Google announced Gemini at I/O 2023 as its next-generation foundation model. According to a report today, Google was originally going to launch Gemini next week, but that has now been delayed until January.
The Information reported how Gemini was supposed to get a big reveal next week with a series of events in California, New York, and Washington, which is aimed at politicians and policymakers.
Sundar Pichai decided to delay the launch after Google “found the AI didn’t reliably handle some non-English queries.” Global language support is top of mind, with Google ultimately aiming to match or surpass OpenAI’s GPT-4. Sources in today’s article say the company has “met that standard in some respects.”
The CEO said in November that the company is “focused on getting Gemini 1.0 out as soon as possible, make sure it’s competitive, state of the art, and we’ll build from there on.” At the moment, Google is said to still be “finalizing the primary, biggest version of Gemini.”
At I/O, Google said Gemini was “seeing impressive multimodal capabilities not seen in prior models.” Besides understanding text and images, another aim is to be “highly efficient at tool and API integrations” as the company works to make Gemini an attractive offering to third-party developers with various sizes — PaLM 2 has four, including the lightweight “Gecko” for mobile — offered. Finally, Google said Gemini is built to “enable future innovations, like memory and planning.”
The big question is how soon Gemini will be integrated into Google services, like Bard, Search, and Workspace.
FTC: We use income earning auto affiliate links. More.
Comments