For over a decade, talking about searching on the Internet has meant talking about searching on Google... at least until Generation Z arrived. Increasingly, young people (and not so young) are using social networks like TikTok or Instagram as search engines, leading Google to revolutionize its search engine to reconquer this demographic with a recipe that is starting to sound familiar: artificial intelligence and short videos.
The strategy is working, and it is the young people aged 18 to 24 who are using the new AI search functions the most, which are starting to arrive in Europe little by little. In fact, on some Spanish phones, the so-called "AI summaries" are beginning to appear, a feature that generates a search view combining images, articles, and even videos, allowing for better resolution of complex searches.
"We know that people increasingly demand to hear the perspective of others, especially the younger ones (...) That's why there is a mix of different types of videos, short videos, texts, and content from more generalist, independent, or individual sources," explains Google's Search Vice President, Elizabeth Reid, in a media meeting at the tech company's headquarters where EL MUNDO was the only Spanish media present.
The head of the search group clarified that this will be applied differently in each case since "it is not the same for political news as it is for sports," but she defended that this was the company's vision: "We have to ensure that high-quality information is obtained from trusted sources. And increasingly, this means combining traditional sources and new influencers," she stated.
As one can easily imagine, transforming the world's largest information search engine is not easy, but the complexity increases when considering the variety of languages and records that exist globally in an environment where users are demanding more. For example, Japan stands out for the need for a formal record, while in India, it is the opposite, as explained by the executive.
If in the first year of AI, anything surprising was indulged, now the expected standards are approaching perfection, as the company knows well, having withdrawn its first AI search experiment due to criticism.
But in less than a year, they have managed to turn the tables, and the new functions are gaining momentum. "AI summaries have been extremely well received. They are proving very useful for over a billion users, and what we find is that once the user starts seeing them, they end up doing many more searches," explains Google's Search Quality Vice President, Srinivasan Venkatachary.
This reception has led the company to take a step further and create the 'AI mode,' a feature currently only available in a trial version in the United States, but it turns the experience into a hybrid between search and using an assistant. In fact, Google's Product Vice President, Robby Stein, complements his colleague with one of the reasons that led to the creation of this mode: many users have started to include AI in their searches to try to always get the summary.
Now, the other key part of the strategy is another buzzword in the tech world: multimodality, or in other words, integrating everything with voice and camera, an area where Google's new big step is connecting Lens with its AI models.
The app that allows focusing on any object and searching on Google already generates 20 billion searches per month, most of them also from young people, and now the group aims to boost this function both for the search engine and for Gemini itself, the chatbot that now 'sees' and interacts in real-time, a novelty that will have to wait a bit longer to be seen in Spain due to the need to train the model with large amounts of Spanish data.
As in any business, another essential point is the monetization of this entire world. AI has infiltrated the management of advertising campaigns on the platform, and in fact, 80% of its clients already use some of the functions of this technology, ranging from targeting specific audiences to designing creatives, improving their returns and efficiency.
However, another question arises related to the financial aspect. How much does the cost of operating a Google that relies mostly on AI for searches increase? "We have a strong infrastructure and the ability to choose the right model for each question. This makes it very, very efficient," points out the company's Search Business Vice President, Sashi Takur. The next stop for more news will come in May with the new edition of Google I/O.