Google on Thursday revealed that its AI chatbot Google Bard will now be called Gemini.
The tech giant also launched Gemini Advanced, a new AI assistant that provides users access to Ultra 1.0, the largest of its Gemini 1.0 foundation models.
Google introduced Gemini in December as a model with multimodal capabilities that can train on and generate text, images, audio, and video.
With the rebranding of Bard as Gemini and the release of Gemini Advanced, Google appears to be moving quickly to stay abreast of OpenAI and its exclusive cloud partner, Microsoft.
Gemini Advanced can be a personal tutor, do advanced coding and help creators go from idea to creation by generating fresh content, according to Google.
It is generally available now in 150 countries and territories in English.
Gemini Advanced is available as part of the Google One AI Premium Plan for $19.99 after an initial free two-month cost.
It is also available as an app on Android and in the Google app on iOS.
Google also plans to bring Gemini to more products by switching its generative AI tool Duet AI to Gemini for Workspace. Consumers with the Google One AI Premium plan can soon use Gemini in Gmail, Docs, Sheets, Slides and Meet, according to the vendor.
Google Cloud customers will also notice a rebranding as Duet AI becomes Gemini in the coming weeks, Google said.
Google's rebranding of Bard and its Gemini Advanced launch is a move to catch up in the generative AI race against archrivals Microsoft and OpenAI.
Last year saw Microsoft make aggressive competitive headway by infusing OpenAI's GPT technology into its applications, mainly in the form of Copilots.
Meanwhile, Google, with its deep roots in AI and specialized AI unit DeepMind, tried to keep up as it rolled out Bard and its Search Generative Experience platform.
An answer to Microsoft
Gemini Advanced looks like a direct challenge to Microsoft Copilot, said Paul Baier, CEO of analyst firm GAI Insights. But the offering takes the approach of focusing on the user data within Gmail, Docs or Slides as opposed to a copilot that is tethered to an application.
"They are rolling more advanced models out for a data-centric copilot view, which is very different from the Microsoft app-centric view," Baier said.
Paul BaierCEO, GAI Insights
While Microsoft's Copilot flagship is Microsoft 365 Copilot, the vendor started with Copilots for Excel, Word and PowerPoint, which caused some confusion among enterprise customers, he said.
"All of Microsoft products started as an on-premises-based software product and then moved to the cloud," Baier continued. Meanwhile, Google started with a cloud-based architecture, which will help it have a more integrated experience for generative AI tools, he added.
"Google is saying, 'Hey, we have all your data. We are going to do a copilot across all your emails, all your slide presentations, all your Google docs,'" Baier said. "They're going to be data-centric, and they're going to be more powerful over time for individual productivity."
For enterprises that use Google Cloud or Workspace, the addition of Gemini will enable easy access to data from spreadsheets, email and Word documents, he continued.
Despite the apparent similarities, Microsoft Copilot and Gemini Advanced are different, Futurum Group analyst Mark Beccue said.
"Where Copilot is different is [Microsoft] is incorporating it into other things that they do that Google necessarily doesn't do," Beccue said. An example is Windows OS.
The GenAI competition
As for the generative AI competition between Google and Microsoft, Google's launch of Gemini Advanced shows that it's not far behind, Beccue said. Microsoft just made Copilot 365 generally available a few months ago, though it previewed it months earlier.
"Microsoft had a big jump," he said.
Moreover, with numerous generative AI products that vendors launched in 2023, cloud giants such as Google, Microsoft and AWS can be expected to start rebranding some of them in the coming months, Gartner analyst Chirag Dekate said.
"What was inadvertently happening last year was as the technology innovators were innovating the technology in their urgency to get the technology out the door quickly, what they were also doing was inadvertently showcasing the complexity of the technology stack," he said. "What you're now seeing is a simplification of the branding."
Google and multimodality
Google is showcasing its native generative AI strategy with Gemini's multimodal capabilities, Dekate added.
"Google's peers will have to figure out how to compete with and how to measure up to mandatory multimodal at a time when many of them do not have the same engines like DeepMind or Google research to drive creation of their own native multimodal stacks," Dekate continued.
This will likely lead to a "Frankensteinian" version of multimodality, he added, referring to the classic literary monster that destroys its inventor.
"Building a native multimodal model does not happen overnight," he said. Google's competitors will probably layer unimodal models and create simulated multimodality to try to keep up. That is likely to be costly and limited in terms of performance, he added.
For enterprises, the challenge will come in using Gemini to create applications that are beyond just large language model chatbots and generative AI-defined summarization and text-based apps, he continued.
Google will likely continue to develop and incorporate Gemini into its stack leading up to Google Next, the tech giant's big user conference in April.
Esther Ajao is a TechTarget Editorial news writer covering artificial intelligence software and systems.