Why don't you like AI with artificial intelligence?

Artificial intelligence The future is already here

Good Morning! Here is Alexa, your personal assistant! Time to get up! I've already switched on the coffee machine. When you get out of the shower, the cappuccino is ready, just the way you like it. You have an early meeting today, so please listen to the news in the self-driving car that picks you up. While you're at work, I'll instruct the household robot to clean up. Milk, butter, and eggs are all - I'll order new ones for tonight. How many degrees should the apartment have when you come home?

THAT is the future as researchers see it for the year 2030. And a large part of it already exists - even if often not yet so mature. We are already talking to our cell phones or to Alexa. We can now switch on the heating or the light from work using the app. When shopping online, the providers already remember what we like and suggest similar products to us. The basis of all of this? Artificial intelligence, or AI for short.

What does artificial intelligence mean?

It is difficult to say exactly, because even "human intelligence" is not firmly defined as a term. AI is about building a computer that can solve problems on its own. It is programmed with the help of algorithms, i.e. with instructions on how to solve a specific problem. A good comparison from the net:

If you imagine a board game, the written instructions for it are the algorithm. The program represents the real play and application of the rules described in the algorithm above.


So it's about teaching a machine enough that it can solve problems without a human. The prerequisite for this, however, is that she learns independently.

Why do we need artificial intelligence?

AI is already in an incredible number of areas: from search engines like Google to vacuum cleaner robots to diagnosing diseases, monitoring public spaces and calculating stock prices. Self-learning computers are therefore already present in almost all areas of our life. And artificial intelligence will radically change our lives over the next few years and decades. Many scientists are convinced of this - including those at Stanford University in California / USA. Researchers there published a comprehensive report on artificial intelligence in 2016. You said back then that artificial intelligence will influence our lives in the following five important areas:

help at home

Wouldn't it be nice if you never had to do the laundry or mop the floor again? There are already little helpers who vacuum or the intelligent refrigerator that reports when the milk is empty. In the future, however, the service sector will be about networking. The goal: a "Smart Home" - a fully networked digital home in which an intelligence, for example the personal assistant, coordinates everything. In addition, the researchers see great potential in robots or drones that deliver parcels or clean offices. This requires more powerful hardware such as computer chips and sensors and further developments, among other things, in speech recognition.

Who is treated and who is not?

It cannot be dismissed out of hand: On average, a computer can do far more calculations per second than a human - you could say that the machine thinks faster. In situations where people's health is a concern, this can make a world of difference. There are already diagnostic programs that determine the ideal treatment option based on a patient's medical history - and much faster than anyone could. The problem: a diagnostic program only knows the data it is fed with. And if it is instructed to pay attention to the costs, such a program could theoretically decide that a seriously ill person no longer receives a certain expensive treatment - because it is statistically not worthwhile. On the other hand, care robots could perhaps compensate for a lack of skilled nursing staff or rehabilitation robots could help people with their therapy. In 2016, for example, a speaking rehab robot called Roreas was presented in Erfurt, which is supposed to help stroke patients get back on their feet. Dresden researchers at the HTW call their robot "August the Smart". He can perceive his surroundings and interact with people.

safety first

Artificial intelligence is already being used for camera surveillance, among other things. In the future, it should be possible to precisely calculate the danger emanating from a place or person. The problem: The data that humans provide to the program can contain a "bias", that is, a distortion. Karen Ullrich recently warned against this. The Leipzig native is currently doing her doctorate on machine learning in Amsterdam. The research sub-area is about how an artificial system learns from examples and can then draw general conclusions from them. At the congress of the Chaos Computer Club, which took place in Leipzig at the end of the year, Ullrich warns of distortions in the data and uses the COMPAS program from the USA as an example. The purpose of this is to determine which offender will relapse after a conviction and which will not. The basis for the program is an algorithm that is fed with data provided by the judicial officers. This is where this bias comes into play: Significantly more Afro-Americans are condemned in the US compared to their proportion of the total population. This fact alone leads to distortions. However, the algorithm does not see the sociological or economic context for crime. For him it simply means: black skin color = risk factor. That is of course highly discriminatory.

Relaxation on the street

Another area in which the researchers also expect major changes is traffic. Self-driving cars will make a large part of private cars obsolete. The researchers' vision: In the cities, residents will share the self-driving cars and will always be able to call one out when necessary. If there are fewer cars on the streets, this naturally has an impact on urban planning and general infrastructure. The downside: truck drivers are also likely to be replaced by self-driving vans and as a result lose their jobs en masse.

Brave new (work) world

And that is the point that worries most critics: The advancement of artificial intelligence will lead to a profound change in the world of work. Professions will die out and many people will probably simply no longer be needed - from parcel deliverers to cleaning specialists to factory workers. How are those affected then supposed to earn a living? For many, work also has something to do with meaning - how will the future world of work do justice to this? These are questions that the experts have to discuss - including at the CES in Las Vegas, the "Consumer Electronics Show" (January 9-12, 2018). The industry there has already become more self-critical, say observers. Critical points such as data protection are also discussed. Around 170,000 visitors are expected at CES. They hope that technical progress will not only make it easier for users to communicate in the future, but will also enable a more uncomplicated, longer and more productive life. The technology fans see entire cities that are networked and automatically take care of waste disposal, recycling or repairs. And are therefore fully in line with the trend. In a preliminary report by CES it says: People are open to new developments and also spend money on them. And according to a study, one in eight in Germany uses so-called "Smart Home" applications - and the trend is rising.