Google Presents Gemini To Bard, Superior To GPT-3.5

Bard will be supported by Gemini language model (photo: Google)

JAKARTA - In the middle of last year, in June to be precise at the I/O 2023 developer conference, Google CEO Sundar Pichai introduced Gemini, Google's latest major language model.

Now, in its latest blog, Pichai announced that Gemini language models will be present at Bard in Bard's biggest upgrade since its first launch.

Designed for flexibility, Gemini is optimized for three different measures, namely Gemini Ultra, Pro, and Nano. Thus, this language model can run in everything from data centers to mobile devices.

Gemini, launched to Bard, will go through two stages. First, starting today, Bard will use a specially tuned version of the Gemini Pro in English for reasoning, planning, further understanding, and more.

Then, earlier next year, the search giant promised to introduce Bard Advanced, which gives you first access to Google's most advanced models and capabilities starting with Gemini Ultra.

Before introducing it to the public, we run the Gemini Pro through a number of industrial standard benchmarks. In six of the eight benchmarks, the Gemini Pro outperforms the GPT-3.5, including in the MMLU (Massive Multitask Language Understanding), one of the leading standards for measuring large AI models, "said Pichai.

In addition, Pichai admitted that specifically, Google has set the Gemini Pro in Bard to be more capable in matters such as understanding, summary, reasoning, coding, and planning.

You can try Bard with Gemini Pro today for text-based commands, with support for other modalities coming soon.

This capability will be available in English in more than 170 countries and territories initially, and will be available in more languages and places, such as Europe, in the near future.