Google's Gemini AI to offer real-time object and event recognition
What's the story
Google has given a sneak peek into its new camera-powered artificial intelligence (AI) feature, hours before the commencement of its I/O developer conference later today.
The tech giant shared a short video on X, demonstrating Gemini AI's capability to recognize objects and events in real time
The video featured what appeared to be a Pixel device capturing the keynote stage at the I/O conference.
AI interaction
AI feature also respond to voice commands in real time
In the video, a voice asks the camera, "hey, what do you think is happening here?"
The AI feature responds by saying, "it looks like people are setting up for a large event, perhaps a conference or presentation."
It also identifies the "I/O" letters as being associated with Google's developer conference and mentions "new advancements in artificial intelligence."
This interaction suggests that the new feature operates in real-time and can respond to voice commands.
Feature release
New feature resembles Google Lens but powered by LLM
The new AI feature bears a striking resemblance to Google Lens, the company's camera-powered search feature.
However, the teaser video indicates that this new feature operates in real-time and responds to voice commands, courtesy of an AI LLM model with vision understanding.
The demonstration on a Pixel device suggests that Google may initially release this advanced AI-powered feature on its Pixel 9 lineup.
Strategic timing
Google's unusual preview coincides with OpenAI's live event
Google's decision to preview this new AI feature just before its keynote is an unusual move, but likely not coincidental.
The teaser was released as OpenAI showcased similar capabilities with its new GPT-4o model during a live event.
More details about Google's innovative feature will be revealed at the I/O conference, which kicks off tonight at 10:30pm IST.