The rhythm has suddenly been turned upside down in Google's announcement schedule on artificial intelligence (AI).
Taken aback by the ChatGPT tornado, an AI tool made available to the public, the Californian giant had not necessarily planned to talk this Wednesday about its own conversational robot, Bard, during a global event organized at its Paris headquarters.
On the official agenda: innovative updates to the Maps or Lens services enriched by the magic word of the moment: AI.
But impossible to ignore the phenomenon signed OpenAI and funded by the great rival Microsoft.
“We have been working on AI for years,” tried to dodge Prabhakar Raghavan, senior vice president of Google.
“No particular event has prompted us to make its announcements now,” he assured.
The Mountain View firm had however a few minutes earlier lifted the veil a little more on this chatbot still in the test phase and which will be integrated into the famous search engine.
Before quickly engaging on its news and new features enriched by AI for its home applications.
PODCAST.
ChatGPT: the robot that is revolutionizing artificial intelligence
For the Google Maps mapping service, powerful algorithms pave the way for the Live View search function, i.e. a more precise exploration of places.
They will above all help to better situate oneself in space, especially indoors, in an airport or a shopping center.
The application will indicate, for example, the presence of the toilets or a vending machine.
The Immersive View feature has the most potential in everyday life.
Billions of images and aerial views were ingested and fused together by state-of-the-art AI – Neural Radiation Fields (NeRF) for the purists – to generate a 3D representation of a city.
You can then fly over it before visiting it and inspect the places to discover.
“You will even be able to simulate a place, like a castle or a park, with rain or at a specific time of day to see the sunshine, thanks to AI,” explains Justin Quimby, product manager of Google Maps.
The Live View feature is coming to Google Maps 🤩
It's never been easier to find your way around ➡️ thanks to our new augmented reality interface, you can easily find the establishments around you.#googlelivefromParis pic.twitter.com/bfAP82J9oT
— Google FR (@GoogleFR) February 8, 2023
Launched in 2017, image recognition by Lens takes a new step with the upcoming possibility, on Android and in the Google application on iOS, of launching a multiple search on the text and the image contained in a photo. or more or less whatever happens on your screen.
Loud announcements, but not powerful enough to overshadow the launch of ChatGPT, a sudden advance in consumer AI made possible by a deep learning model theorized… by Google in 2017.