Google unveiled its take on the AI chatbot — Google Bard — last February. Since its release, Bard has been powered by two different large language models (LLMs) with the promise of making it a more formidable rival to ChatGPT. Now Bard gets its most significant LLM upgrade yet — Gemini.
On Wednesday, Google released Gemini, the company’s most capable and advanced large language model to date, first announced in May at Google I/O. Paired with the release of the LLM was the announcement that– starting today — users will be able to experience a specially tuned version of Gemini Pro for English in Google Bard.
Also: The best AI chatbots of 2023: ChatGPT and alternatives
Gemini Pro refers to the size of Gemini, which was released in three different sizes: Gemini Ultra, Gemini Pro, and Gemini Nano. Each size makes Gemini suitable for different tasks.
For example, Gemini Nano is the most efficient for on-device tasks; Gemini Pro is the best model for scaling across a wide range of tasks; and Gemini Ultra is the most advanced, making it the most suitable version for complex tasks, according to Google.
With the infusion of Gemini Pro, Google Bard is claimed to have more advanced, reasoning, planning, and understanding capabilities.
To show just how efficient Gemini Pro is, Google tested it against a number of industry-standard benchmarks. The results showed that in six of the eight benchmarks, Gemini Pro outperformed GPT-3.5, OpenAI’s LLM that powers the accessible version of ChatGPT.
Compared to leading alternatives, Google goes on to claim, Bard with Gemini Pro was the preferred free chatbot in blind evaluations with third-party testers.
Users can test Bard with Gemini Pro for text-based prompts for themselves — starting today and for free — simply by visiting the chatbot like they regularly would.
For now, Bard with Gemini Pro is only available in English and in 170 countries and territories, but Google plans to expand to more languages and locations soon.
Also: Microsoft Seeing AI app lands on Android to help blind and visually impaired users
The Bard upgrades don’t stop there. Early next year, Google plans on introducing Bard Advanced, which will be powered by the most advanced version of Gemini — Gemini Ultra — which is capable of quickly understanding and acting on multimodal inputs such as text, images, audio, video, and code, according to Google.
Before launching Bard Advanced, Google said, the company will first complete extensive safety checks and launch a trusted tester program.
Before Gemini, when Bard first launched, it was supported by a lightweight model version of Google’s Language Model for Dialogue Applications (LaMDA). Then, in May, Bard became powered by PaLM 2, a more advanced version of PaLM meant to significantly improve the chatbot experience; however, the improvements seemed minimal.
