Writing Robots

Artificial intelligence can do a lot of things. From simple chat bots to linking complex knowledge, AI is used in many different ways. One field in which artificial intelligence already plays an important role today is the automated creation of texts. Does this mean that there soon will be no more need for media professionals and communication experts?

Writing good articles and exciting online content takes time. In the digital age, artificial intelligence can help. AI technologies are able to analyse large amounts of data, understand connections and filter out the central points of a story. Artificial intelligence then uses this information to create texts and graphics. The texts are on such a high level that neither a person nor a machine can distinguish a text created by a robot from one written by a human hand.

Human or Machine?

We were able to make this experience ourselves on the event of the Berner PR Gesellschaft (BPRG) on the subject of robot journalism. Saim Alkan, CEO of AX Semantics, gave us insight into a world in which artificial intelligence produces texts. Right at the beginning of the presentation we played the game "bot or not" – and failed miserably at keeping man and machine apart. If even we as communication professionals do not recognize the difference, do we really still need humans on the keyboard?

If you take a closer look, however, it quickly gets clear that AI texts do not get by completely without the human factor: the software first has to be fed with the information it is supposed to process. In addition, the output also needs to be checked again by a human being, because artificial intelligence, for example, cannot immediately take cultural differences and personal nuances into account.

It's all about the Scope

That is why it makes sense to use artificial intelligence in journalism and communication when there is a large amount of data and many uniform texts are needed. AX Semantics is used, for example, for reporting on particulate matter values, weather and animal populations in local media – hardly topics that journalists are keen to cover.

Texts written by robots also get interesting for companies as soon as large amounts of data need to be processed. This is the case, for example, with online platforms that require text descriptions for many different products and services. Information about the model, colour, style and size of a piece of clothing, for example, is sufficient for a software to produce a product description that a person would not have been able to write in that time.

Personal Notes

But scope alone is not enough either. Saim Alkan shows, for example, that it is not possible yet to write entire books using artificial intelligence. However, he can imagine writing a book concept with software and then having it written by an experienced author. And finally, we all still like to read texts – no matter if journalism or literature – by well-known personalities and not by software. If we know the person behind a story, we can better put it in context, for example by assessing its truthfulness or its importance.

Wherever similar texts in large quantities based on a lot of data are needed, AI can help. However, good journalism and professional communication work by people are still needed. Algorithms and codes can capture or contextualize a lot – but not the finer nuances of interpersonal communication. Therefore, more than ever people are needed who are confident in technology and communication and have the appropriate specialist knowledge.

This is exactly where our strategic communication consulting applies. We advise at the interface between business, tech and communication and can thus offer our customers efficient and integrated solutions.