Google announced that their Bard chatbot will now be called Gemini. The company also announced the launch of Gemini Advanced, the largest version of their Gemini language model, along with two new mobile apps for interacting with the model.
When Google first released their Gemini model in 2023, it came in three sizes: Nano, Pro, and Ultra. They also revealed that the Pro version would be the model powering Bard, replacing PaLM. However, Google held back Ultra, the largest and most powerful, from general availability, to complete safety checks and fine-tuning. In this latest announcement, Google renamed Bard to Gemini, to reflect the chatbot's "mission" to provide access to their "most capable family of models." The company also announced that the Ultra model will be released as Gemini Advanced and will be available as part of their Google One AI Premium plan. According to CEO Sundar Pichai:
The largest model Ultra 1.0 is the first to outperform human experts on MMLU (massive multitask language understanding), which uses a combination of 57 subjects — including math, physics, history, law, medicine and ethics — to test knowledge and problem-solving abilities. Today we’re taking our next step and bringing Ultra to our products and the world.
Pichai also wrote that Gemini models would soon be available in Google's products that currently feature Duet AI, a collaborative AI assistant. In Workspace, where the assistant can help users compose documents and emails, Duet AI will be renamed Gemini for Workspace. Google Cloud offers a Duet AI assistant for coding, which will also be rebranded as Gemini.
Google also released two new methods for accessing Gemini via mobile devices: one for Android and one for iOS. On Android, Gemini is available as an "overlay experience" for Google Assistant, and users can invoke it the same way; for example, by saying "Hey Google." On iOS, Gemini is available via the Google app, via an opt-in toggle.
In a discussion on Hacker News, several users wondered about the reason behind changing Bard's name. One wrote:
I'd suspect they're just trying to start over - Bard kinda got pegged as Google's dumb, not-as-good, late-to-the-party clone of ChatGPT. It got teased a lot. I'm not sure Gemini is much different (still more impressed with GPT 3.5 & 4) but I guess the name change gives them a bit of a reset.
Other Reddit users expressed disappointment with the Gemini mobile experience. One said:
Not a fan, I switched back to regular assistant. Gemini won't answer questions unless you unlock your phone, it often requires you to hit the send button after asking your question, and it takes forever to get simple answers. From your screen off asking Gemini [a question] takes about 10 seconds before receiving and answer, where assistant is almost instant. Gemini has potential for being the better android assistant, but it just isn't 'there' yet.
Several Hacker News users commented positively on Gemini's performance, calling it as good as or even better than GPT-4. Similarly in a thread on Reddit, one user reported Gemini was better than GPT-4 "specifically for copywriting, blogs, emails, etc." However, that user noted that GPT-4 was better for data analysis.