For a moment, it looks as if Prof Dr Frank Kirchner is opening his shirt and opening a flap on his stomach. "I'm going to take the batteries out now," he says with a grin. The expert from the German Research Centre for Artificial Intelligence (DFKI) - in reality a humanoid robot? It's just a joke, of course, but it still makes the audience in Oldenburg's Schlosssaal wonder: What if human-like robots really did live among us? Is that even possible and if so, do we want it?
These questions were at the centre of the seventh Oldenburg Castle Talks, hosted by the University of Oldenburg, the EWE Foundation and the Hanse-Wissenschaftskolleg Delmenhorst on Monday evening. Under the motto "Who is afraid of Robo Sapiens? - Possibilities and limits of intelligent machines and how they influence our lives", around 250 guests followed a lively expert discussion. This was expertly chaired by journalist and RBB presenter Jörg Thadeusz, who was standing in for Maybrit Illner, who was unable to come to Oldenburg due to illness.
Four Computing Science experts and a theologian
Several computer scientists who have been researching the possibilities of artificial intelligence (AI) for years took part in the panel discussion: In addition to Kirchner from DFKI, Prof Dr Dr h.c. H.-Jürgen Apppelrath, Professor of Practical Computing Science at the University of Oldenburg and Director of its affiliated Institute OFFIS, as well as AI experts Prof Dr Ulrich Furbach from the University of Koblenz-Landau and Prof Dr Raúl Rojas from the Free University of Berlin. The only non-Computing Science expert on the panel was the theologian Prof Dr Dirk Evers from Martin Luther University Halle-Wittenberg.
It quickly became clear to the panel of experts that a human-like robot is currently pure science fiction. "So far, we have only been able to substitute a pitiful fraction of human intelligence with machines. Our brain is far too complex to be able to replicate its full capabilities," said Appelrath and Rojas added that humans are naturally able to do many things without thinking, such as recognising faces in a fraction of a second or speaking in their native language. Not thinking, on the other hand, is difficult for robots. Furthermore, a robot has no feelings. Although it can perceive them in its counterpart, it cannot really empathise with them. "In a way, the robot is a psychopath," he clarified. Of course, there are jobs that a robot can do better than a human. "Some academic appointments will die out in five to ten years," predicted Rojas. However, in his opinion, certain tasks should remain the preserve of humans, such as caring for the elderly and sick. "The care robot is a horror idea for me," he said.
Empathic care robot
Appelrath saw things differently in such absolute terms. In his opinion, even simple intelligent technologies could support people in need of help in the early stages and allow them to remain in their own home for longer and with greater autonomy. For example, an intelligent system could register when conspicuous changes occur in the use of lighting and household appliances and notify neighbours, relatives or the care service in a flexibly agreed procedure. "We shouldn't taboo such assistive technologies in care," says Appelrath. Furbach even went one step further: "Why shouldn't a robot do more, for example wash and moisturise a person in need of care?" he asked. Tests had shown that a robot head that is able to react empathically to its counterpart was well received by the test subjects. "We should develop more imagination," he demanded.
Presenter Thadeusz questioned whether it was really that easy. After all, someone would have to decide how much responsibility should be handed over to machines, for example whether aeroplanes should soon be allowed to take off without pilots. "We need a social consensus for this," replied theologian Evers, pointing out that ultimately a human always bears the responsibility, for example the person who programmes the robot or regularly checks its systems.
Countering fears with education
"Well, we should be honest about that," interjected Furbach. Humans have long since relinquished control in certain areas, for example in weapons technology or aircraft control, where the pilot cannot necessarily take countermeasures if the machine receives faulty data. "But we have always learnt to deal with new technologies," summarised Kirchner. He is positive about further developments. "There are many ways in which a robot can make life more pleasant for people," he says. Possible fears should be countered with education - and at an early stage, so that nobody feels threatened by the technology.