Google updates Bard Chatbot with ‘Gemini’ AI while chasing ChatGPT

For more than a year, Google has been racing to build technology that can match ChatGPT, the eye-opening chatbot offered by San Francisco AI startup OpenAI.

On Wednesday, the tech giant took another step in the ongoing race, launching a new version of its chatbot, Google Bard. The updated bot is available to English speakers in more than 170 regions and countries, including the US, starting immediately, and is powered by new AI technology called Gemini, which the company has been developing since the beginning of the year.

“This is the beginning of the Gemini era,” Google CEO Sundar Pichai said in an interview. “It’s the fulfillment of the vision we had when we created Google DeepMind, the company’s artificial intelligence lab. Google will roll out three different versions of the technology in a wide range of products and services in the coming months,” he said.

Mr. Pichai and Demis Hassabis, who oversee Google DeepMind, said Gemini was more powerful than Google’s previous chatbot technologies, and could generate more accurate responses and come closer to mimicking human reasoning in some situations.

“We are very pleased with Gemini’s performance,” Dr. Hassabis said.

When OpenAI wowed the world with its AI-powered ChatGPT chatbot late last year, Google was stunned. The tech giant has spent years developing similar technology, but like other tech giants — most notably Meta — has been reluctant to release technology that could generate biased, false or toxic information.

In March, Google launched its chatbot, Bard, to average reviews. A month later, the company announced that it had merged its two artificial intelligence labs — Google Brain and DeepMind — bringing together more than 2,000 researchers and engineers. And in May, at the groundbreaking Google I/O conference, it announced that Google’s new DeepMind lab had begun developing Gemini.

After founding Brain Lab in 2011, Google acquired DeepMind in 2014, paying $650 million for the London AI startup. DeepMind has operated largely independently of Brain Lab and the rest of Google for a decade, and even tried to exit the company in 2017. But as Google struggled to catch up with OpenAI, Mr. Pichai merged the two labs under Dr. Hassabis, a neuroscientist. Who co-founded DeepMind.

Google has released benchmark results claiming that the most powerful version of Gemini outperformed the latest OpenAI technology, GPT-4, in several key areas. Pichai said it is better at generating computer codes than previous Google techniques, and can summarize news articles and other text documents more accurately.

Gemini is also designed to analyze images and sounds, but these skills won’t be incorporated into the Bard chatbot until later.

Google has built three versions of Gemini with three different skill sets. The largest, the Ultra, is designed to handle complex tasks and will debut next year. Pro, the mid-tier offering, will be rolled out to several Google services, starting Wednesday, using the Bard chatbot. The Nano, the smaller version, will power some features on the Pixel 8 Pro smartphone, such as summarizing audio recordings and offering suggested text replies in WhatsApp starting Wednesday.

Gemini is what scientists call a large language model, or LLM, a complex mathematical system that can learn skills by analyzing massive amounts of data, including digital books, Wikipedia articles and online bulletin boards. By identifying patterns in all this text, the LLM learns to create text on its own. This means he can write research papers, create computer code, and even carry on a conversation.

Through Gemini, Google has also trained technology on digital images and sounds. This is what researchers call a “multimodal” system, meaning it can analyze and respond to images and sounds. If you give him a math problem involving lines, shapes, and other images, for example, he might answer the same way a high school student might.

However, this piece of technology won’t be available to consumers until sometime next year. Google also acknowledged that Gemini, like similar systems, is vulnerable to errors. Can get facts wrong or even “hallucinate” – make things up.

Google Cloud, which provides AI and computing services to other companies, has been eager to provide customers with Gemini as it competes for deals against OpenAI and Microsoft. After OpenAI briefly forced Sam Altman, its CEO, to resign last month, leaving the company in limbo, Google Cloud created a migration plan in an attempt to poach its rival’s customers.

Customers can pay Google the same price as their current OpenAI price and receive cloud credits or discounts.

Google said cloud customers will have access to Gemini Pro — the middle-tier offering — on December 13. Mr Pichai said some strangers were now testing Gemini Ultra – the most powerful version of the technology.

Although Google has spent the past year racing to regain AI leadership from OpenAI, Mr. Pichai said there is enough room in the market for all AI providers.

“It’s far from a zero-sum game,” Mr. Pichai said. “We have a sense of excitement about what we are launching. We also realize that it is very early days because we can see the progress we are making in the follow-up.