"The industry must not miss the AI train".

Here is the interview I recently gave to the Swiss magazine "Schweizer Journalist:in". The news has already caught up with us with the go-live of the new version of Bing...?

Schweizer Journalist:in: When did you first hear about ChatGPT, what was your firstimpression?

In late fall. I have rarely seen a new development being hyped so fast, so widely in industry publications and, a little later, in general publications. Of course, I wanted to familiarize myself with itright away, so I started chatting to ChatGPT.

Is there this one example,this one experience with ChatGPT, where you knew: This is really a breakthrough now,this will change a lot?

I experimented a lot, so I had many "conversations" with ChatGPT. Once I asked how to build a successful media business - the answer was amazingly precise. But the penny really dropped when our CTO explained to me that ChatGPT solves highly complex IT questions within a short time and provides the code forthe answerright away. That's when I understood that my trivial question-answer games (although often a lot of nonsense comes as an answer)impact of this AI application does not even begin to do it’s potential justice. The application answers banal concerns as quickly as the most complex academic questions.

With which digital innovation ofthe past years or decades can this current development step in AI be compared (in terms of potential impact on society, business and media in particular)?

I would use the relationship between writing and the invention of printing as a comparison: Writing and books existed before 1440, but it was printing that brought knowledge to the general public. Analogous: Al has been around for a long time. Thanks to ChatGPT and similar user-friendly applications, this intelligence is becoming accessible to the general public. But while it took centuries to bring books into every home, it's happening fastertoday: ChatGPT reached 57 million users after one month. Technologically - and with a special focus on media - I would assess the relevance of AI as follows compared with the introduction of broadly effective search engines in the 90s. All the knowledge of world history was suddenly accessible at the push of a button.

On your internal list of innovation projects at Ringier: Where is AI atthe moment? Are there already project groups? Do you also network internationally?

ChatGPT is currently in full focus. Other models will probably also become important for media companies - probably more important than ChatGPT. On the question: Use cases are currently being tested in all our markets so that we can find out as quickly as possible where the use of ChatGPT or other Al applications makes sense, and where not. At the moment, this topic is certainly in pole position. It's important to note that it's not just the tech or product teams that are dealing with this. It is also the editors who are currently dealing intensively with Al. What is very important to us is that we are in the process of developing guidelines for dealing with AI, especially ChatGPT. The problem with ChatGPT is, after all, that he or she orit sometimes very self-consciously generates false information. Al code can have a huge impact, they currently often move in a space with little or no regulation! Forthis very reason, we as media companies are all the more called upon to use these instruments in an absolutely responsible manner and with a great deal of circumspection and foresight.

Please dare to take a look into the future: In which areas will Al have a particularly strong impact on workflows in the future, i.e. on the everyday life of journalists (in general in the media industry, but also specifically related to Ringier)?

Creating texts from tables, correcting texts, possibly improving them, storing meta information, setting alternatives to titles. One can imagine that infoboxes will be created with the help of AI, but also that artificial intelligence knows exactly what interests me as a user and primarily displays such content. By the way, this has been happening at our Swiss media company Blick for several years - even if it has not (yet) been broken down to the individual. But it is very likely that I see a different Blick homepage than you do - simply because I am not that interested in sports, for example.

Are there areas or even media brands that will perhaps be less affected by changes?

I think "affected" is the wrong word, as it suggests something negative. It's an opportunity that the industry needs to seize. The media industry has slept through many technological trends in the past - and paid the price. For example, the media industry failed to adopt login as a standard when all other industries did. The industry should not miss the AI train - regardless of the respective state of digitization, whether consumer or reach media.

There is a scenario in which an AI engine is "put over" the entire content database of a publishing house. AI then "learns", for example, how things are written in Blick, how things are written in Sunday edition "SonntagsBlick," what the style of a story looks like in the maganzine "Bilanz" - and can then really create individual contentin the style of that medium. Do you see that as a potential scenario?

In my opinion, this would already be possible. One of my first experiences with ChatGPT was: I asked it to write an essay about flowers in the style of Goethe (sometimes the Germanist in me still shimmers through), then the same in the style of Kleist. It was incredible! Why shouldn't AI be able to write stories in the style of Blick ...? In addition, issues like privacy, data security, and the ability to sit on top of models like ChatGPT or GPT and adjust on its own will make a difference in depth. But: Even if the style is perfectly replicable: That alone does not make it a journalistic work. The human intelligence of journalists will coexist with artificial intelligence. Wisely combined, it will make us better as a media company.

When you test systems like ChatGPT for a while, you’ll notice that you very quickly find it very "normal" to have dialogs with a machine, some of which are extremely complex. Will we really get used to the new technology so quickly? Willthe dialog with the machine become common?

If I want music at home orthe latest news, I ask my smart speakerforit. We've been communicating with machines for a long time. It will become the most normal thing in the world. There will be endless ethical questions that we as a society will have to answer. For example, when Siri plays my favorite song, I have the reflex to say thank you. Children learn to say thank you when they get something. Will children say thank you to the machine? If not: will we also stop saying thank you to people and become more and more rude in this respect? We are confronted with completely new questions here.

