On December 13, the German media group Springer signed a landmark contract with OpenAI, the American artificial intelligence giant. With this partnership, Sam Altman’s company intends to enable ChatGPT, its conversational agent boosted by machine learning, to respond to current topics. This approach testifies to the rapid transformation of the media ecosystem with the rise of this technology.
On paper, this partnership will address one of ChatGPT‘s most significant shortcomings. Until very recently, it was only trained on data prior to 2021. This deadline was recently extended to spring 2023, but the problem remains the same: the algorithm is not able to react to fresh news. You always have to wait for its designers to collect data, check its relevance as best they can, and then integrate it into the public version of the model when an update is made.
But from now on, the chatbot will be much better able to answer these topical questions. It will be able to rely directly on the publications of the group’s various media. For the sake of transparency, the chatbot will systematically include a link to the source used.
This is an essential precaution to allow the public to take into account the specificities of each medium. Indeed, the Axel Springer Group is a patchwork of publications with very varied political affiliations and editorial lines. These include Die Welt, one of Germany’s largest newspapers that is generally considered a reliable source, but also Bild, a sulphurous tabloid that has an extremely bad reputation on the other side of the Rhine. More recently, Axel Springer has also acquired Politico and Business Insider.
In return, Germany’s largest media group will receive a huge windfall. According to a source close to the matter quoted by the Financial Times, he will be paid tens of millions of dollars a year, in addition to a lump sum for the exploitation of all the articles already published — a very large financial windfall that could completely change the way journalists work.
The Axel Springer Group does not hide its intentions in this regard. It is one of the structures that has already begun to experiment with artificial intelligence in the context of journalistic activity. “We want to explore the opportunities of AI-powered journalism to take the quality, social relevance, and business model of journalism to the next level,” the group’s CEO said in the same statement.
This change in philosophy has already had very concrete consequences within the group. As Le Temps reminds us, Axel Springer has already started to reduce its workforce with the aim of automating more and more tasks thanks to AI, such as proofreading or layout.
This isn’t the first time OpenAI has tried to forge ties with the media world. Last July, the company already reached an agreement with the Associated Press for the use of its content. According to L’Usine Digitale, this contract did not give the company the right to produce content based on this material. At this level, the agreement between Axel Springer and Sam Altman’s team is a first – but it is probably only a matter of time before it is emulated.
For example, this fall, the parent company of the famous Wall Street Journal revealed that it had also begun negotiating the use of its products by AI companies, but did not reveal whether it was OpenAI or a competitor. For CEO Robert Thomson, this is a pragmatic measure to prevent journalistic content from being shamelessly plundered — and without compensation — by AIs before the situation escalates. “The intellectual property of the media is under threat and we should vigorously demand compensation. “
But not everyone is on board with this way of doing things. Other media giants, such as the New York Times, Reuters and Radio France, are beginning to erect barriers to defend themselves against the devouring appetite for language models for content. Also according to L’Usine Digitale, they have updated their legal notices so that their articles “are not used in the training of artificial intelligence models”.
Institutions, for their part, are also preparing for this major paradigm shift. Earlier this month, the Council, the Commission, and the European Parliament finally agreed on the content of the so-called AI Act, a very important text that aims to regulate the use of artificial intelligence.
But everyone, from the media, to the regulators and the public, still lacks perspective on the impact of this technological revolution that is already starting to disrupt the way we consume information. It will therefore be necessary to pay particular attention to these issues in the future, and in particular in the coming years during which humanity will have to negotiate an extremely technical shift.
"Because of the Google update, I, like many other blogs, lost a lot of traffic."
Join the Newsletter
Please, subscribe to get our latest content by email.
