Google Bard, the company’s generative AI chatbot and ChatGPT rival, is receiving an update today that the company says will significantly improve its capabilities. The company says Bard will now be powered by Gemini, Google’s newest and most advanced AI model, giving the chatbot more advanced reasoning, planning, understanding, and other capabilities.
Gemini comes in three sizes, Ultra, Pro and Nano, allowing it to run on anything from mobile devices to data centers.
Gemini’s launch to Bard will take place in two phases. Initially, Bard will be updated with a specifically tailored version of Gemini Pro. Next year, Google will introduce Bard Advanced, which will give users access to the best AI model, starting with Gemini Ultra.
The version of Bard with Gemini Pro will be available first in English in more than 170 countries and territories worldwide, with more languages ​​and countries coming soon, including the EU and the UK.
Before being released to the public, Gemini Pro went through a series of industry-standard tests, and in 6 out of 8 of those tests, Gemini outperformed GPT-3.5, Google says. That includes better performance on MMLU, or massively multitasking language understanding tasks, which is one of the key standards for measuring large AI models. He also scored better on GSM8K, which measures elementary school mathematical reasoning. However, as TechCrucnh’s Kyle Wiggers pointed out, GPT-3.5 is over a year old, making this release seem more like a catch-up than a top-performer.
The improvements will make Bard more capable in terms of understanding and summarizing content, reasoning, brainstorming, writing and planning, the company notes.
“This is the biggest quality improvement to Bard since its launch,” said Sissie Hsiao, vice president and general manager of Assistant and Bard at Google, introducing the Bard update at a press conference.
To start, Gemini Pro will power text-based prompts in Bard, Hsiao said, but will expand to multimodal support — that is, texts and images or other modalities — in the coming months.
In 2024, Bard Advanced will debut, which will be a new experience powered by Gemini’s most capable model. With Gemini Ultra, as it is called, AI can understand and act on different types of information, including text, images, audio, video and code, and has multimodal reasoning capabilities. Gemini Ultra can also understand, explain and generate high-quality code in popular programming languages, Google says, in addition to understanding audio and video content. This update seems to be the one to wait for, in that case.
The company says it will launch a reliable testing program for Bard Advanced before opening it up more widely to users early next year. Additionally, Google will subject Bard Advanced to additional security checks before its release.
The update follows a series of other improvements to Bard, since its debut just eight months ago. In recent months, the AI ​​experience has been enhanced with features such as the ability to answer questions about YouTube videos, as well as access users’ Google apps such as Gmail, Docs, Drive, and more, as well as other services. from Google such as Google Flights and hotels. You can also check your answers to help determine if the AI ​​is “hallucinating,” meaning when it provides an answer based on false information.
“Now, with Gemini, we are one step closer to offering you the best AI collaborator in the world,” Hsiao said. That at least seems more honest, since it’s an admission that Bard isn’t there yet.
Bard with Gemini is available today.
