(Credit: PCMag/Michael Kan)
Along with the Gemini chatbot, Google is working on a next-generation AI assistant that can view the world around you, and provide advice and tips throughout the day.
At Google I/O, the company debuted Project Astra, which is focused on creating a "universal AI agent helpful in everyday life," according to Google DeepMind CEO Demis Hassabis.
Astra stands out from Google's Gemini chatbot via its ability to see and converse with users. In a pre-recorded demo, the company showed an employee accessing the AI assistant on a smartphone, pointing its camera at various objects, and asking Gemini to identify or provide details about them.
Astra correctly identifies what it's looking at and provides intelligent responses through voice interactions with the user. For example, Astra looks at some computer code on a monitor and explains how it works. The AI assistant also recalls where the user placed her glasses in the office after viewing her workspace. It also looks at a diagram of an IT system on a whiteboard and offers suggestions on how to improve it.
Hassabis described Astra as a “multimodal” AI assistant, meaning it can respond to various inputs, such as text, images, audio, and video, making it work more like a human person.
The company plans on rolling out parts of Project Astra later this year through the Gemini app.
Project Astra looks poised to compete with OpenAI’s new GPT-4o AI model, which can also speak and view the world through the user’s smartphone camera. But in GPT-4o’s case, the AI was also able to show degrees of human-like emotion, including laughing at human responses.