If you know the background (I'll summarize it in a super concise way for the very few still in the dark), Google has accelerated its projects for the diffusion of the artificial intelligence it is developing. He did it, regardless of what experts and wannabes say, for only two reasons: first, counter the rapid rise of OpenAI (and Microsoft) through the launch of Chat GPT and its integration into the Bing search engine. Second, read the first one. During its annual I/O launch event in San Francisco, the search engine giant presented its vision of AI-integrated search to the world. A vision that, it seems, could undermine the digital journalism industry at its roots.
The big shot takes the field
Google's new search interface, powered by AI and called “Search Generative Experience” (SGE), introduces a feature called “AI Snapshot”. Basically, it's a huge summary displayed at the top of the search results page. Ask, for example, “Why is the glazed donut still so popular?” (take note: never write articles at lunchtime) and before you get to the usual blue links, Google will provide you with a summary generated by a large language model (LLM, to be precise).
What is at stake for journalism?
If this innovation seems harmless to you, know that often all users are looking for is a simple summary or snippet of information. I'm not saying it, they say it several studies on the field. However, with Google which hosts about 91% of all search traffic, (source: SimilarWeb in April), there is a risk that this you become a colossus synonymous with the internet. The Internet is a market, and Google could once again become not only the biggest regulator, but the deus ex machina.
What does journalism risk? A lot. If Google starts shredding original work to provide its users with a distilled version, without ever linking them to the source, how will publishers monetize their work? It's currently unclear whether Google plans to compensate publishers for the content in any way summarized and revised by its artificial intelligence.
What can publishers do?
To address this issue, publishers may need to adopt more sophisticated SEO strategies, diversify their traffic sources, or work with Google to find an equitable solution. The world of publishing is no stranger to similar challenges: In the past, tensions between Google and the publishing industry have led to new policies and practices. I think it's necessary: the second phase of such a destructive process would involve no one running a news site anymore, apart from a few interested groups, since they will never be able to maintain it financially. And at that point Google wouldn't do anything with a model that finds any information... in the absence of information.
Meanwhile, in a recent statement from a spokesperson Google said the company is introducing this new experience as an experiment to “help us iterate and improve as we incorporate feedback from users and other stakeholders.”
What about publishers' fees?
“We have no plans to share on this, but we will continue to work with the broader ecosystem.” If you are capable, interpret this answer for me. For me it's a non-answer. The future of the industry of journalism in an increasingly AI-dominated world is uncertain. But one thing is clear: fair and transparent access to information is a fundamental right. If Google wants to remain true to its goal of maximizing access to information, it must find a way to balance technological innovation with respect for the work of publishers and the right of readers to access high-quality information.
If not, OpenAI will be the least worst of its problems.
