A chatbot can answer questions, help a user book a trip, or simply provide entertainment. There was a first version of the text-based dialogue robot as early as 1966 – it was called “Eliza” and, as a virtual psychotherapist, answered questions from its users. Since then, the technology has continued to develop. Chatbots will support people in many areas of life. You can find out what a chatbot can do, how to use it and how safe it is.

• Messenger bots offer numerous possibilities in one application.
• Well-known chatbots are, for example, Siri and Amazon Alexa.
• Modern communication robots work on the basis of artificial intelligence (AI) and machine learning.

What is a chatbot?

Chatbots are text-based dialog systems with which human users can communicate in natural language. There are also voicebots such as Siri or Amazon Alexa , with which you can interact by voice and which are often included when talking about chatbots. Modern communication robots work on the basis of artificial intelligence and machine learning.

The beginnings of the chatbots

“Eliza” worked according to the keyword principle, so it recognized individual words in the users’ questions and then played out the prepared answers linked to these words. Since the bot could not understand the entire message of the user, the conversations with “Eliza” and her successors were rather bumpy.

Initially, chatbots were primarily used for entertainment – or, like the “Smarterbot” program that appeared in 2001, could find and output information for their users on their devices. In 2011, however, it became clear what progress chatbot technology had made: The “Watson” bot won the “Jeopardy” quiz show against human participants. “Watson” works on the basis of natural language processing, so it can process human language by machine.

Chatbots today

The public has now recognized the potential of chatbots – and programs such as Siri or Alexa are now quite natural to us. In 2016, the messenger bots introduced by Facebook came into play. Users can chat with them via Facebook Messenger, WhatsApp, Telegram or similar applications without having to download the robot as an extra program.

Today’s chatbots understand their users a lot better than the very first versions. They can break whole sentences down into their components, understand grammar and punctuation marks and correct typing errors. With the help of complex recognition patterns, they decipher the content or subject of a message. Various sources and databases are available for the answers.

Machine learning enables them to expand their knowledge independently: On the basis of previous experience and examples, they recognize certain patterns and transfer them to other problems, i.e. they generalize. Incidentally, the “Tay” chatbot from Microsoft proved in 2016 that this also harbors dangers, when it developed and spread racist attitudes by manipulating Twitter users within a short period of time.

The developers are constantly working on improving and developing chatbots so that they can understand human language even better in the future.

What can a chatbot be used for?

Numerous companies are already using them in customer service. A chatbot can answer simple concerns independently – for example, if someone asks about the opening times of the store or requests a document. He can also find out what problem a user has and then forward it to the person responsible. The advantage: Chatbots can be reached around the clock and immediately provide initial answers.

A chatbot can also provide its user with messages on a topic that interests them. So if you only briefly hear about a current event, you can ask the bot what it’s all about. A chatbot can also serve as a live ticker. A major strength of the programs is the information service. They can easily answer questions about the weather, nearby restaurants or event tips by using databases or the Internet.

Users can not only ask a chatbot for information, but also ask it to take specific actions. For example, they can order pizza, book a trip or reserve a seat in their favorite restaurant. In the future, messenger bots will contain entire shop systems and replace separate apps for shopping, weather, news or travel bookings. With chatbots for all topics and processes that interest users, combined in a single messenger app – people should be able to access content from the Internet much more conveniently. You no longer need to search for information or products yourself, but simply instruct a chatbot to do it, which replies in understandable human language and in real time.

Chatbots collect a large amount of data with the help of which they can get to know their users better and provide them with personalized answers to their questions. For this purpose, not only the data that the user enters in a message to the chatbot is used, but also, depending on the request, for example the location or the Facebook profile (when using the Facebook Messenger).

Accordingly, many interested parties are concerned about data protection. However, there are guidelines that chatbot providers must adhere to. It should always be clear to the user whether he is talking to a bot or a real person. It should also be able to be forwarded to a human communication partner if necessary. However, there is still no final clarity on the topic of chatbot security. In case of doubt, users can take a look at the data protection declaration of the respective provider.

Can a chatbot replace humans?

Chatbots are not as popular in Germany as they are in the USA or China, for example. One reason for this is that many of the programs communicate in English. But the Germans’ skepticism about bots is also a cause: Many do not want to talk to a machine instead of a human or do not trust its capabilities. The question arises: Can chatbots replace communication with people?

It all depends on the area. Chatbots offer valuable support in customer service or with authorities, for example when filling out forms. They succeed extremely well in providing information that users would otherwise have to search for themselves. And they will probably also compete with the previously app-based shop and booking systems. However, when it comes to complex issues that are difficult to describe, bots reach their limits. Therefore, in customer service, they often only serve to select issues that they then pass on to a human advisor.

Chatbots can take over or facilitate many simple processes or processes that have to be carried out very often in the same way. Human service staff also remain important for individual advice on complex problems.