One forecast is that the flood of content, of whatever kind, will increase exponentially in the coming years. What does that mean for your media company, your brands?

A medium has the chance to become even more the place of orientation. The times when a media company simply had to inform and "entertain" are over. A medium is increasingly becoming a trusted companion for all situations in life. Incidentally, the future role of search engines such as Google and Bing will also play an important role: How much attention will be paid to quality? What contribution do big tech companies make to ensure that a user doesn't get lost in fake news?

Will the lower hurdles in content creation possibly bring new players to the market?

It is very likely that new dynamics will emerge in the market. And as with everything: There will be socially friendly developments (e.g., in the fight against hate speech) as well as the opposite: It is also possible to steer a machine so that manipulation becomes even more perfidious. But even if the entry hurdle forfuture digital publishers becomes lower: Building a media product, gaining and retaining reach and subscribers on a large scale, establishing the brand, etc. will not be a walk in the park, even with AI. Because - and this is a good thing - journalism will always be more than what a machine can produce. Journalism is weighing, classifying, using your experience, instinct and also creating something completely new. Journalism discusses social standards again and again. Journalism is also creativity and the expression of human creative power. Journalism is a people's business: by people for people.

Some critics think it's all overhype, citing thatthe system often produces hair-raising misinformation and can therefore never be relied upon. Do you share this opinion?

Innovations often follow a similar course: At the beginning, the hype boils up massively, then leads to a phase of disillusionment, and finally - after enough experience has been gained - the hype dies down to a realistic level. Currently, we are moving from the disillusionment phase to the realistic level. Uses and commercial offerings will play a big role in whether ChatGPT - but especially new versions of GPT (like those used by Bing) - live up to the hype. What I am particularly interested in: At the moment, there is also still a lack of transparency regarding the sources from which ChatGPT draws. Microsoft has already announced that the Bing search engine will be able to make sources transparent. For us as media companies, this is an absolute must.

Assuming the forecasts are correct and AI searches will replace Google searches, atleastin part: What does that actually mean for the planned ancillary copyright?

In the case of ChatGPT's results, it is completely unclear where the information that was processed came from. Isn't the planned ancillary copyright already lost before it could even come into force? I asked ChatGPT the other day, "Are many of your sources media content?" - and ChatGPT replied: Yes, many of my sources are media content." Of course - an anecdote to smile about. But not only. ChatGPT and similar offers will have to solve the problem of non-transparent sources, otherwise society will never seriously accept these offers. As soon as the sources are noted, the ancillary copyright is as current and necessary as never before. The extent to which Google search is replaced, or whether and how Google adapts the search and the business model of search engines in general, also has a clearimpact here. Microsoft's CEO Satya Nadella has so far presented an optimistic scenario in which media houses are to have an importantrole and also a functioning business model. The release of the new Bing search could provide us with a first insight here.

When I talk to publishing managers aboutthe topic,there is a palpable enthusiasm. When I talk about it with editors,there's automatically a certain amount of concern. Can you understand that?

I experience a lot of interest and curiosity in our newsrooms. That may also have something to do with the fact that the Ringier media titles are very much digital titles that launched the transformation early. Of course, I understand the fear of being replaced by a machine. But the more you embrace technology as a journalist, the more unfounded the fear becomes.

Are certain skills perhaps no longer as relevantin the future as they might have been just a shorttime ago? Does a news journalist still need to be able to write a quick newsflash when the automatic machine can do this perfectly well anyway?

The machine will take overthe busywork - that's nothing new. Which is a good thing: It leaves more time forresearch and in-depth study. In the end, it makes media better. Media companies with a strong digital focus now have a lot ofroutine in constantly adapting skills to innovations.

If you take a very sober view ofthe current stage of development, Al will probably open up enormous potential for savings on many levels in the future of media companies. What has to happen for this notto become the only impact of AI in media companies?

We see Al less as a way to save money than as a tool to increase quality and promote diversity. To put it bluntly, Al is not the biggest threat to journalism but the lack of consistency and a too slow adaptation of the digital transformation of media houses, often combined with the hope that the golden days of print will magically return. Digital media are delicate clockworks. You have to really be willing to understand them in detail, to maintain them, to adjust loose screws immediately or replace them if necessary. Media companies must position themselves robustly and drive growth consistently in orderto be profitable in the long term. If media companies now simply cut costs and replace natural intelligence with artificial intelligence, this will certainly not bring them success.

Is there anything journalists should do in the future to avoid being among the losers ofthe upheaval?

See technology as a tool, not an enemy. And pursue their vocation with passion.

When do you expectthe first AI systems to be in use at your company? And in whatform?

Al has long been in use at Ringier. One example is EqualVoice, an algorithm thatrecognizes and tracks women and men and helps us increase the presence of women in the media. Newer applications like ChatGPT are in the testing phase. As I said, we want to use these technologies 100 percent responsibly.

Is there anyone in media houses who doesn't have to worry about AI because he or she simply won't be affected?

Our CEO's dog.