Artificial Intelligence

What exactly do we understand by AI and what is the difference to machine learning? We enlighten you on the concepts.

Artificial intelligence, abbreviated AI, is the generic term for all technology dealing with the intelligence previously reserved only for humans.

A subfield of artificial intelligence is machine learning. This form describes the mathematical possiblities for a system to independently generate knowledge from experience and data collections.

Another subfield is deep learning. This research area deals more profoundly with machine learning on basis of neural networks. Big Data with its large amount of data plays an important role from which the systems learn independently. In some cases patterns are recognized which would be never accessible by humans.

To sum up, computers with artificial intelligence solve tasks and problems for which a human would have to use his human mind and intelligence.

Artificial intelligence imitates human abilities in order to make faster decisions based on data. These abilities include:

• Recognizing
  
• Learning
  
• Planning
  
• Logical thinking
  
• Problem solving
  

In comparison to previously static programs, machines handle data in a completely different way. Connections resp. patterns are established between individual data sets so that they can be interpreted and reused.

For algorithms this works on a visual basis as well as with text and sound.

When using Augmented Reality applications artificial intelligence determines for example the environment and can thus display objects in reality.

With these abilities machines will be increasingly able to make good and fast decisions in the future.

The following developments are among the best known examples. 

Since the Christmas sales 2017 Smart Home and voice assistants have made their way into the living rooms of the Germans. These devices listen to people and react on voice commands.

Siri, for example, can be asked for the exchange rate, Google Home provides information about traffic to work and with Alexa products can be bought at Amazon.

All this works by the user talking to the device and then saying a voice command. 

“Hey Siri, how much are 380 Dollars in Euros?” 

For quite some time it is already possible to chat with bots and programs. In such conversations these can then react to the statement depending on the written keywords.

Therefore those chatbots are often used as sources of information or to solve simple problems in the support.

The user often does not realise that he does not write back and forth with a real person.

With improved AI such chatbots learn new references between words, which increases quality and application areas.

A very nice use case for Artificial Intelligence is the tool DeepL. With DeepL texts can be translated into eight different languages. The tool learns with artificial intelligence and improves with every application.

The tool is fed with data from numerous books which have been translated into different languages. This provides a good basis for making translations more efficient and which are better adapted to language use.

Facial recognition is a feature that has for example been integrated into the iPhone by Apple. In principle, artificial intelligence is applied here for security.

Thus, the cellphone can only be unlocked visually when the owner of the cellphone is looking at it. The camera activates automatically and scans the face on numerous points.

Another example for face recognition is used by Facebook. When a photo is uploaded, it is scanned by faces and therefore making it easier to tag people.