Google on Thursday announced the launch of a new version of its generative artificial intelligence (AI) smartphone application Bard, renamed Gemini, and its most advanced interface to date, Ultra 1.0.
This new application, which will be downloadable on Android and usable from the Google application on iOS (Apple),
“is an important first step in the design of a real AI assistant
,” the group said in a press release.
Google cites the example of a motorist taking a photo of his flat tire and asking Gemini for instructions on how to change it, or of a user generating, thanks to the application, a caption for a photo he has just taken.
To discover
PODCAST - Listen to the latest episode of our Tech Questions series
Several of the features of Google's voice assistant will be available through the Gemini app, including making calls or controlling connected devices.
Gemini is available starting Thursday in the United States and will be usable starting next week
“in more countries
,” in English, Japanese and Korean.
A Google spokesperson indicated that the application was currently not downloadable in Europe.
The member states of the European Union have just agreed on new legislation to regulate AI, unprecedented at the global level.
Read alsoArtificial intelligence: Google launches Gemini, its response to GPT-4 from OpenAI
The tool marketed in 150 countries
At the same time, the Alphabet subsidiary unveiled Thursday Gemini Advanced, a paid offer which allows access to Ultra 1.0, a more advanced version of its language model (LLM), the program which allows a user to generate content on request in everyday language.
It allows, among other things, to code, always on request in ordinary language.
Gemini Advanced is sold in 150 countries, with a subscription of $19.99 per month in the United States.
It will integrate the Duet AI functionality, which already made it possible to use generative AI to write or rewrite emails, documents, transparencies or spreadsheets.
By making the use of generative AI possible for its utility software, such as Gmail, Google Docs or Google Meet, the Mountain View (California) group is positioning itself against the offer of Microsoft, its major AI competitor. .
With its Copilot interface, Microsoft already makes it possible to use generative AI as an assistant in the use of software such as Outlook (emails), Word (word processor) or Teams (communication platform).
Microsoft presented a new, more ergonomic version of Copilot on Wednesday.
The giant from Redmond (Washington State) also offered an advertising spot, broadcast on Sunday during the Super Bowl, the final of the professional American football league NFL, which traditionally brings together more than 100 million viewers.