Whether it’s voice assistants or a video streaming portal, artificial intelligence (AI) can be found in an increasing number of applications, not only in the home, but also in industry. But just what is artificial intelligence and what role will it play for people in the future?
There seems to be no problem at the moment that can’t be solved with artificial intelligence (AI). AI systems are brilliant chess and go players, answer users’ questions in the form of voice assistants, or develop effective targeted campaigns in marketing and retail.
But just what is artificial intelligence exactly? It’s a field of computer science that investigates how intelligent human behavior can be applied to mechanical systems.
For that to be possible, a machine must be able to exhibit intelligent behavior equivalent to that of a human. That is determined by what is known as the Turing test. The test was invented in 1950 by Alan Turing, who is also regarded as the father of modern computer science, and involves a human evaluator conducting an interview with another human and a machine without audio and visual contact. Both interviewees try to convince the evaluator that they are thinking human beings. If the evaluator cannot reliably tell which is which, the machine is said to have passed the test and its intelligence is regarded as equal to that of a human.
One of the present-day uses of the Turing test is to detect and repel spam. The CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) method is frequently used on websites: For example, a web user is asked to read a simple mathematical equation from an image, work it out and then type in the result in a box. While humans usually solve these tasks without any difficulty, artificial intelligence largely fails to do so. The reasons: Numerous human behavioral patterns are simulated. They include mouse movements, clicks and keyboard actions with realistic delays, jerks and gaps – things a bot as a form of artificial intelligence cannot generally imitate.
As a result, bots are by and large denied access to websites and hence cannot obtain user data, such as e-mail addresses, for advertising purposes.
Yet that is not the only finding that can be gleaned from the history of artificial intelligence (AI). A look at its development also reveals that there is no such thing as an all-encompassing artificial intelligence. Instead, artificial intelligence describes a collection of various methods and technologies that are closely related and linked to each other.
The two most well-known technologies and methods are machine learning and deep learning, which, in turn, is a sub-area of machine learning:
Artificial intelligence has a broad range of uses. Its application areas extend from cashless payment to various types of security procedure and industrial production.
One example of the use of artificial intelligence (AI) is face recognition, which is becoming part of our everyday life to a greater and greater extent. Hardly any user will give a moment’s thought to the fact that unlocking their smartphone by camera would not be conceivable without AI. That process appears trivial to humans, but it initially entails a lot of effort for a machine. The reason: Face recognition does not compare the current camera picture with the points of a photo, but develops a computational model from a large number of pictures. It is this model that then recognizes the user – even under difficult conditions, such as different lighting or if the user is wearing glasses.
Insurance is one of the industries that is strongly committed to developing artificial intelligence. The Zurich Group has slashed personnel costs by using AI to handle routine cases as part of claim processing. While an experienced claims handler needs 52 minutes for a case, self-learning software can accomplish it in just five seconds.
Another application area in insurance is detecting fraud. Artificial intelligence can check the plausibility of customers’ claims. It compares fraud patterns and relieves claim handlers of arduous work.
Initial insurers have already launched usage-based rates in car insurance. These rates are based on considerable quantities of telematics data collected by the sensors installed in the car. Tailored risk profiles are created to reflect the mileage and the driver’s individual habits. It’s difficult to imagine analyzing the data without AI.
Humanoid robots in retail or care facilities are still the exception. Such robots might be used increasingly in caring for the sick or elderly in the future in view of rising costs and a shortage of nursing staff.
Artificial intelligence creates the foundation for interacting with patients by speech and gestures. The greatest challenge lies in reliably identifying a person’s emotions from their facial expression and the tone of their voice. Only once that is ensured can humanoid robots respond appropriately to humans. Scientists are now conducting research on suitable solutions.
Humans have worked with robots for decades in production and at warehouses. A collaborative robot (or cobot for short) can repeat a movement virtually countless times over without getting tired. Cobots work together with humans in production processes and are no longer kept separate from their human co-workers with protective devices like typical industrial robots. Compared to the latter, collaborative robots are smaller, can be used more flexibly, and are easier to program.
Robots supported by AI-driven image detection systems can dispatch ordered items, for example. In the U.S., cobots determine how fresh fruit and vegetables are on a trial basis and only send out goods that meet the previously defined quality requirements.
Retailers and parcel services need to further increase the efficiency of their processes to ensure that goods ordered online are delivered as soon as possible. Autonomous vehicles in warehouses are therefore often part of everyday life. Artificial intelligence not only helps the vehicles navigate safely, but also take the quickest or shortest route on their own.
AI can simplify our life by accomplishing routine tasks faster. However, the development of artificial intelligence is also the subject of controversial debate.
Face recognition is convenient, and using biometric identifiers to authorize payments is more secure than entering codes. If the transactions are collected at a single point, the data also makes consumers transparent. They have to trust that the stored data is not used for other purposes or combined with other information.
When smart machines take over more and more tasks from humans, skeptics raise the question of how secure jobs will be in such a future. At the end of the day, only one thing seems certain: AI will transform the world of work. That now even applies to professions that were thought to be “immune” to it, such as controlling and marketing. AI will also assume routine tasks in these fields in the future. Instead of humans crunching numbers and extracting key points from them in lengthy, tedious processes, AI will take over the task.
As tasks disappear from working life on the one hand, new challenges, opportunities and hence professions will be created for people on the other. After all, at the start of the chain is always a question or task that AI has to solve – and it’s humans who have to formulate that first. So artificial intelligence will not replace human creativity. The most probable scenario for the world of work moving ahead is that AI will relieve us of risky work and routine tasks.