Artificial intelligence - does our own creation replace us humans?
I read this anxiety again and again between the lines when I hear on the news that artificial intelligence makes thousands of jobs superfluous, so that we as humans are no longer needed. Or I follow the almost mythical stories about the possibilities of thinking and even feeling of objects equipped with artificial intelligence.
This fear is by no means a creation of the digital age. We have experienced it since the invention of and the fascination with automata in the 19th century. For example in E.T.A. Hoffmann's story "The Sandman", in which Olimpia, a wooden doll, is so deceptively "real" human that its true nature only becomes apparent after its destruction. The fear of being completely replaceable and even threatened by our own inventions is also the subject of contemporary films such as "A.I. - Artificial Intelligence "(2001) or" Ex Machina "(2015). It is thus deeply rooted in our culture.
However, the very emotionality of the discourse on artificial intelligence often obscures t what artificial intelligence is and can do at this point in time. Artificial intelligence is used primarily in the areas of speech recognition and translation, image perception and recognition, text supplementation, data analysis and the optimization of automatable processes. AI is phenomenally successful in these areas. Central to this are two aspects: on the one hand, completely new data sources become available, such as pictures, sound recordings and videos. In image recognition, artificial intelligence has been beating human capabilities for two years with an error rate of just 5 percent. On the other hand, the success of artificial intelligence is based on "machine learning", that is, the ability of computers to learn experientially. They are therefore no longer programmed for specific, pre-defined tasks, but "learn" from examples, which they then generalize. That's why diversity is so important to computers equipped with artificial intelligence. The more different data you get about the same topic, the better your hit rate will be. Otherwise, computers, like us humans, suffer from "cognitive bias".
In fact, simplified to this degree computers really seem to have similar mental capacities as human beings. But we should not aloe ourselves to be carried away. In a conversation with a Google expert, I asked how adavanced the capabilities of transfer are developed when it comes to artificial intelligence. His answer: At the moment, we are at the level of a 2-3-year-old, in terms of image recognition, pattern recognition and speech recognition. To achieve this degree of artificial intelligence, Google has built its own data center for about 9 billion dollars because the computing capacity was not enough. Artificial intelligence, even at this level, therefore needs great hard ware capacities and can only be used in a narrowly defined way.
In fact, our imagination plays tricks on us and prevents the beneficial use of artificial intelligence. Even if a computer can generalize from examples in a particular field, the device can not handle this type of tranferential activity in all other fields like humans.
But what's the use of it, if thousands of jobs are lost? And those jobs will fall away. Artificial intelligence will revolutionize our working world. Even more repetitive tasks that many people are doing right now, without knowing exactly for what purpose, will be taken over by computers and robots that can do them better than humans. We, on the other hand, have a chance to re-enter our workplace more creatively and comprehensively as the highly complex emotional and intelligent beings that we are. But what does that mean exactly and what do these new activities look like? Actively shaping them will be our challenge.
We are happy to support you either with our advisory unit or with our educational offers and services of our Campus on this journey into the future. We look forward to your call!