Google CEO On Why Gemini AI Is The ‘Most Capable’ GenAI Model And LLM

‘This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company,’ says CEO Sundar Pichai when unveiled Google’s new AI model Gemini.

The beginning of a new era has arrived at Google with the unveiling of Gemini, the tech giant’s new AI large language model that looks to eventually power many Google products and services, CEO Sundar Pichai said Wednesday.

“Gemini [is] our most capable and general model yet, with state-of-the-art performance across many leading benchmarks,” said Pichai in a blog post. “It was built from the ground up to be multimodal, which means it can generalize and seamlessly understand, operate across and combine different types of information including text, code, audio, image and video.”

The Mountain View, Calif.-based company plans to license Gemini to customers through Google Cloud for them to use in their own workloads and applications.

[Related: Thomas Kurian On Google Cloud’s AI Differentiators Vs. Rivals AWS, Microsoft]

Google’s new Gemini LLM will power its generative AI chatbot, Bard, to help with advanced understanding, reasoning, planning and other abilities. It will also be injected into Google’s Search Generative Experience, which answers search queries with conventional texts.

In 2024, Google said Gemini will become available in more of its products and services including Search, Chrome and Duet AI.

“This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company,” said Pichai. “I’m genuinely excited for what’s ahead, and for the opportunities Gemini will unlock for people everywhere.”

The launching of Gemini today comes approximately one-year after Microsoft-backed OpenAI launched ChatGPT in late November 2023. Versions of Google’s Gemini and OpenAI’s ChatGPT technology will likely compete in the market.

Gemini Flavors: Ultra, Pro And Nano

There are three flavors to Google’s new LLM: Nano, Pro and Ultra.

Gemini Ultra is Google’s largest LLM, which the company says can outperform human experts on massive multitask language understanding (MMLU) that leverages dozens of subjects—such as physics, math, law, and history—for problem solving and knowledge.

“Gemini Ultra’s performance exceeds current state-of-the-art results on 30 of the 32 widely-used academic benchmarks. With a score of 90 percent, Gemini Ultra is the first model to outperform human experts on MMLU,” said CEO Sundar Pichai.

Next in line is Gemini Pro, which will power many Google AI services going forward and has now become the backbone of Google Bard. Pro is Google’s best model for scaling across a wide range of tasks, the company said.

Gemini Nano is the lightweight flavor of Gemini that is meant to be used directly on devices and on-device tasks, including in edge environments. It can also run natively and offline on Android devices. Google Pixel 8 users will get new features this year via Nano.

“Gemini is also our most flexible model yet — able to efficiently run on everything from data centers to mobile devices,” said Demis Hassabis, CEO and Co-Founder of Google AI business unit DeepMind in a blog post. “Its state-of-the-art capabilities will significantly enhance the way developers and enterprise customers build and scale with AI.”

Beginning Dec. 13, enterprise customers and developers can get access to Gemini Pro via the Gemini API in Google Cloud Vertex AI or in Google AI Studio, the company’s free web-based developer tool to prototype and launch applications quickly.

Advance Coding Abilities

One key use case for Gemini is advanced coding.

Gemini can understand, explain and generate high-quality code in the world’s most popular programming languages such as Python, Java, C++, and Go.

Its ability to work across languages and reason about complex information makes it one of the leading foundation models for coding in the world, said Hassabis.

“Gemini can be used as the engine for more advanced coding systems,” he said. “We’re excited for programmers to increasingly use highly capable AI models as collaborative tools that can help them reason about the problems, propose code designs and assist with implementation — so they can release apps and design better services, faster.”

Google CEO’s On AI Future

Google’s CEO says his company worked hard to ensure Gemini’s safety and responsibility, both through internal and external testing as well as via rigorous security risk assessments.

“AI has the potential to create opportunities — from the everyday to the extraordinary — for people everywhere,” said Pichai. “It will bring new waves of innovation and economic progress and drive knowledge, learning, creativity and productivity on a scale we haven’t seen before.”

Google AI goal with Gemini is to make AI more helpful for everyone in the world.

“This is incredible momentum, and yet, we’re only beginning to scratch the surface of what’s possible,” Pichai said. “I believe the transition we are seeing right now with AI will be the most profound in our lifetimes, far bigger than the shift to mobile or to the web before it.”