The technology giant Google has announced a new version of the well-known and still most used search engine. The results of a search are displayed in the form of AI-generated answers. As is so often the case, the USA home base wins. Other parts of the world will follow suit later this year.

A day after OpenAI announced the latest GPT model, Alphabet is now also releasing (big) AI news. According to the Belga news agency and other media, Google’s parent company has been testing AI-generated answers for some time. However, users had to register in advance to test the experimental version. During the annual I/O developer conference, CEO Sundar Pichai noted that the new version leads to “increased searches and satisfaction.”

Narrative answers

Anyone who enters a search query will be presented with more narrative answers in the new version. This means that the user would have to click through different links less often in search of the most appropriate answer. The AI-generated answers appear at the top of the results page with suggestions for additional questions below. Further down the page, as it is now, there are links to websites based on the search criteria.

For the AI-generated answers, Google uses the recently proposed AI model Gemini. The new feature has been tested in several countries, including the United Kingdom. But Google is initially introducing the latest search engine in the USA. According to Liz Reid, head of Google search, the AI ​​overviews will be available to more than a billion people by the end of the year.

AI developments

During the I/O developer conference, Google took a look at some “works in progress.” For example, the company is testing a new AI model for text-to-video conversion. Users can create computer-generated images via text commands. Google appears to be competing with OpenAI’s Sora. Veo has similar features and is expected to launch later this year.

Further development is taking place as part of the Astra project. This is said to be an AI assistant that can use a smartphone camera to identify locations, read code and form alliterative sentences. According to Sundar Pichai, the “pretty magical” tool is expected to debut in the second half of the year. Pichai also said that a faster version of Gemini (A.5 Flash) will come with an updated (and less biased) image generator model.