percent of AI is machine learning and based on data. That is why data security is especially important. In view of big players like Google or Amazon, who accumulate and use huge masses of data in one place, I would here make a case for a decentralized solution. In order to be able to protect data, which in industry or in health care can in some cases be quite personal, they should not get into the hands of only a few players. This is where blockchain technology could become especially important.

JOACHIM WOLF: What is more, the quality of AI hinges on the quality of the data. You cannot make sensible decisions with manipulated data. For this reason, the question of data security is of vital significance. Companies in all lines of business should always keep their systems up to date.

JOACHIM WOLF: The business value must have priority from the very outset. I have to know how a data analyst thinks and at the same time have the tool set of a data engineer if I want to be able to implement the application. One person alone can hardly cover the whole range nowadays. So a team of experts will always be needed. 

DR. LARS SCHWABE: Companies should just get started and gain some experience. They will have to get their employees to go along and break down the fear of the “computerized colleague”. AI should be seen as an aid rather than a threat. Building up expertise is just as important. This does not mean that every company has to hire AI researchers. But now especially, it is important to build up a certain level of in-house expertise for companies to be able to assess and challenge things from a technological point of view. And they should not get too dependent on individual products, service providers or cloud services, otherwise they will end up as a mere passive consumer.

DR. LARS SCHWABE: People will be working more closely together with machines. What is important is that people do not become simply agents of execution, with machines constantly issuing instructions.

JOACHIM WOLF: I see a hybrid model, where man and machine work together intelligently. People cannot be replaced so easily. Improvisational talent, creativity, innovation and empathy will still be theirs alone. By contrast, machines are superior when it comes to processing large amounts of information at high speed. For instance, they can prepare decisions and thus improve the quality of human decision-making abilities. The demands placed on the workplace of the future are going to rise drastically, but the ease of using this technology will also advance. 

DR. LARS SCHWABE: That’s just about how I see it. Many of the big problems facing humanity can only be solved when people and machines work together. There is no alternative here.