At an event in Paris, Google unveiled a range of new features for “Search”, “Lens” and “Maps”.
Prabhakar Raghavan, senior vice president of the company, продемонстрировал how, in the future, the tech giant will use generative AI to summarise information from the internet. One such application is NORA — a question-answering feature for questions that have multiple viewpoints.
In a demo, the company demonstrated ‘which constellations are best to search for while observing the stars?’. The AI-generated answer highlighted several key options and ways to identify them.
“The new generative AI features will help us organise complex information and diverse viewpoints directly in the search results. This will enable you to quickly understand the overall picture and then explore the various aspects,” Raghuvan said.
He did not specify when the feature would be available to users.
Raghavan also showed how generative AI can create 3D models of objects from several photographs taken from different angles.
Earlier, announced “Multisearch” in Google Lens became available worldwide, and “Multisearch near me” started in the US. Later the feature will reach other countries.
Google Lens also learned to keep the background in photos during live translation. Previously, the algorithms blurred the image area under the captions and overlaid the translation on top.
The company announced the return of object-search from the smartphone screen using Lens and “Assistant” in any app, photo or video.
For Google Maps, developers представили Immersive view — a photorealistic 3D rendering of cities. In creating the feature, engineers combined “thousands of Street View” images to create accurate replicas of various locations.
In addition, users will be able to select 3D models of individual buildings, search for entrances and exits on them, and see how the object has changed over time.
Immersive View is available in London, Los Angeles, New York, San Francisco and Tokyo.
The AR navigation feature Live View will gain support for new cities. According to a Google spokesperson, the company will add information for thousands of establishments in Barcelona, Berlin, London, Madrid, Paris, Sydney and others.
According to company representatives, the feature will help users navigate complex venues such as airports, train stations and shopping malls.
During the presentation, Raghavan also mentioned the conversational AI Bard, built on a lighter version of the LaMDA language model. However, he did not provide new details.
Earlier, представила “new Bing” with ChatGPT.
A day earlier, Google announced its own chat-bot for Bard search.