We are permanently surrounded by it. Often it helps us in everyday life without us even realizing: artificial intelligence. It mows the lawn, vacuums the apartment, and in the evening, on the couch, it selects our favorite series for us. It makes language assistants like Siri and Alexa reliable little helpers. We integrate it into our everyday life like a good friend and companion. We trust it a lot; we know that it makes our lives easier, faster and more evolved. It is very close to us and at the same time very strange. What can artificial intelligence actually do? Is it perhaps even more intelligent than we are? What can we do with it - and, perhaps even more importantly, what does it do to us?
Hey Alexa, what is artificial intelligence? Alexa: "Artificial intelligence is a branch of computer science that deals with the automation of intelligent behavior." An intelligent machine's explanation of what AI is is still a bit vague. Therefore, in this text we try to do it with traditional, journalistic intelligence.
Explained in a simple way, AI is based on an algorithm, that is, a computer code. This makes it possible for machines to process and complete tasks independently. The programmer decides on what tasks these are. Nevertheless, we speak of (artificial) intelligence because researchers develop algorithms in such a way that they can classify and assign data, for example, via so-called neural networks or decision trees.
This enables the machines to perform assigned tasks "cleverly". In other words, machines can use artificial intelligence to deduce, learn and think - just like us humans. Or just almost like us. Because the truth is: Even "a four-year-old child is still far superior to even the most artificial artificial intelligence," ERGO Chief Digital Officer Mark Klein writes in his first blog post here on //next. That's why the machine intelligence that prevails today refers to a "weak" AI that has been trained for narrowly defined tasks. In contrast to a strong AI, which actually imitates human intelligence. The latter is still dreams of the future - quite clearly!
Nevertheless, AI can already achieve astonishing things in many areas. While service robots or voice control in smartphones still seem like little gadgets, AI systems can already be a reliable "partner" for us in other areas. Whether in radiology, dermatology or oncology; whether AI-based chatbots such as Ada for Doctors or intelligent image processing with radiomics: medical professionals are increasingly being supported by AI systems in anamnesis or diagnosis. Keyword: "Decision Support". Artificial intelligence can save lives in this way.
However, sometimes it also ensures that people can participate in social life at all. Microsoft's Seeing Al App, for example, supports people with disabilities in their everyday lives by describing and sorting the environment for blind people. The Arya app, on the other hand, helps people suffering from depression to take their lives back into their own hands. How? By learning from behavioral patterns and suggesting an activity that is good for the user at the exact right moment.
A revolutionary in professional life
AI systems are also real specialists in industry: they enable production processes to be planned more efficiently and flexibly than ever before. They also detect possible errors and incorrect processes faster than any human employee could. Take the Smart Factory, for example: Here, all machines are networked with each other and can analyze and compare incoming data during production - in real time. Regardless of whether it's a factory or an office job, the working world today is already full of "employees" equipped with artificial intelligence.
It is estimated that more than half of all companies use AI in some form (LINK). Completely new products and business areas are emerging. Many celebrate this as a technical revolution. Others have doubts or even fear - for example, of losing their jobs. "AI is a job creator, not a job killer," says the Institute for the Future. (LINK) Because AI only shifts the job spectrum towards greater complexity. In fact, machines have so far mainly taken on tasks that follow a certain pattern - and which we humans therefore often gratefully hand over.
The insurance industry, for example, is relying more and more on Insight Engines: AI systems that sort and categorize incoming customer inquiries at lightning speed and summarize all relevant data. ERGO is no exception. More than 90 percent of incoming e-mails that are not assigned to a specific recipient are now forwarded directly to the correct target addresses by a trained AI system. This saves time and nerves for the clerks and enables them to get in touch with customers more quickly.
Not only on the job, but also otherwise: without AI many things would not be possible or would be slower and less accurate. AI opens new ways and visions for the future. But can we keep up? The important thing is: AI shouldn't escape our control. The use of "clever" machines must be regulated and limited by legal frameworks. In other words: AI systems should act independently, but only under our supervision.
What sounds contradictory is extremely important. Why does the music streaming service know which songs I like? How does the navigation system know what address I call home? How does the fitness bracelet know that today's running distance was longer than last week's? All of that is only possible because of the vast amount of data that AI systems collect about us or that they access in order to function. We need to disclose our data to take advantage of the potential of advanced technology.
But we also need to be confident that our data is safe - safe on our cell phone, on our navigation system and on our fitness wristband. The philosopher Richard David Precht warns: "Artificial intelligence may be intelligent in the broadest sense, but it doesn't actually have any sanity at all and not remotely reason!" (Source: ARD, Titel Thesen Temperamente: Richard David Precht über künstliche Intelligenz). Access to the data from our applications therefore has to remain with us. It is important not to take the results of machines for granted.
And: We need ethical regulations that ensure people's privacy, create transparency and prevent discrimination. Formulating such a catalog of ethical rules and regulations is a major challenge. After all, the possibilities opened up by "smart" computers are becoming increasingly versatile and attractive. Even "weak AI with the intelligence of a worm can do harm", says ERGO CDO Mark Klein. ERGO has therefore given itself an AI code that consistently sets the company guidelines.
Using the full potential of AI systems and integrating them into society at the same pace "in a humane way" - a difficult balancing act, but one that we cannot - and do not want to - run away from. After all, it is clear that artificial intelligence has already changed our society for the long term. And AI research is confidently making predictions for the future:
Artificial intelligence will therefore help to find solutions to the fundamental problems of mankind and to implement them: In the field of medicine, AI could play a decisive role in the development of new and important drugs. According to the drug company Merck, a great opportunity for the future here is the acceleration of drug development, so that more new active ingredients reach the market faster.
In the fight against the climate crisis, AI is already a powerful weapon: sophisticated home energy management systems can optimize electricity consumption. Intelligent forecasting methods and networks can guarantee the electricity demand and also the electricity production. According to the German Energy Agency, AI is a key technology for increasing energy efficiency worldwide.
Currently, mainly unpopular and monotonous work is given to AI systems. In the long run, however, these "clever" machines will also be used to perform dangerous tasks for us, for example in rescue operations. Furthermore: They can avert dangers. AI can already warn against disasters, for example, by quickly evaluating vast amounts of weather data. Over time, the machines become more and more efficient here as well. But no one knows exactly what the future will look like today. Not even Alexa.
Text: Ron Voigt