We encounter algorithms just about everywhere on the Internet, for example when shopping online or using search engines, even when posting on social media. But: offline as well. atechbook explains where algorithms come from, what role they play in our daily lives – and why the whole thing is also a bit scary…
You’ve been searching the web for an e-bike – and all of a sudden numerous websites you visit are teeming with corresponding ads? Then a sales algorithm has probably struck. But algorithms can do much more than determine our buying behavior.
What are algorithms?
By definition, algorithms are a series of operations performed step by step with the aim of solving a task or a problem. In our everyday lives, we encounter them everywhere – even offline. For example, in baking, the algorithm is a baking recipe. However, we mainly want to explain the technical side here.
In computer science, algorithms are the basis of programming. They automate processes. Simplified said, as consequence on a certain input, a concrete answer is spent.
The following characteristics apply to algorithms:
- They are determined (= fixed). That means that always the same output must take place on an input.
- Algorithms are deterministic, thus are subject to restrictions. Means: The consistent calculation step is always predetermined.
- Furthermore, algorithms are finite. This “dynamic finiteness” prevents their execution from taking up an unlimited amount of memory.
- Algorithms are terminated. This means that the achieved result is reached after a predetermined number of inputs/steps.
- Last, algorithms are effective. Means that the effect of their instruction is fixed.
Algorithms in online shopping
With the help of algorithms, it is possible, among other things, to tailor advertising specifically to Internet users. With the help of cookies and on the basis of our click behavior and search terms, algorithms create a kind of digital personal profile. Based on this, content and purchase offers can be placed in a targeted manner.
Also interesting: Changes in online shopping
Algorithms also make decisions in numerous other areas that specifically affect us. This idea may not be palatable to everyone. Other examples where technical algorithms are used include movie suggestions on Netflix, partner suggestions in online dating, and spelling programs. Also interesting: as a study by the Bertelsmann Foundation found out, people with a higher level of education tend to see more advantages in automated decision-making.
Human resources, navigation systems, elevators …
While the examples just given are probably not new to many, some uses of algorithms may still come as a surprise. For example, more and more companies are using algorithms in the area of personnel management. Among other things, they scan resumes and application letters according to certain criteria.
The same applies to navigation systems: These make decisions virtually over our heads – with the help of algorithms that calculate the shortest route at lightning speed.
This type of artificial intelligence (AI) is also important for elevators, for example. They signal to it which system should be used to process the elevator calls. Or to put it another way: Without algorithms, the elevator would not “know” whether it should first transport a passenger from the first floor to the 12th floor without any intermediate stops or whether it should let other passengers get on or off in between. Algorithms can also register which call signal is the longest in the past, i.e. which floor has to be reached first.
Controversial topic: predictive policing.
The police also use algorithms to supplement their practice. The goal: to stay “one step ahead” of criminals. For example, software such as Precobs are designed to detect locations where a crime may soon occur, or to identify potential criminals. Through this method of crime prediction, officials hope to succeed in thwarting unlawful acts before they occur.
However, this use of AI is controversial. Among other things, critics fear that a) people will be unfairly targeted if they fit the pattern of suspicion because of any criminal history. And b) this also threatens “normal,” innocent guarantors who have slipped into the pre-programmed pattern in the course of random behavior.
- Bertelsmann Foundation